diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,297152 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999941101981918,
+  "eval_steps": 500,
+  "global_step": 42446,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8460450258712872,
+      "learning_rate": 1.569858712715856e-08,
+      "loss": 0.7462,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7206315324613786,
+      "learning_rate": 3.139717425431712e-08,
+      "loss": 1.0016,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5907194231701847,
+      "learning_rate": 4.709576138147567e-08,
+      "loss": 0.9054,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.541331082403491,
+      "learning_rate": 6.279434850863424e-08,
+      "loss": 1.1022,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.718668193001434,
+      "learning_rate": 7.849293563579278e-08,
+      "loss": 0.9501,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3587551795455832,
+      "learning_rate": 9.419152276295134e-08,
+      "loss": 1.0958,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.1976386712588525,
+      "learning_rate": 1.098901098901099e-07,
+      "loss": 0.8728,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.029345675046727,
+      "learning_rate": 1.2558869701726847e-07,
+      "loss": 0.9346,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7479437013628525,
+      "learning_rate": 1.4128728414442702e-07,
+      "loss": 0.9687,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.087996495219446,
+      "learning_rate": 1.5698587127158556e-07,
+      "loss": 1.0847,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6271521995407747,
+      "learning_rate": 1.726844583987441e-07,
+      "loss": 1.0091,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.555699396272173,
+      "learning_rate": 1.8838304552590267e-07,
+      "loss": 1.0548,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.286935746581951,
+      "learning_rate": 2.0408163265306121e-07,
+      "loss": 0.9347,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4497662275346257,
+      "learning_rate": 2.197802197802198e-07,
+      "loss": 0.993,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4855440534570863,
+      "learning_rate": 2.3547880690737838e-07,
+      "loss": 0.9765,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2289745808852293,
+      "learning_rate": 2.5117739403453695e-07,
+      "loss": 0.9495,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8126900605507084,
+      "learning_rate": 2.6687598116169546e-07,
+      "loss": 0.9735,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8745560687184897,
+      "learning_rate": 2.8257456828885403e-07,
+      "loss": 1.0236,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7569851455966737,
+      "learning_rate": 2.982731554160126e-07,
+      "loss": 0.7269,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.053087902636629,
+      "learning_rate": 3.139717425431711e-07,
+      "loss": 0.9729,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.702960319581147,
+      "learning_rate": 3.296703296703297e-07,
+      "loss": 0.8906,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.430907403431023,
+      "learning_rate": 3.453689167974882e-07,
+      "loss": 1.0532,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.759494237825174,
+      "learning_rate": 3.610675039246468e-07,
+      "loss": 1.0936,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6146025905493517,
+      "learning_rate": 3.7676609105180534e-07,
+      "loss": 0.9863,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.148262685372617,
+      "learning_rate": 3.924646781789639e-07,
+      "loss": 1.161,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.246788262975213,
+      "learning_rate": 4.0816326530612243e-07,
+      "loss": 1.0982,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.653684425476009,
+      "learning_rate": 4.2386185243328105e-07,
+      "loss": 0.888,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8934230930506697,
+      "learning_rate": 4.395604395604396e-07,
+      "loss": 0.9424,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.972844157140064,
+      "learning_rate": 4.5525902668759813e-07,
+      "loss": 1.0147,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.155966209171864,
+      "learning_rate": 4.7095761381475676e-07,
+      "loss": 0.9899,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1733700890065992,
+      "learning_rate": 4.866562009419152e-07,
+      "loss": 1.1162,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2931948564844338,
+      "learning_rate": 5.023547880690739e-07,
+      "loss": 0.9439,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.477577194518864,
+      "learning_rate": 5.180533751962324e-07,
+      "loss": 1.117,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.035736590965567,
+      "learning_rate": 5.337519623233909e-07,
+      "loss": 0.9963,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0507900431606405,
+      "learning_rate": 5.494505494505495e-07,
+      "loss": 0.9973,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.079686883445612,
+      "learning_rate": 5.651491365777081e-07,
+      "loss": 0.9219,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8294794435498594,
+      "learning_rate": 5.808477237048666e-07,
+      "loss": 0.9901,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7915001455978623,
+      "learning_rate": 5.965463108320252e-07,
+      "loss": 0.7264,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2000074612297214,
+      "learning_rate": 6.122448979591837e-07,
+      "loss": 1.1214,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7934605158582612,
+      "learning_rate": 6.279434850863422e-07,
+      "loss": 0.7103,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7589719712326537,
+      "learning_rate": 6.436420722135009e-07,
+      "loss": 0.7456,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4718912769576247,
+      "learning_rate": 6.593406593406594e-07,
+      "loss": 1.0092,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0157824522258347,
+      "learning_rate": 6.750392464678179e-07,
+      "loss": 1.0511,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7941235689101007,
+      "learning_rate": 6.907378335949764e-07,
+      "loss": 0.8714,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8530071170115088,
+      "learning_rate": 7.064364207221351e-07,
+      "loss": 0.9359,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.200276151165166,
+      "learning_rate": 7.221350078492937e-07,
+      "loss": 0.9888,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7394228125963971,
+      "learning_rate": 7.378335949764521e-07,
+      "loss": 0.7091,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0631782893422637,
+      "learning_rate": 7.535321821036107e-07,
+      "loss": 1.015,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.840163268292435,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 0.9105,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9239730616981694,
+      "learning_rate": 7.849293563579278e-07,
+      "loss": 1.0043,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9063248435213376,
+      "learning_rate": 8.006279434850864e-07,
+      "loss": 0.9356,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.975841324744863,
+      "learning_rate": 8.163265306122449e-07,
+      "loss": 0.9368,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.991080124890634,
+      "learning_rate": 8.320251177394035e-07,
+      "loss": 1.1174,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.32648371720962,
+      "learning_rate": 8.477237048665621e-07,
+      "loss": 0.9855,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9138841486938496,
+      "learning_rate": 8.634222919937206e-07,
+      "loss": 0.9902,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7014612242023528,
+      "learning_rate": 8.791208791208792e-07,
+      "loss": 0.9094,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2952413496889017,
+      "learning_rate": 8.948194662480378e-07,
+      "loss": 0.853,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9543466273132275,
+      "learning_rate": 9.105180533751963e-07,
+      "loss": 0.9171,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7587772239054165,
+      "learning_rate": 9.262166405023548e-07,
+      "loss": 0.8071,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.08404504957823,
+      "learning_rate": 9.419152276295135e-07,
+      "loss": 1.0576,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9395298505067329,
+      "learning_rate": 9.57613814756672e-07,
+      "loss": 0.9717,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.883754912344211,
+      "learning_rate": 9.733124018838304e-07,
+      "loss": 0.9536,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7898296484676324,
+      "learning_rate": 9.890109890109891e-07,
+      "loss": 1.0393,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9611608592178549,
+      "learning_rate": 1.0047095761381478e-06,
+      "loss": 1.1481,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6285926648646845,
+      "learning_rate": 1.0204081632653063e-06,
+      "loss": 0.9601,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.195871832495232,
+      "learning_rate": 1.0361067503924647e-06,
+      "loss": 1.0662,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7278512398515504,
+      "learning_rate": 1.0518053375196232e-06,
+      "loss": 0.8972,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7762503367951534,
+      "learning_rate": 1.0675039246467819e-06,
+      "loss": 0.837,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.705873223671486,
+      "learning_rate": 1.0832025117739405e-06,
+      "loss": 1.015,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7915422632883975,
+      "learning_rate": 1.098901098901099e-06,
+      "loss": 0.9931,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.402673188935797,
+      "learning_rate": 1.1145996860282575e-06,
+      "loss": 1.0415,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8141277220212206,
+      "learning_rate": 1.1302982731554161e-06,
+      "loss": 0.8986,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7665111790855551,
+      "learning_rate": 1.1459968602825746e-06,
+      "loss": 0.6675,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6754193729594724,
+      "learning_rate": 1.1616954474097333e-06,
+      "loss": 0.9112,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9329373913702053,
+      "learning_rate": 1.1773940345368917e-06,
+      "loss": 1.0213,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7408619251823372,
+      "learning_rate": 1.1930926216640504e-06,
+      "loss": 0.8736,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6713491802181992,
+      "learning_rate": 1.2087912087912089e-06,
+      "loss": 0.806,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8077162871497467,
+      "learning_rate": 1.2244897959183673e-06,
+      "loss": 0.99,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0057531262392994,
+      "learning_rate": 1.240188383045526e-06,
+      "loss": 0.9605,
+      "step": 79
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7579569165698241,
+      "learning_rate": 1.2558869701726845e-06,
+      "loss": 0.7157,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6073771959902685,
+      "learning_rate": 1.2715855572998431e-06,
+      "loss": 0.9095,
+      "step": 81
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8125456570324496,
+      "learning_rate": 1.2872841444270018e-06,
+      "loss": 1.0042,
+      "step": 82
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7395101315155517,
+      "learning_rate": 1.30298273155416e-06,
+      "loss": 1.0877,
+      "step": 83
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0690020560886797,
+      "learning_rate": 1.3186813186813187e-06,
+      "loss": 0.9171,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7975260209469566,
+      "learning_rate": 1.3343799058084774e-06,
+      "loss": 0.7143,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6954591419593534,
+      "learning_rate": 1.3500784929356359e-06,
+      "loss": 0.9295,
+      "step": 86
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.734049189114901,
+      "learning_rate": 1.3657770800627946e-06,
+      "loss": 0.7282,
+      "step": 87
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4594636281623798,
+      "learning_rate": 1.3814756671899528e-06,
+      "loss": 0.7326,
+      "step": 88
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5138813385316199,
+      "learning_rate": 1.3971742543171115e-06,
+      "loss": 0.9081,
+      "step": 89
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4682298034558556,
+      "learning_rate": 1.4128728414442702e-06,
+      "loss": 0.8502,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3479757008806175,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 1.0367,
+      "step": 91
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.678037611345193,
+      "learning_rate": 1.4442700156985873e-06,
+      "loss": 1.0462,
+      "step": 92
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6512471322965054,
+      "learning_rate": 1.459968602825746e-06,
+      "loss": 0.9309,
+      "step": 93
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5816262585916547,
+      "learning_rate": 1.4756671899529042e-06,
+      "loss": 0.9327,
+      "step": 94
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7576066714766521,
+      "learning_rate": 1.491365777080063e-06,
+      "loss": 0.6746,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6173833876735633,
+      "learning_rate": 1.5070643642072214e-06,
+      "loss": 0.8179,
+      "step": 96
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.774749687476227,
+      "learning_rate": 1.52276295133438e-06,
+      "loss": 0.8882,
+      "step": 97
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7314245799758982,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 0.7265,
+      "step": 98
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4927029300662777,
+      "learning_rate": 1.554160125588697e-06,
+      "loss": 1.0968,
+      "step": 99
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.699018466624318,
+      "learning_rate": 1.5698587127158556e-06,
+      "loss": 0.9592,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6594505226532759,
+      "learning_rate": 1.5855572998430143e-06,
+      "loss": 0.9846,
+      "step": 101
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6078773090310121,
+      "learning_rate": 1.6012558869701728e-06,
+      "loss": 1.0441,
+      "step": 102
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8527288271828772,
+      "learning_rate": 1.6169544740973315e-06,
+      "loss": 0.9879,
+      "step": 103
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1075417607260483,
+      "learning_rate": 1.6326530612244897e-06,
+      "loss": 0.9529,
+      "step": 104
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0131089043183827,
+      "learning_rate": 1.6483516483516484e-06,
+      "loss": 0.8095,
+      "step": 105
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2238330331570135,
+      "learning_rate": 1.664050235478807e-06,
+      "loss": 0.9625,
+      "step": 106
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7549909316632963,
+      "learning_rate": 1.6797488226059655e-06,
+      "loss": 0.94,
+      "step": 107
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.492889219757501,
+      "learning_rate": 1.6954474097331242e-06,
+      "loss": 0.6414,
+      "step": 108
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5117545506939705,
+      "learning_rate": 1.7111459968602829e-06,
+      "loss": 0.8253,
+      "step": 109
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2042562492151574,
+      "learning_rate": 1.7268445839874411e-06,
+      "loss": 0.8855,
+      "step": 110
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7020564620730672,
+      "learning_rate": 1.7425431711145998e-06,
+      "loss": 0.8439,
+      "step": 111
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7478540769836374,
+      "learning_rate": 1.7582417582417585e-06,
+      "loss": 0.7285,
+      "step": 112
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6842481832451077,
+      "learning_rate": 1.773940345368917e-06,
+      "loss": 0.9207,
+      "step": 113
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.772397367957122,
+      "learning_rate": 1.7896389324960756e-06,
+      "loss": 1.05,
+      "step": 114
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7350741739056468,
+      "learning_rate": 1.8053375196232339e-06,
+      "loss": 0.7228,
+      "step": 115
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9184481855301636,
+      "learning_rate": 1.8210361067503925e-06,
+      "loss": 0.932,
+      "step": 116
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5653614808057856,
+      "learning_rate": 1.8367346938775512e-06,
+      "loss": 0.9541,
+      "step": 117
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7013769294451977,
+      "learning_rate": 1.8524332810047097e-06,
+      "loss": 1.0359,
+      "step": 118
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.584859911678946,
+      "learning_rate": 1.8681318681318684e-06,
+      "loss": 0.9004,
+      "step": 119
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3687242090772394,
+      "learning_rate": 1.883830455259027e-06,
+      "loss": 0.9944,
+      "step": 120
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5158104361417704,
+      "learning_rate": 1.8995290423861853e-06,
+      "loss": 0.9589,
+      "step": 121
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5824724685911926,
+      "learning_rate": 1.915227629513344e-06,
+      "loss": 0.9672,
+      "step": 122
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6871851528991626,
+      "learning_rate": 1.9309262166405024e-06,
+      "loss": 0.9294,
+      "step": 123
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6885536497836706,
+      "learning_rate": 1.946624803767661e-06,
+      "loss": 0.7682,
+      "step": 124
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.810625080118532,
+      "learning_rate": 1.9623233908948198e-06,
+      "loss": 1.0624,
+      "step": 125
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5159343456480203,
+      "learning_rate": 1.9780219780219782e-06,
+      "loss": 0.9702,
+      "step": 126
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.728083171937014,
+      "learning_rate": 1.9937205651491367e-06,
+      "loss": 0.9291,
+      "step": 127
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7078504105436638,
+      "learning_rate": 2.0094191522762956e-06,
+      "loss": 0.9571,
+      "step": 128
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6397345312913376,
+      "learning_rate": 2.0251177394034536e-06,
+      "loss": 1.0045,
+      "step": 129
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6950767280422028,
+      "learning_rate": 2.0408163265306125e-06,
+      "loss": 1.1177,
+      "step": 130
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7214510732274362,
+      "learning_rate": 2.056514913657771e-06,
+      "loss": 0.8886,
+      "step": 131
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4504670668382724,
+      "learning_rate": 2.0722135007849294e-06,
+      "loss": 0.8589,
+      "step": 132
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5327517987728936,
+      "learning_rate": 2.0879120879120883e-06,
+      "loss": 0.914,
+      "step": 133
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9125018041621102,
+      "learning_rate": 2.1036106750392464e-06,
+      "loss": 1.0105,
+      "step": 134
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.829521029585462,
+      "learning_rate": 2.1193092621664052e-06,
+      "loss": 1.0047,
+      "step": 135
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0220068567921463,
+      "learning_rate": 2.1350078492935637e-06,
+      "loss": 0.8653,
+      "step": 136
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8101402512040083,
+      "learning_rate": 2.150706436420722e-06,
+      "loss": 0.8834,
+      "step": 137
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8941164564708524,
+      "learning_rate": 2.166405023547881e-06,
+      "loss": 0.9635,
+      "step": 138
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6776282657665365,
+      "learning_rate": 2.1821036106750395e-06,
+      "loss": 0.9051,
+      "step": 139
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.767795325914773,
+      "learning_rate": 2.197802197802198e-06,
+      "loss": 0.8733,
+      "step": 140
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.847015926327031,
+      "learning_rate": 2.2135007849293564e-06,
+      "loss": 0.8734,
+      "step": 141
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6623607259593434,
+      "learning_rate": 2.229199372056515e-06,
+      "loss": 0.9349,
+      "step": 142
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7384461782736158,
+      "learning_rate": 2.244897959183674e-06,
+      "loss": 0.9628,
+      "step": 143
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7328028731002392,
+      "learning_rate": 2.2605965463108323e-06,
+      "loss": 0.9883,
+      "step": 144
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7638154271263282,
+      "learning_rate": 2.2762951334379907e-06,
+      "loss": 0.9478,
+      "step": 145
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6253891112074832,
+      "learning_rate": 2.291993720565149e-06,
+      "loss": 1.0515,
+      "step": 146
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9424593559569008,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 1.014,
+      "step": 147
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.88394764250397,
+      "learning_rate": 2.3233908948194665e-06,
+      "loss": 0.7674,
+      "step": 148
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.618040908951974,
+      "learning_rate": 2.339089481946625e-06,
+      "loss": 1.0285,
+      "step": 149
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8222977985259934,
+      "learning_rate": 2.3547880690737835e-06,
+      "loss": 0.7573,
+      "step": 150
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6525221171183473,
+      "learning_rate": 2.370486656200942e-06,
+      "loss": 0.8527,
+      "step": 151
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8573905588254263,
+      "learning_rate": 2.386185243328101e-06,
+      "loss": 1.0047,
+      "step": 152
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9174043313584799,
+      "learning_rate": 2.4018838304552593e-06,
+      "loss": 0.8877,
+      "step": 153
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7827435199934967,
+      "learning_rate": 2.4175824175824177e-06,
+      "loss": 0.7475,
+      "step": 154
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7306191781177225,
+      "learning_rate": 2.4332810047095766e-06,
+      "loss": 0.9294,
+      "step": 155
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7631544800057204,
+      "learning_rate": 2.4489795918367347e-06,
+      "loss": 0.7263,
+      "step": 156
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6750653013048458,
+      "learning_rate": 2.4646781789638936e-06,
+      "loss": 0.8758,
+      "step": 157
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8709548560987914,
+      "learning_rate": 2.480376766091052e-06,
+      "loss": 0.9347,
+      "step": 158
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5910997511034637,
+      "learning_rate": 2.4960753532182105e-06,
+      "loss": 0.8752,
+      "step": 159
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5698992036809458,
+      "learning_rate": 2.511773940345369e-06,
+      "loss": 0.9143,
+      "step": 160
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.809011535310256,
+      "learning_rate": 2.5274725274725274e-06,
+      "loss": 0.8518,
+      "step": 161
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.014545689184869,
+      "learning_rate": 2.5431711145996863e-06,
+      "loss": 0.9683,
+      "step": 162
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8785002431577622,
+      "learning_rate": 2.5588697017268448e-06,
+      "loss": 0.973,
+      "step": 163
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7304554700361456,
+      "learning_rate": 2.5745682888540036e-06,
+      "loss": 0.7104,
+      "step": 164
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5439234443774095,
+      "learning_rate": 2.5902668759811617e-06,
+      "loss": 0.9302,
+      "step": 165
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.670121053653117,
+      "learning_rate": 2.60596546310832e-06,
+      "loss": 0.9057,
+      "step": 166
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7170684452888467,
+      "learning_rate": 2.621664050235479e-06,
+      "loss": 0.7883,
+      "step": 167
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6122008900167564,
+      "learning_rate": 2.6373626373626375e-06,
+      "loss": 0.8676,
+      "step": 168
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6536853433821481,
+      "learning_rate": 2.6530612244897964e-06,
+      "loss": 0.9309,
+      "step": 169
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6393646046540775,
+      "learning_rate": 2.668759811616955e-06,
+      "loss": 0.913,
+      "step": 170
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6285782523533185,
+      "learning_rate": 2.684458398744113e-06,
+      "loss": 0.9212,
+      "step": 171
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5028069400057655,
+      "learning_rate": 2.7001569858712718e-06,
+      "loss": 0.8576,
+      "step": 172
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.771363510844233,
+      "learning_rate": 2.7158555729984302e-06,
+      "loss": 0.9444,
+      "step": 173
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6441260398946904,
+      "learning_rate": 2.731554160125589e-06,
+      "loss": 0.8884,
+      "step": 174
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7394358297462535,
+      "learning_rate": 2.7472527472527476e-06,
+      "loss": 0.9386,
+      "step": 175
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9145554778044618,
+      "learning_rate": 2.7629513343799056e-06,
+      "loss": 0.9091,
+      "step": 176
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8326651414385802,
+      "learning_rate": 2.7786499215070645e-06,
+      "loss": 0.9776,
+      "step": 177
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7471843373009295,
+      "learning_rate": 2.794348508634223e-06,
+      "loss": 0.6777,
+      "step": 178
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0451486069816194,
+      "learning_rate": 2.810047095761382e-06,
+      "loss": 0.8991,
+      "step": 179
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9441789786333548,
+      "learning_rate": 2.8257456828885403e-06,
+      "loss": 0.9645,
+      "step": 180
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2369967472038206,
+      "learning_rate": 2.8414442700156984e-06,
+      "loss": 1.03,
+      "step": 181
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7342641696765728,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 1.0984,
+      "step": 182
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8759492069799892,
+      "learning_rate": 2.8728414442700157e-06,
+      "loss": 0.7885,
+      "step": 183
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.723160257481204,
+      "learning_rate": 2.8885400313971746e-06,
+      "loss": 0.9367,
+      "step": 184
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.950821190106223,
+      "learning_rate": 2.904238618524333e-06,
+      "loss": 0.9736,
+      "step": 185
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8394725566786159,
+      "learning_rate": 2.919937205651492e-06,
+      "loss": 1.0015,
+      "step": 186
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6245286496773275,
+      "learning_rate": 2.93563579277865e-06,
+      "loss": 0.9073,
+      "step": 187
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7782529671411222,
+      "learning_rate": 2.9513343799058085e-06,
+      "loss": 0.9411,
+      "step": 188
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6361229159886466,
+      "learning_rate": 2.9670329670329673e-06,
+      "loss": 0.8528,
+      "step": 189
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0049419242809945,
+      "learning_rate": 2.982731554160126e-06,
+      "loss": 0.9455,
+      "step": 190
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4037348937300853,
+      "learning_rate": 2.9984301412872847e-06,
+      "loss": 0.8488,
+      "step": 191
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.100201160123796,
+      "learning_rate": 3.0141287284144427e-06,
+      "loss": 1.0515,
+      "step": 192
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5454708269000643,
+      "learning_rate": 3.029827315541601e-06,
+      "loss": 0.9684,
+      "step": 193
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9096297880195001,
+      "learning_rate": 3.04552590266876e-06,
+      "loss": 0.8743,
+      "step": 194
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7024083988491807,
+      "learning_rate": 3.0612244897959185e-06,
+      "loss": 0.9903,
+      "step": 195
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8200337962394724,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.6807,
+      "step": 196
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7853739712417193,
+      "learning_rate": 3.092621664050236e-06,
+      "loss": 0.7325,
+      "step": 197
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6447243669836342,
+      "learning_rate": 3.108320251177394e-06,
+      "loss": 0.9179,
+      "step": 198
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8081000617802951,
+      "learning_rate": 3.124018838304553e-06,
+      "loss": 0.6872,
+      "step": 199
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.771853774534688,
+      "learning_rate": 3.1397174254317113e-06,
+      "loss": 0.9024,
+      "step": 200
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6803031573850955,
+      "learning_rate": 3.15541601255887e-06,
+      "loss": 0.8962,
+      "step": 201
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7906948812064489,
+      "learning_rate": 3.1711145996860286e-06,
+      "loss": 0.7009,
+      "step": 202
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4494396326397498,
+      "learning_rate": 3.1868131868131867e-06,
+      "loss": 0.9252,
+      "step": 203
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5724446495384325,
+      "learning_rate": 3.2025117739403456e-06,
+      "loss": 0.9953,
+      "step": 204
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6269575814737576,
+      "learning_rate": 3.218210361067504e-06,
+      "loss": 0.9114,
+      "step": 205
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5711430360594705,
+      "learning_rate": 3.233908948194663e-06,
+      "loss": 0.8782,
+      "step": 206
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4364746265041175,
+      "learning_rate": 3.2496075353218214e-06,
+      "loss": 0.9665,
+      "step": 207
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9166162974817362,
+      "learning_rate": 3.2653061224489794e-06,
+      "loss": 0.9278,
+      "step": 208
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7513039446031353,
+      "learning_rate": 3.2810047095761383e-06,
+      "loss": 0.6957,
+      "step": 209
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9833174269997278,
+      "learning_rate": 3.2967032967032968e-06,
+      "loss": 0.9993,
+      "step": 210
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6029471604798737,
+      "learning_rate": 3.3124018838304557e-06,
+      "loss": 0.988,
+      "step": 211
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6761423484477576,
+      "learning_rate": 3.328100470957614e-06,
+      "loss": 0.9565,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.489073628190942,
+      "learning_rate": 3.343799058084773e-06,
+      "loss": 0.8434,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.925049597474539,
+      "learning_rate": 3.359497645211931e-06,
+      "loss": 0.8494,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6516982721230924,
+      "learning_rate": 3.3751962323390895e-06,
+      "loss": 0.8999,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6832253765739942,
+      "learning_rate": 3.3908948194662484e-06,
+      "loss": 1.041,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.878088181030581,
+      "learning_rate": 3.406593406593407e-06,
+      "loss": 0.9722,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6209465098766387,
+      "learning_rate": 3.4222919937205657e-06,
+      "loss": 0.9235,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6759573366278044,
+      "learning_rate": 3.4379905808477238e-06,
+      "loss": 0.8312,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6243046240364065,
+      "learning_rate": 3.4536891679748822e-06,
+      "loss": 0.9062,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9718848666847388,
+      "learning_rate": 3.469387755102041e-06,
+      "loss": 0.9293,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6755479856284274,
+      "learning_rate": 3.4850863422291996e-06,
+      "loss": 0.8377,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.72514412053546,
+      "learning_rate": 3.5007849293563585e-06,
+      "loss": 0.729,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6120278563160824,
+      "learning_rate": 3.516483516483517e-06,
+      "loss": 1.0568,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6064230993258377,
+      "learning_rate": 3.532182103610675e-06,
+      "loss": 0.8876,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6514253990630532,
+      "learning_rate": 3.547880690737834e-06,
+      "loss": 0.8577,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.096162614190356,
+      "learning_rate": 3.5635792778649923e-06,
+      "loss": 0.9161,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1467725585521187,
+      "learning_rate": 3.5792778649921512e-06,
+      "loss": 0.9189,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.543616921394075,
+      "learning_rate": 3.5949764521193097e-06,
+      "loss": 0.8943,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6380255184072785,
+      "learning_rate": 3.6106750392464677e-06,
+      "loss": 0.9599,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1271476897097386,
+      "learning_rate": 3.6263736263736266e-06,
+      "loss": 0.879,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7765652746857012,
+      "learning_rate": 3.642072213500785e-06,
+      "loss": 0.7191,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.027936855342829,
+      "learning_rate": 3.657770800627944e-06,
+      "loss": 0.9105,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8023499918124267,
+      "learning_rate": 3.6734693877551024e-06,
+      "loss": 0.6709,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7920551618514433,
+      "learning_rate": 3.6891679748822605e-06,
+      "loss": 1.0095,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6919185442341989,
+      "learning_rate": 3.7048665620094194e-06,
+      "loss": 1.0194,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8164655554130533,
+      "learning_rate": 3.720565149136578e-06,
+      "loss": 0.8882,
+      "step": 237
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9212758289926553,
+      "learning_rate": 3.7362637362637367e-06,
+      "loss": 1.0315,
+      "step": 238
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7486040750788994,
+      "learning_rate": 3.751962323390895e-06,
+      "loss": 0.6983,
+      "step": 239
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7744658159793874,
+      "learning_rate": 3.767660910518054e-06,
+      "loss": 0.7872,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7430606915557436,
+      "learning_rate": 3.783359497645212e-06,
+      "loss": 0.7293,
+      "step": 241
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.742196612971735,
+      "learning_rate": 3.7990580847723706e-06,
+      "loss": 0.9294,
+      "step": 242
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4637875207024553,
+      "learning_rate": 3.814756671899529e-06,
+      "loss": 0.8535,
+      "step": 243
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5209887827395956,
+      "learning_rate": 3.830455259026688e-06,
+      "loss": 1.0744,
+      "step": 244
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7540364955999892,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.9274,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5184510117815146,
+      "learning_rate": 3.861852433281005e-06,
+      "loss": 0.9343,
+      "step": 246
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6134621034645797,
+      "learning_rate": 3.877551020408164e-06,
+      "loss": 0.8976,
+      "step": 247
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7837154318671415,
+      "learning_rate": 3.893249607535322e-06,
+      "loss": 0.9135,
+      "step": 248
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3334639119339546,
+      "learning_rate": 3.908948194662481e-06,
+      "loss": 1.0111,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0878860250385327,
+      "learning_rate": 3.9246467817896395e-06,
+      "loss": 0.9491,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.092826715169234,
+      "learning_rate": 3.940345368916798e-06,
+      "loss": 0.9012,
+      "step": 251
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6266810841166537,
+      "learning_rate": 3.9560439560439565e-06,
+      "loss": 0.8658,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5318370804290868,
+      "learning_rate": 3.9717425431711145e-06,
+      "loss": 1.0041,
+      "step": 253
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7784513910320596,
+      "learning_rate": 3.987441130298273e-06,
+      "loss": 0.8168,
+      "step": 254
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7629587684239383,
+      "learning_rate": 4.003139717425432e-06,
+      "loss": 0.8658,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5462106558788018,
+      "learning_rate": 4.018838304552591e-06,
+      "loss": 0.7858,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4480463248415534,
+      "learning_rate": 4.034536891679749e-06,
+      "loss": 0.8903,
+      "step": 257
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7684774651827488,
+      "learning_rate": 4.050235478806907e-06,
+      "loss": 0.8704,
+      "step": 258
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.712890917840925,
+      "learning_rate": 4.065934065934066e-06,
+      "loss": 0.7651,
+      "step": 259
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.332095165221455,
+      "learning_rate": 4.081632653061225e-06,
+      "loss": 0.9048,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5427886247370297,
+      "learning_rate": 4.097331240188384e-06,
+      "loss": 0.7422,
+      "step": 261
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7451199445807843,
+      "learning_rate": 4.113029827315542e-06,
+      "loss": 0.694,
+      "step": 262
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8217157710315472,
+      "learning_rate": 4.1287284144427e-06,
+      "loss": 0.9077,
+      "step": 263
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5982581971989942,
+      "learning_rate": 4.144427001569859e-06,
+      "loss": 0.8924,
+      "step": 264
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4157570588262873,
+      "learning_rate": 4.160125588697018e-06,
+      "loss": 0.891,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.682598289365908,
+      "learning_rate": 4.175824175824177e-06,
+      "loss": 0.878,
+      "step": 266
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6562511875237174,
+      "learning_rate": 4.191522762951335e-06,
+      "loss": 1.0662,
+      "step": 267
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6373652526047278,
+      "learning_rate": 4.207221350078493e-06,
+      "loss": 0.9789,
+      "step": 268
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7113294000872286,
+      "learning_rate": 4.222919937205652e-06,
+      "loss": 0.8882,
+      "step": 269
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.819980126979845,
+      "learning_rate": 4.2386185243328105e-06,
+      "loss": 0.9415,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7532702339073831,
+      "learning_rate": 4.254317111459969e-06,
+      "loss": 0.6899,
+      "step": 271
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7666392997942506,
+      "learning_rate": 4.270015698587127e-06,
+      "loss": 0.9009,
+      "step": 272
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.638456596091315,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.9071,
+      "step": 273
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6075986740545918,
+      "learning_rate": 4.301412872841444e-06,
+      "loss": 0.9549,
+      "step": 274
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0882329952256953,
+      "learning_rate": 4.317111459968603e-06,
+      "loss": 0.9511,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6905906676245892,
+      "learning_rate": 4.332810047095762e-06,
+      "loss": 0.9485,
+      "step": 276
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.788442944099477,
+      "learning_rate": 4.34850863422292e-06,
+      "loss": 0.91,
+      "step": 277
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7145751512497558,
+      "learning_rate": 4.364207221350079e-06,
+      "loss": 0.9663,
+      "step": 278
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8281627593117127,
+      "learning_rate": 4.379905808477237e-06,
+      "loss": 0.8963,
+      "step": 279
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.573997823359264,
+      "learning_rate": 4.395604395604396e-06,
+      "loss": 0.9589,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7768596940410974,
+      "learning_rate": 4.411302982731555e-06,
+      "loss": 0.9299,
+      "step": 281
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.599625503906451,
+      "learning_rate": 4.427001569858713e-06,
+      "loss": 0.9293,
+      "step": 282
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9318794385457432,
+      "learning_rate": 4.442700156985872e-06,
+      "loss": 0.9173,
+      "step": 283
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6651527670102098,
+      "learning_rate": 4.45839874411303e-06,
+      "loss": 0.8385,
+      "step": 284
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7115752325635438,
+      "learning_rate": 4.474097331240189e-06,
+      "loss": 1.0471,
+      "step": 285
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7704220938992883,
+      "learning_rate": 4.489795918367348e-06,
+      "loss": 1.0104,
+      "step": 286
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1452804228108158,
+      "learning_rate": 4.505494505494506e-06,
+      "loss": 0.9042,
+      "step": 287
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.688137776079435,
+      "learning_rate": 4.5211930926216645e-06,
+      "loss": 0.9548,
+      "step": 288
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.776597626211792,
+      "learning_rate": 4.5368916797488226e-06,
+      "loss": 0.8996,
+      "step": 289
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.16966773010081,
+      "learning_rate": 4.5525902668759815e-06,
+      "loss": 0.8292,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6010916798527417,
+      "learning_rate": 4.56828885400314e-06,
+      "loss": 0.8489,
+      "step": 291
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5654750931988162,
+      "learning_rate": 4.583987441130298e-06,
+      "loss": 0.7755,
+      "step": 292
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.92934619014978,
+      "learning_rate": 4.599686028257457e-06,
+      "loss": 0.9909,
+      "step": 293
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5551036176452149,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.9644,
+      "step": 294
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.894422498204267,
+      "learning_rate": 4.631083202511774e-06,
+      "loss": 0.8673,
+      "step": 295
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7618097524507628,
+      "learning_rate": 4.646781789638933e-06,
+      "loss": 0.699,
+      "step": 296
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.182682577802841,
+      "learning_rate": 4.662480376766091e-06,
+      "loss": 0.8305,
+      "step": 297
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6195424252925041,
+      "learning_rate": 4.67817896389325e-06,
+      "loss": 1.0153,
+      "step": 298
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5413560056633067,
+      "learning_rate": 4.693877551020409e-06,
+      "loss": 0.9142,
+      "step": 299
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6567484755628235,
+      "learning_rate": 4.709576138147567e-06,
+      "loss": 1.0291,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2964409016814478,
+      "learning_rate": 4.725274725274726e-06,
+      "loss": 0.8101,
+      "step": 301
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.47705994492833,
+      "learning_rate": 4.740973312401884e-06,
+      "loss": 0.9427,
+      "step": 302
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.605601988116758,
+      "learning_rate": 4.756671899529043e-06,
+      "loss": 0.8259,
+      "step": 303
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7259669258069992,
+      "learning_rate": 4.772370486656202e-06,
+      "loss": 0.883,
+      "step": 304
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.029412894390522,
+      "learning_rate": 4.7880690737833605e-06,
+      "loss": 0.8337,
+      "step": 305
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0437928231224447,
+      "learning_rate": 4.8037676609105186e-06,
+      "loss": 0.935,
+      "step": 306
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.006487590905564,
+      "learning_rate": 4.819466248037677e-06,
+      "loss": 0.9722,
+      "step": 307
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.693866428756853,
+      "learning_rate": 4.8351648351648355e-06,
+      "loss": 0.8315,
+      "step": 308
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6727025944998986,
+      "learning_rate": 4.850863422291994e-06,
+      "loss": 0.9968,
+      "step": 309
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.133481931406222,
+      "learning_rate": 4.866562009419153e-06,
+      "loss": 0.8555,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8825642583761812,
+      "learning_rate": 4.882260596546311e-06,
+      "loss": 0.8249,
+      "step": 311
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7540573797275143,
+      "learning_rate": 4.897959183673469e-06,
+      "loss": 0.9443,
+      "step": 312
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.77261676438318,
+      "learning_rate": 4.913657770800628e-06,
+      "loss": 0.8461,
+      "step": 313
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7795614751688495,
+      "learning_rate": 4.929356357927787e-06,
+      "loss": 0.7131,
+      "step": 314
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7355444692714257,
+      "learning_rate": 4.945054945054946e-06,
+      "loss": 0.86,
+      "step": 315
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5732227210427612,
+      "learning_rate": 4.960753532182104e-06,
+      "loss": 0.7521,
+      "step": 316
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.751257632586913,
+      "learning_rate": 4.976452119309262e-06,
+      "loss": 0.7922,
+      "step": 317
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7876222106388955,
+      "learning_rate": 4.992150706436421e-06,
+      "loss": 0.8256,
+      "step": 318
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4251481807948407,
+      "learning_rate": 5.007849293563579e-06,
+      "loss": 0.9798,
+      "step": 319
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6928938746398037,
+      "learning_rate": 5.023547880690738e-06,
+      "loss": 1.0037,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6715218308265323,
+      "learning_rate": 5.039246467817897e-06,
+      "loss": 0.9846,
+      "step": 321
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9044995442772477,
+      "learning_rate": 5.054945054945055e-06,
+      "loss": 0.9574,
+      "step": 322
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6434512459440722,
+      "learning_rate": 5.070643642072214e-06,
+      "loss": 0.8586,
+      "step": 323
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4308099073674434,
+      "learning_rate": 5.086342229199373e-06,
+      "loss": 0.838,
+      "step": 324
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.872784633931917,
+      "learning_rate": 5.1020408163265315e-06,
+      "loss": 0.7793,
+      "step": 325
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8141130389098534,
+      "learning_rate": 5.1177394034536895e-06,
+      "loss": 1.0594,
+      "step": 326
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.426048243035347,
+      "learning_rate": 5.133437990580848e-06,
+      "loss": 0.8959,
+      "step": 327
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9439358753209357,
+      "learning_rate": 5.149136577708007e-06,
+      "loss": 0.9045,
+      "step": 328
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7685856935780896,
+      "learning_rate": 5.164835164835166e-06,
+      "loss": 0.6985,
+      "step": 329
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.672385298103698,
+      "learning_rate": 5.180533751962323e-06,
+      "loss": 1.0852,
+      "step": 330
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5242373098568602,
+      "learning_rate": 5.196232339089482e-06,
+      "loss": 0.8751,
+      "step": 331
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.650507275631921,
+      "learning_rate": 5.21193092621664e-06,
+      "loss": 0.972,
+      "step": 332
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.533098945392548,
+      "learning_rate": 5.227629513343799e-06,
+      "loss": 0.926,
+      "step": 333
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5340885753957647,
+      "learning_rate": 5.243328100470958e-06,
+      "loss": 0.9785,
+      "step": 334
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7063529867269156,
+      "learning_rate": 5.259026687598117e-06,
+      "loss": 0.7026,
+      "step": 335
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.009637859048855,
+      "learning_rate": 5.274725274725275e-06,
+      "loss": 0.8365,
+      "step": 336
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9150161392857206,
+      "learning_rate": 5.290423861852434e-06,
+      "loss": 0.9464,
+      "step": 337
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.793073046516515,
+      "learning_rate": 5.306122448979593e-06,
+      "loss": 0.718,
+      "step": 338
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.582929474339108,
+      "learning_rate": 5.321821036106752e-06,
+      "loss": 1.0954,
+      "step": 339
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7200586617303322,
+      "learning_rate": 5.33751962323391e-06,
+      "loss": 0.8951,
+      "step": 340
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7986647600905398,
+      "learning_rate": 5.353218210361068e-06,
+      "loss": 0.713,
+      "step": 341
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7655799088988134,
+      "learning_rate": 5.368916797488226e-06,
+      "loss": 0.7075,
+      "step": 342
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7439014048033346,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 0.7225,
+      "step": 343
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8400821077079543,
+      "learning_rate": 5.4003139717425436e-06,
+      "loss": 0.9038,
+      "step": 344
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7199174365268866,
+      "learning_rate": 5.4160125588697024e-06,
+      "loss": 0.8496,
+      "step": 345
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.433901682820075,
+      "learning_rate": 5.4317111459968605e-06,
+      "loss": 0.9082,
+      "step": 346
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8272818658450655,
+      "learning_rate": 5.447409733124019e-06,
+      "loss": 0.692,
+      "step": 347
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.909943263111769,
+      "learning_rate": 5.463108320251178e-06,
+      "loss": 0.7879,
+      "step": 348
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6206476552873772,
+      "learning_rate": 5.478806907378337e-06,
+      "loss": 1.0187,
+      "step": 349
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.741741113944881,
+      "learning_rate": 5.494505494505495e-06,
+      "loss": 0.9185,
+      "step": 350
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4516218249761,
+      "learning_rate": 5.510204081632653e-06,
+      "loss": 0.8543,
+      "step": 351
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5721089312782939,
+      "learning_rate": 5.525902668759811e-06,
+      "loss": 0.8993,
+      "step": 352
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8759386735297725,
+      "learning_rate": 5.54160125588697e-06,
+      "loss": 0.755,
+      "step": 353
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.619287696178784,
+      "learning_rate": 5.557299843014129e-06,
+      "loss": 0.8892,
+      "step": 354
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6756210773183873,
+      "learning_rate": 5.572998430141288e-06,
+      "loss": 0.914,
+      "step": 355
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.195814257292635,
+      "learning_rate": 5.588697017268446e-06,
+      "loss": 0.8845,
+      "step": 356
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5823767947323466,
+      "learning_rate": 5.604395604395605e-06,
+      "loss": 0.8918,
+      "step": 357
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7508455233506104,
+      "learning_rate": 5.620094191522764e-06,
+      "loss": 0.8704,
+      "step": 358
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0368147622539543,
+      "learning_rate": 5.635792778649923e-06,
+      "loss": 0.8236,
+      "step": 359
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6543836591975936,
+      "learning_rate": 5.651491365777081e-06,
+      "loss": 0.806,
+      "step": 360
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8875510535409736,
+      "learning_rate": 5.6671899529042395e-06,
+      "loss": 0.9953,
+      "step": 361
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.589079057225448,
+      "learning_rate": 5.682888540031397e-06,
+      "loss": 0.966,
+      "step": 362
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.320606615781571,
+      "learning_rate": 5.698587127158556e-06,
+      "loss": 0.8925,
+      "step": 363
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7930569556009263,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.7166,
+      "step": 364
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.092620613788071,
+      "learning_rate": 5.729984301412873e-06,
+      "loss": 0.9238,
+      "step": 365
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.685326633345394,
+      "learning_rate": 5.7456828885400314e-06,
+      "loss": 0.8661,
+      "step": 366
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7738979187718995,
+      "learning_rate": 5.76138147566719e-06,
+      "loss": 0.8343,
+      "step": 367
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2645852669789943,
+      "learning_rate": 5.777080062794349e-06,
+      "loss": 0.8303,
+      "step": 368
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8063381613856929,
+      "learning_rate": 5.792778649921508e-06,
+      "loss": 0.7363,
+      "step": 369
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5873720158616547,
+      "learning_rate": 5.808477237048666e-06,
+      "loss": 0.9178,
+      "step": 370
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7283231793948184,
+      "learning_rate": 5.824175824175825e-06,
+      "loss": 1.0091,
+      "step": 371
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8230443402337704,
+      "learning_rate": 5.839874411302984e-06,
+      "loss": 0.8739,
+      "step": 372
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6965537277903686,
+      "learning_rate": 5.855572998430141e-06,
+      "loss": 0.8856,
+      "step": 373
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7291942903240501,
+      "learning_rate": 5.8712715855573e-06,
+      "loss": 0.6803,
+      "step": 374
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5775537284170178,
+      "learning_rate": 5.886970172684459e-06,
+      "loss": 0.7486,
+      "step": 375
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5828036417872406,
+      "learning_rate": 5.902668759811617e-06,
+      "loss": 0.9675,
+      "step": 376
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7620527353481854,
+      "learning_rate": 5.918367346938776e-06,
+      "loss": 0.7114,
+      "step": 377
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6843804689391384,
+      "learning_rate": 5.934065934065935e-06,
+      "loss": 0.7973,
+      "step": 378
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5921379320995936,
+      "learning_rate": 5.949764521193094e-06,
+      "loss": 0.9484,
+      "step": 379
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6797777582054125,
+      "learning_rate": 5.965463108320252e-06,
+      "loss": 0.9288,
+      "step": 380
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8078914040002656,
+      "learning_rate": 5.9811616954474105e-06,
+      "loss": 0.8683,
+      "step": 381
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7639320396071694,
+      "learning_rate": 5.996860282574569e-06,
+      "loss": 0.7368,
+      "step": 382
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5626642479846586,
+      "learning_rate": 6.012558869701728e-06,
+      "loss": 0.9693,
+      "step": 383
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.526458969925365,
+      "learning_rate": 6.0282574568288855e-06,
+      "loss": 0.8736,
+      "step": 384
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.350325005678273,
+      "learning_rate": 6.043956043956044e-06,
+      "loss": 0.8508,
+      "step": 385
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8490473402271852,
+      "learning_rate": 6.059654631083202e-06,
+      "loss": 0.7377,
+      "step": 386
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5425303419179717,
+      "learning_rate": 6.075353218210361e-06,
+      "loss": 0.9885,
+      "step": 387
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4143159487553159,
+      "learning_rate": 6.09105180533752e-06,
+      "loss": 0.9118,
+      "step": 388
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8263138890667472,
+      "learning_rate": 6.106750392464679e-06,
+      "loss": 0.8325,
+      "step": 389
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4986753044931902,
+      "learning_rate": 6.122448979591837e-06,
+      "loss": 0.8593,
+      "step": 390
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7649679864974134,
+      "learning_rate": 6.138147566718996e-06,
+      "loss": 0.8088,
+      "step": 391
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4032479582020456,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.848,
+      "step": 392
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.584868108700286,
+      "learning_rate": 6.169544740973314e-06,
+      "loss": 0.9129,
+      "step": 393
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.740978390545799,
+      "learning_rate": 6.185243328100472e-06,
+      "loss": 0.8711,
+      "step": 394
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8093697143818332,
+      "learning_rate": 6.20094191522763e-06,
+      "loss": 0.9143,
+      "step": 395
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6417134105286975,
+      "learning_rate": 6.216640502354788e-06,
+      "loss": 0.7881,
+      "step": 396
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0646920177031713,
+      "learning_rate": 6.232339089481947e-06,
+      "loss": 0.7462,
+      "step": 397
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6601798697852992,
+      "learning_rate": 6.248037676609106e-06,
+      "loss": 0.9624,
+      "step": 398
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6608988171100363,
+      "learning_rate": 6.2637362637362645e-06,
+      "loss": 0.8987,
+      "step": 399
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4580975721363358,
+      "learning_rate": 6.279434850863423e-06,
+      "loss": 0.7943,
+      "step": 400
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6097142630303243,
+      "learning_rate": 6.2951334379905815e-06,
+      "loss": 0.8632,
+      "step": 401
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5420474408462392,
+      "learning_rate": 6.31083202511774e-06,
+      "loss": 0.9384,
+      "step": 402
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6564119668943802,
+      "learning_rate": 6.326530612244899e-06,
+      "loss": 0.8469,
+      "step": 403
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7320364887443198,
+      "learning_rate": 6.342229199372057e-06,
+      "loss": 0.9235,
+      "step": 404
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5646628404158445,
+      "learning_rate": 6.357927786499215e-06,
+      "loss": 0.8423,
+      "step": 405
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9266573214632146,
+      "learning_rate": 6.373626373626373e-06,
+      "loss": 0.9803,
+      "step": 406
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4995094172054182,
+      "learning_rate": 6.389324960753532e-06,
+      "loss": 0.9555,
+      "step": 407
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5503886125189712,
+      "learning_rate": 6.405023547880691e-06,
+      "loss": 0.8772,
+      "step": 408
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6886464994940635,
+      "learning_rate": 6.42072213500785e-06,
+      "loss": 0.942,
+      "step": 409
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7054310694304058,
+      "learning_rate": 6.436420722135008e-06,
+      "loss": 0.9968,
+      "step": 410
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9840890169572238,
+      "learning_rate": 6.452119309262167e-06,
+      "loss": 0.7848,
+      "step": 411
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7569541272820437,
+      "learning_rate": 6.467817896389326e-06,
+      "loss": 0.9221,
+      "step": 412
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.549123471190071,
+      "learning_rate": 6.483516483516485e-06,
+      "loss": 0.9586,
+      "step": 413
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.186664611268925,
+      "learning_rate": 6.499215070643643e-06,
+      "loss": 0.9596,
+      "step": 414
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7752301488990921,
+      "learning_rate": 6.514913657770802e-06,
+      "loss": 0.7226,
+      "step": 415
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.776127166382774,
+      "learning_rate": 6.530612244897959e-06,
+      "loss": 0.8986,
+      "step": 416
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6249152451608622,
+      "learning_rate": 6.546310832025118e-06,
+      "loss": 1.0708,
+      "step": 417
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6233021668098317,
+      "learning_rate": 6.562009419152277e-06,
+      "loss": 0.9114,
+      "step": 418
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.238035972864371,
+      "learning_rate": 6.5777080062794355e-06,
+      "loss": 0.9271,
+      "step": 419
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6713050851916749,
+      "learning_rate": 6.5934065934065935e-06,
+      "loss": 1.023,
+      "step": 420
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8259208904366752,
+      "learning_rate": 6.609105180533752e-06,
+      "loss": 0.7053,
+      "step": 421
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3865246902635486,
+      "learning_rate": 6.624803767660911e-06,
+      "loss": 0.9517,
+      "step": 422
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7048355220597637,
+      "learning_rate": 6.64050235478807e-06,
+      "loss": 0.744,
+      "step": 423
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8396980479215188,
+      "learning_rate": 6.656200941915228e-06,
+      "loss": 0.8431,
+      "step": 424
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7173456572204306,
+      "learning_rate": 6.671899529042387e-06,
+      "loss": 0.8804,
+      "step": 425
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.837877768340835,
+      "learning_rate": 6.687598116169546e-06,
+      "loss": 0.8395,
+      "step": 426
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.763323002776545,
+      "learning_rate": 6.703296703296703e-06,
+      "loss": 0.8501,
+      "step": 427
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7594775560718325,
+      "learning_rate": 6.718995290423862e-06,
+      "loss": 0.9765,
+      "step": 428
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.569919479321543,
+      "learning_rate": 6.734693877551021e-06,
+      "loss": 0.8075,
+      "step": 429
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0444447618248476,
+      "learning_rate": 6.750392464678179e-06,
+      "loss": 0.8606,
+      "step": 430
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.535133805085101,
+      "learning_rate": 6.766091051805338e-06,
+      "loss": 0.8853,
+      "step": 431
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7724684051289312,
+      "learning_rate": 6.781789638932497e-06,
+      "loss": 0.8539,
+      "step": 432
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.556333674304132,
+      "learning_rate": 6.797488226059656e-06,
+      "loss": 0.8697,
+      "step": 433
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7700353583253088,
+      "learning_rate": 6.813186813186814e-06,
+      "loss": 0.77,
+      "step": 434
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.771833833039703,
+      "learning_rate": 6.828885400313973e-06,
+      "loss": 0.8227,
+      "step": 435
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7139180100863313,
+      "learning_rate": 6.8445839874411315e-06,
+      "loss": 0.6666,
+      "step": 436
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7848408246559564,
+      "learning_rate": 6.8602825745682895e-06,
+      "loss": 0.6858,
+      "step": 437
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.416169600959562,
+      "learning_rate": 6.8759811616954476e-06,
+      "loss": 1.0374,
+      "step": 438
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8184287783546358,
+      "learning_rate": 6.8916797488226065e-06,
+      "loss": 0.8749,
+      "step": 439
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8361032096034418,
+      "learning_rate": 6.9073783359497645e-06,
+      "loss": 1.0192,
+      "step": 440
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9153853544348582,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 0.8951,
+      "step": 441
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6647157913189576,
+      "learning_rate": 6.938775510204082e-06,
+      "loss": 1.067,
+      "step": 442
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0059656563778585,
+      "learning_rate": 6.954474097331241e-06,
+      "loss": 0.8625,
+      "step": 443
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.152819156851572,
+      "learning_rate": 6.970172684458399e-06,
+      "loss": 0.8457,
+      "step": 444
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7657508558152963,
+      "learning_rate": 6.985871271585558e-06,
+      "loss": 0.9861,
+      "step": 445
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.656904461837766,
+      "learning_rate": 7.001569858712717e-06,
+      "loss": 0.7998,
+      "step": 446
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.504413267459723,
+      "learning_rate": 7.017268445839875e-06,
+      "loss": 0.933,
+      "step": 447
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9210551796392688,
+      "learning_rate": 7.032967032967034e-06,
+      "loss": 0.9589,
+      "step": 448
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5772166404208434,
+      "learning_rate": 7.048665620094192e-06,
+      "loss": 0.8203,
+      "step": 449
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.724448500466887,
+      "learning_rate": 7.06436420722135e-06,
+      "loss": 0.9983,
+      "step": 450
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7309950373772045,
+      "learning_rate": 7.080062794348509e-06,
+      "loss": 0.8968,
+      "step": 451
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.4129981313874005,
+      "learning_rate": 7.095761381475668e-06,
+      "loss": 0.9872,
+      "step": 452
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.752995586634231,
+      "learning_rate": 7.111459968602827e-06,
+      "loss": 0.976,
+      "step": 453
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.725323907055808,
+      "learning_rate": 7.127158555729985e-06,
+      "loss": 1.0006,
+      "step": 454
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7664192151496099,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.9574,
+      "step": 455
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7163768249761857,
+      "learning_rate": 7.1585557299843024e-06,
+      "loss": 0.951,
+      "step": 456
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.598374402093358,
+      "learning_rate": 7.174254317111461e-06,
+      "loss": 0.852,
+      "step": 457
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8208244036384786,
+      "learning_rate": 7.189952904238619e-06,
+      "loss": 0.6907,
+      "step": 458
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6760768984050174,
+      "learning_rate": 7.205651491365777e-06,
+      "loss": 0.8569,
+      "step": 459
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0092925740333394,
+      "learning_rate": 7.2213500784929355e-06,
+      "loss": 0.995,
+      "step": 460
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7883736604863099,
+      "learning_rate": 7.237048665620094e-06,
+      "loss": 0.7286,
+      "step": 461
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6280329033826084,
+      "learning_rate": 7.252747252747253e-06,
+      "loss": 0.8743,
+      "step": 462
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7164379893860724,
+      "learning_rate": 7.268445839874412e-06,
+      "loss": 0.8919,
+      "step": 463
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5995369906533583,
+      "learning_rate": 7.28414442700157e-06,
+      "loss": 0.84,
+      "step": 464
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7463267711777104,
+      "learning_rate": 7.299843014128729e-06,
+      "loss": 0.9305,
+      "step": 465
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8346393468604667,
+      "learning_rate": 7.315541601255888e-06,
+      "loss": 0.7129,
+      "step": 466
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4946373415783387,
+      "learning_rate": 7.331240188383047e-06,
+      "loss": 0.872,
+      "step": 467
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.464557671248544,
+      "learning_rate": 7.346938775510205e-06,
+      "loss": 0.8593,
+      "step": 468
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7324480908160413,
+      "learning_rate": 7.362637362637364e-06,
+      "loss": 0.9436,
+      "step": 469
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.655798328393443,
+      "learning_rate": 7.378335949764521e-06,
+      "loss": 0.8686,
+      "step": 470
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7361639325551224,
+      "learning_rate": 7.39403453689168e-06,
+      "loss": 0.9504,
+      "step": 471
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.745242458158335,
+      "learning_rate": 7.409733124018839e-06,
+      "loss": 0.9657,
+      "step": 472
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7470337070017374,
+      "learning_rate": 7.425431711145998e-06,
+      "loss": 0.8286,
+      "step": 473
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2049139483870195,
+      "learning_rate": 7.441130298273156e-06,
+      "loss": 0.9803,
+      "step": 474
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5451164361023129,
+      "learning_rate": 7.4568288854003145e-06,
+      "loss": 0.7562,
+      "step": 475
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7607492456824902,
+      "learning_rate": 7.472527472527473e-06,
+      "loss": 0.7093,
+      "step": 476
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7298273045124486,
+      "learning_rate": 7.488226059654632e-06,
+      "loss": 0.9133,
+      "step": 477
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7383071590418817,
+      "learning_rate": 7.50392464678179e-06,
+      "loss": 1.0171,
+      "step": 478
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7980482671725755,
+      "learning_rate": 7.519623233908949e-06,
+      "loss": 0.8436,
+      "step": 479
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.62840733369174,
+      "learning_rate": 7.535321821036108e-06,
+      "loss": 0.932,
+      "step": 480
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5341660864261726,
+      "learning_rate": 7.551020408163265e-06,
+      "loss": 0.9077,
+      "step": 481
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5946373275613839,
+      "learning_rate": 7.566718995290424e-06,
+      "loss": 0.7736,
+      "step": 482
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.78559828429967,
+      "learning_rate": 7.582417582417583e-06,
+      "loss": 0.98,
+      "step": 483
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1055305364598147,
+      "learning_rate": 7.598116169544741e-06,
+      "loss": 0.8495,
+      "step": 484
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9031031810282559,
+      "learning_rate": 7.6138147566719e-06,
+      "loss": 0.6999,
+      "step": 485
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.698038769274121,
+      "learning_rate": 7.629513343799058e-06,
+      "loss": 1.1025,
+      "step": 486
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9317007444942966,
+      "learning_rate": 7.645211930926217e-06,
+      "loss": 0.8246,
+      "step": 487
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5210357983043155,
+      "learning_rate": 7.660910518053376e-06,
+      "loss": 0.9141,
+      "step": 488
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.14680756301141,
+      "learning_rate": 7.676609105180535e-06,
+      "loss": 0.9251,
+      "step": 489
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7661383316773386,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.6701,
+      "step": 490
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6264064833675997,
+      "learning_rate": 7.708006279434852e-06,
+      "loss": 0.9385,
+      "step": 491
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8934511987280895,
+      "learning_rate": 7.72370486656201e-06,
+      "loss": 0.9457,
+      "step": 492
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7584166162063173,
+      "learning_rate": 7.739403453689169e-06,
+      "loss": 0.9823,
+      "step": 493
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9182679732803625,
+      "learning_rate": 7.755102040816327e-06,
+      "loss": 0.9242,
+      "step": 494
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7587178211340441,
+      "learning_rate": 7.770800627943486e-06,
+      "loss": 0.6752,
+      "step": 495
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.765042632182092,
+      "learning_rate": 7.786499215070644e-06,
+      "loss": 0.9512,
+      "step": 496
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9299977444747187,
+      "learning_rate": 7.802197802197802e-06,
+      "loss": 0.6753,
+      "step": 497
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.794205256340073,
+      "learning_rate": 7.817896389324961e-06,
+      "loss": 0.9878,
+      "step": 498
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8731546120827038,
+      "learning_rate": 7.83359497645212e-06,
+      "loss": 0.9227,
+      "step": 499
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4943180925885915,
+      "learning_rate": 7.849293563579279e-06,
+      "loss": 0.9621,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6646050681318452,
+      "learning_rate": 7.864992150706438e-06,
+      "loss": 1.0709,
+      "step": 501
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.352973402870089,
+      "learning_rate": 7.880690737833597e-06,
+      "loss": 0.8173,
+      "step": 502
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6607238582303445,
+      "learning_rate": 7.896389324960754e-06,
+      "loss": 0.9773,
+      "step": 503
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.893593127049003,
+      "learning_rate": 7.912087912087913e-06,
+      "loss": 0.9377,
+      "step": 504
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7804591893787941,
+      "learning_rate": 7.927786499215072e-06,
+      "loss": 0.6811,
+      "step": 505
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.997556881809457,
+      "learning_rate": 7.943485086342229e-06,
+      "loss": 0.8894,
+      "step": 506
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.584119180205275,
+      "learning_rate": 7.959183673469388e-06,
+      "loss": 0.9232,
+      "step": 507
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5724653392431813,
+      "learning_rate": 7.974882260596547e-06,
+      "loss": 0.9246,
+      "step": 508
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7381060401476487,
+      "learning_rate": 7.990580847723706e-06,
+      "loss": 0.8955,
+      "step": 509
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5928208391699712,
+      "learning_rate": 8.006279434850865e-06,
+      "loss": 0.9379,
+      "step": 510
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5809674397971334,
+      "learning_rate": 8.021978021978023e-06,
+      "loss": 0.9242,
+      "step": 511
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8621850977430828,
+      "learning_rate": 8.037676609105182e-06,
+      "loss": 0.7012,
+      "step": 512
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6600539395418399,
+      "learning_rate": 8.05337519623234e-06,
+      "loss": 0.8961,
+      "step": 513
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.847102838159444,
+      "learning_rate": 8.069073783359498e-06,
+      "loss": 0.9949,
+      "step": 514
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7764552450136993,
+      "learning_rate": 8.084772370486657e-06,
+      "loss": 0.7456,
+      "step": 515
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.528900285576021,
+      "learning_rate": 8.100470957613814e-06,
+      "loss": 0.8459,
+      "step": 516
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7267994365632204,
+      "learning_rate": 8.116169544740973e-06,
+      "loss": 1.1412,
+      "step": 517
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6945218025522784,
+      "learning_rate": 8.131868131868132e-06,
+      "loss": 0.9885,
+      "step": 518
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8357194121671145,
+      "learning_rate": 8.147566718995291e-06,
+      "loss": 0.8011,
+      "step": 519
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8753842562179937,
+      "learning_rate": 8.16326530612245e-06,
+      "loss": 0.8505,
+      "step": 520
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6197259978686163,
+      "learning_rate": 8.178963893249609e-06,
+      "loss": 0.9445,
+      "step": 521
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.008305828118052,
+      "learning_rate": 8.194662480376768e-06,
+      "loss": 0.8333,
+      "step": 522
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7021338281439498,
+      "learning_rate": 8.210361067503925e-06,
+      "loss": 0.8684,
+      "step": 523
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8061075361300492,
+      "learning_rate": 8.226059654631084e-06,
+      "loss": 0.671,
+      "step": 524
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7721356080280418,
+      "learning_rate": 8.241758241758243e-06,
+      "loss": 0.6738,
+      "step": 525
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7551241999943503,
+      "learning_rate": 8.2574568288854e-06,
+      "loss": 0.668,
+      "step": 526
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7636597017324074,
+      "learning_rate": 8.273155416012559e-06,
+      "loss": 0.9509,
+      "step": 527
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5611216730986053,
+      "learning_rate": 8.288854003139718e-06,
+      "loss": 0.8016,
+      "step": 528
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.725184759870583,
+      "learning_rate": 8.304552590266877e-06,
+      "loss": 0.8585,
+      "step": 529
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8220371871242451,
+      "learning_rate": 8.320251177394036e-06,
+      "loss": 0.8002,
+      "step": 530
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.897685164456066,
+      "learning_rate": 8.335949764521194e-06,
+      "loss": 0.9254,
+      "step": 531
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5927531570823628,
+      "learning_rate": 8.351648351648353e-06,
+      "loss": 0.9869,
+      "step": 532
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6954332201545355,
+      "learning_rate": 8.36734693877551e-06,
+      "loss": 1.0514,
+      "step": 533
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.62497536594888,
+      "learning_rate": 8.38304552590267e-06,
+      "loss": 0.9918,
+      "step": 534
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5873590281098113,
+      "learning_rate": 8.398744113029828e-06,
+      "loss": 0.8855,
+      "step": 535
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8283073712778697,
+      "learning_rate": 8.414442700156985e-06,
+      "loss": 0.9472,
+      "step": 536
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6125298369716612,
+      "learning_rate": 8.430141287284144e-06,
+      "loss": 0.9423,
+      "step": 537
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7481888228594835,
+      "learning_rate": 8.445839874411303e-06,
+      "loss": 0.9483,
+      "step": 538
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7661350957231459,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.9292,
+      "step": 539
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5782359258965057,
+      "learning_rate": 8.477237048665621e-06,
+      "loss": 0.86,
+      "step": 540
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.802124557380052,
+      "learning_rate": 8.49293563579278e-06,
+      "loss": 0.9411,
+      "step": 541
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.290082176548404,
+      "learning_rate": 8.508634222919939e-06,
+      "loss": 0.9422,
+      "step": 542
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.011572238589142,
+      "learning_rate": 8.524332810047096e-06,
+      "loss": 0.8607,
+      "step": 543
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7658557253179936,
+      "learning_rate": 8.540031397174255e-06,
+      "loss": 1.0086,
+      "step": 544
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.150213637179915,
+      "learning_rate": 8.555729984301414e-06,
+      "loss": 0.9659,
+      "step": 545
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7272666902054692,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.9548,
+      "step": 546
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7416279341799001,
+      "learning_rate": 8.58712715855573e-06,
+      "loss": 0.8756,
+      "step": 547
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6999591910819427,
+      "learning_rate": 8.602825745682889e-06,
+      "loss": 1.0288,
+      "step": 548
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6050377314176134,
+      "learning_rate": 8.618524332810048e-06,
+      "loss": 0.9281,
+      "step": 549
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.524974104198204,
+      "learning_rate": 8.634222919937206e-06,
+      "loss": 0.944,
+      "step": 550
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6840002816597013,
+      "learning_rate": 8.649921507064365e-06,
+      "loss": 0.8472,
+      "step": 551
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.739747072433568,
+      "learning_rate": 8.665620094191524e-06,
+      "loss": 0.8641,
+      "step": 552
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7855446440003131,
+      "learning_rate": 8.681318681318681e-06,
+      "loss": 0.717,
+      "step": 553
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.865064089052186,
+      "learning_rate": 8.69701726844584e-06,
+      "loss": 0.9664,
+      "step": 554
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7729105283050764,
+      "learning_rate": 8.712715855573e-06,
+      "loss": 0.7365,
+      "step": 555
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6370706716537786,
+      "learning_rate": 8.728414442700158e-06,
+      "loss": 0.9142,
+      "step": 556
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8064247893751555,
+      "learning_rate": 8.744113029827315e-06,
+      "loss": 1.0017,
+      "step": 557
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7095998919557753,
+      "learning_rate": 8.759811616954474e-06,
+      "loss": 0.9939,
+      "step": 558
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7786156741653569,
+      "learning_rate": 8.775510204081633e-06,
+      "loss": 0.7083,
+      "step": 559
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.704319213474312,
+      "learning_rate": 8.791208791208792e-06,
+      "loss": 0.7969,
+      "step": 560
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3487608953246037,
+      "learning_rate": 8.80690737833595e-06,
+      "loss": 0.753,
+      "step": 561
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.785803839421554,
+      "learning_rate": 8.82260596546311e-06,
+      "loss": 0.9363,
+      "step": 562
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7229176782163969,
+      "learning_rate": 8.838304552590269e-06,
+      "loss": 0.9327,
+      "step": 563
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5198339300869335,
+      "learning_rate": 8.854003139717426e-06,
+      "loss": 0.9624,
+      "step": 564
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8810798693506836,
+      "learning_rate": 8.869701726844585e-06,
+      "loss": 0.8733,
+      "step": 565
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7335084916993988,
+      "learning_rate": 8.885400313971744e-06,
+      "loss": 0.7066,
+      "step": 566
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0188803158113395,
+      "learning_rate": 8.9010989010989e-06,
+      "loss": 0.8495,
+      "step": 567
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7096797301634759,
+      "learning_rate": 8.91679748822606e-06,
+      "loss": 0.8534,
+      "step": 568
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9376800517679587,
+      "learning_rate": 8.932496075353219e-06,
+      "loss": 0.9367,
+      "step": 569
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.664721655831306,
+      "learning_rate": 8.948194662480377e-06,
+      "loss": 0.8498,
+      "step": 570
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7960714593143557,
+      "learning_rate": 8.963893249607536e-06,
+      "loss": 0.7565,
+      "step": 571
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9232395760611778,
+      "learning_rate": 8.979591836734695e-06,
+      "loss": 0.9612,
+      "step": 572
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6877195503222526,
+      "learning_rate": 8.995290423861854e-06,
+      "loss": 0.8487,
+      "step": 573
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0712367607731825,
+      "learning_rate": 9.010989010989011e-06,
+      "loss": 0.9176,
+      "step": 574
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7860686090616789,
+      "learning_rate": 9.02668759811617e-06,
+      "loss": 0.9292,
+      "step": 575
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.057613777486309,
+      "learning_rate": 9.042386185243329e-06,
+      "loss": 0.7435,
+      "step": 576
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8753793510871055,
+      "learning_rate": 9.058084772370488e-06,
+      "loss": 0.9771,
+      "step": 577
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8219091369458924,
+      "learning_rate": 9.073783359497645e-06,
+      "loss": 0.9468,
+      "step": 578
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8745626720483641,
+      "learning_rate": 9.089481946624804e-06,
+      "loss": 0.9258,
+      "step": 579
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8103442993864878,
+      "learning_rate": 9.105180533751963e-06,
+      "loss": 0.7582,
+      "step": 580
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5892727147479648,
+      "learning_rate": 9.120879120879122e-06,
+      "loss": 0.9593,
+      "step": 581
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8761300501580898,
+      "learning_rate": 9.13657770800628e-06,
+      "loss": 0.8909,
+      "step": 582
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.702051418241563,
+      "learning_rate": 9.15227629513344e-06,
+      "loss": 0.948,
+      "step": 583
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8450408946393135,
+      "learning_rate": 9.167974882260597e-06,
+      "loss": 0.889,
+      "step": 584
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8199034357728308,
+      "learning_rate": 9.183673469387756e-06,
+      "loss": 0.8401,
+      "step": 585
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9952762031897984,
+      "learning_rate": 9.199372056514915e-06,
+      "loss": 0.9097,
+      "step": 586
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7097397568419486,
+      "learning_rate": 9.215070643642073e-06,
+      "loss": 0.8959,
+      "step": 587
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0479601466780846,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.9188,
+      "step": 588
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.646902997127537,
+      "learning_rate": 9.24646781789639e-06,
+      "loss": 0.9438,
+      "step": 589
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8622725329701275,
+      "learning_rate": 9.262166405023548e-06,
+      "loss": 0.866,
+      "step": 590
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5251311127300358,
+      "learning_rate": 9.277864992150707e-06,
+      "loss": 0.7382,
+      "step": 591
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8469817722246833,
+      "learning_rate": 9.293563579277866e-06,
+      "loss": 1.052,
+      "step": 592
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5482754223074802,
+      "learning_rate": 9.309262166405025e-06,
+      "loss": 0.8695,
+      "step": 593
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.551913115133504,
+      "learning_rate": 9.324960753532182e-06,
+      "loss": 0.7833,
+      "step": 594
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8818214672801212,
+      "learning_rate": 9.340659340659341e-06,
+      "loss": 1.0383,
+      "step": 595
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.47937331066311,
+      "learning_rate": 9.3563579277865e-06,
+      "loss": 0.8123,
+      "step": 596
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8771892070731555,
+      "learning_rate": 9.372056514913659e-06,
+      "loss": 0.9037,
+      "step": 597
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8413088594498161,
+      "learning_rate": 9.387755102040818e-06,
+      "loss": 0.747,
+      "step": 598
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.609872939270583,
+      "learning_rate": 9.403453689167977e-06,
+      "loss": 0.8763,
+      "step": 599
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1728295606131436,
+      "learning_rate": 9.419152276295134e-06,
+      "loss": 0.8816,
+      "step": 600
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6318842960542583,
+      "learning_rate": 9.434850863422293e-06,
+      "loss": 0.8991,
+      "step": 601
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.785653289476675,
+      "learning_rate": 9.450549450549452e-06,
+      "loss": 0.8985,
+      "step": 602
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.971048406907077,
+      "learning_rate": 9.46624803767661e-06,
+      "loss": 0.7507,
+      "step": 603
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8561546107578277,
+      "learning_rate": 9.481946624803768e-06,
+      "loss": 0.8182,
+      "step": 604
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6296081720134037,
+      "learning_rate": 9.497645211930927e-06,
+      "loss": 0.8974,
+      "step": 605
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5612294966097597,
+      "learning_rate": 9.513343799058085e-06,
+      "loss": 0.8808,
+      "step": 606
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3510426309347863,
+      "learning_rate": 9.529042386185244e-06,
+      "loss": 0.9465,
+      "step": 607
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.275243146477986,
+      "learning_rate": 9.544740973312403e-06,
+      "loss": 0.8372,
+      "step": 608
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6622897056716104,
+      "learning_rate": 9.560439560439562e-06,
+      "loss": 0.932,
+      "step": 609
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5695538107714757,
+      "learning_rate": 9.576138147566721e-06,
+      "loss": 0.9826,
+      "step": 610
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6584489156190314,
+      "learning_rate": 9.591836734693878e-06,
+      "loss": 0.8776,
+      "step": 611
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6982911707578157,
+      "learning_rate": 9.607535321821037e-06,
+      "loss": 0.8542,
+      "step": 612
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7942281248556102,
+      "learning_rate": 9.623233908948196e-06,
+      "loss": 0.9432,
+      "step": 613
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.888862545195381,
+      "learning_rate": 9.638932496075353e-06,
+      "loss": 0.8074,
+      "step": 614
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7674383625374355,
+      "learning_rate": 9.654631083202512e-06,
+      "loss": 0.9058,
+      "step": 615
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4645705448097655,
+      "learning_rate": 9.670329670329671e-06,
+      "loss": 1.1416,
+      "step": 616
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7113773988382346,
+      "learning_rate": 9.68602825745683e-06,
+      "loss": 0.8347,
+      "step": 617
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6423599727514364,
+      "learning_rate": 9.701726844583989e-06,
+      "loss": 1.0244,
+      "step": 618
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5995073963167783,
+      "learning_rate": 9.717425431711148e-06,
+      "loss": 0.8275,
+      "step": 619
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.605113478882301,
+      "learning_rate": 9.733124018838307e-06,
+      "loss": 0.8929,
+      "step": 620
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7745799953096707,
+      "learning_rate": 9.748822605965464e-06,
+      "loss": 0.7041,
+      "step": 621
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.054745787048315,
+      "learning_rate": 9.764521193092623e-06,
+      "loss": 0.9183,
+      "step": 622
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6696062974077617,
+      "learning_rate": 9.780219780219781e-06,
+      "loss": 0.9189,
+      "step": 623
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.618096797505277,
+      "learning_rate": 9.795918367346939e-06,
+      "loss": 0.8715,
+      "step": 624
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2421876762179065,
+      "learning_rate": 9.811616954474098e-06,
+      "loss": 0.924,
+      "step": 625
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.234466951675743,
+      "learning_rate": 9.827315541601256e-06,
+      "loss": 0.9957,
+      "step": 626
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6164487297722485,
+      "learning_rate": 9.843014128728415e-06,
+      "loss": 0.795,
+      "step": 627
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7964806738212102,
+      "learning_rate": 9.858712715855574e-06,
+      "loss": 0.7828,
+      "step": 628
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0643042255641215,
+      "learning_rate": 9.874411302982733e-06,
+      "loss": 0.8833,
+      "step": 629
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.522523554344303,
+      "learning_rate": 9.890109890109892e-06,
+      "loss": 0.8294,
+      "step": 630
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.763137648600385,
+      "learning_rate": 9.90580847723705e-06,
+      "loss": 0.9382,
+      "step": 631
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9601445733321623,
+      "learning_rate": 9.921507064364208e-06,
+      "loss": 0.8453,
+      "step": 632
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0234299260247637,
+      "learning_rate": 9.937205651491367e-06,
+      "loss": 0.8457,
+      "step": 633
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4667730647750183,
+      "learning_rate": 9.952904238618524e-06,
+      "loss": 0.8317,
+      "step": 634
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.452586276083112,
+      "learning_rate": 9.968602825745683e-06,
+      "loss": 0.8379,
+      "step": 635
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6405841747863734,
+      "learning_rate": 9.984301412872842e-06,
+      "loss": 0.9103,
+      "step": 636
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9332977377803082,
+      "learning_rate": 1e-05,
+      "loss": 0.9155,
+      "step": 637
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6389472394245084,
+      "learning_rate": 1.0015698587127158e-05,
+      "loss": 0.7657,
+      "step": 638
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1302738615602004,
+      "learning_rate": 1.0031397174254319e-05,
+      "loss": 0.8999,
+      "step": 639
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7019042372618038,
+      "learning_rate": 1.0047095761381476e-05,
+      "loss": 0.7464,
+      "step": 640
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6034876161080915,
+      "learning_rate": 1.0062794348508635e-05,
+      "loss": 0.9165,
+      "step": 641
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.753497405518127,
+      "learning_rate": 1.0078492935635794e-05,
+      "loss": 0.9022,
+      "step": 642
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8054143317647386,
+      "learning_rate": 1.0094191522762952e-05,
+      "loss": 0.9318,
+      "step": 643
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6675520843483518,
+      "learning_rate": 1.010989010989011e-05,
+      "loss": 0.8472,
+      "step": 644
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5807680887426643,
+      "learning_rate": 1.012558869701727e-05,
+      "loss": 0.8723,
+      "step": 645
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6767568314704402,
+      "learning_rate": 1.0141287284144427e-05,
+      "loss": 0.9032,
+      "step": 646
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6352344225341837,
+      "learning_rate": 1.0156985871271588e-05,
+      "loss": 1.0247,
+      "step": 647
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6065320107869088,
+      "learning_rate": 1.0172684458398745e-05,
+      "loss": 0.8405,
+      "step": 648
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7823561742598357,
+      "learning_rate": 1.0188383045525902e-05,
+      "loss": 0.8871,
+      "step": 649
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5920715455972898,
+      "learning_rate": 1.0204081632653063e-05,
+      "loss": 0.9287,
+      "step": 650
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5926341182490724,
+      "learning_rate": 1.021978021978022e-05,
+      "loss": 0.9799,
+      "step": 651
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.766861210543239,
+      "learning_rate": 1.0235478806907379e-05,
+      "loss": 0.933,
+      "step": 652
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.671626513148731,
+      "learning_rate": 1.0251177394034538e-05,
+      "loss": 0.9523,
+      "step": 653
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7119320888339993,
+      "learning_rate": 1.0266875981161697e-05,
+      "loss": 0.6892,
+      "step": 654
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5048995715038331,
+      "learning_rate": 1.0282574568288854e-05,
+      "loss": 1.0159,
+      "step": 655
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5677748271029295,
+      "learning_rate": 1.0298273155416015e-05,
+      "loss": 0.9711,
+      "step": 656
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6702330529070082,
+      "learning_rate": 1.0313971742543172e-05,
+      "loss": 0.8788,
+      "step": 657
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0982739005232887,
+      "learning_rate": 1.0329670329670332e-05,
+      "loss": 0.9036,
+      "step": 658
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9200774857798155,
+      "learning_rate": 1.034536891679749e-05,
+      "loss": 0.9498,
+      "step": 659
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8912641529785001,
+      "learning_rate": 1.0361067503924647e-05,
+      "loss": 0.9181,
+      "step": 660
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7089229642358827,
+      "learning_rate": 1.0376766091051806e-05,
+      "loss": 0.7837,
+      "step": 661
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2495290725847923,
+      "learning_rate": 1.0392464678178965e-05,
+      "loss": 0.985,
+      "step": 662
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7139664787449378,
+      "learning_rate": 1.0408163265306123e-05,
+      "loss": 0.8392,
+      "step": 663
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6072592712253382,
+      "learning_rate": 1.042386185243328e-05,
+      "loss": 0.9326,
+      "step": 664
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.015628950377708,
+      "learning_rate": 1.0439560439560441e-05,
+      "loss": 0.9109,
+      "step": 665
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.108458081249242,
+      "learning_rate": 1.0455259026687598e-05,
+      "loss": 0.7885,
+      "step": 666
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6501759303941352,
+      "learning_rate": 1.0470957613814759e-05,
+      "loss": 0.9109,
+      "step": 667
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0382660418124376,
+      "learning_rate": 1.0486656200941916e-05,
+      "loss": 0.8646,
+      "step": 668
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.620788321802992,
+      "learning_rate": 1.0502354788069075e-05,
+      "loss": 0.895,
+      "step": 669
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9542639030514497,
+      "learning_rate": 1.0518053375196234e-05,
+      "loss": 1.0031,
+      "step": 670
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8742719079156822,
+      "learning_rate": 1.0533751962323391e-05,
+      "loss": 0.8766,
+      "step": 671
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8997371718697715,
+      "learning_rate": 1.054945054945055e-05,
+      "loss": 0.9136,
+      "step": 672
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.803133847548563,
+      "learning_rate": 1.0565149136577709e-05,
+      "loss": 0.9153,
+      "step": 673
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.661876496208474,
+      "learning_rate": 1.0580847723704868e-05,
+      "loss": 0.8659,
+      "step": 674
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5546211883810626,
+      "learning_rate": 1.0596546310832025e-05,
+      "loss": 0.7539,
+      "step": 675
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.818865518504498,
+      "learning_rate": 1.0612244897959186e-05,
+      "loss": 0.7735,
+      "step": 676
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.606172567551492,
+      "learning_rate": 1.0627943485086343e-05,
+      "loss": 0.9729,
+      "step": 677
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8637354540713988,
+      "learning_rate": 1.0643642072213503e-05,
+      "loss": 0.9983,
+      "step": 678
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6858206907425506,
+      "learning_rate": 1.065934065934066e-05,
+      "loss": 0.9535,
+      "step": 679
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7645334657297347,
+      "learning_rate": 1.067503924646782e-05,
+      "loss": 0.8649,
+      "step": 680
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8781449563212127,
+      "learning_rate": 1.0690737833594977e-05,
+      "loss": 1.0092,
+      "step": 681
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.238974953062295,
+      "learning_rate": 1.0706436420722135e-05,
+      "loss": 0.835,
+      "step": 682
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.828725997576094,
+      "learning_rate": 1.0722135007849294e-05,
+      "loss": 0.915,
+      "step": 683
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.646893666485531,
+      "learning_rate": 1.0737833594976452e-05,
+      "loss": 0.9376,
+      "step": 684
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7511646895848352,
+      "learning_rate": 1.0753532182103612e-05,
+      "loss": 0.9489,
+      "step": 685
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8146951286773436,
+      "learning_rate": 1.076923076923077e-05,
+      "loss": 0.9233,
+      "step": 686
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8037570810056633,
+      "learning_rate": 1.078492935635793e-05,
+      "loss": 0.6787,
+      "step": 687
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.464079752529571,
+      "learning_rate": 1.0800627943485087e-05,
+      "loss": 0.9376,
+      "step": 688
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3960339160644273,
+      "learning_rate": 1.0816326530612246e-05,
+      "loss": 0.9534,
+      "step": 689
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.725932709634153,
+      "learning_rate": 1.0832025117739405e-05,
+      "loss": 0.867,
+      "step": 690
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6983744287600993,
+      "learning_rate": 1.0847723704866562e-05,
+      "loss": 0.8968,
+      "step": 691
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6827401688185513,
+      "learning_rate": 1.0863422291993721e-05,
+      "loss": 1.0619,
+      "step": 692
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8086948702455796,
+      "learning_rate": 1.087912087912088e-05,
+      "loss": 0.7514,
+      "step": 693
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.231151555527415,
+      "learning_rate": 1.0894819466248039e-05,
+      "loss": 1.0178,
+      "step": 694
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.913672104365935,
+      "learning_rate": 1.0910518053375196e-05,
+      "loss": 0.9129,
+      "step": 695
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0591420704146386,
+      "learning_rate": 1.0926216640502357e-05,
+      "loss": 0.7516,
+      "step": 696
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7164951800094832,
+      "learning_rate": 1.0941915227629514e-05,
+      "loss": 0.9808,
+      "step": 697
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7303751878391527,
+      "learning_rate": 1.0957613814756674e-05,
+      "loss": 0.8973,
+      "step": 698
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7100914853427964,
+      "learning_rate": 1.0973312401883831e-05,
+      "loss": 0.8474,
+      "step": 699
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7829106828271576,
+      "learning_rate": 1.098901098901099e-05,
+      "loss": 0.7011,
+      "step": 700
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7349706463595538,
+      "learning_rate": 1.1004709576138148e-05,
+      "loss": 0.918,
+      "step": 701
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6862607074585254,
+      "learning_rate": 1.1020408163265306e-05,
+      "loss": 1.0434,
+      "step": 702
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7355236921437933,
+      "learning_rate": 1.1036106750392465e-05,
+      "loss": 0.7862,
+      "step": 703
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.892090411016933,
+      "learning_rate": 1.1051805337519623e-05,
+      "loss": 0.903,
+      "step": 704
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7554037638635003,
+      "learning_rate": 1.1067503924646783e-05,
+      "loss": 0.6653,
+      "step": 705
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7915847074994495,
+      "learning_rate": 1.108320251177394e-05,
+      "loss": 0.8992,
+      "step": 706
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.859164242591878,
+      "learning_rate": 1.1098901098901101e-05,
+      "loss": 0.9364,
+      "step": 707
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9273619920329883,
+      "learning_rate": 1.1114599686028258e-05,
+      "loss": 0.8627,
+      "step": 708
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7032888594522984,
+      "learning_rate": 1.1130298273155417e-05,
+      "loss": 0.9389,
+      "step": 709
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.721121321899293,
+      "learning_rate": 1.1145996860282576e-05,
+      "loss": 0.8232,
+      "step": 710
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8765580435574094,
+      "learning_rate": 1.1161695447409735e-05,
+      "loss": 0.8311,
+      "step": 711
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.721965742902074,
+      "learning_rate": 1.1177394034536892e-05,
+      "loss": 0.6737,
+      "step": 712
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.833164106077372,
+      "learning_rate": 1.119309262166405e-05,
+      "loss": 0.9096,
+      "step": 713
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.882269513214081,
+      "learning_rate": 1.120879120879121e-05,
+      "loss": 0.9826,
+      "step": 714
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.095453434892582,
+      "learning_rate": 1.1224489795918367e-05,
+      "loss": 0.9619,
+      "step": 715
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5599180888465762,
+      "learning_rate": 1.1240188383045527e-05,
+      "loss": 0.8557,
+      "step": 716
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.789741031302047,
+      "learning_rate": 1.1255886970172685e-05,
+      "loss": 0.6836,
+      "step": 717
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7104019284040652,
+      "learning_rate": 1.1271585557299845e-05,
+      "loss": 0.7372,
+      "step": 718
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7519638576879533,
+      "learning_rate": 1.1287284144427002e-05,
+      "loss": 0.6326,
+      "step": 719
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7521395158707554,
+      "learning_rate": 1.1302982731554161e-05,
+      "loss": 0.6499,
+      "step": 720
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.032726468211759,
+      "learning_rate": 1.131868131868132e-05,
+      "loss": 0.9818,
+      "step": 721
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9391304499019766,
+      "learning_rate": 1.1334379905808479e-05,
+      "loss": 0.884,
+      "step": 722
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7713573876046262,
+      "learning_rate": 1.1350078492935636e-05,
+      "loss": 0.6616,
+      "step": 723
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4079278777651547,
+      "learning_rate": 1.1365777080062793e-05,
+      "loss": 0.7939,
+      "step": 724
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8081581720513715,
+      "learning_rate": 1.1381475667189954e-05,
+      "loss": 0.6592,
+      "step": 725
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7761988215652376,
+      "learning_rate": 1.1397174254317111e-05,
+      "loss": 0.9474,
+      "step": 726
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7526479385891982,
+      "learning_rate": 1.1412872841444272e-05,
+      "loss": 0.7415,
+      "step": 727
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8046648775087506,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 1.0024,
+      "step": 728
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5050014693368219,
+      "learning_rate": 1.1444270015698588e-05,
+      "loss": 0.8928,
+      "step": 729
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9113597526195392,
+      "learning_rate": 1.1459968602825747e-05,
+      "loss": 0.8403,
+      "step": 730
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.574163205219045,
+      "learning_rate": 1.1475667189952906e-05,
+      "loss": 0.9318,
+      "step": 731
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0670924234189436,
+      "learning_rate": 1.1491365777080063e-05,
+      "loss": 0.9754,
+      "step": 732
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6365423031298987,
+      "learning_rate": 1.1507064364207223e-05,
+      "loss": 0.8433,
+      "step": 733
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9376201174811558,
+      "learning_rate": 1.152276295133438e-05,
+      "loss": 0.9377,
+      "step": 734
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8185898077212972,
+      "learning_rate": 1.1538461538461538e-05,
+      "loss": 0.8416,
+      "step": 735
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8663343644385166,
+      "learning_rate": 1.1554160125588698e-05,
+      "loss": 0.9158,
+      "step": 736
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6450699894436667,
+      "learning_rate": 1.1569858712715856e-05,
+      "loss": 0.9937,
+      "step": 737
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7748119802434673,
+      "learning_rate": 1.1585557299843016e-05,
+      "loss": 0.6664,
+      "step": 738
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9969396778280746,
+      "learning_rate": 1.1601255886970173e-05,
+      "loss": 0.903,
+      "step": 739
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6981466077170164,
+      "learning_rate": 1.1616954474097332e-05,
+      "loss": 0.7879,
+      "step": 740
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.67286796169989,
+      "learning_rate": 1.1632653061224491e-05,
+      "loss": 0.8956,
+      "step": 741
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6566995474574937,
+      "learning_rate": 1.164835164835165e-05,
+      "loss": 0.9561,
+      "step": 742
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7753943449750962,
+      "learning_rate": 1.1664050235478807e-05,
+      "loss": 0.9659,
+      "step": 743
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9346313510711484,
+      "learning_rate": 1.1679748822605968e-05,
+      "loss": 0.9068,
+      "step": 744
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8582500396415398,
+      "learning_rate": 1.1695447409733125e-05,
+      "loss": 1.0011,
+      "step": 745
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7939263192350094,
+      "learning_rate": 1.1711145996860282e-05,
+      "loss": 0.8821,
+      "step": 746
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.634260381893241,
+      "learning_rate": 1.1726844583987443e-05,
+      "loss": 0.8632,
+      "step": 747
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8497150653502055,
+      "learning_rate": 1.17425431711146e-05,
+      "loss": 0.9358,
+      "step": 748
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.069861412248486,
+      "learning_rate": 1.1758241758241759e-05,
+      "loss": 1.0366,
+      "step": 749
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7679564419424186,
+      "learning_rate": 1.1773940345368918e-05,
+      "loss": 0.6767,
+      "step": 750
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9368485275503156,
+      "learning_rate": 1.1789638932496077e-05,
+      "loss": 1.0094,
+      "step": 751
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.009842129030666,
+      "learning_rate": 1.1805337519623234e-05,
+      "loss": 0.8219,
+      "step": 752
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7423216000609834,
+      "learning_rate": 1.1821036106750394e-05,
+      "loss": 0.799,
+      "step": 753
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.047777881590659,
+      "learning_rate": 1.1836734693877552e-05,
+      "loss": 0.886,
+      "step": 754
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6847556544014277,
+      "learning_rate": 1.1852433281004712e-05,
+      "loss": 0.9185,
+      "step": 755
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6095243552205807,
+      "learning_rate": 1.186813186813187e-05,
+      "loss": 0.8469,
+      "step": 756
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6848804579345231,
+      "learning_rate": 1.1883830455259027e-05,
+      "loss": 0.783,
+      "step": 757
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.843915117030274,
+      "learning_rate": 1.1899529042386187e-05,
+      "loss": 1.0188,
+      "step": 758
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.657207234395801,
+      "learning_rate": 1.1915227629513344e-05,
+      "loss": 0.8597,
+      "step": 759
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6148764328172156,
+      "learning_rate": 1.1930926216640503e-05,
+      "loss": 1.0842,
+      "step": 760
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3643791742695366,
+      "learning_rate": 1.1946624803767662e-05,
+      "loss": 0.9937,
+      "step": 761
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9721394897727136,
+      "learning_rate": 1.1962323390894821e-05,
+      "loss": 0.9085,
+      "step": 762
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.26865607447993,
+      "learning_rate": 1.1978021978021978e-05,
+      "loss": 0.8054,
+      "step": 763
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.990375145561411,
+      "learning_rate": 1.1993720565149139e-05,
+      "loss": 1.0114,
+      "step": 764
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.34135475232029,
+      "learning_rate": 1.2009419152276296e-05,
+      "loss": 0.8244,
+      "step": 765
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.922955771208112,
+      "learning_rate": 1.2025117739403457e-05,
+      "loss": 0.9868,
+      "step": 766
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7697906470460132,
+      "learning_rate": 1.2040816326530614e-05,
+      "loss": 1.0206,
+      "step": 767
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8588205962878077,
+      "learning_rate": 1.2056514913657771e-05,
+      "loss": 0.9353,
+      "step": 768
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6125666825997544,
+      "learning_rate": 1.207221350078493e-05,
+      "loss": 0.8118,
+      "step": 769
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5577287945068259,
+      "learning_rate": 1.2087912087912089e-05,
+      "loss": 0.7796,
+      "step": 770
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5272776336792728,
+      "learning_rate": 1.2103610675039248e-05,
+      "loss": 0.9614,
+      "step": 771
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.756062718593308,
+      "learning_rate": 1.2119309262166405e-05,
+      "loss": 0.82,
+      "step": 772
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3948894588266385,
+      "learning_rate": 1.2135007849293565e-05,
+      "loss": 0.8731,
+      "step": 773
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.879943114496335,
+      "learning_rate": 1.2150706436420723e-05,
+      "loss": 0.8761,
+      "step": 774
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7761830532123035,
+      "learning_rate": 1.2166405023547883e-05,
+      "loss": 0.8669,
+      "step": 775
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0003369593813805,
+      "learning_rate": 1.218210361067504e-05,
+      "loss": 0.9316,
+      "step": 776
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8201716962002137,
+      "learning_rate": 1.21978021978022e-05,
+      "loss": 1.0139,
+      "step": 777
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.8500936069368885,
+      "learning_rate": 1.2213500784929358e-05,
+      "loss": 0.8937,
+      "step": 778
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9068938403866882,
+      "learning_rate": 1.2229199372056515e-05,
+      "loss": 0.8725,
+      "step": 779
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5878624999363777,
+      "learning_rate": 1.2244897959183674e-05,
+      "loss": 0.8246,
+      "step": 780
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.454652636718264,
+      "learning_rate": 1.2260596546310833e-05,
+      "loss": 0.9519,
+      "step": 781
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3592374364752065,
+      "learning_rate": 1.2276295133437992e-05,
+      "loss": 0.9276,
+      "step": 782
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7999639604603799,
+      "learning_rate": 1.229199372056515e-05,
+      "loss": 0.9318,
+      "step": 783
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3776835924489723,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.8295,
+      "step": 784
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.878493544943101,
+      "learning_rate": 1.2323390894819467e-05,
+      "loss": 0.8847,
+      "step": 785
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.839212280762879,
+      "learning_rate": 1.2339089481946628e-05,
+      "loss": 1.0239,
+      "step": 786
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4976203781022204,
+      "learning_rate": 1.2354788069073785e-05,
+      "loss": 0.8998,
+      "step": 787
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6555029189565227,
+      "learning_rate": 1.2370486656200944e-05,
+      "loss": 0.8635,
+      "step": 788
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8127601065237622,
+      "learning_rate": 1.23861852433281e-05,
+      "loss": 0.9003,
+      "step": 789
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1768611761855325,
+      "learning_rate": 1.240188383045526e-05,
+      "loss": 0.8961,
+      "step": 790
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.023405277320002,
+      "learning_rate": 1.2417582417582419e-05,
+      "loss": 1.0067,
+      "step": 791
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.73860268962193,
+      "learning_rate": 1.2433281004709576e-05,
+      "loss": 1.0124,
+      "step": 792
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8625457913248604,
+      "learning_rate": 1.2448979591836736e-05,
+      "loss": 0.9154,
+      "step": 793
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7745176818880476,
+      "learning_rate": 1.2464678178963894e-05,
+      "loss": 0.7038,
+      "step": 794
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3229410647513884,
+      "learning_rate": 1.2480376766091054e-05,
+      "loss": 0.8687,
+      "step": 795
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1306650241545326,
+      "learning_rate": 1.2496075353218211e-05,
+      "loss": 0.9543,
+      "step": 796
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.788799612928587,
+      "learning_rate": 1.251177394034537e-05,
+      "loss": 0.9677,
+      "step": 797
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8040205538611442,
+      "learning_rate": 1.2527472527472529e-05,
+      "loss": 0.9856,
+      "step": 798
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6191861567721524,
+      "learning_rate": 1.2543171114599686e-05,
+      "loss": 0.9462,
+      "step": 799
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4222405531836453,
+      "learning_rate": 1.2558869701726845e-05,
+      "loss": 0.944,
+      "step": 800
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0681877228508534,
+      "learning_rate": 1.2574568288854004e-05,
+      "loss": 0.8986,
+      "step": 801
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6398716410634526,
+      "learning_rate": 1.2590266875981163e-05,
+      "loss": 0.9795,
+      "step": 802
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6693773081472965,
+      "learning_rate": 1.260596546310832e-05,
+      "loss": 0.968,
+      "step": 803
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8243403372376288,
+      "learning_rate": 1.262166405023548e-05,
+      "loss": 1.0221,
+      "step": 804
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8151354440342997,
+      "learning_rate": 1.2637362637362638e-05,
+      "loss": 1.0077,
+      "step": 805
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.421090335812189,
+      "learning_rate": 1.2653061224489798e-05,
+      "loss": 0.8316,
+      "step": 806
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.589142045767219,
+      "learning_rate": 1.2668759811616956e-05,
+      "loss": 0.8433,
+      "step": 807
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7163934255624083,
+      "learning_rate": 1.2684458398744115e-05,
+      "loss": 0.8963,
+      "step": 808
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8458419341728567,
+      "learning_rate": 1.2700156985871272e-05,
+      "loss": 0.9494,
+      "step": 809
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8205841225310502,
+      "learning_rate": 1.271585557299843e-05,
+      "loss": 0.8545,
+      "step": 810
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0190295657470667,
+      "learning_rate": 1.273155416012559e-05,
+      "loss": 1.0285,
+      "step": 811
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.821542750484262,
+      "learning_rate": 1.2747252747252747e-05,
+      "loss": 0.848,
+      "step": 812
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.290341560138557,
+      "learning_rate": 1.2762951334379907e-05,
+      "loss": 1.002,
+      "step": 813
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7684162921623348,
+      "learning_rate": 1.2778649921507064e-05,
+      "loss": 0.7983,
+      "step": 814
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6540371606065456,
+      "learning_rate": 1.2794348508634225e-05,
+      "loss": 1.03,
+      "step": 815
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8657211135638492,
+      "learning_rate": 1.2810047095761382e-05,
+      "loss": 0.8478,
+      "step": 816
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7931960266753744,
+      "learning_rate": 1.2825745682888541e-05,
+      "loss": 0.7528,
+      "step": 817
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5157777671089612,
+      "learning_rate": 1.28414442700157e-05,
+      "loss": 0.9079,
+      "step": 818
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6060862753054737,
+      "learning_rate": 1.2857142857142859e-05,
+      "loss": 0.7614,
+      "step": 819
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6102125808452996,
+      "learning_rate": 1.2872841444270016e-05,
+      "loss": 0.8796,
+      "step": 820
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6464565051736062,
+      "learning_rate": 1.2888540031397175e-05,
+      "loss": 0.961,
+      "step": 821
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7227044786732406,
+      "learning_rate": 1.2904238618524334e-05,
+      "loss": 0.8263,
+      "step": 822
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9744019324523172,
+      "learning_rate": 1.2919937205651491e-05,
+      "loss": 0.8441,
+      "step": 823
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6917702573834423,
+      "learning_rate": 1.2935635792778652e-05,
+      "loss": 0.9015,
+      "step": 824
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.750420672648657,
+      "learning_rate": 1.2951334379905809e-05,
+      "loss": 0.8404,
+      "step": 825
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3073199150827604,
+      "learning_rate": 1.296703296703297e-05,
+      "loss": 1.0261,
+      "step": 826
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6670854348049013,
+      "learning_rate": 1.2982731554160127e-05,
+      "loss": 0.8418,
+      "step": 827
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0728252182668654,
+      "learning_rate": 1.2998430141287286e-05,
+      "loss": 0.9455,
+      "step": 828
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7641556028155023,
+      "learning_rate": 1.3014128728414443e-05,
+      "loss": 0.9082,
+      "step": 829
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7653287281571128,
+      "learning_rate": 1.3029827315541603e-05,
+      "loss": 0.8711,
+      "step": 830
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.722450133863031,
+      "learning_rate": 1.304552590266876e-05,
+      "loss": 0.9774,
+      "step": 831
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6625441366143656,
+      "learning_rate": 1.3061224489795918e-05,
+      "loss": 0.8653,
+      "step": 832
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.664922660645558,
+      "learning_rate": 1.3076923076923078e-05,
+      "loss": 0.8118,
+      "step": 833
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7197014415480485,
+      "learning_rate": 1.3092621664050235e-05,
+      "loss": 1.0268,
+      "step": 834
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6344713565897375,
+      "learning_rate": 1.3108320251177396e-05,
+      "loss": 0.9329,
+      "step": 835
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7398654178556763,
+      "learning_rate": 1.3124018838304553e-05,
+      "loss": 1.0188,
+      "step": 836
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5370469699422253,
+      "learning_rate": 1.3139717425431712e-05,
+      "loss": 0.7834,
+      "step": 837
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5221466241384312,
+      "learning_rate": 1.3155416012558871e-05,
+      "loss": 0.8467,
+      "step": 838
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9542935878219252,
+      "learning_rate": 1.317111459968603e-05,
+      "loss": 0.8457,
+      "step": 839
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.648728426579854,
+      "learning_rate": 1.3186813186813187e-05,
+      "loss": 0.9097,
+      "step": 840
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1324306854927557,
+      "learning_rate": 1.3202511773940348e-05,
+      "loss": 0.9258,
+      "step": 841
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.771119512218935,
+      "learning_rate": 1.3218210361067505e-05,
+      "loss": 0.7355,
+      "step": 842
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6613155591189244,
+      "learning_rate": 1.3233908948194662e-05,
+      "loss": 0.885,
+      "step": 843
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5478259860672356,
+      "learning_rate": 1.3249607535321823e-05,
+      "loss": 0.9714,
+      "step": 844
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6454076806104396,
+      "learning_rate": 1.326530612244898e-05,
+      "loss": 1.0079,
+      "step": 845
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.830630498452458,
+      "learning_rate": 1.328100470957614e-05,
+      "loss": 0.867,
+      "step": 846
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9425351122791696,
+      "learning_rate": 1.3296703296703298e-05,
+      "loss": 0.9297,
+      "step": 847
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9317583378596523,
+      "learning_rate": 1.3312401883830456e-05,
+      "loss": 0.7996,
+      "step": 848
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8789324801357878,
+      "learning_rate": 1.3328100470957614e-05,
+      "loss": 0.8841,
+      "step": 849
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7237683458034148,
+      "learning_rate": 1.3343799058084774e-05,
+      "loss": 0.8438,
+      "step": 850
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.712745006939898,
+      "learning_rate": 1.3359497645211931e-05,
+      "loss": 0.9592,
+      "step": 851
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.670452039443604,
+      "learning_rate": 1.3375196232339092e-05,
+      "loss": 0.9566,
+      "step": 852
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6364102885547047,
+      "learning_rate": 1.339089481946625e-05,
+      "loss": 0.891,
+      "step": 853
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7857497440230614,
+      "learning_rate": 1.3406593406593406e-05,
+      "loss": 0.7758,
+      "step": 854
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7754047592115068,
+      "learning_rate": 1.3422291993720567e-05,
+      "loss": 0.9193,
+      "step": 855
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7747058683906485,
+      "learning_rate": 1.3437990580847724e-05,
+      "loss": 0.914,
+      "step": 856
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.616097890672518,
+      "learning_rate": 1.3453689167974883e-05,
+      "loss": 0.9094,
+      "step": 857
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6124050781095873,
+      "learning_rate": 1.3469387755102042e-05,
+      "loss": 0.8182,
+      "step": 858
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5617060163549739,
+      "learning_rate": 1.34850863422292e-05,
+      "loss": 0.8426,
+      "step": 859
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.832380481996329,
+      "learning_rate": 1.3500784929356358e-05,
+      "loss": 0.796,
+      "step": 860
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.945374041585282,
+      "learning_rate": 1.3516483516483519e-05,
+      "loss": 0.9582,
+      "step": 861
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9128789956325258,
+      "learning_rate": 1.3532182103610676e-05,
+      "loss": 0.7087,
+      "step": 862
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.560924481696143,
+      "learning_rate": 1.3547880690737836e-05,
+      "loss": 0.875,
+      "step": 863
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6809669151247162,
+      "learning_rate": 1.3563579277864994e-05,
+      "loss": 0.7381,
+      "step": 864
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.057546628761549,
+      "learning_rate": 1.357927786499215e-05,
+      "loss": 0.8212,
+      "step": 865
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2211010196472185,
+      "learning_rate": 1.3594976452119311e-05,
+      "loss": 0.7943,
+      "step": 866
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.539511702981035,
+      "learning_rate": 1.3610675039246469e-05,
+      "loss": 0.8928,
+      "step": 867
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.824515249343796,
+      "learning_rate": 1.3626373626373627e-05,
+      "loss": 0.7715,
+      "step": 868
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9314425002852345,
+      "learning_rate": 1.3642072213500786e-05,
+      "loss": 1.0304,
+      "step": 869
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8277441100749736,
+      "learning_rate": 1.3657770800627945e-05,
+      "loss": 0.8415,
+      "step": 870
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4760089283932198,
+      "learning_rate": 1.3673469387755102e-05,
+      "loss": 0.8496,
+      "step": 871
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.778650266328723,
+      "learning_rate": 1.3689167974882263e-05,
+      "loss": 0.8524,
+      "step": 872
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8074529601787053,
+      "learning_rate": 1.370486656200942e-05,
+      "loss": 0.9091,
+      "step": 873
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8637883806356061,
+      "learning_rate": 1.3720565149136579e-05,
+      "loss": 0.9901,
+      "step": 874
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6924839987981315,
+      "learning_rate": 1.3736263736263738e-05,
+      "loss": 0.9689,
+      "step": 875
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6669125109987133,
+      "learning_rate": 1.3751962323390895e-05,
+      "loss": 0.9332,
+      "step": 876
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1649214852068392,
+      "learning_rate": 1.3767660910518054e-05,
+      "loss": 0.9606,
+      "step": 877
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5848220827281474,
+      "learning_rate": 1.3783359497645213e-05,
+      "loss": 0.9648,
+      "step": 878
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8008673341809671,
+      "learning_rate": 1.3799058084772372e-05,
+      "loss": 0.8802,
+      "step": 879
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6927096009140987,
+      "learning_rate": 1.3814756671899529e-05,
+      "loss": 0.9202,
+      "step": 880
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6747768091850967,
+      "learning_rate": 1.383045525902669e-05,
+      "loss": 0.7927,
+      "step": 881
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6717030980516339,
+      "learning_rate": 1.3846153846153847e-05,
+      "loss": 0.9068,
+      "step": 882
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7947286786905883,
+      "learning_rate": 1.3861852433281007e-05,
+      "loss": 0.6712,
+      "step": 883
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5944749683474013,
+      "learning_rate": 1.3877551020408165e-05,
+      "loss": 0.9979,
+      "step": 884
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.279418098762317,
+      "learning_rate": 1.3893249607535323e-05,
+      "loss": 0.9038,
+      "step": 885
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7351063584345507,
+      "learning_rate": 1.3908948194662482e-05,
+      "loss": 0.8879,
+      "step": 886
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4933604642546074,
+      "learning_rate": 1.392464678178964e-05,
+      "loss": 0.9805,
+      "step": 887
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8482767365052222,
+      "learning_rate": 1.3940345368916798e-05,
+      "loss": 0.93,
+      "step": 888
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.883951833095372,
+      "learning_rate": 1.3956043956043957e-05,
+      "loss": 0.8929,
+      "step": 889
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5153373596707775,
+      "learning_rate": 1.3971742543171116e-05,
+      "loss": 0.8691,
+      "step": 890
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8669363168316966,
+      "learning_rate": 1.3987441130298273e-05,
+      "loss": 0.9391,
+      "step": 891
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.767778177015158,
+      "learning_rate": 1.4003139717425434e-05,
+      "loss": 0.8818,
+      "step": 892
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.646026631856072,
+      "learning_rate": 1.4018838304552591e-05,
+      "loss": 1.0106,
+      "step": 893
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7265996785362994,
+      "learning_rate": 1.403453689167975e-05,
+      "loss": 0.9005,
+      "step": 894
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9462393895742705,
+      "learning_rate": 1.4050235478806909e-05,
+      "loss": 0.8767,
+      "step": 895
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8702921574023539,
+      "learning_rate": 1.4065934065934068e-05,
+      "loss": 0.8146,
+      "step": 896
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.186144174274527,
+      "learning_rate": 1.4081632653061225e-05,
+      "loss": 0.9102,
+      "step": 897
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7892801422320197,
+      "learning_rate": 1.4097331240188384e-05,
+      "loss": 0.8567,
+      "step": 898
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7577703324035203,
+      "learning_rate": 1.4113029827315543e-05,
+      "loss": 1.0051,
+      "step": 899
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6017280379487115,
+      "learning_rate": 1.41287284144427e-05,
+      "loss": 1.0016,
+      "step": 900
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5512040906708515,
+      "learning_rate": 1.414442700156986e-05,
+      "loss": 0.9238,
+      "step": 901
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6884368693389444,
+      "learning_rate": 1.4160125588697018e-05,
+      "loss": 0.8871,
+      "step": 902
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.754368362575317,
+      "learning_rate": 1.4175824175824178e-05,
+      "loss": 0.9092,
+      "step": 903
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7503751161502085,
+      "learning_rate": 1.4191522762951335e-05,
+      "loss": 0.746,
+      "step": 904
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6553087266076199,
+      "learning_rate": 1.4207221350078494e-05,
+      "loss": 0.9754,
+      "step": 905
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7220556955085202,
+      "learning_rate": 1.4222919937205653e-05,
+      "loss": 0.8779,
+      "step": 906
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8514080151034118,
+      "learning_rate": 1.423861852433281e-05,
+      "loss": 0.9343,
+      "step": 907
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6728211785911167,
+      "learning_rate": 1.425431711145997e-05,
+      "loss": 1.0052,
+      "step": 908
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5810086334422075,
+      "learning_rate": 1.4270015698587128e-05,
+      "loss": 0.8153,
+      "step": 909
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7000012029848353,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 0.9279,
+      "step": 910
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5907980253109988,
+      "learning_rate": 1.4301412872841444e-05,
+      "loss": 0.8447,
+      "step": 911
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9071066981723257,
+      "learning_rate": 1.4317111459968605e-05,
+      "loss": 0.7728,
+      "step": 912
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8311011160686315,
+      "learning_rate": 1.4332810047095762e-05,
+      "loss": 0.9187,
+      "step": 913
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6537316335638867,
+      "learning_rate": 1.4348508634222923e-05,
+      "loss": 0.7163,
+      "step": 914
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7715049208384865,
+      "learning_rate": 1.436420722135008e-05,
+      "loss": 0.8711,
+      "step": 915
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.690592117882765,
+      "learning_rate": 1.4379905808477239e-05,
+      "loss": 0.8193,
+      "step": 916
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.692423006946277,
+      "learning_rate": 1.4395604395604396e-05,
+      "loss": 0.8061,
+      "step": 917
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8081298903770209,
+      "learning_rate": 1.4411302982731555e-05,
+      "loss": 0.9648,
+      "step": 918
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1926527526658566,
+      "learning_rate": 1.4427001569858714e-05,
+      "loss": 0.9064,
+      "step": 919
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.544071793729708,
+      "learning_rate": 1.4442700156985871e-05,
+      "loss": 0.9356,
+      "step": 920
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8361552486937287,
+      "learning_rate": 1.4458398744113031e-05,
+      "loss": 1.031,
+      "step": 921
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6380260499791304,
+      "learning_rate": 1.4474097331240189e-05,
+      "loss": 0.8854,
+      "step": 922
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3947911721930415,
+      "learning_rate": 1.448979591836735e-05,
+      "loss": 0.9251,
+      "step": 923
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8728641814899163,
+      "learning_rate": 1.4505494505494506e-05,
+      "loss": 0.9825,
+      "step": 924
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5592554332101076,
+      "learning_rate": 1.4521193092621665e-05,
+      "loss": 0.9223,
+      "step": 925
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5793540507937909,
+      "learning_rate": 1.4536891679748824e-05,
+      "loss": 0.962,
+      "step": 926
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7865291581285963,
+      "learning_rate": 1.4552590266875983e-05,
+      "loss": 0.7159,
+      "step": 927
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5669686299960275,
+      "learning_rate": 1.456828885400314e-05,
+      "loss": 0.8926,
+      "step": 928
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7701884399603238,
+      "learning_rate": 1.45839874411303e-05,
+      "loss": 0.6692,
+      "step": 929
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9393693606277542,
+      "learning_rate": 1.4599686028257458e-05,
+      "loss": 0.8441,
+      "step": 930
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2919572060492297,
+      "learning_rate": 1.4615384615384615e-05,
+      "loss": 0.918,
+      "step": 931
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0834145558500734,
+      "learning_rate": 1.4631083202511776e-05,
+      "loss": 0.9826,
+      "step": 932
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6986391341951037,
+      "learning_rate": 1.4646781789638933e-05,
+      "loss": 0.9442,
+      "step": 933
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0547700522194323,
+      "learning_rate": 1.4662480376766094e-05,
+      "loss": 1.0023,
+      "step": 934
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7720634680854919,
+      "learning_rate": 1.467817896389325e-05,
+      "loss": 0.6602,
+      "step": 935
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.610263481288216,
+      "learning_rate": 1.469387755102041e-05,
+      "loss": 0.906,
+      "step": 936
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9366454122346748,
+      "learning_rate": 1.4709576138147567e-05,
+      "loss": 0.937,
+      "step": 937
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.786915992667344,
+      "learning_rate": 1.4725274725274727e-05,
+      "loss": 0.7582,
+      "step": 938
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8512165688844928,
+      "learning_rate": 1.4740973312401885e-05,
+      "loss": 0.9203,
+      "step": 939
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7855132081541556,
+      "learning_rate": 1.4756671899529042e-05,
+      "loss": 0.8472,
+      "step": 940
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8321941719328425,
+      "learning_rate": 1.4772370486656202e-05,
+      "loss": 0.7928,
+      "step": 941
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8486888130021448,
+      "learning_rate": 1.478806907378336e-05,
+      "loss": 0.933,
+      "step": 942
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8657436023731557,
+      "learning_rate": 1.480376766091052e-05,
+      "loss": 0.6187,
+      "step": 943
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.759295082320484,
+      "learning_rate": 1.4819466248037677e-05,
+      "loss": 0.9139,
+      "step": 944
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6749622047075854,
+      "learning_rate": 1.4835164835164836e-05,
+      "loss": 0.8467,
+      "step": 945
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9298633636651932,
+      "learning_rate": 1.4850863422291995e-05,
+      "loss": 0.8862,
+      "step": 946
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.830384559106838,
+      "learning_rate": 1.4866562009419154e-05,
+      "loss": 0.8969,
+      "step": 947
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7518886347271028,
+      "learning_rate": 1.4882260596546311e-05,
+      "loss": 1.0322,
+      "step": 948
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.549401920512933,
+      "learning_rate": 1.4897959183673472e-05,
+      "loss": 0.8492,
+      "step": 949
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8266313608773426,
+      "learning_rate": 1.4913657770800629e-05,
+      "loss": 0.806,
+      "step": 950
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7836277905273198,
+      "learning_rate": 1.4929356357927786e-05,
+      "loss": 0.744,
+      "step": 951
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0720840477861238,
+      "learning_rate": 1.4945054945054947e-05,
+      "loss": 0.9288,
+      "step": 952
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7765066527488338,
+      "learning_rate": 1.4960753532182104e-05,
+      "loss": 0.9146,
+      "step": 953
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8712680093323557,
+      "learning_rate": 1.4976452119309265e-05,
+      "loss": 0.9833,
+      "step": 954
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8913559374594782,
+      "learning_rate": 1.4992150706436422e-05,
+      "loss": 0.9385,
+      "step": 955
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.570508758324839,
+      "learning_rate": 1.500784929356358e-05,
+      "loss": 0.8962,
+      "step": 956
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5828059534827035,
+      "learning_rate": 1.5023547880690738e-05,
+      "loss": 0.8382,
+      "step": 957
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7097908776387847,
+      "learning_rate": 1.5039246467817898e-05,
+      "loss": 0.9609,
+      "step": 958
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9217560801025584,
+      "learning_rate": 1.5054945054945056e-05,
+      "loss": 0.9156,
+      "step": 959
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.716422979170672,
+      "learning_rate": 1.5070643642072216e-05,
+      "loss": 1.0063,
+      "step": 960
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.356833357425073,
+      "learning_rate": 1.5086342229199373e-05,
+      "loss": 0.9778,
+      "step": 961
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8561807795581893,
+      "learning_rate": 1.510204081632653e-05,
+      "loss": 0.8908,
+      "step": 962
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.638926777011614,
+      "learning_rate": 1.5117739403453691e-05,
+      "loss": 1.035,
+      "step": 963
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.247453707263664,
+      "learning_rate": 1.5133437990580848e-05,
+      "loss": 0.8497,
+      "step": 964
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8121689962693821,
+      "learning_rate": 1.5149136577708007e-05,
+      "loss": 0.7191,
+      "step": 965
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7792809099975225,
+      "learning_rate": 1.5164835164835166e-05,
+      "loss": 0.6918,
+      "step": 966
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.731553701616306,
+      "learning_rate": 1.5180533751962325e-05,
+      "loss": 0.8454,
+      "step": 967
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7043900896589461,
+      "learning_rate": 1.5196232339089482e-05,
+      "loss": 0.8931,
+      "step": 968
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8521262705420087,
+      "learning_rate": 1.5211930926216643e-05,
+      "loss": 0.8287,
+      "step": 969
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5934088973018792,
+      "learning_rate": 1.52276295133438e-05,
+      "loss": 0.7748,
+      "step": 970
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7763731713709159,
+      "learning_rate": 1.524332810047096e-05,
+      "loss": 0.7463,
+      "step": 971
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5718075786790666,
+      "learning_rate": 1.5259026687598116e-05,
+      "loss": 0.8677,
+      "step": 972
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5202217933064097,
+      "learning_rate": 1.5274725274725277e-05,
+      "loss": 0.8853,
+      "step": 973
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.908496148722421,
+      "learning_rate": 1.5290423861852434e-05,
+      "loss": 0.9847,
+      "step": 974
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9983693973342627,
+      "learning_rate": 1.530612244897959e-05,
+      "loss": 0.8575,
+      "step": 975
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0642760468507175,
+      "learning_rate": 1.532182103610675e-05,
+      "loss": 0.918,
+      "step": 976
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.593752122281861,
+      "learning_rate": 1.533751962323391e-05,
+      "loss": 0.7715,
+      "step": 977
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8055136604337234,
+      "learning_rate": 1.535321821036107e-05,
+      "loss": 0.7821,
+      "step": 978
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1811879711046704,
+      "learning_rate": 1.5368916797488227e-05,
+      "loss": 0.9018,
+      "step": 979
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.849110772973131,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.9642,
+      "step": 980
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6023184393156416,
+      "learning_rate": 1.5400313971742544e-05,
+      "loss": 0.9845,
+      "step": 981
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5886188167258046,
+      "learning_rate": 1.5416012558869705e-05,
+      "loss": 0.8563,
+      "step": 982
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8636124937380523,
+      "learning_rate": 1.5431711145996862e-05,
+      "loss": 0.8673,
+      "step": 983
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7978116497536768,
+      "learning_rate": 1.544740973312402e-05,
+      "loss": 0.7033,
+      "step": 984
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6985055139952756,
+      "learning_rate": 1.546310832025118e-05,
+      "loss": 0.9896,
+      "step": 985
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.466706155487227,
+      "learning_rate": 1.5478806907378337e-05,
+      "loss": 0.9341,
+      "step": 986
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9638623399410284,
+      "learning_rate": 1.5494505494505498e-05,
+      "loss": 0.9557,
+      "step": 987
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.566167363383221,
+      "learning_rate": 1.5510204081632655e-05,
+      "loss": 0.8163,
+      "step": 988
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8444569095378358,
+      "learning_rate": 1.5525902668759812e-05,
+      "loss": 0.9071,
+      "step": 989
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6870823686486605,
+      "learning_rate": 1.5541601255886973e-05,
+      "loss": 0.8416,
+      "step": 990
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9822997512161977,
+      "learning_rate": 1.555729984301413e-05,
+      "loss": 0.8912,
+      "step": 991
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7725416479191867,
+      "learning_rate": 1.5572998430141287e-05,
+      "loss": 0.6754,
+      "step": 992
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8072860657359135,
+      "learning_rate": 1.5588697017268448e-05,
+      "loss": 0.8156,
+      "step": 993
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5576205142203892,
+      "learning_rate": 1.5604395604395605e-05,
+      "loss": 0.8398,
+      "step": 994
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6690601497494257,
+      "learning_rate": 1.5620094191522762e-05,
+      "loss": 0.8301,
+      "step": 995
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9934777403042514,
+      "learning_rate": 1.5635792778649923e-05,
+      "loss": 0.9317,
+      "step": 996
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1023885711939876,
+      "learning_rate": 1.565149136577708e-05,
+      "loss": 0.9179,
+      "step": 997
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.785734485921509,
+      "learning_rate": 1.566718995290424e-05,
+      "loss": 0.92,
+      "step": 998
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8294315221442281,
+      "learning_rate": 1.5682888540031398e-05,
+      "loss": 0.7694,
+      "step": 999
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7408536905329814,
+      "learning_rate": 1.5698587127158558e-05,
+      "loss": 0.9368,
+      "step": 1000
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9223460048449295,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 0.8862,
+      "step": 1001
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9664699097755245,
+      "learning_rate": 1.5729984301412876e-05,
+      "loss": 0.9045,
+      "step": 1002
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7476226561679135,
+      "learning_rate": 1.5745682888540033e-05,
+      "loss": 0.7125,
+      "step": 1003
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8267768635624893,
+      "learning_rate": 1.5761381475667194e-05,
+      "loss": 0.8591,
+      "step": 1004
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7811885088697144,
+      "learning_rate": 1.577708006279435e-05,
+      "loss": 0.7476,
+      "step": 1005
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7319867445824124,
+      "learning_rate": 1.5792778649921508e-05,
+      "loss": 0.8611,
+      "step": 1006
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8284784111941506,
+      "learning_rate": 1.580847723704867e-05,
+      "loss": 0.857,
+      "step": 1007
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7476582120222033,
+      "learning_rate": 1.5824175824175826e-05,
+      "loss": 1.0608,
+      "step": 1008
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7597688918574248,
+      "learning_rate": 1.5839874411302983e-05,
+      "loss": 0.8661,
+      "step": 1009
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.835705509581894,
+      "learning_rate": 1.5855572998430144e-05,
+      "loss": 0.9551,
+      "step": 1010
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9686796739702284,
+      "learning_rate": 1.58712715855573e-05,
+      "loss": 0.9866,
+      "step": 1011
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6025160067417745,
+      "learning_rate": 1.5886970172684458e-05,
+      "loss": 0.8566,
+      "step": 1012
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.711576904713869,
+      "learning_rate": 1.590266875981162e-05,
+      "loss": 0.8671,
+      "step": 1013
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.757345083232871,
+      "learning_rate": 1.5918367346938776e-05,
+      "loss": 0.94,
+      "step": 1014
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.301717888181273,
+      "learning_rate": 1.5934065934065933e-05,
+      "loss": 0.8783,
+      "step": 1015
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7211219556980095,
+      "learning_rate": 1.5949764521193094e-05,
+      "loss": 0.8207,
+      "step": 1016
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9524587765135006,
+      "learning_rate": 1.596546310832025e-05,
+      "loss": 0.9684,
+      "step": 1017
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6286439794728733,
+      "learning_rate": 1.598116169544741e-05,
+      "loss": 0.9469,
+      "step": 1018
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5837955888210196,
+      "learning_rate": 1.599686028257457e-05,
+      "loss": 1.065,
+      "step": 1019
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8306852508740852,
+      "learning_rate": 1.601255886970173e-05,
+      "loss": 0.9815,
+      "step": 1020
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8477042154277004,
+      "learning_rate": 1.6028257456828886e-05,
+      "loss": 0.9336,
+      "step": 1021
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4521635996279358,
+      "learning_rate": 1.6043956043956047e-05,
+      "loss": 0.8523,
+      "step": 1022
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6595412708483115,
+      "learning_rate": 1.6059654631083204e-05,
+      "loss": 0.9234,
+      "step": 1023
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7731929813782725,
+      "learning_rate": 1.6075353218210365e-05,
+      "loss": 0.9614,
+      "step": 1024
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5391720480201125,
+      "learning_rate": 1.6091051805337522e-05,
+      "loss": 0.941,
+      "step": 1025
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9966204881200489,
+      "learning_rate": 1.610675039246468e-05,
+      "loss": 0.9925,
+      "step": 1026
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7465428487799362,
+      "learning_rate": 1.612244897959184e-05,
+      "loss": 0.983,
+      "step": 1027
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7249965916826475,
+      "learning_rate": 1.6138147566718997e-05,
+      "loss": 0.9284,
+      "step": 1028
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6630510644972063,
+      "learning_rate": 1.6153846153846154e-05,
+      "loss": 0.93,
+      "step": 1029
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5269607399806173,
+      "learning_rate": 1.6169544740973315e-05,
+      "loss": 0.9028,
+      "step": 1030
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5053971735431786,
+      "learning_rate": 1.6185243328100472e-05,
+      "loss": 0.9301,
+      "step": 1031
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9194636476291969,
+      "learning_rate": 1.620094191522763e-05,
+      "loss": 0.9461,
+      "step": 1032
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.890033225054018,
+      "learning_rate": 1.621664050235479e-05,
+      "loss": 0.9252,
+      "step": 1033
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7805666783698804,
+      "learning_rate": 1.6232339089481947e-05,
+      "loss": 0.9268,
+      "step": 1034
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8527872554121435,
+      "learning_rate": 1.6248037676609107e-05,
+      "loss": 0.6964,
+      "step": 1035
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.695598459466154,
+      "learning_rate": 1.6263736263736265e-05,
+      "loss": 0.8768,
+      "step": 1036
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.598418410374014,
+      "learning_rate": 1.6279434850863422e-05,
+      "loss": 0.8476,
+      "step": 1037
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8368816733800333,
+      "learning_rate": 1.6295133437990582e-05,
+      "loss": 0.8653,
+      "step": 1038
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7387642535611887,
+      "learning_rate": 1.631083202511774e-05,
+      "loss": 0.8871,
+      "step": 1039
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.739681194156606,
+      "learning_rate": 1.63265306122449e-05,
+      "loss": 0.896,
+      "step": 1040
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5581706314961674,
+      "learning_rate": 1.6342229199372057e-05,
+      "loss": 0.9042,
+      "step": 1041
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7725639671273452,
+      "learning_rate": 1.6357927786499218e-05,
+      "loss": 0.8966,
+      "step": 1042
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2897587989891757,
+      "learning_rate": 1.6373626373626375e-05,
+      "loss": 0.9421,
+      "step": 1043
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7006733497804785,
+      "learning_rate": 1.6389324960753536e-05,
+      "loss": 0.9132,
+      "step": 1044
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8141404921074853,
+      "learning_rate": 1.6405023547880693e-05,
+      "loss": 0.97,
+      "step": 1045
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.641131023121253,
+      "learning_rate": 1.642072213500785e-05,
+      "loss": 0.9962,
+      "step": 1046
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.732638089049135,
+      "learning_rate": 1.643642072213501e-05,
+      "loss": 0.99,
+      "step": 1047
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.598505650199421,
+      "learning_rate": 1.6452119309262168e-05,
+      "loss": 0.903,
+      "step": 1048
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.69738142188311,
+      "learning_rate": 1.6467817896389325e-05,
+      "loss": 0.8792,
+      "step": 1049
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2465636070950437,
+      "learning_rate": 1.6483516483516486e-05,
+      "loss": 0.9293,
+      "step": 1050
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6340687130723448,
+      "learning_rate": 1.6499215070643643e-05,
+      "loss": 0.9264,
+      "step": 1051
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9686730191849424,
+      "learning_rate": 1.65149136577708e-05,
+      "loss": 1.0199,
+      "step": 1052
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8472297079203757,
+      "learning_rate": 1.653061224489796e-05,
+      "loss": 0.8137,
+      "step": 1053
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.328817698761475,
+      "learning_rate": 1.6546310832025118e-05,
+      "loss": 0.8238,
+      "step": 1054
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5934781077148255,
+      "learning_rate": 1.6562009419152278e-05,
+      "loss": 0.9649,
+      "step": 1055
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.282451742788982,
+      "learning_rate": 1.6577708006279435e-05,
+      "loss": 0.8592,
+      "step": 1056
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7707026151558081,
+      "learning_rate": 1.6593406593406596e-05,
+      "loss": 0.7066,
+      "step": 1057
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.831371824468732,
+      "learning_rate": 1.6609105180533753e-05,
+      "loss": 0.9231,
+      "step": 1058
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6663891312401276,
+      "learning_rate": 1.662480376766091e-05,
+      "loss": 0.8664,
+      "step": 1059
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6267823004793067,
+      "learning_rate": 1.664050235478807e-05,
+      "loss": 0.9282,
+      "step": 1060
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8460454810294507,
+      "learning_rate": 1.6656200941915228e-05,
+      "loss": 0.9155,
+      "step": 1061
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5167022337641614,
+      "learning_rate": 1.667189952904239e-05,
+      "loss": 0.9527,
+      "step": 1062
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.098286807679052,
+      "learning_rate": 1.6687598116169546e-05,
+      "loss": 0.9412,
+      "step": 1063
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8815803544498542,
+      "learning_rate": 1.6703296703296707e-05,
+      "loss": 1.0518,
+      "step": 1064
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7817009056195549,
+      "learning_rate": 1.6718995290423864e-05,
+      "loss": 0.7361,
+      "step": 1065
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.931493191141776,
+      "learning_rate": 1.673469387755102e-05,
+      "loss": 0.7894,
+      "step": 1066
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.74129109113232,
+      "learning_rate": 1.675039246467818e-05,
+      "loss": 0.8199,
+      "step": 1067
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6442415174466112,
+      "learning_rate": 1.676609105180534e-05,
+      "loss": 0.725,
+      "step": 1068
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8313686642602083,
+      "learning_rate": 1.6781789638932496e-05,
+      "loss": 0.7638,
+      "step": 1069
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6348538789553688,
+      "learning_rate": 1.6797488226059656e-05,
+      "loss": 0.8237,
+      "step": 1070
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.061351592462051,
+      "learning_rate": 1.6813186813186814e-05,
+      "loss": 0.8823,
+      "step": 1071
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8916593024413737,
+      "learning_rate": 1.682888540031397e-05,
+      "loss": 0.9101,
+      "step": 1072
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9439064675822377,
+      "learning_rate": 1.684458398744113e-05,
+      "loss": 0.8654,
+      "step": 1073
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.012929904503322,
+      "learning_rate": 1.686028257456829e-05,
+      "loss": 0.9729,
+      "step": 1074
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7152719426923648,
+      "learning_rate": 1.687598116169545e-05,
+      "loss": 0.8923,
+      "step": 1075
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6851528824302304,
+      "learning_rate": 1.6891679748822606e-05,
+      "loss": 0.776,
+      "step": 1076
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5522108375865935,
+      "learning_rate": 1.6907378335949767e-05,
+      "loss": 0.8671,
+      "step": 1077
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.921161063626231,
+      "learning_rate": 1.6923076923076924e-05,
+      "loss": 0.8312,
+      "step": 1078
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.493900920847617,
+      "learning_rate": 1.6938775510204085e-05,
+      "loss": 0.8418,
+      "step": 1079
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6965066065331833,
+      "learning_rate": 1.6954474097331242e-05,
+      "loss": 0.8934,
+      "step": 1080
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.761065476488696,
+      "learning_rate": 1.69701726844584e-05,
+      "loss": 0.9618,
+      "step": 1081
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8736001025749119,
+      "learning_rate": 1.698587127158556e-05,
+      "loss": 0.7787,
+      "step": 1082
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4821005283278925,
+      "learning_rate": 1.7001569858712717e-05,
+      "loss": 0.9948,
+      "step": 1083
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7993909578879006,
+      "learning_rate": 1.7017268445839878e-05,
+      "loss": 0.8724,
+      "step": 1084
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.746389627790212,
+      "learning_rate": 1.7032967032967035e-05,
+      "loss": 0.8467,
+      "step": 1085
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7939984856872485,
+      "learning_rate": 1.7048665620094192e-05,
+      "loss": 0.7164,
+      "step": 1086
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5804588610811858,
+      "learning_rate": 1.7064364207221352e-05,
+      "loss": 0.789,
+      "step": 1087
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9896474011641219,
+      "learning_rate": 1.708006279434851e-05,
+      "loss": 0.9485,
+      "step": 1088
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.629114531833108,
+      "learning_rate": 1.7095761381475667e-05,
+      "loss": 0.8357,
+      "step": 1089
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7858607886452438,
+      "learning_rate": 1.7111459968602827e-05,
+      "loss": 0.8863,
+      "step": 1090
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4254863909121622,
+      "learning_rate": 1.7127158555729985e-05,
+      "loss": 0.8703,
+      "step": 1091
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.768838923476117,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.9257,
+      "step": 1092
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7023632149808434,
+      "learning_rate": 1.7158555729984302e-05,
+      "loss": 0.9413,
+      "step": 1093
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5010766473253463,
+      "learning_rate": 1.717425431711146e-05,
+      "loss": 0.8407,
+      "step": 1094
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8856666610686947,
+      "learning_rate": 1.718995290423862e-05,
+      "loss": 0.9097,
+      "step": 1095
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7635552193735515,
+      "learning_rate": 1.7205651491365777e-05,
+      "loss": 1.0318,
+      "step": 1096
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6367608431052798,
+      "learning_rate": 1.7221350078492938e-05,
+      "loss": 0.9007,
+      "step": 1097
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9502825884315724,
+      "learning_rate": 1.7237048665620095e-05,
+      "loss": 0.954,
+      "step": 1098
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7776157281583593,
+      "learning_rate": 1.7252747252747256e-05,
+      "loss": 0.9629,
+      "step": 1099
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7035968879292,
+      "learning_rate": 1.7268445839874413e-05,
+      "loss": 1.0787,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.210318233089424,
+      "learning_rate": 1.7284144427001574e-05,
+      "loss": 0.9031,
+      "step": 1101
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5088504816373227,
+      "learning_rate": 1.729984301412873e-05,
+      "loss": 0.9137,
+      "step": 1102
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9745636417896977,
+      "learning_rate": 1.7315541601255888e-05,
+      "loss": 0.8309,
+      "step": 1103
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5940498471622886,
+      "learning_rate": 1.733124018838305e-05,
+      "loss": 1.0742,
+      "step": 1104
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8149212644026143,
+      "learning_rate": 1.7346938775510206e-05,
+      "loss": 0.9549,
+      "step": 1105
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7670450271463543,
+      "learning_rate": 1.7362637362637363e-05,
+      "loss": 0.6759,
+      "step": 1106
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6721168301318938,
+      "learning_rate": 1.7378335949764523e-05,
+      "loss": 0.915,
+      "step": 1107
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7715492743812231,
+      "learning_rate": 1.739403453689168e-05,
+      "loss": 0.939,
+      "step": 1108
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.94186325038456,
+      "learning_rate": 1.7409733124018838e-05,
+      "loss": 0.9972,
+      "step": 1109
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6448047206793759,
+      "learning_rate": 1.7425431711146e-05,
+      "loss": 0.9478,
+      "step": 1110
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6481457468396103,
+      "learning_rate": 1.7441130298273156e-05,
+      "loss": 0.9524,
+      "step": 1111
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7334628357476378,
+      "learning_rate": 1.7456828885400316e-05,
+      "loss": 0.8066,
+      "step": 1112
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8750234494847737,
+      "learning_rate": 1.7472527472527473e-05,
+      "loss": 0.9117,
+      "step": 1113
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8175527826179987,
+      "learning_rate": 1.748822605965463e-05,
+      "loss": 0.9825,
+      "step": 1114
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8169257606901688,
+      "learning_rate": 1.750392464678179e-05,
+      "loss": 0.7427,
+      "step": 1115
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.813581058034854,
+      "learning_rate": 1.751962323390895e-05,
+      "loss": 1.0587,
+      "step": 1116
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9571936648468327,
+      "learning_rate": 1.753532182103611e-05,
+      "loss": 0.9193,
+      "step": 1117
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.61440990528249,
+      "learning_rate": 1.7551020408163266e-05,
+      "loss": 0.7104,
+      "step": 1118
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7516861061768723,
+      "learning_rate": 1.7566718995290427e-05,
+      "loss": 0.9463,
+      "step": 1119
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.750254853597998,
+      "learning_rate": 1.7582417582417584e-05,
+      "loss": 0.8284,
+      "step": 1120
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5428730127655985,
+      "learning_rate": 1.7598116169544744e-05,
+      "loss": 1.0019,
+      "step": 1121
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1617631520203435,
+      "learning_rate": 1.76138147566719e-05,
+      "loss": 0.8648,
+      "step": 1122
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.80223562332589,
+      "learning_rate": 1.762951334379906e-05,
+      "loss": 0.6946,
+      "step": 1123
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.757019485550833,
+      "learning_rate": 1.764521193092622e-05,
+      "loss": 0.9696,
+      "step": 1124
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.713278852551737,
+      "learning_rate": 1.7660910518053377e-05,
+      "loss": 0.7207,
+      "step": 1125
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.916770080317809,
+      "learning_rate": 1.7676609105180537e-05,
+      "loss": 0.988,
+      "step": 1126
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4112031048576616,
+      "learning_rate": 1.7692307692307694e-05,
+      "loss": 0.9548,
+      "step": 1127
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8044979764858745,
+      "learning_rate": 1.770800627943485e-05,
+      "loss": 0.699,
+      "step": 1128
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8068883019221884,
+      "learning_rate": 1.772370486656201e-05,
+      "loss": 0.8292,
+      "step": 1129
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5183194784403788,
+      "learning_rate": 1.773940345368917e-05,
+      "loss": 0.9028,
+      "step": 1130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6878869451395755,
+      "learning_rate": 1.7755102040816327e-05,
+      "loss": 0.9346,
+      "step": 1131
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8907619048648676,
+      "learning_rate": 1.7770800627943487e-05,
+      "loss": 0.8079,
+      "step": 1132
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5185417896900544,
+      "learning_rate": 1.7786499215070644e-05,
+      "loss": 0.9469,
+      "step": 1133
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0589855908812464,
+      "learning_rate": 1.78021978021978e-05,
+      "loss": 0.7822,
+      "step": 1134
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8567627729614733,
+      "learning_rate": 1.7817896389324962e-05,
+      "loss": 1.011,
+      "step": 1135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8016376845323346,
+      "learning_rate": 1.783359497645212e-05,
+      "loss": 0.6987,
+      "step": 1136
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7570276131462177,
+      "learning_rate": 1.784929356357928e-05,
+      "loss": 1.0067,
+      "step": 1137
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.677966758908019,
+      "learning_rate": 1.7864992150706437e-05,
+      "loss": 0.882,
+      "step": 1138
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6198741128949623,
+      "learning_rate": 1.7880690737833598e-05,
+      "loss": 0.8348,
+      "step": 1139
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6936785121424633,
+      "learning_rate": 1.7896389324960755e-05,
+      "loss": 0.9207,
+      "step": 1140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.955299119997513,
+      "learning_rate": 1.7912087912087915e-05,
+      "loss": 0.93,
+      "step": 1141
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.755262477339357,
+      "learning_rate": 1.7927786499215073e-05,
+      "loss": 0.6656,
+      "step": 1142
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3311813552153122,
+      "learning_rate": 1.7943485086342233e-05,
+      "loss": 0.8952,
+      "step": 1143
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0836813438884083,
+      "learning_rate": 1.795918367346939e-05,
+      "loss": 0.7772,
+      "step": 1144
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8506119195169772,
+      "learning_rate": 1.7974882260596548e-05,
+      "loss": 0.8447,
+      "step": 1145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.352171725863756,
+      "learning_rate": 1.7990580847723708e-05,
+      "loss": 0.8483,
+      "step": 1146
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.622612272916359,
+      "learning_rate": 1.8006279434850865e-05,
+      "loss": 0.9515,
+      "step": 1147
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.081121602077935,
+      "learning_rate": 1.8021978021978023e-05,
+      "loss": 1.0211,
+      "step": 1148
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3898030396034198,
+      "learning_rate": 1.803767660910518e-05,
+      "loss": 0.9911,
+      "step": 1149
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1311499949963095,
+      "learning_rate": 1.805337519623234e-05,
+      "loss": 0.9874,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2189962435026276,
+      "learning_rate": 1.8069073783359498e-05,
+      "loss": 0.8518,
+      "step": 1151
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5316187144730498,
+      "learning_rate": 1.8084772370486658e-05,
+      "loss": 0.8599,
+      "step": 1152
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9780264226512982,
+      "learning_rate": 1.8100470957613815e-05,
+      "loss": 1.0885,
+      "step": 1153
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.687085059265193,
+      "learning_rate": 1.8116169544740976e-05,
+      "loss": 0.7666,
+      "step": 1154
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.996912300194023,
+      "learning_rate": 1.8131868131868133e-05,
+      "loss": 0.9141,
+      "step": 1155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.566437268336489,
+      "learning_rate": 1.814756671899529e-05,
+      "loss": 0.8344,
+      "step": 1156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5672350801792496,
+      "learning_rate": 1.816326530612245e-05,
+      "loss": 1.0537,
+      "step": 1157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8575512156783247,
+      "learning_rate": 1.8178963893249608e-05,
+      "loss": 0.8998,
+      "step": 1158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7559543810234852,
+      "learning_rate": 1.819466248037677e-05,
+      "loss": 1.0244,
+      "step": 1159
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8730211900233062,
+      "learning_rate": 1.8210361067503926e-05,
+      "loss": 1.1257,
+      "step": 1160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7945693191788292,
+      "learning_rate": 1.8226059654631086e-05,
+      "loss": 0.8524,
+      "step": 1161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.888204810286345,
+      "learning_rate": 1.8241758241758244e-05,
+      "loss": 0.9743,
+      "step": 1162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.041279838696801,
+      "learning_rate": 1.8257456828885404e-05,
+      "loss": 0.9562,
+      "step": 1163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5786752555345875,
+      "learning_rate": 1.827315541601256e-05,
+      "loss": 0.9275,
+      "step": 1164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9853177734739746,
+      "learning_rate": 1.828885400313972e-05,
+      "loss": 0.9203,
+      "step": 1165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.740944376156382,
+      "learning_rate": 1.830455259026688e-05,
+      "loss": 1.0018,
+      "step": 1166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7288784282133334,
+      "learning_rate": 1.8320251177394036e-05,
+      "loss": 0.9078,
+      "step": 1167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9163275405802203,
+      "learning_rate": 1.8335949764521194e-05,
+      "loss": 0.8576,
+      "step": 1168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7425145398696071,
+      "learning_rate": 1.835164835164835e-05,
+      "loss": 0.8877,
+      "step": 1169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5701034825135922,
+      "learning_rate": 1.836734693877551e-05,
+      "loss": 0.9163,
+      "step": 1170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7169508861211003,
+      "learning_rate": 1.838304552590267e-05,
+      "loss": 0.7594,
+      "step": 1171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0052056106700102,
+      "learning_rate": 1.839874411302983e-05,
+      "loss": 0.8714,
+      "step": 1172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7440105672379582,
+      "learning_rate": 1.8414442700156986e-05,
+      "loss": 0.9107,
+      "step": 1173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6818949027274406,
+      "learning_rate": 1.8430141287284147e-05,
+      "loss": 0.7311,
+      "step": 1174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3430022924804446,
+      "learning_rate": 1.8445839874411304e-05,
+      "loss": 0.8828,
+      "step": 1175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6585730099519778,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.8203,
+      "step": 1176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6547109176240111,
+      "learning_rate": 1.8477237048665622e-05,
+      "loss": 0.9512,
+      "step": 1177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0684489291244805,
+      "learning_rate": 1.849293563579278e-05,
+      "loss": 0.8674,
+      "step": 1178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6678597215498396,
+      "learning_rate": 1.850863422291994e-05,
+      "loss": 0.9503,
+      "step": 1179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7005610935104392,
+      "learning_rate": 1.8524332810047097e-05,
+      "loss": 0.8669,
+      "step": 1180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7305614666616425,
+      "learning_rate": 1.8540031397174257e-05,
+      "loss": 0.8211,
+      "step": 1181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7861661590810303,
+      "learning_rate": 1.8555729984301415e-05,
+      "loss": 0.845,
+      "step": 1182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7405164318259554,
+      "learning_rate": 1.8571428571428575e-05,
+      "loss": 0.8337,
+      "step": 1183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1711316068882116,
+      "learning_rate": 1.8587127158555732e-05,
+      "loss": 0.9917,
+      "step": 1184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9918309056059893,
+      "learning_rate": 1.860282574568289e-05,
+      "loss": 1.0146,
+      "step": 1185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9234396274022607,
+      "learning_rate": 1.861852433281005e-05,
+      "loss": 0.8527,
+      "step": 1186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6085900139300648,
+      "learning_rate": 1.8634222919937207e-05,
+      "loss": 0.8574,
+      "step": 1187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6806887818640415,
+      "learning_rate": 1.8649921507064364e-05,
+      "loss": 0.8368,
+      "step": 1188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.707492387996025,
+      "learning_rate": 1.866562009419152e-05,
+      "loss": 0.9028,
+      "step": 1189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7895973101878129,
+      "learning_rate": 1.8681318681318682e-05,
+      "loss": 0.6782,
+      "step": 1190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.063816965702635,
+      "learning_rate": 1.869701726844584e-05,
+      "loss": 0.894,
+      "step": 1191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7390977014850284,
+      "learning_rate": 1.8712715855573e-05,
+      "loss": 0.9468,
+      "step": 1192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7703475733779297,
+      "learning_rate": 1.8728414442700157e-05,
+      "loss": 0.7231,
+      "step": 1193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8739158418774882,
+      "learning_rate": 1.8744113029827318e-05,
+      "loss": 0.8819,
+      "step": 1194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.737580740762265,
+      "learning_rate": 1.8759811616954475e-05,
+      "loss": 0.8957,
+      "step": 1195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7041583481456903,
+      "learning_rate": 1.8775510204081636e-05,
+      "loss": 0.9181,
+      "step": 1196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.809602733901579,
+      "learning_rate": 1.8791208791208793e-05,
+      "loss": 0.7942,
+      "step": 1197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6110804724752572,
+      "learning_rate": 1.8806907378335953e-05,
+      "loss": 0.9586,
+      "step": 1198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3152876604092687,
+      "learning_rate": 1.882260596546311e-05,
+      "loss": 0.9108,
+      "step": 1199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5924013812300344,
+      "learning_rate": 1.8838304552590268e-05,
+      "loss": 0.8681,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6887102845938957,
+      "learning_rate": 1.885400313971743e-05,
+      "loss": 0.8598,
+      "step": 1201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.569675493510987,
+      "learning_rate": 1.8869701726844586e-05,
+      "loss": 0.9944,
+      "step": 1202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4911324378838438,
+      "learning_rate": 1.8885400313971746e-05,
+      "loss": 0.9116,
+      "step": 1203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.067467331059619,
+      "learning_rate": 1.8901098901098903e-05,
+      "loss": 1.0619,
+      "step": 1204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5709214132508795,
+      "learning_rate": 1.891679748822606e-05,
+      "loss": 0.8059,
+      "step": 1205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.529084288441918,
+      "learning_rate": 1.893249607535322e-05,
+      "loss": 0.8705,
+      "step": 1206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.638136447716283,
+      "learning_rate": 1.8948194662480378e-05,
+      "loss": 0.8296,
+      "step": 1207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.559373680616587,
+      "learning_rate": 1.8963893249607535e-05,
+      "loss": 1.0566,
+      "step": 1208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9357455190823294,
+      "learning_rate": 1.8979591836734696e-05,
+      "loss": 0.9935,
+      "step": 1209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.012555344069866,
+      "learning_rate": 1.8995290423861853e-05,
+      "loss": 0.915,
+      "step": 1210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2446725609960114,
+      "learning_rate": 1.901098901098901e-05,
+      "loss": 1.0375,
+      "step": 1211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6713800353547097,
+      "learning_rate": 1.902668759811617e-05,
+      "loss": 1.0152,
+      "step": 1212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7676849971123423,
+      "learning_rate": 1.9042386185243328e-05,
+      "loss": 0.9323,
+      "step": 1213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8293956067334562,
+      "learning_rate": 1.905808477237049e-05,
+      "loss": 0.9304,
+      "step": 1214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.430152477776286,
+      "learning_rate": 1.9073783359497646e-05,
+      "loss": 0.9531,
+      "step": 1215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1666425978965314,
+      "learning_rate": 1.9089481946624807e-05,
+      "loss": 0.8836,
+      "step": 1216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7428529890906133,
+      "learning_rate": 1.9105180533751964e-05,
+      "loss": 0.9245,
+      "step": 1217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.915758423720003,
+      "learning_rate": 1.9120879120879124e-05,
+      "loss": 0.9989,
+      "step": 1218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.310725892782091,
+      "learning_rate": 1.913657770800628e-05,
+      "loss": 0.8562,
+      "step": 1219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8169842376620158,
+      "learning_rate": 1.9152276295133442e-05,
+      "loss": 1.1397,
+      "step": 1220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6748581017590196,
+      "learning_rate": 1.91679748822606e-05,
+      "loss": 0.8112,
+      "step": 1221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9530031733814046,
+      "learning_rate": 1.9183673469387756e-05,
+      "loss": 0.9722,
+      "step": 1222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7165674057188265,
+      "learning_rate": 1.9199372056514917e-05,
+      "loss": 0.9473,
+      "step": 1223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.19117208074782,
+      "learning_rate": 1.9215070643642074e-05,
+      "loss": 0.9643,
+      "step": 1224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7794256492332413,
+      "learning_rate": 1.923076923076923e-05,
+      "loss": 0.7344,
+      "step": 1225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7385545374453708,
+      "learning_rate": 1.9246467817896392e-05,
+      "loss": 1.0897,
+      "step": 1226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7332437182269864,
+      "learning_rate": 1.926216640502355e-05,
+      "loss": 0.7487,
+      "step": 1227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.816406539867619,
+      "learning_rate": 1.9277864992150706e-05,
+      "loss": 0.8797,
+      "step": 1228
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8278453166840234,
+      "learning_rate": 1.9293563579277867e-05,
+      "loss": 0.7135,
+      "step": 1229
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8025344572012048,
+      "learning_rate": 1.9309262166405024e-05,
+      "loss": 0.8513,
+      "step": 1230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8051474782212327,
+      "learning_rate": 1.932496075353218e-05,
+      "loss": 0.9048,
+      "step": 1231
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1225371796826753,
+      "learning_rate": 1.9340659340659342e-05,
+      "loss": 0.9114,
+      "step": 1232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6788167620205474,
+      "learning_rate": 1.93563579277865e-05,
+      "loss": 0.9201,
+      "step": 1233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8999017396152391,
+      "learning_rate": 1.937205651491366e-05,
+      "loss": 0.8679,
+      "step": 1234
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.901529018288656,
+      "learning_rate": 1.9387755102040817e-05,
+      "loss": 0.925,
+      "step": 1235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.133672616097761,
+      "learning_rate": 1.9403453689167977e-05,
+      "loss": 0.8762,
+      "step": 1236
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7268283999332945,
+      "learning_rate": 1.9419152276295135e-05,
+      "loss": 1.0006,
+      "step": 1237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7617993586801066,
+      "learning_rate": 1.9434850863422295e-05,
+      "loss": 0.9924,
+      "step": 1238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8053610725777911,
+      "learning_rate": 1.9450549450549452e-05,
+      "loss": 0.6373,
+      "step": 1239
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.210392223097973,
+      "learning_rate": 1.9466248037676613e-05,
+      "loss": 0.9383,
+      "step": 1240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.06762781429194,
+      "learning_rate": 1.948194662480377e-05,
+      "loss": 0.9986,
+      "step": 1241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.568915982035608,
+      "learning_rate": 1.9497645211930927e-05,
+      "loss": 0.8937,
+      "step": 1242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6055743475579207,
+      "learning_rate": 1.9513343799058088e-05,
+      "loss": 0.8853,
+      "step": 1243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7963915300837879,
+      "learning_rate": 1.9529042386185245e-05,
+      "loss": 0.7263,
+      "step": 1244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.790613605571093,
+      "learning_rate": 1.9544740973312402e-05,
+      "loss": 0.9934,
+      "step": 1245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.060987799540306,
+      "learning_rate": 1.9560439560439563e-05,
+      "loss": 0.8705,
+      "step": 1246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6747518949209554,
+      "learning_rate": 1.957613814756672e-05,
+      "loss": 0.8405,
+      "step": 1247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.831060177540496,
+      "learning_rate": 1.9591836734693877e-05,
+      "loss": 0.9969,
+      "step": 1248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.993645280477595,
+      "learning_rate": 1.9607535321821038e-05,
+      "loss": 0.9378,
+      "step": 1249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6711536377150034,
+      "learning_rate": 1.9623233908948195e-05,
+      "loss": 0.8358,
+      "step": 1250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7378135747859558,
+      "learning_rate": 1.9638932496075356e-05,
+      "loss": 0.6993,
+      "step": 1251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0635761553959835,
+      "learning_rate": 1.9654631083202513e-05,
+      "loss": 0.9667,
+      "step": 1252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.662249722691596,
+      "learning_rate": 1.967032967032967e-05,
+      "loss": 0.9054,
+      "step": 1253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.081438049333332,
+      "learning_rate": 1.968602825745683e-05,
+      "loss": 0.7978,
+      "step": 1254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1612184848968536,
+      "learning_rate": 1.9701726844583988e-05,
+      "loss": 0.9431,
+      "step": 1255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6679304346438526,
+      "learning_rate": 1.971742543171115e-05,
+      "loss": 0.8802,
+      "step": 1256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.970666622133577,
+      "learning_rate": 1.9733124018838306e-05,
+      "loss": 0.8864,
+      "step": 1257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5368589205116034,
+      "learning_rate": 1.9748822605965466e-05,
+      "loss": 0.9217,
+      "step": 1258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8319439436207463,
+      "learning_rate": 1.9764521193092623e-05,
+      "loss": 0.8881,
+      "step": 1259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8067770575223223,
+      "learning_rate": 1.9780219780219784e-05,
+      "loss": 1.0611,
+      "step": 1260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8092477465160812,
+      "learning_rate": 1.979591836734694e-05,
+      "loss": 0.8917,
+      "step": 1261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7844192192050278,
+      "learning_rate": 1.98116169544741e-05,
+      "loss": 0.7656,
+      "step": 1262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.135927165979381,
+      "learning_rate": 1.982731554160126e-05,
+      "loss": 0.9785,
+      "step": 1263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7406502719010106,
+      "learning_rate": 1.9843014128728416e-05,
+      "loss": 0.9722,
+      "step": 1264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5681681178644271,
+      "learning_rate": 1.9858712715855573e-05,
+      "loss": 0.8911,
+      "step": 1265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7963507073980087,
+      "learning_rate": 1.9874411302982734e-05,
+      "loss": 0.9635,
+      "step": 1266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7379779123224548,
+      "learning_rate": 1.989010989010989e-05,
+      "loss": 0.9396,
+      "step": 1267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8088977433300133,
+      "learning_rate": 1.990580847723705e-05,
+      "loss": 0.7006,
+      "step": 1268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6449911118672644,
+      "learning_rate": 1.992150706436421e-05,
+      "loss": 0.9576,
+      "step": 1269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.612635922083334,
+      "learning_rate": 1.9937205651491366e-05,
+      "loss": 0.9764,
+      "step": 1270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6688117062705208,
+      "learning_rate": 1.9952904238618527e-05,
+      "loss": 0.891,
+      "step": 1271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7512099461239371,
+      "learning_rate": 1.9968602825745684e-05,
+      "loss": 0.7364,
+      "step": 1272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.83312549698183,
+      "learning_rate": 1.9984301412872844e-05,
+      "loss": 0.9053,
+      "step": 1273
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7237460529955437,
+      "learning_rate": 2e-05,
+      "loss": 0.9179,
+      "step": 1274
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8055459609151507,
+      "learning_rate": 1.9999999970888418e-05,
+      "loss": 0.8502,
+      "step": 1275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.78795696293218,
+      "learning_rate": 1.999999988355367e-05,
+      "loss": 1.0046,
+      "step": 1276
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2881925542171713,
+      "learning_rate": 1.9999999737995757e-05,
+      "loss": 0.8896,
+      "step": 1277
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8483354411132249,
+      "learning_rate": 1.9999999534214676e-05,
+      "loss": 0.7272,
+      "step": 1278
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7759633103374364,
+      "learning_rate": 1.9999999272210435e-05,
+      "loss": 0.8595,
+      "step": 1279
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8666786598831615,
+      "learning_rate": 1.9999998951983033e-05,
+      "loss": 0.8552,
+      "step": 1280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5452173195960548,
+      "learning_rate": 1.999999857353247e-05,
+      "loss": 0.9306,
+      "step": 1281
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2615187320309,
+      "learning_rate": 1.9999998136858748e-05,
+      "loss": 0.9212,
+      "step": 1282
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.663300049676059,
+      "learning_rate": 1.9999997641961873e-05,
+      "loss": 0.9038,
+      "step": 1283
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3764913427471694,
+      "learning_rate": 1.9999997088841844e-05,
+      "loss": 0.839,
+      "step": 1284
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5515500753950633,
+      "learning_rate": 1.9999996477498667e-05,
+      "loss": 1.0344,
+      "step": 1285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6830968068430903,
+      "learning_rate": 1.9999995807932346e-05,
+      "loss": 0.8692,
+      "step": 1286
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5449540158486963,
+      "learning_rate": 1.999999508014288e-05,
+      "loss": 0.9641,
+      "step": 1287
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5451356927334725,
+      "learning_rate": 1.999999429413028e-05,
+      "loss": 0.8979,
+      "step": 1288
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.390813112010595,
+      "learning_rate": 1.9999993449894542e-05,
+      "loss": 0.7521,
+      "step": 1289
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6301763453184386,
+      "learning_rate": 1.9999992547435683e-05,
+      "loss": 1.0668,
+      "step": 1290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7875288214434405,
+      "learning_rate": 1.99999915867537e-05,
+      "loss": 0.6625,
+      "step": 1291
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3679906031162394,
+      "learning_rate": 1.9999990567848595e-05,
+      "loss": 0.911,
+      "step": 1292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.765061229988411,
+      "learning_rate": 1.9999989490720384e-05,
+      "loss": 1.0167,
+      "step": 1293
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.083460205144846,
+      "learning_rate": 1.9999988355369067e-05,
+      "loss": 0.8563,
+      "step": 1294
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9477995687300131,
+      "learning_rate": 1.999998716179465e-05,
+      "loss": 0.9548,
+      "step": 1295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6802964731336179,
+      "learning_rate": 1.9999985909997145e-05,
+      "loss": 0.9447,
+      "step": 1296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.433852360203671,
+      "learning_rate": 1.9999984599976555e-05,
+      "loss": 0.8781,
+      "step": 1297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.54813838940723,
+      "learning_rate": 1.9999983231732888e-05,
+      "loss": 0.9462,
+      "step": 1298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.906665031962177,
+      "learning_rate": 1.9999981805266152e-05,
+      "loss": 1.0306,
+      "step": 1299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4192470000018316,
+      "learning_rate": 1.9999980320576356e-05,
+      "loss": 0.9131,
+      "step": 1300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9010549050452201,
+      "learning_rate": 1.999997877766351e-05,
+      "loss": 1.0812,
+      "step": 1301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6007327584623916,
+      "learning_rate": 1.9999977176527625e-05,
+      "loss": 1.0172,
+      "step": 1302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7013110902539161,
+      "learning_rate": 1.99999755171687e-05,
+      "loss": 1.0818,
+      "step": 1303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5978400731923896,
+      "learning_rate": 1.9999973799586754e-05,
+      "loss": 0.7972,
+      "step": 1304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6299024703968552,
+      "learning_rate": 1.9999972023781797e-05,
+      "loss": 0.9011,
+      "step": 1305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6643129954105533,
+      "learning_rate": 1.999997018975383e-05,
+      "loss": 1.0686,
+      "step": 1306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.914657190111075,
+      "learning_rate": 1.9999968297502878e-05,
+      "loss": 0.8725,
+      "step": 1307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5901593666376677,
+      "learning_rate": 1.9999966347028943e-05,
+      "loss": 0.8399,
+      "step": 1308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6506368578251676,
+      "learning_rate": 1.9999964338332037e-05,
+      "loss": 0.93,
+      "step": 1309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8321291545739704,
+      "learning_rate": 1.9999962271412173e-05,
+      "loss": 0.8462,
+      "step": 1310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.761534887665067,
+      "learning_rate": 1.9999960146269365e-05,
+      "loss": 1.0669,
+      "step": 1311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8312314732086554,
+      "learning_rate": 1.999995796290362e-05,
+      "loss": 0.8251,
+      "step": 1312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6801789640414782,
+      "learning_rate": 1.9999955721314954e-05,
+      "loss": 1.0093,
+      "step": 1313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1405180441755403,
+      "learning_rate": 1.999995342150338e-05,
+      "loss": 0.8493,
+      "step": 1314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7365619055979145,
+      "learning_rate": 1.999995106346891e-05,
+      "loss": 0.8691,
+      "step": 1315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.842154655006893,
+      "learning_rate": 1.9999948647211562e-05,
+      "loss": 0.9796,
+      "step": 1316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7996765148350433,
+      "learning_rate": 1.9999946172731347e-05,
+      "loss": 0.7474,
+      "step": 1317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6179263092461493,
+      "learning_rate": 1.999994364002828e-05,
+      "loss": 0.9595,
+      "step": 1318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.321949983469984,
+      "learning_rate": 1.999994104910237e-05,
+      "loss": 0.8381,
+      "step": 1319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5601916658631263,
+      "learning_rate": 1.9999938399953645e-05,
+      "loss": 0.9527,
+      "step": 1320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.003003401744781,
+      "learning_rate": 1.9999935692582107e-05,
+      "loss": 0.8294,
+      "step": 1321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7063491684129937,
+      "learning_rate": 1.9999932926987778e-05,
+      "loss": 0.9855,
+      "step": 1322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.012991283239903,
+      "learning_rate": 1.9999930103170676e-05,
+      "loss": 1.0241,
+      "step": 1323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7965348768164054,
+      "learning_rate": 1.9999927221130814e-05,
+      "loss": 0.7066,
+      "step": 1324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7438563731670171,
+      "learning_rate": 1.9999924280868208e-05,
+      "loss": 0.9197,
+      "step": 1325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0645131858716246,
+      "learning_rate": 1.999992128238288e-05,
+      "loss": 0.8416,
+      "step": 1326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9748685955351848,
+      "learning_rate": 1.9999918225674843e-05,
+      "loss": 0.8398,
+      "step": 1327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7361673629926917,
+      "learning_rate": 1.9999915110744116e-05,
+      "loss": 0.9138,
+      "step": 1328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6425686953196839,
+      "learning_rate": 1.9999911937590715e-05,
+      "loss": 0.9655,
+      "step": 1329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7883493509653676,
+      "learning_rate": 1.999990870621466e-05,
+      "loss": 0.998,
+      "step": 1330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7495922845748095,
+      "learning_rate": 1.9999905416615976e-05,
+      "loss": 0.9569,
+      "step": 1331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.589984364904814,
+      "learning_rate": 1.9999902068794674e-05,
+      "loss": 0.7552,
+      "step": 1332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6576143969608639,
+      "learning_rate": 1.9999898662750778e-05,
+      "loss": 0.8892,
+      "step": 1333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.732686102634992,
+      "learning_rate": 1.9999895198484303e-05,
+      "loss": 0.9303,
+      "step": 1334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5485585888481197,
+      "learning_rate": 1.9999891675995274e-05,
+      "loss": 0.806,
+      "step": 1335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5608636821299573,
+      "learning_rate": 1.9999888095283707e-05,
+      "loss": 0.8415,
+      "step": 1336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.826260384286386,
+      "learning_rate": 1.999988445634963e-05,
+      "loss": 0.98,
+      "step": 1337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7560839767773144,
+      "learning_rate": 1.999988075919306e-05,
+      "loss": 0.9687,
+      "step": 1338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7561886680888505,
+      "learning_rate": 1.9999877003814015e-05,
+      "loss": 0.8489,
+      "step": 1339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7604388322440225,
+      "learning_rate": 1.999987319021252e-05,
+      "loss": 0.6811,
+      "step": 1340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5992233522494064,
+      "learning_rate": 1.99998693183886e-05,
+      "loss": 0.8196,
+      "step": 1341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8694712439781578,
+      "learning_rate": 1.9999865388342274e-05,
+      "loss": 1.0149,
+      "step": 1342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6978491435242584,
+      "learning_rate": 1.9999861400073564e-05,
+      "loss": 1.0608,
+      "step": 1343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7290516695308171,
+      "learning_rate": 1.9999857353582496e-05,
+      "loss": 0.8099,
+      "step": 1344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8898080583879201,
+      "learning_rate": 1.9999853248869094e-05,
+      "loss": 0.7569,
+      "step": 1345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.532993889809439,
+      "learning_rate": 1.999984908593338e-05,
+      "loss": 0.8315,
+      "step": 1346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8347579803114902,
+      "learning_rate": 1.999984486477538e-05,
+      "loss": 0.9421,
+      "step": 1347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6765066684231091,
+      "learning_rate": 1.9999840585395113e-05,
+      "loss": 0.9472,
+      "step": 1348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7824335397273179,
+      "learning_rate": 1.9999836247792614e-05,
+      "loss": 0.9345,
+      "step": 1349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3872748663458254,
+      "learning_rate": 1.99998318519679e-05,
+      "loss": 0.8964,
+      "step": 1350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6943321808229184,
+      "learning_rate": 1.9999827397920993e-05,
+      "loss": 0.9568,
+      "step": 1351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5262253314826262,
+      "learning_rate": 1.9999822885651935e-05,
+      "loss": 0.9501,
+      "step": 1352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0546489968659505,
+      "learning_rate": 1.9999818315160736e-05,
+      "loss": 0.9136,
+      "step": 1353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6845066612077102,
+      "learning_rate": 1.999981368644743e-05,
+      "loss": 0.8465,
+      "step": 1354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9826432454777239,
+      "learning_rate": 1.9999808999512044e-05,
+      "loss": 0.8643,
+      "step": 1355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.532192982759424,
+      "learning_rate": 1.99998042543546e-05,
+      "loss": 0.9828,
+      "step": 1356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6891179402760401,
+      "learning_rate": 1.9999799450975136e-05,
+      "loss": 0.7931,
+      "step": 1357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.775096642207369,
+      "learning_rate": 1.999979458937367e-05,
+      "loss": 0.7856,
+      "step": 1358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.734630974236628,
+      "learning_rate": 1.9999789669550235e-05,
+      "loss": 0.8447,
+      "step": 1359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5215301679878555,
+      "learning_rate": 1.999978469150486e-05,
+      "loss": 0.8313,
+      "step": 1360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.258947368873347,
+      "learning_rate": 1.999977965523757e-05,
+      "loss": 0.8869,
+      "step": 1361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5509865751968293,
+      "learning_rate": 1.9999774560748398e-05,
+      "loss": 0.9318,
+      "step": 1362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6230551022670818,
+      "learning_rate": 1.9999769408037377e-05,
+      "loss": 0.9578,
+      "step": 1363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.027948826654006,
+      "learning_rate": 1.9999764197104526e-05,
+      "loss": 1.0228,
+      "step": 1364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9120505745247853,
+      "learning_rate": 1.9999758927949882e-05,
+      "loss": 0.9147,
+      "step": 1365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.190077803530714,
+      "learning_rate": 1.999975360057348e-05,
+      "loss": 1.0101,
+      "step": 1366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6700380238565435,
+      "learning_rate": 1.9999748214975345e-05,
+      "loss": 0.9743,
+      "step": 1367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7051026759457077,
+      "learning_rate": 1.999974277115551e-05,
+      "loss": 0.9852,
+      "step": 1368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9494237368102796,
+      "learning_rate": 1.9999737269114008e-05,
+      "loss": 0.9347,
+      "step": 1369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7919907074039958,
+      "learning_rate": 1.9999731708850868e-05,
+      "loss": 0.9474,
+      "step": 1370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7539306183235406,
+      "learning_rate": 1.9999726090366122e-05,
+      "loss": 0.9169,
+      "step": 1371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8034604968620267,
+      "learning_rate": 1.9999720413659808e-05,
+      "loss": 0.8996,
+      "step": 1372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7167833378604471,
+      "learning_rate": 1.9999714678731955e-05,
+      "loss": 0.9185,
+      "step": 1373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5966846471820293,
+      "learning_rate": 1.99997088855826e-05,
+      "loss": 0.855,
+      "step": 1374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.814895659847525,
+      "learning_rate": 1.9999703034211772e-05,
+      "loss": 0.6922,
+      "step": 1375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2252313654760165,
+      "learning_rate": 1.9999697124619505e-05,
+      "loss": 0.8374,
+      "step": 1376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.255208555472756,
+      "learning_rate": 1.9999691156805836e-05,
+      "loss": 0.9882,
+      "step": 1377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9937153440320092,
+      "learning_rate": 1.99996851307708e-05,
+      "loss": 0.9056,
+      "step": 1378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7206931839516668,
+      "learning_rate": 1.999967904651443e-05,
+      "loss": 0.9407,
+      "step": 1379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.700392118693385,
+      "learning_rate": 1.9999672904036768e-05,
+      "loss": 0.9643,
+      "step": 1380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8312579832708082,
+      "learning_rate": 1.9999666703337838e-05,
+      "loss": 0.7477,
+      "step": 1381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7004162069321407,
+      "learning_rate": 1.9999660444417687e-05,
+      "loss": 0.9408,
+      "step": 1382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7840457870007238,
+      "learning_rate": 1.9999654127276345e-05,
+      "loss": 0.7072,
+      "step": 1383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6553767369855001,
+      "learning_rate": 1.999964775191385e-05,
+      "loss": 0.8843,
+      "step": 1384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8627196927801022,
+      "learning_rate": 1.999964131833024e-05,
+      "loss": 1.0156,
+      "step": 1385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9352168746226124,
+      "learning_rate": 1.9999634826525555e-05,
+      "loss": 1.0067,
+      "step": 1386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2892808811031764,
+      "learning_rate": 1.999962827649983e-05,
+      "loss": 0.7505,
+      "step": 1387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7669353618185726,
+      "learning_rate": 1.9999621668253103e-05,
+      "loss": 0.8556,
+      "step": 1388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7622507622504442,
+      "learning_rate": 1.9999615001785412e-05,
+      "loss": 0.8319,
+      "step": 1389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6927002510999636,
+      "learning_rate": 1.9999608277096796e-05,
+      "loss": 0.9931,
+      "step": 1390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5113502487326713,
+      "learning_rate": 1.99996014941873e-05,
+      "loss": 0.9794,
+      "step": 1391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5816597489102697,
+      "learning_rate": 1.9999594653056952e-05,
+      "loss": 0.866,
+      "step": 1392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0073816684686356,
+      "learning_rate": 1.99995877537058e-05,
+      "loss": 0.9212,
+      "step": 1393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5908323371885287,
+      "learning_rate": 1.999958079613388e-05,
+      "loss": 0.8632,
+      "step": 1394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8170805463344433,
+      "learning_rate": 1.9999573780341236e-05,
+      "loss": 0.9175,
+      "step": 1395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.503438613012848,
+      "learning_rate": 1.9999566706327908e-05,
+      "loss": 0.7736,
+      "step": 1396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.682270818473176,
+      "learning_rate": 1.999955957409394e-05,
+      "loss": 0.8164,
+      "step": 1397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.029499946168032,
+      "learning_rate": 1.9999552383639368e-05,
+      "loss": 0.8721,
+      "step": 1398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0672164534100785,
+      "learning_rate": 1.9999545134964235e-05,
+      "loss": 1.0177,
+      "step": 1399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0074601835040826,
+      "learning_rate": 1.9999537828068586e-05,
+      "loss": 0.8496,
+      "step": 1400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.842079952269603,
+      "learning_rate": 1.999953046295246e-05,
+      "loss": 0.9144,
+      "step": 1401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7338333191486808,
+      "learning_rate": 1.9999523039615904e-05,
+      "loss": 0.8797,
+      "step": 1402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8732725831960348,
+      "learning_rate": 1.9999515558058958e-05,
+      "loss": 0.9706,
+      "step": 1403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.947320312450198,
+      "learning_rate": 1.9999508018281667e-05,
+      "loss": 0.8921,
+      "step": 1404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7648595638844227,
+      "learning_rate": 1.999950042028407e-05,
+      "loss": 0.8096,
+      "step": 1405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.637296720202476,
+      "learning_rate": 1.9999492764066225e-05,
+      "loss": 0.9087,
+      "step": 1406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7322175618499802,
+      "learning_rate": 1.999948504962816e-05,
+      "loss": 0.9164,
+      "step": 1407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9224639994244364,
+      "learning_rate": 1.9999477276969932e-05,
+      "loss": 0.7728,
+      "step": 1408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.685811146837617,
+      "learning_rate": 1.9999469446091578e-05,
+      "loss": 0.8656,
+      "step": 1409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7614650038558395,
+      "learning_rate": 1.999946155699315e-05,
+      "loss": 0.9449,
+      "step": 1410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5864808102792947,
+      "learning_rate": 1.999945360967469e-05,
+      "loss": 0.8466,
+      "step": 1411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6356735030208591,
+      "learning_rate": 1.999944560413624e-05,
+      "loss": 1.005,
+      "step": 1412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.586282171842174,
+      "learning_rate": 1.999943754037786e-05,
+      "loss": 0.9782,
+      "step": 1413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8775359522138397,
+      "learning_rate": 1.9999429418399585e-05,
+      "loss": 0.8972,
+      "step": 1414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8010949642147447,
+      "learning_rate": 1.9999421238201465e-05,
+      "loss": 0.9848,
+      "step": 1415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9452892248154923,
+      "learning_rate": 1.9999412999783555e-05,
+      "loss": 1.1291,
+      "step": 1416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0678948230322076,
+      "learning_rate": 1.999940470314589e-05,
+      "loss": 0.8302,
+      "step": 1417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6462008183937162,
+      "learning_rate": 1.9999396348288526e-05,
+      "loss": 0.9639,
+      "step": 1418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8717744000743302,
+      "learning_rate": 1.9999387935211515e-05,
+      "loss": 0.9174,
+      "step": 1419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4028665047255733,
+      "learning_rate": 1.99993794639149e-05,
+      "loss": 0.994,
+      "step": 1420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5352434179911576,
+      "learning_rate": 1.999937093439873e-05,
+      "loss": 0.8337,
+      "step": 1421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6593469545049992,
+      "learning_rate": 1.9999362346663056e-05,
+      "loss": 0.8312,
+      "step": 1422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.588931320363836,
+      "learning_rate": 1.999935370070793e-05,
+      "loss": 1.03,
+      "step": 1423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7935755084856027,
+      "learning_rate": 1.9999344996533402e-05,
+      "loss": 1.0602,
+      "step": 1424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.768051463829479,
+      "learning_rate": 1.9999336234139522e-05,
+      "loss": 0.9504,
+      "step": 1425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4995067729343572,
+      "learning_rate": 1.9999327413526337e-05,
+      "loss": 0.7677,
+      "step": 1426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.393337954189917,
+      "learning_rate": 1.9999318534693904e-05,
+      "loss": 0.8718,
+      "step": 1427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5937235043464373,
+      "learning_rate": 1.9999309597642275e-05,
+      "loss": 0.8522,
+      "step": 1428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.775227607123528,
+      "learning_rate": 1.9999300602371495e-05,
+      "loss": 1.0009,
+      "step": 1429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.741566011712476,
+      "learning_rate": 1.9999291548881628e-05,
+      "loss": 0.8349,
+      "step": 1430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7888989236631134,
+      "learning_rate": 1.9999282437172716e-05,
+      "loss": 0.8804,
+      "step": 1431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.629196794587314,
+      "learning_rate": 1.9999273267244813e-05,
+      "loss": 0.9492,
+      "step": 1432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.003497928521289,
+      "learning_rate": 1.9999264039097982e-05,
+      "loss": 0.8213,
+      "step": 1433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.674977458186919,
+      "learning_rate": 1.9999254752732265e-05,
+      "loss": 1.0149,
+      "step": 1434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8512887354015832,
+      "learning_rate": 1.999924540814772e-05,
+      "loss": 0.903,
+      "step": 1435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7165907212469085,
+      "learning_rate": 1.999923600534441e-05,
+      "loss": 1.0528,
+      "step": 1436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1150143058727613,
+      "learning_rate": 1.9999226544322377e-05,
+      "loss": 0.8813,
+      "step": 1437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.020171080078068,
+      "learning_rate": 1.999921702508168e-05,
+      "loss": 0.8778,
+      "step": 1438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5836392180492114,
+      "learning_rate": 1.9999207447622376e-05,
+      "loss": 1.0649,
+      "step": 1439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9652270444526372,
+      "learning_rate": 1.999919781194452e-05,
+      "loss": 0.9963,
+      "step": 1440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4857118185508065,
+      "learning_rate": 1.9999188118048173e-05,
+      "loss": 0.9806,
+      "step": 1441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.647642051114784,
+      "learning_rate": 1.9999178365933388e-05,
+      "loss": 0.9229,
+      "step": 1442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3329957329350317,
+      "learning_rate": 1.9999168555600215e-05,
+      "loss": 0.9896,
+      "step": 1443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7506275291868922,
+      "learning_rate": 1.999915868704872e-05,
+      "loss": 0.8364,
+      "step": 1444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8194406184501263,
+      "learning_rate": 1.9999148760278956e-05,
+      "loss": 0.6996,
+      "step": 1445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6382216849745121,
+      "learning_rate": 1.9999138775290985e-05,
+      "loss": 0.8989,
+      "step": 1446
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.588297993110073,
+      "learning_rate": 1.999912873208486e-05,
+      "loss": 0.9394,
+      "step": 1447
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.788512179929603,
+      "learning_rate": 1.9999118630660643e-05,
+      "loss": 0.9354,
+      "step": 1448
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.629410802433003,
+      "learning_rate": 1.999910847101839e-05,
+      "loss": 1.0229,
+      "step": 1449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7923015887392627,
+      "learning_rate": 1.9999098253158166e-05,
+      "loss": 0.7765,
+      "step": 1450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.868828429768284,
+      "learning_rate": 1.999908797708002e-05,
+      "loss": 0.9,
+      "step": 1451
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.755183547705761,
+      "learning_rate": 1.9999077642784024e-05,
+      "loss": 1.0062,
+      "step": 1452
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.440901989716646,
+      "learning_rate": 1.999906725027023e-05,
+      "loss": 0.9134,
+      "step": 1453
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6025874220406402,
+      "learning_rate": 1.9999056799538698e-05,
+      "loss": 0.9009,
+      "step": 1454
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9455871938800848,
+      "learning_rate": 1.9999046290589495e-05,
+      "loss": 0.9875,
+      "step": 1455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4806548032740054,
+      "learning_rate": 1.9999035723422677e-05,
+      "loss": 0.9725,
+      "step": 1456
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7749600118237399,
+      "learning_rate": 1.999902509803831e-05,
+      "loss": 1.1077,
+      "step": 1457
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.586271790997382,
+      "learning_rate": 1.9999014414436454e-05,
+      "loss": 0.9747,
+      "step": 1458
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.961056988978891,
+      "learning_rate": 1.9999003672617165e-05,
+      "loss": 0.8958,
+      "step": 1459
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.751416554232621,
+      "learning_rate": 1.9998992872580516e-05,
+      "loss": 0.9229,
+      "step": 1460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1395486217260267,
+      "learning_rate": 1.9998982014326563e-05,
+      "loss": 0.8493,
+      "step": 1461
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7935840840219506,
+      "learning_rate": 1.9998971097855372e-05,
+      "loss": 0.8106,
+      "step": 1462
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9848963451176704,
+      "learning_rate": 1.9998960123167003e-05,
+      "loss": 0.8339,
+      "step": 1463
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.695569079540921,
+      "learning_rate": 1.9998949090261525e-05,
+      "loss": 0.9318,
+      "step": 1464
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.713950042576206,
+      "learning_rate": 1.9998937999138998e-05,
+      "loss": 0.9253,
+      "step": 1465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.66054198305059,
+      "learning_rate": 1.999892684979949e-05,
+      "loss": 0.9811,
+      "step": 1466
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6298485824680107,
+      "learning_rate": 1.9998915642243064e-05,
+      "loss": 0.9266,
+      "step": 1467
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8238264801293618,
+      "learning_rate": 1.9998904376469784e-05,
+      "loss": 0.8686,
+      "step": 1468
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.854211571487225,
+      "learning_rate": 1.999889305247972e-05,
+      "loss": 0.9313,
+      "step": 1469
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.06915039103315,
+      "learning_rate": 1.9998881670272933e-05,
+      "loss": 0.8659,
+      "step": 1470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8334175430337417,
+      "learning_rate": 1.9998870229849492e-05,
+      "loss": 0.762,
+      "step": 1471
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.547016969942756,
+      "learning_rate": 1.9998858731209463e-05,
+      "loss": 0.8111,
+      "step": 1472
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6625394645735785,
+      "learning_rate": 1.999884717435291e-05,
+      "loss": 0.879,
+      "step": 1473
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.767574400841889,
+      "learning_rate": 1.999883555927991e-05,
+      "loss": 0.8987,
+      "step": 1474
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8942690081284013,
+      "learning_rate": 1.999882388599052e-05,
+      "loss": 1.0245,
+      "step": 1475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0101872875523514,
+      "learning_rate": 1.999881215448481e-05,
+      "loss": 1.0233,
+      "step": 1476
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5620882991152458,
+      "learning_rate": 1.999880036476285e-05,
+      "loss": 1.0089,
+      "step": 1477
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7751302325543967,
+      "learning_rate": 1.999878851682471e-05,
+      "loss": 0.8431,
+      "step": 1478
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6849426279473525,
+      "learning_rate": 1.999877661067046e-05,
+      "loss": 0.9655,
+      "step": 1479
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8992789827693541,
+      "learning_rate": 1.9998764646300163e-05,
+      "loss": 0.9477,
+      "step": 1480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0107260940946623,
+      "learning_rate": 1.9998752623713894e-05,
+      "loss": 0.9564,
+      "step": 1481
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7403069496906588,
+      "learning_rate": 1.9998740542911722e-05,
+      "loss": 0.701,
+      "step": 1482
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8033963364588245,
+      "learning_rate": 1.9998728403893717e-05,
+      "loss": 0.7181,
+      "step": 1483
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6614803885676852,
+      "learning_rate": 1.9998716206659948e-05,
+      "loss": 0.9117,
+      "step": 1484
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5695916935058816,
+      "learning_rate": 1.9998703951210492e-05,
+      "loss": 0.9938,
+      "step": 1485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7651952391062283,
+      "learning_rate": 1.9998691637545417e-05,
+      "loss": 0.7258,
+      "step": 1486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7558482809261289,
+      "learning_rate": 1.9998679265664787e-05,
+      "loss": 0.9513,
+      "step": 1487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.666042257546932,
+      "learning_rate": 1.9998666835568684e-05,
+      "loss": 0.7734,
+      "step": 1488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5335965623650047,
+      "learning_rate": 1.999865434725718e-05,
+      "loss": 0.8874,
+      "step": 1489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7762499952485452,
+      "learning_rate": 1.999864180073034e-05,
+      "loss": 0.7119,
+      "step": 1490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7624620555329306,
+      "learning_rate": 1.9998629195988243e-05,
+      "loss": 0.6759,
+      "step": 1491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6527321149800849,
+      "learning_rate": 1.9998616533030965e-05,
+      "loss": 1.0532,
+      "step": 1492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8029262738546272,
+      "learning_rate": 1.9998603811858573e-05,
+      "loss": 0.7859,
+      "step": 1493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.077062084856052,
+      "learning_rate": 1.9998591032471144e-05,
+      "loss": 0.8635,
+      "step": 1494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6949781373472532,
+      "learning_rate": 1.9998578194868754e-05,
+      "loss": 0.8637,
+      "step": 1495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.010589464468235,
+      "learning_rate": 1.9998565299051473e-05,
+      "loss": 0.8057,
+      "step": 1496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8193715235244443,
+      "learning_rate": 1.9998552345019383e-05,
+      "loss": 0.9466,
+      "step": 1497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8048706368805647,
+      "learning_rate": 1.9998539332772555e-05,
+      "loss": 0.9819,
+      "step": 1498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.763698066772919,
+      "learning_rate": 1.9998526262311063e-05,
+      "loss": 0.955,
+      "step": 1499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8646658753163778,
+      "learning_rate": 1.999851313363499e-05,
+      "loss": 0.8593,
+      "step": 1500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6197082592742362,
+      "learning_rate": 1.9998499946744404e-05,
+      "loss": 0.8847,
+      "step": 1501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9017749646150517,
+      "learning_rate": 1.999848670163939e-05,
+      "loss": 0.9317,
+      "step": 1502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6276425148377602,
+      "learning_rate": 1.9998473398320016e-05,
+      "loss": 0.9697,
+      "step": 1503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.56969611194336,
+      "learning_rate": 1.9998460036786365e-05,
+      "loss": 0.8779,
+      "step": 1504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.9505916758326745,
+      "learning_rate": 1.999844661703852e-05,
+      "loss": 0.8917,
+      "step": 1505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8248102748664454,
+      "learning_rate": 1.9998433139076547e-05,
+      "loss": 0.9633,
+      "step": 1506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2474524792700548,
+      "learning_rate": 1.9998419602900535e-05,
+      "loss": 0.8413,
+      "step": 1507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.933333651732959,
+      "learning_rate": 1.9998406008510555e-05,
+      "loss": 0.7916,
+      "step": 1508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8309633307395905,
+      "learning_rate": 1.999839235590669e-05,
+      "loss": 0.9061,
+      "step": 1509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.522295994522353,
+      "learning_rate": 1.999837864508902e-05,
+      "loss": 1.0033,
+      "step": 1510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7933039792377061,
+      "learning_rate": 1.999836487605763e-05,
+      "loss": 0.9584,
+      "step": 1511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7427234882615552,
+      "learning_rate": 1.9998351048812585e-05,
+      "loss": 0.9477,
+      "step": 1512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7815953804156186,
+      "learning_rate": 1.999833716335398e-05,
+      "loss": 0.873,
+      "step": 1513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.108917600547362,
+      "learning_rate": 1.999832321968189e-05,
+      "loss": 0.8618,
+      "step": 1514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1232973409451907,
+      "learning_rate": 1.9998309217796395e-05,
+      "loss": 0.834,
+      "step": 1515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6766169503758983,
+      "learning_rate": 1.9998295157697583e-05,
+      "loss": 0.7813,
+      "step": 1516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.806991785853015,
+      "learning_rate": 1.9998281039385527e-05,
+      "loss": 0.9058,
+      "step": 1517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8194707951932942,
+      "learning_rate": 1.9998266862860315e-05,
+      "loss": 0.9123,
+      "step": 1518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.975272281946664,
+      "learning_rate": 1.9998252628122025e-05,
+      "loss": 0.9228,
+      "step": 1519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6255805663520544,
+      "learning_rate": 1.9998238335170746e-05,
+      "loss": 0.9024,
+      "step": 1520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8151774208076794,
+      "learning_rate": 1.9998223984006558e-05,
+      "loss": 0.9107,
+      "step": 1521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9182117677177706,
+      "learning_rate": 1.9998209574629542e-05,
+      "loss": 0.9303,
+      "step": 1522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7110322137517988,
+      "learning_rate": 1.9998195107039785e-05,
+      "loss": 0.9237,
+      "step": 1523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7973138467836174,
+      "learning_rate": 1.9998180581237372e-05,
+      "loss": 1.0314,
+      "step": 1524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0129945138632284,
+      "learning_rate": 1.9998165997222387e-05,
+      "loss": 0.8287,
+      "step": 1525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7691751252928318,
+      "learning_rate": 1.9998151354994914e-05,
+      "loss": 0.863,
+      "step": 1526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.007335409429243,
+      "learning_rate": 1.999813665455504e-05,
+      "loss": 0.9078,
+      "step": 1527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5702192724544928,
+      "learning_rate": 1.9998121895902843e-05,
+      "loss": 0.8806,
+      "step": 1528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.108703287725206,
+      "learning_rate": 1.999810707903842e-05,
+      "loss": 0.9006,
+      "step": 1529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.73364990079322,
+      "learning_rate": 1.999809220396185e-05,
+      "loss": 0.8638,
+      "step": 1530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7036372061028298,
+      "learning_rate": 1.9998077270673222e-05,
+      "loss": 0.836,
+      "step": 1531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4275580880911933,
+      "learning_rate": 1.9998062279172624e-05,
+      "loss": 0.7806,
+      "step": 1532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9734394383776088,
+      "learning_rate": 1.9998047229460144e-05,
+      "loss": 0.9534,
+      "step": 1533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6196158232920965,
+      "learning_rate": 1.9998032121535863e-05,
+      "loss": 0.9294,
+      "step": 1534
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0553148510894816,
+      "learning_rate": 1.9998016955399877e-05,
+      "loss": 0.8754,
+      "step": 1535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0560003895612873,
+      "learning_rate": 1.999800173105227e-05,
+      "loss": 0.8575,
+      "step": 1536
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.603086846636104,
+      "learning_rate": 1.9997986448493133e-05,
+      "loss": 0.8922,
+      "step": 1537
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6360053649247897,
+      "learning_rate": 1.999797110772255e-05,
+      "loss": 0.9472,
+      "step": 1538
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8012764961257188,
+      "learning_rate": 1.9997955708740616e-05,
+      "loss": 0.9125,
+      "step": 1539
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9017997029132736,
+      "learning_rate": 1.999794025154742e-05,
+      "loss": 0.9339,
+      "step": 1540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7579455530951853,
+      "learning_rate": 1.999792473614305e-05,
+      "loss": 0.8463,
+      "step": 1541
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8165492157919578,
+      "learning_rate": 1.9997909162527596e-05,
+      "loss": 0.9686,
+      "step": 1542
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7647158128325473,
+      "learning_rate": 1.999789353070115e-05,
+      "loss": 0.875,
+      "step": 1543
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.957738098483419,
+      "learning_rate": 1.99978778406638e-05,
+      "loss": 0.9897,
+      "step": 1544
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9744696442405332,
+      "learning_rate": 1.9997862092415645e-05,
+      "loss": 0.806,
+      "step": 1545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7513590068927223,
+      "learning_rate": 1.9997846285956767e-05,
+      "loss": 0.936,
+      "step": 1546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.717749680428828,
+      "learning_rate": 1.9997830421287265e-05,
+      "loss": 0.9243,
+      "step": 1547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.810231352486841,
+      "learning_rate": 1.999781449840723e-05,
+      "loss": 1.0337,
+      "step": 1548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.426438712913837,
+      "learning_rate": 1.999779851731675e-05,
+      "loss": 0.8869,
+      "step": 1549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8221564061175555,
+      "learning_rate": 1.9997782478015923e-05,
+      "loss": 1.0458,
+      "step": 1550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6848371383048808,
+      "learning_rate": 1.9997766380504842e-05,
+      "loss": 0.8653,
+      "step": 1551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8363568573356044,
+      "learning_rate": 1.9997750224783605e-05,
+      "loss": 0.978,
+      "step": 1552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7350089630439867,
+      "learning_rate": 1.9997734010852294e-05,
+      "loss": 0.8738,
+      "step": 1553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7296886202290918,
+      "learning_rate": 1.9997717738711013e-05,
+      "loss": 0.8152,
+      "step": 1554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6263215281507495,
+      "learning_rate": 1.9997701408359852e-05,
+      "loss": 0.9566,
+      "step": 1555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.525434971470341,
+      "learning_rate": 1.9997685019798913e-05,
+      "loss": 0.9424,
+      "step": 1556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7299833401336657,
+      "learning_rate": 1.999766857302828e-05,
+      "loss": 0.8608,
+      "step": 1557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7939543020065232,
+      "learning_rate": 1.9997652068048057e-05,
+      "loss": 1.0935,
+      "step": 1558
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8908489355506306,
+      "learning_rate": 1.999763550485834e-05,
+      "loss": 0.917,
+      "step": 1559
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8526105783983153,
+      "learning_rate": 1.9997618883459225e-05,
+      "loss": 1.003,
+      "step": 1560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5329836884263954,
+      "learning_rate": 1.9997602203850807e-05,
+      "loss": 1.1003,
+      "step": 1561
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.873104676116411,
+      "learning_rate": 1.9997585466033186e-05,
+      "loss": 0.972,
+      "step": 1562
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7185456217925166,
+      "learning_rate": 1.999756867000645e-05,
+      "loss": 0.8875,
+      "step": 1563
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8498812246783523,
+      "learning_rate": 1.9997551815770712e-05,
+      "loss": 0.8421,
+      "step": 1564
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7088133168953146,
+      "learning_rate": 1.9997534903326058e-05,
+      "loss": 0.8538,
+      "step": 1565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8012396818974414,
+      "learning_rate": 1.9997517932672592e-05,
+      "loss": 0.6977,
+      "step": 1566
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.000212910122706,
+      "learning_rate": 1.9997500903810412e-05,
+      "loss": 0.8698,
+      "step": 1567
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1260208049356617,
+      "learning_rate": 1.9997483816739613e-05,
+      "loss": 0.9389,
+      "step": 1568
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.289478853750021,
+      "learning_rate": 1.99974666714603e-05,
+      "loss": 0.8702,
+      "step": 1569
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.702976560467114,
+      "learning_rate": 1.9997449467972574e-05,
+      "loss": 1.1071,
+      "step": 1570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7941613579568396,
+      "learning_rate": 1.9997432206276528e-05,
+      "loss": 0.9141,
+      "step": 1571
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5809466680076563,
+      "learning_rate": 1.9997414886372272e-05,
+      "loss": 1.0225,
+      "step": 1572
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.872046857816528,
+      "learning_rate": 1.9997397508259895e-05,
+      "loss": 0.9035,
+      "step": 1573
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2466005042178496,
+      "learning_rate": 1.9997380071939506e-05,
+      "loss": 1.0322,
+      "step": 1574
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.601035117842701,
+      "learning_rate": 1.999736257741121e-05,
+      "loss": 0.8961,
+      "step": 1575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7501577553580256,
+      "learning_rate": 1.9997345024675095e-05,
+      "loss": 0.9695,
+      "step": 1576
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8907141071005917,
+      "learning_rate": 1.999732741373128e-05,
+      "loss": 0.7128,
+      "step": 1577
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.106841884531087,
+      "learning_rate": 1.9997309744579857e-05,
+      "loss": 1.023,
+      "step": 1578
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3247168530130695,
+      "learning_rate": 1.9997292017220933e-05,
+      "loss": 1.079,
+      "step": 1579
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0378886959290914,
+      "learning_rate": 1.9997274231654605e-05,
+      "loss": 1.0422,
+      "step": 1580
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6612068993161904,
+      "learning_rate": 1.9997256387880987e-05,
+      "loss": 0.8916,
+      "step": 1581
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.769912542626668,
+      "learning_rate": 1.9997238485900175e-05,
+      "loss": 0.6628,
+      "step": 1582
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8859246735541089,
+      "learning_rate": 1.9997220525712276e-05,
+      "loss": 0.8138,
+      "step": 1583
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2450242854518567,
+      "learning_rate": 1.999720250731739e-05,
+      "loss": 1.0137,
+      "step": 1584
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.796344761310367,
+      "learning_rate": 1.9997184430715632e-05,
+      "loss": 0.7315,
+      "step": 1585
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.880272714971701,
+      "learning_rate": 1.9997166295907097e-05,
+      "loss": 0.8958,
+      "step": 1586
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7009923791106216,
+      "learning_rate": 1.9997148102891897e-05,
+      "loss": 0.8727,
+      "step": 1587
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.835358744930637,
+      "learning_rate": 1.9997129851670136e-05,
+      "loss": 0.8525,
+      "step": 1588
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6273374861414052,
+      "learning_rate": 1.999711154224192e-05,
+      "loss": 1.0051,
+      "step": 1589
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8255767277996657,
+      "learning_rate": 1.9997093174607354e-05,
+      "loss": 0.8514,
+      "step": 1590
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7430382454436006,
+      "learning_rate": 1.999707474876655e-05,
+      "loss": 0.853,
+      "step": 1591
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8121297232703355,
+      "learning_rate": 1.9997056264719607e-05,
+      "loss": 1.0873,
+      "step": 1592
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.80664953538914,
+      "learning_rate": 1.999703772246664e-05,
+      "loss": 0.925,
+      "step": 1593
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8158035662833767,
+      "learning_rate": 1.999701912200775e-05,
+      "loss": 0.9956,
+      "step": 1594
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7969107054189783,
+      "learning_rate": 1.9997000463343057e-05,
+      "loss": 0.8839,
+      "step": 1595
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7284447433026606,
+      "learning_rate": 1.9996981746472657e-05,
+      "loss": 1.0609,
+      "step": 1596
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0136398836004465,
+      "learning_rate": 1.9996962971396664e-05,
+      "loss": 0.8926,
+      "step": 1597
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8060341116905358,
+      "learning_rate": 1.9996944138115192e-05,
+      "loss": 0.9086,
+      "step": 1598
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7943927411534821,
+      "learning_rate": 1.9996925246628344e-05,
+      "loss": 0.813,
+      "step": 1599
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8327182247667608,
+      "learning_rate": 1.999690629693623e-05,
+      "loss": 0.9137,
+      "step": 1600
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1787475964125416,
+      "learning_rate": 1.9996887289038968e-05,
+      "loss": 0.7948,
+      "step": 1601
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3633427220568244,
+      "learning_rate": 1.999686822293666e-05,
+      "loss": 0.8391,
+      "step": 1602
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7542659109190508,
+      "learning_rate": 1.999684909862942e-05,
+      "loss": 0.6793,
+      "step": 1603
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9034841082211185,
+      "learning_rate": 1.999682991611736e-05,
+      "loss": 1.0203,
+      "step": 1604
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7829133842063671,
+      "learning_rate": 1.9996810675400588e-05,
+      "loss": 0.8651,
+      "step": 1605
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.86784728065355,
+      "learning_rate": 1.9996791376479223e-05,
+      "loss": 0.8188,
+      "step": 1606
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7734131588517876,
+      "learning_rate": 1.9996772019353373e-05,
+      "loss": 0.8784,
+      "step": 1607
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6079090955973507,
+      "learning_rate": 1.9996752604023152e-05,
+      "loss": 0.9449,
+      "step": 1608
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4736986976509536,
+      "learning_rate": 1.999673313048867e-05,
+      "loss": 0.8294,
+      "step": 1609
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.025621531755809,
+      "learning_rate": 1.9996713598750047e-05,
+      "loss": 0.8336,
+      "step": 1610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.572200669724049,
+      "learning_rate": 1.999669400880739e-05,
+      "loss": 0.9807,
+      "step": 1611
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0560829040863338,
+      "learning_rate": 1.9996674360660815e-05,
+      "loss": 0.8765,
+      "step": 1612
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0618233394394436,
+      "learning_rate": 1.999665465431044e-05,
+      "loss": 0.8148,
+      "step": 1613
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.503648004499104,
+      "learning_rate": 1.999663488975637e-05,
+      "loss": 0.9903,
+      "step": 1614
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2369070412065053,
+      "learning_rate": 1.9996615066998733e-05,
+      "loss": 0.7788,
+      "step": 1615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.710550403668236,
+      "learning_rate": 1.9996595186037635e-05,
+      "loss": 0.9662,
+      "step": 1616
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9030608709073793,
+      "learning_rate": 1.9996575246873196e-05,
+      "loss": 0.9993,
+      "step": 1617
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1064773458290933,
+      "learning_rate": 1.9996555249505533e-05,
+      "loss": 0.9161,
+      "step": 1618
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7073800177858274,
+      "learning_rate": 1.999653519393476e-05,
+      "loss": 0.8423,
+      "step": 1619
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7754231273644554,
+      "learning_rate": 1.999651508016099e-05,
+      "loss": 0.9712,
+      "step": 1620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.684155952624959,
+      "learning_rate": 1.999649490818435e-05,
+      "loss": 1.0226,
+      "step": 1621
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.592545594088582,
+      "learning_rate": 1.999647467800495e-05,
+      "loss": 0.9195,
+      "step": 1622
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6639856892841736,
+      "learning_rate": 1.999645438962291e-05,
+      "loss": 0.9812,
+      "step": 1623
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2786946257889484,
+      "learning_rate": 1.9996434043038344e-05,
+      "loss": 0.98,
+      "step": 1624
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6382151557309461,
+      "learning_rate": 1.9996413638251377e-05,
+      "loss": 1.1067,
+      "step": 1625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5817954570943296,
+      "learning_rate": 1.9996393175262123e-05,
+      "loss": 1.0276,
+      "step": 1626
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6184237279409497,
+      "learning_rate": 1.9996372654070708e-05,
+      "loss": 0.946,
+      "step": 1627
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4591008688043743,
+      "learning_rate": 1.999635207467724e-05,
+      "loss": 0.8759,
+      "step": 1628
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8584359932774077,
+      "learning_rate": 1.999633143708185e-05,
+      "loss": 0.9231,
+      "step": 1629
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.743169442219941,
+      "learning_rate": 1.9996310741284652e-05,
+      "loss": 0.9554,
+      "step": 1630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7966076914817208,
+      "learning_rate": 1.999628998728577e-05,
+      "loss": 0.7356,
+      "step": 1631
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6529635733257062,
+      "learning_rate": 1.999626917508532e-05,
+      "loss": 0.9274,
+      "step": 1632
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.788537115868023,
+      "learning_rate": 1.999624830468343e-05,
+      "loss": 0.7104,
+      "step": 1633
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6112829277895648,
+      "learning_rate": 1.9996227376080213e-05,
+      "loss": 0.9815,
+      "step": 1634
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7787458735542774,
+      "learning_rate": 1.99962063892758e-05,
+      "loss": 0.7144,
+      "step": 1635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8571773725942813,
+      "learning_rate": 1.9996185344270306e-05,
+      "loss": 1.1678,
+      "step": 1636
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6938879903906368,
+      "learning_rate": 1.9996164241063857e-05,
+      "loss": 1.0334,
+      "step": 1637
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9017324895726204,
+      "learning_rate": 1.9996143079656575e-05,
+      "loss": 0.8174,
+      "step": 1638
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9530613717950354,
+      "learning_rate": 1.999612186004858e-05,
+      "loss": 0.9449,
+      "step": 1639
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9502092383577085,
+      "learning_rate": 1.9996100582240002e-05,
+      "loss": 1.0163,
+      "step": 1640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.601630679484191,
+      "learning_rate": 1.9996079246230962e-05,
+      "loss": 0.8761,
+      "step": 1641
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9670127093478988,
+      "learning_rate": 1.9996057852021584e-05,
+      "loss": 1.0223,
+      "step": 1642
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.342410660736773,
+      "learning_rate": 1.9996036399611992e-05,
+      "loss": 0.8872,
+      "step": 1643
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8495535614155498,
+      "learning_rate": 1.999601488900231e-05,
+      "loss": 0.7104,
+      "step": 1644
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8984200223914067,
+      "learning_rate": 1.9995993320192668e-05,
+      "loss": 0.883,
+      "step": 1645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.708466000362097,
+      "learning_rate": 1.9995971693183185e-05,
+      "loss": 0.8957,
+      "step": 1646
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4711489343545425,
+      "learning_rate": 1.999595000797399e-05,
+      "loss": 0.8448,
+      "step": 1647
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.742513045272388,
+      "learning_rate": 1.999592826456521e-05,
+      "loss": 0.9304,
+      "step": 1648
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9775178042199313,
+      "learning_rate": 1.9995906462956973e-05,
+      "loss": 0.8515,
+      "step": 1649
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6043154374951496,
+      "learning_rate": 1.9995884603149403e-05,
+      "loss": 0.9198,
+      "step": 1650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.825452904096607,
+      "learning_rate": 1.999586268514263e-05,
+      "loss": 0.6838,
+      "step": 1651
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8413857979409033,
+      "learning_rate": 1.9995840708936773e-05,
+      "loss": 0.7388,
+      "step": 1652
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5362060888086793,
+      "learning_rate": 1.9995818674531972e-05,
+      "loss": 0.9358,
+      "step": 1653
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.955241049056127,
+      "learning_rate": 1.999579658192835e-05,
+      "loss": 0.9487,
+      "step": 1654
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6096944459916085,
+      "learning_rate": 1.9995774431126036e-05,
+      "loss": 0.9005,
+      "step": 1655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2450022679963255,
+      "learning_rate": 1.999575222212516e-05,
+      "loss": 0.9114,
+      "step": 1656
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.698620937820629,
+      "learning_rate": 1.9995729954925844e-05,
+      "loss": 0.949,
+      "step": 1657
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2385012317810595,
+      "learning_rate": 1.9995707629528227e-05,
+      "loss": 0.9814,
+      "step": 1658
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.527438116024335,
+      "learning_rate": 1.9995685245932436e-05,
+      "loss": 0.8972,
+      "step": 1659
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7945671758755413,
+      "learning_rate": 1.99956628041386e-05,
+      "loss": 0.8689,
+      "step": 1660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.307125002363222,
+      "learning_rate": 1.999564030414685e-05,
+      "loss": 0.9638,
+      "step": 1661
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7191537733294082,
+      "learning_rate": 1.999561774595732e-05,
+      "loss": 1.0026,
+      "step": 1662
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7828921481231372,
+      "learning_rate": 1.999559512957014e-05,
+      "loss": 0.8661,
+      "step": 1663
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.12464995836956,
+      "learning_rate": 1.9995572454985437e-05,
+      "loss": 0.877,
+      "step": 1664
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5186294869954802,
+      "learning_rate": 1.999554972220335e-05,
+      "loss": 1.0036,
+      "step": 1665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6408430902529991,
+      "learning_rate": 1.9995526931224004e-05,
+      "loss": 0.9138,
+      "step": 1666
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.738849955539084,
+      "learning_rate": 1.9995504082047538e-05,
+      "loss": 0.9353,
+      "step": 1667
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8678253160494009,
+      "learning_rate": 1.9995481174674084e-05,
+      "loss": 0.7811,
+      "step": 1668
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6760878253318456,
+      "learning_rate": 1.999545820910377e-05,
+      "loss": 0.9131,
+      "step": 1669
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7094213168629653,
+      "learning_rate": 1.9995435185336735e-05,
+      "loss": 0.9253,
+      "step": 1670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6440810157052996,
+      "learning_rate": 1.9995412103373115e-05,
+      "loss": 0.8838,
+      "step": 1671
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.21547895945985,
+      "learning_rate": 1.999538896321304e-05,
+      "loss": 1.0214,
+      "step": 1672
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8012290274172634,
+      "learning_rate": 1.9995365764856646e-05,
+      "loss": 0.7513,
+      "step": 1673
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6712256672504886,
+      "learning_rate": 1.9995342508304065e-05,
+      "loss": 1.0065,
+      "step": 1674
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6164916389539759,
+      "learning_rate": 1.999531919355544e-05,
+      "loss": 0.9157,
+      "step": 1675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.055178448035698,
+      "learning_rate": 1.9995295820610897e-05,
+      "loss": 0.8986,
+      "step": 1676
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2411913268887274,
+      "learning_rate": 1.9995272389470582e-05,
+      "loss": 0.9478,
+      "step": 1677
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8106475768240603,
+      "learning_rate": 1.9995248900134624e-05,
+      "loss": 0.7364,
+      "step": 1678
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4065544757439858,
+      "learning_rate": 1.9995225352603166e-05,
+      "loss": 0.7729,
+      "step": 1679
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0820284280452563,
+      "learning_rate": 1.9995201746876337e-05,
+      "loss": 0.9148,
+      "step": 1680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7035386971060666,
+      "learning_rate": 1.999517808295428e-05,
+      "loss": 0.9825,
+      "step": 1681
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8150540027785498,
+      "learning_rate": 1.9995154360837135e-05,
+      "loss": 0.9815,
+      "step": 1682
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8485339809961687,
+      "learning_rate": 1.999513058052503e-05,
+      "loss": 0.9405,
+      "step": 1683
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.094305363876127,
+      "learning_rate": 1.999510674201812e-05,
+      "loss": 0.9554,
+      "step": 1684
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1132428665738603,
+      "learning_rate": 1.9995082845316528e-05,
+      "loss": 0.9769,
+      "step": 1685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6960782751634202,
+      "learning_rate": 1.99950588904204e-05,
+      "loss": 0.9203,
+      "step": 1686
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8756003324624029,
+      "learning_rate": 1.9995034877329876e-05,
+      "loss": 0.9651,
+      "step": 1687
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8695691386168074,
+      "learning_rate": 1.9995010806045096e-05,
+      "loss": 0.9923,
+      "step": 1688
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.775734241867895,
+      "learning_rate": 1.9994986676566194e-05,
+      "loss": 0.9739,
+      "step": 1689
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.129760110795647,
+      "learning_rate": 1.9994962488893316e-05,
+      "loss": 0.7703,
+      "step": 1690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5922800676392608,
+      "learning_rate": 1.9994938243026605e-05,
+      "loss": 0.9666,
+      "step": 1691
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8438478340071592,
+      "learning_rate": 1.99949139389662e-05,
+      "loss": 0.7371,
+      "step": 1692
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8093568907913287,
+      "learning_rate": 1.999488957671224e-05,
+      "loss": 0.8207,
+      "step": 1693
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.13018390828409,
+      "learning_rate": 1.9994865156264868e-05,
+      "loss": 0.8764,
+      "step": 1694
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3758335753393616,
+      "learning_rate": 1.9994840677624226e-05,
+      "loss": 0.8032,
+      "step": 1695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8711111507267413,
+      "learning_rate": 1.999481614079046e-05,
+      "loss": 0.9381,
+      "step": 1696
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9131228225919197,
+      "learning_rate": 1.9994791545763707e-05,
+      "loss": 1.0397,
+      "step": 1697
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7712869104982638,
+      "learning_rate": 1.9994766892544113e-05,
+      "loss": 0.9804,
+      "step": 1698
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6393032597589015,
+      "learning_rate": 1.9994742181131824e-05,
+      "loss": 0.8334,
+      "step": 1699
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8985781844992609,
+      "learning_rate": 1.999471741152698e-05,
+      "loss": 0.8985,
+      "step": 1700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6800547934091234,
+      "learning_rate": 1.9994692583729728e-05,
+      "loss": 0.9517,
+      "step": 1701
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6978283538189844,
+      "learning_rate": 1.9994667697740212e-05,
+      "loss": 0.9504,
+      "step": 1702
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8516537619560274,
+      "learning_rate": 1.9994642753558577e-05,
+      "loss": 0.9408,
+      "step": 1703
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0899262900797546,
+      "learning_rate": 1.999461775118497e-05,
+      "loss": 0.9479,
+      "step": 1704
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6872048789962146,
+      "learning_rate": 1.9994592690619526e-05,
+      "loss": 0.9335,
+      "step": 1705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.148396875541959,
+      "learning_rate": 1.9994567571862405e-05,
+      "loss": 1.0208,
+      "step": 1706
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0666840242982647,
+      "learning_rate": 1.9994542394913748e-05,
+      "loss": 0.9991,
+      "step": 1707
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0602601922241695,
+      "learning_rate": 1.99945171597737e-05,
+      "loss": 0.9925,
+      "step": 1708
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5993859960911425,
+      "learning_rate": 1.9994491866442408e-05,
+      "loss": 1.0046,
+      "step": 1709
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7927007816241685,
+      "learning_rate": 1.9994466514920023e-05,
+      "loss": 0.7282,
+      "step": 1710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7212740277352043,
+      "learning_rate": 1.9994441105206684e-05,
+      "loss": 0.911,
+      "step": 1711
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9421573789368154,
+      "learning_rate": 1.9994415637302545e-05,
+      "loss": 0.8953,
+      "step": 1712
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.260311556018887,
+      "learning_rate": 1.999439011120776e-05,
+      "loss": 0.9709,
+      "step": 1713
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.747508346495273,
+      "learning_rate": 1.9994364526922467e-05,
+      "loss": 1.0059,
+      "step": 1714
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8328189070091943,
+      "learning_rate": 1.9994338884446823e-05,
+      "loss": 0.9282,
+      "step": 1715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7943574158745319,
+      "learning_rate": 1.999431318378097e-05,
+      "loss": 0.7258,
+      "step": 1716
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9972491172056661,
+      "learning_rate": 1.9994287424925063e-05,
+      "loss": 0.7942,
+      "step": 1717
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6464202983181129,
+      "learning_rate": 1.9994261607879252e-05,
+      "loss": 0.9438,
+      "step": 1718
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0020573996654334,
+      "learning_rate": 1.9994235732643687e-05,
+      "loss": 0.9386,
+      "step": 1719
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0950613410119656,
+      "learning_rate": 1.9994209799218514e-05,
+      "loss": 0.9186,
+      "step": 1720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8485742721763716,
+      "learning_rate": 1.9994183807603885e-05,
+      "loss": 0.964,
+      "step": 1721
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.782088286934138,
+      "learning_rate": 1.999415775779996e-05,
+      "loss": 0.9029,
+      "step": 1722
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8191624257781012,
+      "learning_rate": 1.999413164980688e-05,
+      "loss": 0.781,
+      "step": 1723
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.689964373428912,
+      "learning_rate": 1.9994105483624807e-05,
+      "loss": 0.8923,
+      "step": 1724
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9077981457611564,
+      "learning_rate": 1.9994079259253883e-05,
+      "loss": 0.8948,
+      "step": 1725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7890378290867603,
+      "learning_rate": 1.9994052976694265e-05,
+      "loss": 0.8009,
+      "step": 1726
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9238600151267815,
+      "learning_rate": 1.999402663594611e-05,
+      "loss": 0.9197,
+      "step": 1727
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5158399926270985,
+      "learning_rate": 1.9994000237009565e-05,
+      "loss": 0.9304,
+      "step": 1728
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.971533134188685,
+      "learning_rate": 1.9993973779884785e-05,
+      "loss": 1.0509,
+      "step": 1729
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6741646491105175,
+      "learning_rate": 1.999394726457193e-05,
+      "loss": 0.9075,
+      "step": 1730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.798810063051662,
+      "learning_rate": 1.9993920691071146e-05,
+      "loss": 0.8858,
+      "step": 1731
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.77663619425195,
+      "learning_rate": 1.9993894059382595e-05,
+      "loss": 0.832,
+      "step": 1732
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7366741783482906,
+      "learning_rate": 1.999386736950643e-05,
+      "loss": 0.843,
+      "step": 1733
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6621498394423972,
+      "learning_rate": 1.9993840621442802e-05,
+      "loss": 0.9733,
+      "step": 1734
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7744420239396255,
+      "learning_rate": 1.9993813815191868e-05,
+      "loss": 0.6979,
+      "step": 1735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6148400186382317,
+      "learning_rate": 1.9993786950753792e-05,
+      "loss": 0.8994,
+      "step": 1736
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8716826797956978,
+      "learning_rate": 1.999376002812872e-05,
+      "loss": 0.8735,
+      "step": 1737
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6371496623830557,
+      "learning_rate": 1.9993733047316814e-05,
+      "loss": 1.0359,
+      "step": 1738
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7666608914079274,
+      "learning_rate": 1.999370600831823e-05,
+      "loss": 0.7491,
+      "step": 1739
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0842686565823856,
+      "learning_rate": 1.9993678911133126e-05,
+      "loss": 0.9329,
+      "step": 1740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.889340212376937,
+      "learning_rate": 1.9993651755761658e-05,
+      "loss": 1.0099,
+      "step": 1741
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.469501517795714,
+      "learning_rate": 1.9993624542203987e-05,
+      "loss": 0.9856,
+      "step": 1742
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6038134007966056,
+      "learning_rate": 1.9993597270460273e-05,
+      "loss": 0.9726,
+      "step": 1743
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.602820894407549,
+      "learning_rate": 1.999356994053067e-05,
+      "loss": 1.0725,
+      "step": 1744
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1083564918117017,
+      "learning_rate": 1.9993542552415334e-05,
+      "loss": 0.9463,
+      "step": 1745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5344776594025435,
+      "learning_rate": 1.9993515106114433e-05,
+      "loss": 0.9902,
+      "step": 1746
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.211808462916918,
+      "learning_rate": 1.9993487601628124e-05,
+      "loss": 0.8054,
+      "step": 1747
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5726324824046345,
+      "learning_rate": 1.9993460038956565e-05,
+      "loss": 0.9494,
+      "step": 1748
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9295214037850947,
+      "learning_rate": 1.999343241809992e-05,
+      "loss": 0.8674,
+      "step": 1749
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.24840314143114,
+      "learning_rate": 1.9993404739058346e-05,
+      "loss": 0.9171,
+      "step": 1750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7641385990406487,
+      "learning_rate": 1.9993377001832005e-05,
+      "loss": 0.7037,
+      "step": 1751
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.897418473547005,
+      "learning_rate": 1.999334920642106e-05,
+      "loss": 0.8534,
+      "step": 1752
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6205222306688642,
+      "learning_rate": 1.999332135282567e-05,
+      "loss": 0.9717,
+      "step": 1753
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.703615490156728,
+      "learning_rate": 1.9993293441046e-05,
+      "loss": 0.8553,
+      "step": 1754
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7181728712836988,
+      "learning_rate": 1.9993265471082213e-05,
+      "loss": 1.0037,
+      "step": 1755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.830518040249583,
+      "learning_rate": 1.999323744293447e-05,
+      "loss": 0.9974,
+      "step": 1756
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.65295750366764,
+      "learning_rate": 1.999320935660294e-05,
+      "loss": 1.0149,
+      "step": 1757
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7665337791698839,
+      "learning_rate": 1.9993181212087775e-05,
+      "loss": 0.7496,
+      "step": 1758
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.106375811407699,
+      "learning_rate": 1.9993153009389145e-05,
+      "loss": 0.8636,
+      "step": 1759
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7721986134811927,
+      "learning_rate": 1.999312474850722e-05,
+      "loss": 0.951,
+      "step": 1760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9137605056499012,
+      "learning_rate": 1.9993096429442153e-05,
+      "loss": 0.9757,
+      "step": 1761
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5965392482054357,
+      "learning_rate": 1.9993068052194115e-05,
+      "loss": 0.7708,
+      "step": 1762
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7166330527515292,
+      "learning_rate": 1.9993039616763276e-05,
+      "loss": 0.8545,
+      "step": 1763
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7609725208715719,
+      "learning_rate": 1.999301112314979e-05,
+      "loss": 0.6911,
+      "step": 1764
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.024244946017813,
+      "learning_rate": 1.9992982571353837e-05,
+      "loss": 0.749,
+      "step": 1765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8190598873732227,
+      "learning_rate": 1.9992953961375573e-05,
+      "loss": 0.7592,
+      "step": 1766
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.829140152730974,
+      "learning_rate": 1.9992925293215164e-05,
+      "loss": 0.7805,
+      "step": 1767
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9539485330655788,
+      "learning_rate": 1.999289656687278e-05,
+      "loss": 0.8965,
+      "step": 1768
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7672762710645644,
+      "learning_rate": 1.9992867782348595e-05,
+      "loss": 0.6865,
+      "step": 1769
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.686543087418232,
+      "learning_rate": 1.9992838939642766e-05,
+      "loss": 0.9606,
+      "step": 1770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9396558516956497,
+      "learning_rate": 1.9992810038755464e-05,
+      "loss": 0.9417,
+      "step": 1771
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.090606140371637,
+      "learning_rate": 1.999278107968686e-05,
+      "loss": 0.8301,
+      "step": 1772
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8806427966204902,
+      "learning_rate": 1.999275206243712e-05,
+      "loss": 0.7791,
+      "step": 1773
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.162852738508095,
+      "learning_rate": 1.9992722987006412e-05,
+      "loss": 0.925,
+      "step": 1774
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.789554115512833,
+      "learning_rate": 1.999269385339491e-05,
+      "loss": 0.7991,
+      "step": 1775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9189560998790565,
+      "learning_rate": 1.9992664661602782e-05,
+      "loss": 0.8752,
+      "step": 1776
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8180311690615619,
+      "learning_rate": 1.9992635411630194e-05,
+      "loss": 0.6957,
+      "step": 1777
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5809234176387517,
+      "learning_rate": 1.9992606103477318e-05,
+      "loss": 0.831,
+      "step": 1778
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.704415431038897,
+      "learning_rate": 1.999257673714433e-05,
+      "loss": 0.8747,
+      "step": 1779
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8078758053228918,
+      "learning_rate": 1.999254731263139e-05,
+      "loss": 0.9278,
+      "step": 1780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.283986706010367,
+      "learning_rate": 1.9992517829938683e-05,
+      "loss": 0.8675,
+      "step": 1781
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.950075795540608,
+      "learning_rate": 1.999248828906637e-05,
+      "loss": 0.9848,
+      "step": 1782
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0568629742212305,
+      "learning_rate": 1.9992458690014625e-05,
+      "loss": 0.8601,
+      "step": 1783
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7966482911407898,
+      "learning_rate": 1.9992429032783625e-05,
+      "loss": 0.7446,
+      "step": 1784
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.841410607995906,
+      "learning_rate": 1.9992399317373537e-05,
+      "loss": 0.9412,
+      "step": 1785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6498601293448216,
+      "learning_rate": 1.999236954378454e-05,
+      "loss": 0.9061,
+      "step": 1786
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2586394801291663,
+      "learning_rate": 1.9992339712016805e-05,
+      "loss": 0.8651,
+      "step": 1787
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3054032233993946,
+      "learning_rate": 1.9992309822070503e-05,
+      "loss": 1.0051,
+      "step": 1788
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8777115352812639,
+      "learning_rate": 1.9992279873945807e-05,
+      "loss": 1.0819,
+      "step": 1789
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8366166231559695,
+      "learning_rate": 1.9992249867642894e-05,
+      "loss": 0.8298,
+      "step": 1790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.652003451909213,
+      "learning_rate": 1.999221980316194e-05,
+      "loss": 1.0027,
+      "step": 1791
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.821092305217102,
+      "learning_rate": 1.9992189680503124e-05,
+      "loss": 0.9163,
+      "step": 1792
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6037405286980435,
+      "learning_rate": 1.999215949966661e-05,
+      "loss": 0.9121,
+      "step": 1793
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5753450677420249,
+      "learning_rate": 1.999212926065258e-05,
+      "loss": 0.8732,
+      "step": 1794
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5973114383941813,
+      "learning_rate": 1.9992098963461213e-05,
+      "loss": 0.8658,
+      "step": 1795
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9475656603058193,
+      "learning_rate": 1.999206860809268e-05,
+      "loss": 0.9184,
+      "step": 1796
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.305892858131083,
+      "learning_rate": 1.9992038194547163e-05,
+      "loss": 0.933,
+      "step": 1797
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5394729303413233,
+      "learning_rate": 1.9992007722824832e-05,
+      "loss": 0.8496,
+      "step": 1798
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4896937515713793,
+      "learning_rate": 1.999197719292587e-05,
+      "loss": 0.8688,
+      "step": 1799
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3681089579546963,
+      "learning_rate": 1.9991946604850456e-05,
+      "loss": 0.8828,
+      "step": 1800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.797997628335516,
+      "learning_rate": 1.9991915958598766e-05,
+      "loss": 0.8559,
+      "step": 1801
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7517103161457859,
+      "learning_rate": 1.9991885254170972e-05,
+      "loss": 1.0107,
+      "step": 1802
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8280227750169045,
+      "learning_rate": 1.9991854491567265e-05,
+      "loss": 1.0545,
+      "step": 1803
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.672391350538439,
+      "learning_rate": 1.9991823670787817e-05,
+      "loss": 1.0083,
+      "step": 1804
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8142182835937212,
+      "learning_rate": 1.9991792791832805e-05,
+      "loss": 0.7634,
+      "step": 1805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7632630712449269,
+      "learning_rate": 1.999176185470241e-05,
+      "loss": 0.8328,
+      "step": 1806
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4375751249540953,
+      "learning_rate": 1.9991730859396822e-05,
+      "loss": 0.9277,
+      "step": 1807
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8196041983689156,
+      "learning_rate": 1.9991699805916207e-05,
+      "loss": 0.7215,
+      "step": 1808
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.064392054951163,
+      "learning_rate": 1.9991668694260756e-05,
+      "loss": 1.0169,
+      "step": 1809
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.919527210728221,
+      "learning_rate": 1.9991637524430642e-05,
+      "loss": 0.8069,
+      "step": 1810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7583960785983828,
+      "learning_rate": 1.9991606296426054e-05,
+      "loss": 0.7319,
+      "step": 1811
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9690299664967081,
+      "learning_rate": 1.9991575010247173e-05,
+      "loss": 0.9014,
+      "step": 1812
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7430279602484067,
+      "learning_rate": 1.9991543665894177e-05,
+      "loss": 0.9386,
+      "step": 1813
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8106303505613058,
+      "learning_rate": 1.999151226336725e-05,
+      "loss": 0.7656,
+      "step": 1814
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8336093139695361,
+      "learning_rate": 1.9991480802666573e-05,
+      "loss": 0.6698,
+      "step": 1815
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.205870769847105,
+      "learning_rate": 1.9991449283792338e-05,
+      "loss": 0.9808,
+      "step": 1816
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4281069434803997,
+      "learning_rate": 1.9991417706744718e-05,
+      "loss": 0.9106,
+      "step": 1817
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7755723638825314,
+      "learning_rate": 1.99913860715239e-05,
+      "loss": 0.9127,
+      "step": 1818
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8177521264628254,
+      "learning_rate": 1.9991354378130074e-05,
+      "loss": 0.8864,
+      "step": 1819
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0953775168878837,
+      "learning_rate": 1.9991322626563413e-05,
+      "loss": 0.9021,
+      "step": 1820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6694401865964548,
+      "learning_rate": 1.9991290816824112e-05,
+      "loss": 1.0071,
+      "step": 1821
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8654386982052629,
+      "learning_rate": 1.999125894891235e-05,
+      "loss": 0.9898,
+      "step": 1822
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8755769778782176,
+      "learning_rate": 1.999122702282832e-05,
+      "loss": 1.1222,
+      "step": 1823
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4819322902832395,
+      "learning_rate": 1.99911950385722e-05,
+      "loss": 0.8935,
+      "step": 1824
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6098322714586295,
+      "learning_rate": 1.999116299614418e-05,
+      "loss": 0.8853,
+      "step": 1825
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.68989430087454,
+      "learning_rate": 1.9991130895544443e-05,
+      "loss": 0.9101,
+      "step": 1826
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6092718573672975,
+      "learning_rate": 1.999109873677318e-05,
+      "loss": 0.8765,
+      "step": 1827
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8204622533399912,
+      "learning_rate": 1.9991066519830584e-05,
+      "loss": 0.8617,
+      "step": 1828
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4596672494692404,
+      "learning_rate": 1.9991034244716826e-05,
+      "loss": 0.8433,
+      "step": 1829
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8202991120134333,
+      "learning_rate": 1.9991001911432108e-05,
+      "loss": 0.7594,
+      "step": 1830
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6470162709869702,
+      "learning_rate": 1.9990969519976614e-05,
+      "loss": 0.851,
+      "step": 1831
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.811587341500778,
+      "learning_rate": 1.999093707035053e-05,
+      "loss": 0.8954,
+      "step": 1832
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7395020873253755,
+      "learning_rate": 1.9990904562554048e-05,
+      "loss": 0.9664,
+      "step": 1833
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5401832480224937,
+      "learning_rate": 1.999087199658736e-05,
+      "loss": 0.9641,
+      "step": 1834
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9321155473071585,
+      "learning_rate": 1.999083937245065e-05,
+      "loss": 0.9595,
+      "step": 1835
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7668976938683696,
+      "learning_rate": 1.999080669014411e-05,
+      "loss": 0.8254,
+      "step": 1836
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3282099020181266,
+      "learning_rate": 1.999077394966793e-05,
+      "loss": 0.9696,
+      "step": 1837
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0442032896703695,
+      "learning_rate": 1.9990741151022302e-05,
+      "loss": 1.046,
+      "step": 1838
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5873729943239343,
+      "learning_rate": 1.9990708294207414e-05,
+      "loss": 0.7851,
+      "step": 1839
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5412388742489698,
+      "learning_rate": 1.999067537922346e-05,
+      "loss": 0.92,
+      "step": 1840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2466940060124196,
+      "learning_rate": 1.9990642406070634e-05,
+      "loss": 0.8015,
+      "step": 1841
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7309841102942163,
+      "learning_rate": 1.999060937474912e-05,
+      "loss": 0.9272,
+      "step": 1842
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.635567140411678,
+      "learning_rate": 1.999057628525912e-05,
+      "loss": 0.9583,
+      "step": 1843
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6344450049032297,
+      "learning_rate": 1.9990543137600816e-05,
+      "loss": 0.9333,
+      "step": 1844
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7985270337312702,
+      "learning_rate": 1.9990509931774412e-05,
+      "loss": 0.9157,
+      "step": 1845
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.607341760576823,
+      "learning_rate": 1.9990476667780097e-05,
+      "loss": 0.9823,
+      "step": 1846
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.747432898172235,
+      "learning_rate": 1.999044334561806e-05,
+      "loss": 0.9616,
+      "step": 1847
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6168044941369357,
+      "learning_rate": 1.9990409965288503e-05,
+      "loss": 0.9381,
+      "step": 1848
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7280295031580672,
+      "learning_rate": 1.999037652679161e-05,
+      "loss": 0.975,
+      "step": 1849
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7704916954852141,
+      "learning_rate": 1.9990343030127588e-05,
+      "loss": 0.9336,
+      "step": 1850
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8262122602461424,
+      "learning_rate": 1.9990309475296623e-05,
+      "loss": 0.7075,
+      "step": 1851
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0963298803018757,
+      "learning_rate": 1.9990275862298916e-05,
+      "loss": 0.86,
+      "step": 1852
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4506284356631673,
+      "learning_rate": 1.9990242191134657e-05,
+      "loss": 0.8452,
+      "step": 1853
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8367404539777796,
+      "learning_rate": 1.999020846180405e-05,
+      "loss": 0.9517,
+      "step": 1854
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8148829586972137,
+      "learning_rate": 1.999017467430728e-05,
+      "loss": 0.8857,
+      "step": 1855
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6210765253869694,
+      "learning_rate": 1.9990140828644553e-05,
+      "loss": 0.9588,
+      "step": 1856
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6736996590199502,
+      "learning_rate": 1.999010692481606e-05,
+      "loss": 0.8957,
+      "step": 1857
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6776888600087196,
+      "learning_rate": 1.999007296282201e-05,
+      "loss": 0.9013,
+      "step": 1858
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7393296133466474,
+      "learning_rate": 1.9990038942662585e-05,
+      "loss": 0.9113,
+      "step": 1859
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6553283421552096,
+      "learning_rate": 1.9990004864337994e-05,
+      "loss": 0.9443,
+      "step": 1860
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8150659801081825,
+      "learning_rate": 1.998997072784843e-05,
+      "loss": 0.8284,
+      "step": 1861
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4098214367889246,
+      "learning_rate": 1.9989936533194093e-05,
+      "loss": 0.8395,
+      "step": 1862
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.812022688059446,
+      "learning_rate": 1.9989902280375185e-05,
+      "loss": 0.9748,
+      "step": 1863
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7217672728100966,
+      "learning_rate": 1.9989867969391904e-05,
+      "loss": 1.0103,
+      "step": 1864
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.794540327248707,
+      "learning_rate": 1.9989833600244446e-05,
+      "loss": 0.8584,
+      "step": 1865
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5301551884800138,
+      "learning_rate": 1.9989799172933018e-05,
+      "loss": 0.973,
+      "step": 1866
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0513925081913316,
+      "learning_rate": 1.9989764687457813e-05,
+      "loss": 1.0756,
+      "step": 1867
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5659089081672786,
+      "learning_rate": 1.9989730143819033e-05,
+      "loss": 0.979,
+      "step": 1868
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2984777801316962,
+      "learning_rate": 1.9989695542016887e-05,
+      "loss": 0.9927,
+      "step": 1869
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8527454673454222,
+      "learning_rate": 1.998966088205157e-05,
+      "loss": 0.7743,
+      "step": 1870
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.582176722647412,
+      "learning_rate": 1.998962616392328e-05,
+      "loss": 0.9075,
+      "step": 1871
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6253285693022521,
+      "learning_rate": 1.998959138763223e-05,
+      "loss": 0.944,
+      "step": 1872
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5675717778210156,
+      "learning_rate": 1.998955655317861e-05,
+      "loss": 1.0528,
+      "step": 1873
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8219602748954724,
+      "learning_rate": 1.998952166056263e-05,
+      "loss": 0.9335,
+      "step": 1874
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6293241419490616,
+      "learning_rate": 1.99894867097845e-05,
+      "loss": 1.0232,
+      "step": 1875
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.687934287702198,
+      "learning_rate": 1.998945170084441e-05,
+      "loss": 0.8905,
+      "step": 1876
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7793214146551521,
+      "learning_rate": 1.998941663374257e-05,
+      "loss": 0.9897,
+      "step": 1877
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.861786439425863,
+      "learning_rate": 1.998938150847918e-05,
+      "loss": 0.9221,
+      "step": 1878
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8788894339283927,
+      "learning_rate": 1.9989346325054454e-05,
+      "loss": 0.8137,
+      "step": 1879
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7931256244779462,
+      "learning_rate": 1.998931108346859e-05,
+      "loss": 0.7056,
+      "step": 1880
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3083347549041218,
+      "learning_rate": 1.998927578372179e-05,
+      "loss": 1.0138,
+      "step": 1881
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.879330807121127,
+      "learning_rate": 1.998924042581427e-05,
+      "loss": 0.9595,
+      "step": 1882
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7025169600038021,
+      "learning_rate": 1.9989205009746227e-05,
+      "loss": 0.9575,
+      "step": 1883
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8241630136235767,
+      "learning_rate": 1.998916953551787e-05,
+      "loss": 0.7421,
+      "step": 1884
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7242731519977696,
+      "learning_rate": 1.9989134003129405e-05,
+      "loss": 0.8312,
+      "step": 1885
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5621970761665378,
+      "learning_rate": 1.998909841258104e-05,
+      "loss": 1.0054,
+      "step": 1886
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.945339436923858,
+      "learning_rate": 1.9989062763872978e-05,
+      "loss": 0.9665,
+      "step": 1887
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.269387039014987,
+      "learning_rate": 1.9989027057005438e-05,
+      "loss": 0.7436,
+      "step": 1888
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8003209345861436,
+      "learning_rate": 1.9988991291978613e-05,
+      "loss": 0.7478,
+      "step": 1889
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8984402425168778,
+      "learning_rate": 1.998895546879272e-05,
+      "loss": 0.8878,
+      "step": 1890
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6821021088955739,
+      "learning_rate": 1.9988919587447967e-05,
+      "loss": 0.8538,
+      "step": 1891
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6861426254368805,
+      "learning_rate": 1.9988883647944563e-05,
+      "loss": 0.9961,
+      "step": 1892
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6471591909777636,
+      "learning_rate": 1.998884765028271e-05,
+      "loss": 0.9159,
+      "step": 1893
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7782287378014427,
+      "learning_rate": 1.9988811594462628e-05,
+      "loss": 1.0014,
+      "step": 1894
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.74197807802034,
+      "learning_rate": 1.9988775480484524e-05,
+      "loss": 0.8871,
+      "step": 1895
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.923160556818599,
+      "learning_rate": 1.9988739308348603e-05,
+      "loss": 1.0206,
+      "step": 1896
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0767965288532446,
+      "learning_rate": 1.998870307805508e-05,
+      "loss": 0.9214,
+      "step": 1897
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6567193588718747,
+      "learning_rate": 1.998866678960417e-05,
+      "loss": 1.0185,
+      "step": 1898
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7792409441744743,
+      "learning_rate": 1.9988630442996072e-05,
+      "loss": 0.903,
+      "step": 1899
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.029656777664901,
+      "learning_rate": 1.9988594038231012e-05,
+      "loss": 1.0211,
+      "step": 1900
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.256503401952045,
+      "learning_rate": 1.9988557575309192e-05,
+      "loss": 0.8699,
+      "step": 1901
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7043733507299654,
+      "learning_rate": 1.9988521054230828e-05,
+      "loss": 0.9277,
+      "step": 1902
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8490598463290335,
+      "learning_rate": 1.9988484474996134e-05,
+      "loss": 0.7693,
+      "step": 1903
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7435098124029085,
+      "learning_rate": 1.9988447837605317e-05,
+      "loss": 1.079,
+      "step": 1904
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.612934169671713,
+      "learning_rate": 1.9988411142058596e-05,
+      "loss": 0.8201,
+      "step": 1905
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8524876403534034,
+      "learning_rate": 1.9988374388356186e-05,
+      "loss": 0.8806,
+      "step": 1906
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6403605594154467,
+      "learning_rate": 1.9988337576498296e-05,
+      "loss": 0.8649,
+      "step": 1907
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5708108701513859,
+      "learning_rate": 1.998830070648514e-05,
+      "loss": 0.8593,
+      "step": 1908
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5995676067363833,
+      "learning_rate": 1.9988263778316937e-05,
+      "loss": 0.9752,
+      "step": 1909
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2352445537041365,
+      "learning_rate": 1.99882267919939e-05,
+      "loss": 0.7852,
+      "step": 1910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7697836398754554,
+      "learning_rate": 1.9988189747516245e-05,
+      "loss": 0.9693,
+      "step": 1911
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6028492090843869,
+      "learning_rate": 1.9988152644884186e-05,
+      "loss": 1.0682,
+      "step": 1912
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.835553788846079,
+      "learning_rate": 1.998811548409794e-05,
+      "loss": 0.9228,
+      "step": 1913
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.181450685970848,
+      "learning_rate": 1.9988078265157727e-05,
+      "loss": 0.8483,
+      "step": 1914
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7264648285877164,
+      "learning_rate": 1.9988040988063755e-05,
+      "loss": 0.9362,
+      "step": 1915
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7200155496165415,
+      "learning_rate": 1.998800365281625e-05,
+      "loss": 0.9505,
+      "step": 1916
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.909902376352857,
+      "learning_rate": 1.9987966259415423e-05,
+      "loss": 0.9659,
+      "step": 1917
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8053370598570532,
+      "learning_rate": 1.9987928807861494e-05,
+      "loss": 0.7394,
+      "step": 1918
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.864503941360688,
+      "learning_rate": 1.9987891298154684e-05,
+      "loss": 0.9378,
+      "step": 1919
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.706248273597259,
+      "learning_rate": 1.998785373029521e-05,
+      "loss": 0.9496,
+      "step": 1920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2967422566709508,
+      "learning_rate": 1.9987816104283284e-05,
+      "loss": 0.8838,
+      "step": 1921
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7904792296881444,
+      "learning_rate": 1.998777842011913e-05,
+      "loss": 0.8856,
+      "step": 1922
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6218463121979974,
+      "learning_rate": 1.9987740677802976e-05,
+      "loss": 0.9846,
+      "step": 1923
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2550286696930626,
+      "learning_rate": 1.998770287733503e-05,
+      "loss": 1.0748,
+      "step": 1924
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9652575540350943,
+      "learning_rate": 1.9987665018715514e-05,
+      "loss": 0.8778,
+      "step": 1925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8298317355567748,
+      "learning_rate": 1.9987627101944652e-05,
+      "loss": 0.8805,
+      "step": 1926
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5251379065933595,
+      "learning_rate": 1.998758912702266e-05,
+      "loss": 0.8356,
+      "step": 1927
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8643693269573027,
+      "learning_rate": 1.9987551093949765e-05,
+      "loss": 0.9598,
+      "step": 1928
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7977750527903023,
+      "learning_rate": 1.9987513002726186e-05,
+      "loss": 0.8322,
+      "step": 1929
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.727404170136656,
+      "learning_rate": 1.9987474853352142e-05,
+      "loss": 0.9057,
+      "step": 1930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.000272583587569,
+      "learning_rate": 1.998743664582786e-05,
+      "loss": 0.9758,
+      "step": 1931
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6811337688630605,
+      "learning_rate": 1.9987398380153556e-05,
+      "loss": 0.852,
+      "step": 1932
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9434607863773299,
+      "learning_rate": 1.9987360056329458e-05,
+      "loss": 1.042,
+      "step": 1933
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.86374524463483,
+      "learning_rate": 1.9987321674355792e-05,
+      "loss": 0.8919,
+      "step": 1934
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6524153693115256,
+      "learning_rate": 1.9987283234232777e-05,
+      "loss": 0.9287,
+      "step": 1935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7606899376832277,
+      "learning_rate": 1.998724473596063e-05,
+      "loss": 0.9441,
+      "step": 1936
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5205690109833685,
+      "learning_rate": 1.9987206179539587e-05,
+      "loss": 0.9842,
+      "step": 1937
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.181953038132647,
+      "learning_rate": 1.9987167564969868e-05,
+      "loss": 0.9766,
+      "step": 1938
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.874130244346317,
+      "learning_rate": 1.9987128892251696e-05,
+      "loss": 0.7878,
+      "step": 1939
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.755881325731854,
+      "learning_rate": 1.99870901613853e-05,
+      "loss": 0.9878,
+      "step": 1940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8793748717420788,
+      "learning_rate": 1.99870513723709e-05,
+      "loss": 0.9003,
+      "step": 1941
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7553035528885379,
+      "learning_rate": 1.9987012525208728e-05,
+      "loss": 0.7242,
+      "step": 1942
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.884690318802054,
+      "learning_rate": 1.9986973619899006e-05,
+      "loss": 1.0321,
+      "step": 1943
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6124617023884833,
+      "learning_rate": 1.9986934656441962e-05,
+      "loss": 0.9336,
+      "step": 1944
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6656774510302077,
+      "learning_rate": 1.998689563483782e-05,
+      "loss": 0.9572,
+      "step": 1945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6383414106133687,
+      "learning_rate": 1.9986856555086812e-05,
+      "loss": 0.978,
+      "step": 1946
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5667912598371163,
+      "learning_rate": 1.9986817417189168e-05,
+      "loss": 0.9213,
+      "step": 1947
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6603229387832872,
+      "learning_rate": 1.9986778221145103e-05,
+      "loss": 0.9676,
+      "step": 1948
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6629259186827692,
+      "learning_rate": 1.9986738966954856e-05,
+      "loss": 0.8805,
+      "step": 1949
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.62393808210454,
+      "learning_rate": 1.9986699654618656e-05,
+      "loss": 0.8496,
+      "step": 1950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6782793003263032,
+      "learning_rate": 1.9986660284136728e-05,
+      "loss": 0.8526,
+      "step": 1951
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.578628777054172,
+      "learning_rate": 1.99866208555093e-05,
+      "loss": 0.9261,
+      "step": 1952
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.816696567577169,
+      "learning_rate": 1.9986581368736607e-05,
+      "loss": 0.8089,
+      "step": 1953
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7880997625726884,
+      "learning_rate": 1.9986541823818872e-05,
+      "loss": 0.7239,
+      "step": 1954
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.09991162575295,
+      "learning_rate": 1.9986502220756333e-05,
+      "loss": 0.9336,
+      "step": 1955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5573723173603227,
+      "learning_rate": 1.998646255954921e-05,
+      "loss": 0.8208,
+      "step": 1956
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.926200928315544,
+      "learning_rate": 1.9986422840197746e-05,
+      "loss": 0.8873,
+      "step": 1957
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.340693518796728,
+      "learning_rate": 1.9986383062702165e-05,
+      "loss": 0.9799,
+      "step": 1958
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6880497005896822,
+      "learning_rate": 1.99863432270627e-05,
+      "loss": 0.9694,
+      "step": 1959
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.07182639673836,
+      "learning_rate": 1.998630333327958e-05,
+      "loss": 0.865,
+      "step": 1960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6437602735508166,
+      "learning_rate": 1.9986263381353045e-05,
+      "loss": 0.8265,
+      "step": 1961
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7699949528049819,
+      "learning_rate": 1.998622337128332e-05,
+      "loss": 0.715,
+      "step": 1962
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8711793957769223,
+      "learning_rate": 1.9986183303070644e-05,
+      "loss": 1.0017,
+      "step": 1963
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.237438405600957,
+      "learning_rate": 1.9986143176715246e-05,
+      "loss": 0.961,
+      "step": 1964
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7913252395103768,
+      "learning_rate": 1.9986102992217356e-05,
+      "loss": 0.6383,
+      "step": 1965
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.850662630308141,
+      "learning_rate": 1.998606274957722e-05,
+      "loss": 1.0042,
+      "step": 1966
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.303280523836632,
+      "learning_rate": 1.9986022448795062e-05,
+      "loss": 1.0942,
+      "step": 1967
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9850250556223588,
+      "learning_rate": 1.998598208987112e-05,
+      "loss": 0.8269,
+      "step": 1968
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9921174955398417,
+      "learning_rate": 1.998594167280563e-05,
+      "loss": 0.8166,
+      "step": 1969
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6305477643765094,
+      "learning_rate": 1.998590119759882e-05,
+      "loss": 0.9433,
+      "step": 1970
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6797283308339406,
+      "learning_rate": 1.998586066425094e-05,
+      "loss": 0.9093,
+      "step": 1971
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.759926090041935,
+      "learning_rate": 1.9985820072762213e-05,
+      "loss": 0.9027,
+      "step": 1972
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.620747886632559,
+      "learning_rate": 1.9985779423132883e-05,
+      "loss": 1.0262,
+      "step": 1973
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.015650316509584,
+      "learning_rate": 1.9985738715363185e-05,
+      "loss": 0.9118,
+      "step": 1974
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.809837419185958,
+      "learning_rate": 1.9985697949453348e-05,
+      "loss": 0.7288,
+      "step": 1975
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3215537259621586,
+      "learning_rate": 1.998565712540362e-05,
+      "loss": 0.7934,
+      "step": 1976
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.823425126191455,
+      "learning_rate": 1.9985616243214236e-05,
+      "loss": 0.9215,
+      "step": 1977
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7098131319150345,
+      "learning_rate": 1.9985575302885435e-05,
+      "loss": 0.8087,
+      "step": 1978
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7821465769208651,
+      "learning_rate": 1.9985534304417448e-05,
+      "loss": 0.748,
+      "step": 1979
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8869699156003956,
+      "learning_rate": 1.998549324781052e-05,
+      "loss": 0.9344,
+      "step": 1980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9740845020855387,
+      "learning_rate": 1.998545213306489e-05,
+      "loss": 0.8757,
+      "step": 1981
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9487697754553728,
+      "learning_rate": 1.9985410960180795e-05,
+      "loss": 0.8037,
+      "step": 1982
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7633616591396675,
+      "learning_rate": 1.9985369729158476e-05,
+      "loss": 0.9093,
+      "step": 1983
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8064765458732723,
+      "learning_rate": 1.9985328439998176e-05,
+      "loss": 0.9092,
+      "step": 1984
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6406338506449942,
+      "learning_rate": 1.998528709270013e-05,
+      "loss": 0.9724,
+      "step": 1985
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8082559398471152,
+      "learning_rate": 1.9985245687264583e-05,
+      "loss": 0.851,
+      "step": 1986
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7604536985015795,
+      "learning_rate": 1.9985204223691773e-05,
+      "loss": 0.728,
+      "step": 1987
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7718278432678227,
+      "learning_rate": 1.9985162701981942e-05,
+      "loss": 0.7173,
+      "step": 1988
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7073247125387665,
+      "learning_rate": 1.9985121122135334e-05,
+      "loss": 0.9492,
+      "step": 1989
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6933892639218937,
+      "learning_rate": 1.998507948415219e-05,
+      "loss": 1.1438,
+      "step": 1990
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.907500382726766,
+      "learning_rate": 1.9985037788032753e-05,
+      "loss": 0.8174,
+      "step": 1991
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9598457025196696,
+      "learning_rate": 1.9984996033777263e-05,
+      "loss": 1.0552,
+      "step": 1992
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.741071102096099,
+      "learning_rate": 1.9984954221385962e-05,
+      "loss": 1.0589,
+      "step": 1993
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6010585704092857,
+      "learning_rate": 1.99849123508591e-05,
+      "loss": 0.9665,
+      "step": 1994
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8237553941409843,
+      "learning_rate": 1.998487042219692e-05,
+      "loss": 0.8309,
+      "step": 1995
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.726727153452555,
+      "learning_rate": 1.9984828435399657e-05,
+      "loss": 0.9263,
+      "step": 1996
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2324841200241266,
+      "learning_rate": 1.9984786390467566e-05,
+      "loss": 0.8104,
+      "step": 1997
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.896090482930794,
+      "learning_rate": 1.9984744287400883e-05,
+      "loss": 1.0385,
+      "step": 1998
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.732262179824477,
+      "learning_rate": 1.9984702126199863e-05,
+      "loss": 0.9131,
+      "step": 1999
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6605686967073674,
+      "learning_rate": 1.9984659906864742e-05,
+      "loss": 1.0415,
+      "step": 2000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9665968876534354,
+      "learning_rate": 1.998461762939577e-05,
+      "loss": 0.9321,
+      "step": 2001
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.835475730415937,
+      "learning_rate": 1.9984575293793198e-05,
+      "loss": 0.9154,
+      "step": 2002
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.394515708378759,
+      "learning_rate": 1.9984532900057263e-05,
+      "loss": 0.9508,
+      "step": 2003
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2939883162017973,
+      "learning_rate": 1.998449044818822e-05,
+      "loss": 0.9362,
+      "step": 2004
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1034246507380274,
+      "learning_rate": 1.998444793818631e-05,
+      "loss": 0.9182,
+      "step": 2005
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8358200940186823,
+      "learning_rate": 1.9984405370051783e-05,
+      "loss": 0.7531,
+      "step": 2006
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0637601627008055,
+      "learning_rate": 1.9984362743784887e-05,
+      "loss": 0.8275,
+      "step": 2007
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0910651148791137,
+      "learning_rate": 1.998432005938587e-05,
+      "loss": 0.9413,
+      "step": 2008
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7957866444882575,
+      "learning_rate": 1.998427731685498e-05,
+      "loss": 0.933,
+      "step": 2009
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8537501838768786,
+      "learning_rate": 1.998423451619247e-05,
+      "loss": 0.9231,
+      "step": 2010
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.048246396667472,
+      "learning_rate": 1.998419165739858e-05,
+      "loss": 0.8453,
+      "step": 2011
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.703861454779081,
+      "learning_rate": 1.998414874047357e-05,
+      "loss": 0.9614,
+      "step": 2012
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4992182955812943,
+      "learning_rate": 1.998410576541768e-05,
+      "loss": 0.91,
+      "step": 2013
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5646896857632946,
+      "learning_rate": 1.9984062732231174e-05,
+      "loss": 1.0068,
+      "step": 2014
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7168798332187063,
+      "learning_rate": 1.9984019640914285e-05,
+      "loss": 0.8245,
+      "step": 2015
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.021692564584993,
+      "learning_rate": 1.9983976491467278e-05,
+      "loss": 0.9114,
+      "step": 2016
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6880711509483342,
+      "learning_rate": 1.9983933283890398e-05,
+      "loss": 0.9256,
+      "step": 2017
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6899056945429853,
+      "learning_rate": 1.9983890018183896e-05,
+      "loss": 0.87,
+      "step": 2018
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6201620851454217,
+      "learning_rate": 1.9983846694348027e-05,
+      "loss": 1.0017,
+      "step": 2019
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6414065220692227,
+      "learning_rate": 1.9983803312383042e-05,
+      "loss": 1.0418,
+      "step": 2020
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7953680787217744,
+      "learning_rate": 1.998375987228919e-05,
+      "loss": 0.7291,
+      "step": 2021
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5573743028421894,
+      "learning_rate": 1.998371637406673e-05,
+      "loss": 0.9117,
+      "step": 2022
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.203145914895803,
+      "learning_rate": 1.9983672817715914e-05,
+      "loss": 0.9314,
+      "step": 2023
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.622925322501368,
+      "learning_rate": 1.9983629203236993e-05,
+      "loss": 1.0255,
+      "step": 2024
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.780068163191732,
+      "learning_rate": 1.998358553063022e-05,
+      "loss": 0.8553,
+      "step": 2025
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8649085274525883,
+      "learning_rate": 1.998354179989585e-05,
+      "loss": 0.9113,
+      "step": 2026
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.754237617966611,
+      "learning_rate": 1.9983498011034143e-05,
+      "loss": 0.9525,
+      "step": 2027
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7893128460469959,
+      "learning_rate": 1.9983454164045352e-05,
+      "loss": 0.7424,
+      "step": 2028
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9662876244370122,
+      "learning_rate": 1.998341025892972e-05,
+      "loss": 0.8286,
+      "step": 2029
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6339094944120278,
+      "learning_rate": 1.9983366295687522e-05,
+      "loss": 0.8927,
+      "step": 2030
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6118152186708024,
+      "learning_rate": 1.9983322274319005e-05,
+      "loss": 0.9186,
+      "step": 2031
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9448423058696902,
+      "learning_rate": 1.998327819482442e-05,
+      "loss": 0.9517,
+      "step": 2032
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.682911555785016,
+      "learning_rate": 1.998323405720403e-05,
+      "loss": 0.8894,
+      "step": 2033
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.792880017362886,
+      "learning_rate": 1.9983189861458093e-05,
+      "loss": 1.0339,
+      "step": 2034
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8094227604457139,
+      "learning_rate": 1.9983145607586862e-05,
+      "loss": 0.7675,
+      "step": 2035
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6312761461286505,
+      "learning_rate": 1.9983101295590597e-05,
+      "loss": 0.9559,
+      "step": 2036
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8420861139458418,
+      "learning_rate": 1.9983056925469558e-05,
+      "loss": 0.9722,
+      "step": 2037
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9967561666213411,
+      "learning_rate": 1.9983012497223994e-05,
+      "loss": 1.0604,
+      "step": 2038
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1431930704165905,
+      "learning_rate": 1.998296801085418e-05,
+      "loss": 0.9188,
+      "step": 2039
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.735422991125576,
+      "learning_rate": 1.9982923466360357e-05,
+      "loss": 0.9824,
+      "step": 2040
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.621811433429304,
+      "learning_rate": 1.9982878863742798e-05,
+      "loss": 0.8495,
+      "step": 2041
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6943687009632478,
+      "learning_rate": 1.9982834203001757e-05,
+      "loss": 0.8448,
+      "step": 2042
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8320126829637504,
+      "learning_rate": 1.9982789484137495e-05,
+      "loss": 0.9534,
+      "step": 2043
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0456828527566495,
+      "learning_rate": 1.9982744707150272e-05,
+      "loss": 0.8581,
+      "step": 2044
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7578806938910918,
+      "learning_rate": 1.9982699872040346e-05,
+      "loss": 0.9599,
+      "step": 2045
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8571639323678432,
+      "learning_rate": 1.9982654978807986e-05,
+      "loss": 0.738,
+      "step": 2046
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7353756105769786,
+      "learning_rate": 1.9982610027453447e-05,
+      "loss": 0.8338,
+      "step": 2047
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7049831313941581,
+      "learning_rate": 1.9982565017976992e-05,
+      "loss": 0.9151,
+      "step": 2048
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8146933739258755,
+      "learning_rate": 1.998251995037888e-05,
+      "loss": 1.0013,
+      "step": 2049
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7544882538571,
+      "learning_rate": 1.9982474824659378e-05,
+      "loss": 0.837,
+      "step": 2050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3045289615098876,
+      "learning_rate": 1.9982429640818747e-05,
+      "loss": 0.9035,
+      "step": 2051
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8267857479020189,
+      "learning_rate": 1.998238439885725e-05,
+      "loss": 0.7798,
+      "step": 2052
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9593956015693412,
+      "learning_rate": 1.998233909877515e-05,
+      "loss": 0.9712,
+      "step": 2053
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.580016252343815,
+      "learning_rate": 1.9982293740572714e-05,
+      "loss": 0.8609,
+      "step": 2054
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.856168223611931,
+      "learning_rate": 1.99822483242502e-05,
+      "loss": 0.9829,
+      "step": 2055
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7910047270667924,
+      "learning_rate": 1.998220284980788e-05,
+      "loss": 1.0188,
+      "step": 2056
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.650925881463684,
+      "learning_rate": 1.9982157317246015e-05,
+      "loss": 0.8833,
+      "step": 2057
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5638771551737491,
+      "learning_rate": 1.9982111726564865e-05,
+      "loss": 0.9633,
+      "step": 2058
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.069820321342172,
+      "learning_rate": 1.9982066077764702e-05,
+      "loss": 0.9429,
+      "step": 2059
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5591730910475499,
+      "learning_rate": 1.998202037084579e-05,
+      "loss": 0.964,
+      "step": 2060
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.083676052194489,
+      "learning_rate": 1.9981974605808398e-05,
+      "loss": 0.9522,
+      "step": 2061
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7061193296450115,
+      "learning_rate": 1.9981928782652788e-05,
+      "loss": 0.9137,
+      "step": 2062
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8186081142587602,
+      "learning_rate": 1.9981882901379228e-05,
+      "loss": 0.9301,
+      "step": 2063
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1253778174853926,
+      "learning_rate": 1.9981836961987987e-05,
+      "loss": 0.9259,
+      "step": 2064
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.768908104916534,
+      "learning_rate": 1.998179096447933e-05,
+      "loss": 0.7547,
+      "step": 2065
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8147586123700017,
+      "learning_rate": 1.9981744908853524e-05,
+      "loss": 0.99,
+      "step": 2066
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7027721353229768,
+      "learning_rate": 1.998169879511084e-05,
+      "loss": 0.8194,
+      "step": 2067
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.71142456711741,
+      "learning_rate": 1.9981652623251546e-05,
+      "loss": 1.0007,
+      "step": 2068
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6632836121340346,
+      "learning_rate": 1.998160639327591e-05,
+      "loss": 0.9,
+      "step": 2069
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8677523252596793,
+      "learning_rate": 1.9981560105184205e-05,
+      "loss": 0.9507,
+      "step": 2070
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9732922154571524,
+      "learning_rate": 1.998151375897669e-05,
+      "loss": 0.9835,
+      "step": 2071
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7528586040830323,
+      "learning_rate": 1.9981467354653645e-05,
+      "loss": 0.7958,
+      "step": 2072
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6026268359992046,
+      "learning_rate": 1.998142089221534e-05,
+      "loss": 0.9827,
+      "step": 2073
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9466604770464788,
+      "learning_rate": 1.9981374371662035e-05,
+      "loss": 0.8503,
+      "step": 2074
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.794636028639653,
+      "learning_rate": 1.9981327792994016e-05,
+      "loss": 0.9756,
+      "step": 2075
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5760589017584077,
+      "learning_rate": 1.9981281156211543e-05,
+      "loss": 0.8177,
+      "step": 2076
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6559358281396,
+      "learning_rate": 1.9981234461314893e-05,
+      "loss": 0.8957,
+      "step": 2077
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.436348759719535,
+      "learning_rate": 1.9981187708304335e-05,
+      "loss": 0.9151,
+      "step": 2078
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6067420744426577,
+      "learning_rate": 1.9981140897180144e-05,
+      "loss": 0.8334,
+      "step": 2079
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.630825425554162,
+      "learning_rate": 1.9981094027942585e-05,
+      "loss": 1.0245,
+      "step": 2080
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7257095853848656,
+      "learning_rate": 1.998104710059194e-05,
+      "loss": 1.0044,
+      "step": 2081
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9343516949977855,
+      "learning_rate": 1.998100011512848e-05,
+      "loss": 0.966,
+      "step": 2082
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8058409642884543,
+      "learning_rate": 1.9980953071552477e-05,
+      "loss": 0.8332,
+      "step": 2083
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.222278373075257,
+      "learning_rate": 1.9980905969864204e-05,
+      "loss": 0.9051,
+      "step": 2084
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9287344964013038,
+      "learning_rate": 1.998085881006394e-05,
+      "loss": 0.9905,
+      "step": 2085
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.009123298526866,
+      "learning_rate": 1.998081159215195e-05,
+      "loss": 0.7847,
+      "step": 2086
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.713678210693012,
+      "learning_rate": 1.998076431612852e-05,
+      "loss": 0.9452,
+      "step": 2087
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2306376212855263,
+      "learning_rate": 1.998071698199392e-05,
+      "loss": 0.9173,
+      "step": 2088
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7875748962644302,
+      "learning_rate": 1.9980669589748426e-05,
+      "loss": 0.9227,
+      "step": 2089
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6996485381388733,
+      "learning_rate": 1.998062213939231e-05,
+      "loss": 1.0114,
+      "step": 2090
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8905677237792744,
+      "learning_rate": 1.9980574630925856e-05,
+      "loss": 0.9262,
+      "step": 2091
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6767078687396744,
+      "learning_rate": 1.9980527064349334e-05,
+      "loss": 0.997,
+      "step": 2092
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6033078177457862,
+      "learning_rate": 1.9980479439663024e-05,
+      "loss": 0.9289,
+      "step": 2093
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6632683221222553,
+      "learning_rate": 1.9980431756867206e-05,
+      "loss": 1.0246,
+      "step": 2094
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.817320012717107,
+      "learning_rate": 1.9980384015962152e-05,
+      "loss": 0.9993,
+      "step": 2095
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.015497632199055,
+      "learning_rate": 1.9980336216948144e-05,
+      "loss": 0.978,
+      "step": 2096
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7173521589357232,
+      "learning_rate": 1.9980288359825457e-05,
+      "loss": 1.0156,
+      "step": 2097
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8320938873264947,
+      "learning_rate": 1.9980240444594372e-05,
+      "loss": 0.9576,
+      "step": 2098
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7230369489393782,
+      "learning_rate": 1.998019247125517e-05,
+      "loss": 1.0096,
+      "step": 2099
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.49814627381044,
+      "learning_rate": 1.998014443980812e-05,
+      "loss": 0.8571,
+      "step": 2100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2159273372396373,
+      "learning_rate": 1.998009635025352e-05,
+      "loss": 1.0578,
+      "step": 2101
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9298423645891196,
+      "learning_rate": 1.9980048202591632e-05,
+      "loss": 0.9834,
+      "step": 2102
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9170159509017182,
+      "learning_rate": 1.9979999996822744e-05,
+      "loss": 0.831,
+      "step": 2103
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.350848703819201,
+      "learning_rate": 1.997995173294714e-05,
+      "loss": 0.973,
+      "step": 2104
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7326798214570005,
+      "learning_rate": 1.9979903410965095e-05,
+      "loss": 0.9396,
+      "step": 2105
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.671486255473249,
+      "learning_rate": 1.9979855030876894e-05,
+      "loss": 0.9661,
+      "step": 2106
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.846164987031269,
+      "learning_rate": 1.997980659268282e-05,
+      "loss": 0.8367,
+      "step": 2107
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7474230909847226,
+      "learning_rate": 1.9979758096383147e-05,
+      "loss": 1.0113,
+      "step": 2108
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8058372377314773,
+      "learning_rate": 1.9979709541978164e-05,
+      "loss": 0.7527,
+      "step": 2109
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7957712741536094,
+      "learning_rate": 1.9979660929468157e-05,
+      "loss": 0.9788,
+      "step": 2110
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0231021471512243,
+      "learning_rate": 1.99796122588534e-05,
+      "loss": 0.9983,
+      "step": 2111
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4238044583040694,
+      "learning_rate": 1.997956353013418e-05,
+      "loss": 0.9186,
+      "step": 2112
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7760916468729397,
+      "learning_rate": 1.9979514743310787e-05,
+      "loss": 0.6961,
+      "step": 2113
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.786428926723108,
+      "learning_rate": 1.9979465898383498e-05,
+      "loss": 0.8427,
+      "step": 2114
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3324934338763454,
+      "learning_rate": 1.9979416995352595e-05,
+      "loss": 0.957,
+      "step": 2115
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5095508781728917,
+      "learning_rate": 1.997936803421837e-05,
+      "loss": 0.9202,
+      "step": 2116
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0554037052555376,
+      "learning_rate": 1.9979319014981105e-05,
+      "loss": 0.9208,
+      "step": 2117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7100559199014491,
+      "learning_rate": 1.9979269937641087e-05,
+      "loss": 0.9247,
+      "step": 2118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8029347275276701,
+      "learning_rate": 1.9979220802198597e-05,
+      "loss": 0.8556,
+      "step": 2119
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8570487425762978,
+      "learning_rate": 1.9979171608653926e-05,
+      "loss": 1.0358,
+      "step": 2120
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7652770259699404,
+      "learning_rate": 1.9979122357007355e-05,
+      "loss": 0.9011,
+      "step": 2121
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.690039064183506,
+      "learning_rate": 1.9979073047259176e-05,
+      "loss": 0.9348,
+      "step": 2122
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8829733550746715,
+      "learning_rate": 1.9979023679409676e-05,
+      "loss": 1.0126,
+      "step": 2123
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6360002307759385,
+      "learning_rate": 1.997897425345914e-05,
+      "loss": 0.7858,
+      "step": 2124
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8598349093165725,
+      "learning_rate": 1.9978924769407858e-05,
+      "loss": 0.8722,
+      "step": 2125
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7823488504545195,
+      "learning_rate": 1.9978875227256115e-05,
+      "loss": 0.7459,
+      "step": 2126
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.784379830389471,
+      "learning_rate": 1.99788256270042e-05,
+      "loss": 0.7346,
+      "step": 2127
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6595618500456737,
+      "learning_rate": 1.9978775968652405e-05,
+      "loss": 0.8715,
+      "step": 2128
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6524455582371995,
+      "learning_rate": 1.9978726252201018e-05,
+      "loss": 0.87,
+      "step": 2129
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6216462701443293,
+      "learning_rate": 1.9978676477650326e-05,
+      "loss": 1.034,
+      "step": 2130
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7505785203229896,
+      "learning_rate": 1.9978626645000617e-05,
+      "loss": 0.9804,
+      "step": 2131
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3968418174692794,
+      "learning_rate": 1.997857675425219e-05,
+      "loss": 0.9226,
+      "step": 2132
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7372925871118883,
+      "learning_rate": 1.997852680540533e-05,
+      "loss": 1.056,
+      "step": 2133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7943262570340939,
+      "learning_rate": 1.9978476798460324e-05,
+      "loss": 0.7205,
+      "step": 2134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1181206641856067,
+      "learning_rate": 1.997842673341747e-05,
+      "loss": 0.8033,
+      "step": 2135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7577319322243259,
+      "learning_rate": 1.9978376610277053e-05,
+      "loss": 0.6227,
+      "step": 2136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.172085224855081,
+      "learning_rate": 1.997832642903937e-05,
+      "loss": 0.9496,
+      "step": 2137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6822186366683938,
+      "learning_rate": 1.9978276189704713e-05,
+      "loss": 0.9993,
+      "step": 2138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0543751964124137,
+      "learning_rate": 1.997822589227337e-05,
+      "loss": 0.9163,
+      "step": 2139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7744512917523578,
+      "learning_rate": 1.9978175536745642e-05,
+      "loss": 0.7525,
+      "step": 2140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7347559188067345,
+      "learning_rate": 1.9978125123121812e-05,
+      "loss": 0.9592,
+      "step": 2141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9056179471985923,
+      "learning_rate": 1.997807465140218e-05,
+      "loss": 0.8311,
+      "step": 2142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7687605668018982,
+      "learning_rate": 1.997802412158704e-05,
+      "loss": 0.7384,
+      "step": 2143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0089775231406053,
+      "learning_rate": 1.997797353367668e-05,
+      "loss": 1.0134,
+      "step": 2144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.642481356914606,
+      "learning_rate": 1.9977922887671404e-05,
+      "loss": 0.8431,
+      "step": 2145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8108088952907724,
+      "learning_rate": 1.9977872183571502e-05,
+      "loss": 1.0288,
+      "step": 2146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.038869749921996,
+      "learning_rate": 1.9977821421377266e-05,
+      "loss": 0.8685,
+      "step": 2147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7567748195271913,
+      "learning_rate": 1.9977770601088994e-05,
+      "loss": 0.9884,
+      "step": 2148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7091387231983062,
+      "learning_rate": 1.9977719722706985e-05,
+      "loss": 0.7884,
+      "step": 2149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8274600301148105,
+      "learning_rate": 1.9977668786231536e-05,
+      "loss": 0.9751,
+      "step": 2150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9461700607905614,
+      "learning_rate": 1.9977617791662935e-05,
+      "loss": 1.0024,
+      "step": 2151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8197088263318465,
+      "learning_rate": 1.9977566739001485e-05,
+      "loss": 0.7341,
+      "step": 2152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5747665710615104,
+      "learning_rate": 1.9977515628247482e-05,
+      "loss": 0.9341,
+      "step": 2153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.697603192430294,
+      "learning_rate": 1.997746445940123e-05,
+      "loss": 1.0474,
+      "step": 2154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1727892131646604,
+      "learning_rate": 1.9977413232463017e-05,
+      "loss": 0.9203,
+      "step": 2155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.655965832373167,
+      "learning_rate": 1.9977361947433144e-05,
+      "loss": 0.9125,
+      "step": 2156
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.588638198235665,
+      "learning_rate": 1.9977310604311912e-05,
+      "loss": 0.8222,
+      "step": 2157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7567993469076326,
+      "learning_rate": 1.9977259203099623e-05,
+      "loss": 0.8649,
+      "step": 2158
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.178563618477594,
+      "learning_rate": 1.9977207743796566e-05,
+      "loss": 0.953,
+      "step": 2159
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8058437899755437,
+      "learning_rate": 1.9977156226403052e-05,
+      "loss": 0.9051,
+      "step": 2160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6916296419221977,
+      "learning_rate": 1.997710465091937e-05,
+      "loss": 0.8287,
+      "step": 2161
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9638576597885322,
+      "learning_rate": 1.9977053017345832e-05,
+      "loss": 0.9472,
+      "step": 2162
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8537495919720488,
+      "learning_rate": 1.997700132568273e-05,
+      "loss": 0.9014,
+      "step": 2163
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.453518580560023,
+      "learning_rate": 1.997694957593037e-05,
+      "loss": 0.9198,
+      "step": 2164
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8075908015450031,
+      "learning_rate": 1.9976897768089047e-05,
+      "loss": 0.7286,
+      "step": 2165
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9055644316719986,
+      "learning_rate": 1.9976845902159066e-05,
+      "loss": 0.9815,
+      "step": 2166
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8607413182942368,
+      "learning_rate": 1.997679397814073e-05,
+      "loss": 0.7181,
+      "step": 2167
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5923334916522043,
+      "learning_rate": 1.9976741996034344e-05,
+      "loss": 0.9981,
+      "step": 2168
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.868911415860727,
+      "learning_rate": 1.9976689955840206e-05,
+      "loss": 0.93,
+      "step": 2169
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7805725666092027,
+      "learning_rate": 1.997663785755862e-05,
+      "loss": 0.7352,
+      "step": 2170
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9599059472414484,
+      "learning_rate": 1.997658570118989e-05,
+      "loss": 1.0254,
+      "step": 2171
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.642189870448357,
+      "learning_rate": 1.997653348673432e-05,
+      "loss": 0.945,
+      "step": 2172
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.597178551992481,
+      "learning_rate": 1.9976481214192215e-05,
+      "loss": 0.9366,
+      "step": 2173
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.866798835300824,
+      "learning_rate": 1.9976428883563876e-05,
+      "loss": 0.8899,
+      "step": 2174
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8547088148656856,
+      "learning_rate": 1.997637649484961e-05,
+      "loss": 0.9249,
+      "step": 2175
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0294761012640468,
+      "learning_rate": 1.997632404804972e-05,
+      "loss": 1.1093,
+      "step": 2176
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.65778003983674,
+      "learning_rate": 1.9976271543164514e-05,
+      "loss": 0.9982,
+      "step": 2177
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.508956771583488,
+      "learning_rate": 1.99762189801943e-05,
+      "loss": 0.9694,
+      "step": 2178
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7806824264146128,
+      "learning_rate": 1.9976166359139375e-05,
+      "loss": 0.8975,
+      "step": 2179
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9770923272676355,
+      "learning_rate": 1.9976113680000055e-05,
+      "loss": 0.9058,
+      "step": 2180
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6667341979353711,
+      "learning_rate": 1.997606094277664e-05,
+      "loss": 0.8924,
+      "step": 2181
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0764012629569657,
+      "learning_rate": 1.997600814746944e-05,
+      "loss": 0.8527,
+      "step": 2182
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0207276596610657,
+      "learning_rate": 1.9975955294078764e-05,
+      "loss": 0.7861,
+      "step": 2183
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7282239443228349,
+      "learning_rate": 1.9975902382604916e-05,
+      "loss": 0.8663,
+      "step": 2184
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.74336813622372,
+      "learning_rate": 1.997584941304821e-05,
+      "loss": 0.9054,
+      "step": 2185
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.312996506500195,
+      "learning_rate": 1.9975796385408948e-05,
+      "loss": 0.8266,
+      "step": 2186
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.943647849175451,
+      "learning_rate": 1.997574329968744e-05,
+      "loss": 0.9338,
+      "step": 2187
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.86121372203266,
+      "learning_rate": 1.9975690155883998e-05,
+      "loss": 0.6944,
+      "step": 2188
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.321749349153583,
+      "learning_rate": 1.997563695399893e-05,
+      "loss": 0.8209,
+      "step": 2189
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9393489390299128,
+      "learning_rate": 1.997558369403254e-05,
+      "loss": 0.9941,
+      "step": 2190
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5922788807513675,
+      "learning_rate": 1.997553037598515e-05,
+      "loss": 0.9171,
+      "step": 2191
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8389553168553994,
+      "learning_rate": 1.997547699985706e-05,
+      "loss": 0.7652,
+      "step": 2192
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8329811892721527,
+      "learning_rate": 1.997542356564859e-05,
+      "loss": 0.7503,
+      "step": 2193
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8719767034436043,
+      "learning_rate": 1.997537007336004e-05,
+      "loss": 0.9798,
+      "step": 2194
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.392387790748327,
+      "learning_rate": 1.997531652299173e-05,
+      "loss": 0.9856,
+      "step": 2195
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4737682231731224,
+      "learning_rate": 1.9975262914543968e-05,
+      "loss": 0.9641,
+      "step": 2196
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6407816071533368,
+      "learning_rate": 1.997520924801707e-05,
+      "loss": 0.8222,
+      "step": 2197
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6966536285217586,
+      "learning_rate": 1.9975155523411343e-05,
+      "loss": 0.9073,
+      "step": 2198
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6725819195753566,
+      "learning_rate": 1.9975101740727104e-05,
+      "loss": 0.8954,
+      "step": 2199
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.633534353191678,
+      "learning_rate": 1.9975047899964663e-05,
+      "loss": 0.9436,
+      "step": 2200
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4652107734633428,
+      "learning_rate": 1.9974994001124336e-05,
+      "loss": 0.974,
+      "step": 2201
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.771030721949784,
+      "learning_rate": 1.9974940044206433e-05,
+      "loss": 0.7076,
+      "step": 2202
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.59783994022666,
+      "learning_rate": 1.9974886029211275e-05,
+      "loss": 0.9545,
+      "step": 2203
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7598713579298318,
+      "learning_rate": 1.9974831956139172e-05,
+      "loss": 0.7333,
+      "step": 2204
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.606280314981976,
+      "learning_rate": 1.997477782499044e-05,
+      "loss": 1.0035,
+      "step": 2205
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6953120075234178,
+      "learning_rate": 1.997472363576539e-05,
+      "loss": 0.9718,
+      "step": 2206
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6570866972668588,
+      "learning_rate": 1.997466938846434e-05,
+      "loss": 0.8774,
+      "step": 2207
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6818241469981994,
+      "learning_rate": 1.9974615083087614e-05,
+      "loss": 0.7734,
+      "step": 2208
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8674039918913938,
+      "learning_rate": 1.9974560719635517e-05,
+      "loss": 0.8577,
+      "step": 2209
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.513105676182982,
+      "learning_rate": 1.997450629810837e-05,
+      "loss": 0.9566,
+      "step": 2210
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7384417995503942,
+      "learning_rate": 1.997445181850649e-05,
+      "loss": 0.9197,
+      "step": 2211
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.872743167509238,
+      "learning_rate": 1.997439728083019e-05,
+      "loss": 0.8878,
+      "step": 2212
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.764551066948508,
+      "learning_rate": 1.9974342685079795e-05,
+      "loss": 0.866,
+      "step": 2213
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9335292258568775,
+      "learning_rate": 1.997428803125562e-05,
+      "loss": 0.8529,
+      "step": 2214
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.788028325850593,
+      "learning_rate": 1.9974233319357977e-05,
+      "loss": 0.8659,
+      "step": 2215
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5945825785533105,
+      "learning_rate": 1.9974178549387195e-05,
+      "loss": 1.0392,
+      "step": 2216
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6864165213080395,
+      "learning_rate": 1.9974123721343587e-05,
+      "loss": 1.0194,
+      "step": 2217
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8953531989467383,
+      "learning_rate": 1.9974068835227468e-05,
+      "loss": 0.8668,
+      "step": 2218
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9517091087802132,
+      "learning_rate": 1.997401389103917e-05,
+      "loss": 1.0143,
+      "step": 2219
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6562943720588232,
+      "learning_rate": 1.9973958888779e-05,
+      "loss": 1.0511,
+      "step": 2220
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7145815023337663,
+      "learning_rate": 1.9973903828447286e-05,
+      "loss": 1.0242,
+      "step": 2221
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0465927645778215,
+      "learning_rate": 1.9973848710044345e-05,
+      "loss": 0.8956,
+      "step": 2222
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0854520601493487,
+      "learning_rate": 1.99737935335705e-05,
+      "loss": 0.8528,
+      "step": 2223
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6980492182383748,
+      "learning_rate": 1.997373829902607e-05,
+      "loss": 0.9832,
+      "step": 2224
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.535447294518126,
+      "learning_rate": 1.997368300641138e-05,
+      "loss": 0.9179,
+      "step": 2225
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.42994176529589,
+      "learning_rate": 1.997362765572675e-05,
+      "loss": 1.0077,
+      "step": 2226
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6645884151343342,
+      "learning_rate": 1.99735722469725e-05,
+      "loss": 0.7826,
+      "step": 2227
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.776801212167214,
+      "learning_rate": 1.9973516780148957e-05,
+      "loss": 0.9835,
+      "step": 2228
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7207169402322393,
+      "learning_rate": 1.9973461255256443e-05,
+      "loss": 1.0255,
+      "step": 2229
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6952348748810824,
+      "learning_rate": 1.9973405672295276e-05,
+      "loss": 0.845,
+      "step": 2230
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.236315683159513,
+      "learning_rate": 1.9973350031265786e-05,
+      "loss": 1.0092,
+      "step": 2231
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6957455755866144,
+      "learning_rate": 1.9973294332168294e-05,
+      "loss": 1.0199,
+      "step": 2232
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.693813096199994,
+      "learning_rate": 1.9973238575003123e-05,
+      "loss": 1.0234,
+      "step": 2233
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7398208814678635,
+      "learning_rate": 1.9973182759770605e-05,
+      "loss": 1.0351,
+      "step": 2234
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6668733093569181,
+      "learning_rate": 1.9973126886471053e-05,
+      "loss": 0.8909,
+      "step": 2235
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.832883278298335,
+      "learning_rate": 1.9973070955104802e-05,
+      "loss": 1.0271,
+      "step": 2236
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0454289683085656,
+      "learning_rate": 1.9973014965672175e-05,
+      "loss": 0.7802,
+      "step": 2237
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7858847291076628,
+      "learning_rate": 1.9972958918173498e-05,
+      "loss": 0.7292,
+      "step": 2238
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.883435796399833,
+      "learning_rate": 1.9972902812609093e-05,
+      "loss": 0.8608,
+      "step": 2239
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8047928996075815,
+      "learning_rate": 1.9972846648979292e-05,
+      "loss": 0.7554,
+      "step": 2240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7230026873008981,
+      "learning_rate": 1.9972790427284424e-05,
+      "loss": 0.8545,
+      "step": 2241
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.711973187297525,
+      "learning_rate": 1.997273414752481e-05,
+      "loss": 0.7668,
+      "step": 2242
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8117439521387189,
+      "learning_rate": 1.997267780970078e-05,
+      "loss": 0.9523,
+      "step": 2243
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7945902522632202,
+      "learning_rate": 1.9972621413812664e-05,
+      "loss": 0.7529,
+      "step": 2244
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8183626897930976,
+      "learning_rate": 1.997256495986079e-05,
+      "loss": 0.801,
+      "step": 2245
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7378282222872514,
+      "learning_rate": 1.9972508447845483e-05,
+      "loss": 0.9763,
+      "step": 2246
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6265320053101155,
+      "learning_rate": 1.9972451877767075e-05,
+      "loss": 0.9461,
+      "step": 2247
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6039944117009604,
+      "learning_rate": 1.9972395249625897e-05,
+      "loss": 0.8811,
+      "step": 2248
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6675314186852088,
+      "learning_rate": 1.9972338563422273e-05,
+      "loss": 0.9617,
+      "step": 2249
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5762228095219375,
+      "learning_rate": 1.997228181915654e-05,
+      "loss": 1.0269,
+      "step": 2250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9284446202579928,
+      "learning_rate": 1.9972225016829023e-05,
+      "loss": 0.8944,
+      "step": 2251
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9957755809143876,
+      "learning_rate": 1.9972168156440054e-05,
+      "loss": 0.9657,
+      "step": 2252
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.809951482520653,
+      "learning_rate": 1.9972111237989968e-05,
+      "loss": 0.9716,
+      "step": 2253
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.533121266703794,
+      "learning_rate": 1.997205426147909e-05,
+      "loss": 0.8832,
+      "step": 2254
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.806488417676861,
+      "learning_rate": 1.9971997226907756e-05,
+      "loss": 0.8686,
+      "step": 2255
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8250788292102054,
+      "learning_rate": 1.9971940134276298e-05,
+      "loss": 0.9109,
+      "step": 2256
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7396303582566315,
+      "learning_rate": 1.9971882983585048e-05,
+      "loss": 0.9334,
+      "step": 2257
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.044759127292159,
+      "learning_rate": 1.9971825774834338e-05,
+      "loss": 0.9114,
+      "step": 2258
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.855696580570756,
+      "learning_rate": 1.99717685080245e-05,
+      "loss": 1.1001,
+      "step": 2259
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6142203531783996,
+      "learning_rate": 1.997171118315587e-05,
+      "loss": 0.9133,
+      "step": 2260
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.790823706367059,
+      "learning_rate": 1.9971653800228782e-05,
+      "loss": 0.7407,
+      "step": 2261
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6981761748955246,
+      "learning_rate": 1.9971596359243562e-05,
+      "loss": 0.8724,
+      "step": 2262
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7261548835825509,
+      "learning_rate": 1.9971538860200556e-05,
+      "loss": 0.8388,
+      "step": 2263
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.834141408932957,
+      "learning_rate": 1.9971481303100093e-05,
+      "loss": 0.95,
+      "step": 2264
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3769510661777846,
+      "learning_rate": 1.997142368794251e-05,
+      "loss": 0.915,
+      "step": 2265
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.980268315683123,
+      "learning_rate": 1.997136601472814e-05,
+      "loss": 0.9592,
+      "step": 2266
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.735280420298666,
+      "learning_rate": 1.997130828345732e-05,
+      "loss": 0.9391,
+      "step": 2267
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8784300110962444,
+      "learning_rate": 1.997125049413039e-05,
+      "loss": 0.9364,
+      "step": 2268
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.940428882072197,
+      "learning_rate": 1.9971192646747678e-05,
+      "loss": 0.8601,
+      "step": 2269
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.512272392207098,
+      "learning_rate": 1.997113474130953e-05,
+      "loss": 0.8718,
+      "step": 2270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.490817663065329,
+      "learning_rate": 1.9971076777816275e-05,
+      "loss": 0.8491,
+      "step": 2271
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0402028102311904,
+      "learning_rate": 1.9971018756268256e-05,
+      "loss": 0.8698,
+      "step": 2272
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5613162236548057,
+      "learning_rate": 1.997096067666581e-05,
+      "loss": 0.9534,
+      "step": 2273
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.652859252792861,
+      "learning_rate": 1.997090253900927e-05,
+      "loss": 0.8052,
+      "step": 2274
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8253129744725839,
+      "learning_rate": 1.997084434329898e-05,
+      "loss": 0.7331,
+      "step": 2275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.731842872449227,
+      "learning_rate": 1.9970786089535283e-05,
+      "loss": 0.872,
+      "step": 2276
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7829274102640468,
+      "learning_rate": 1.9970727777718505e-05,
+      "loss": 0.7689,
+      "step": 2277
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9094951386409211,
+      "learning_rate": 1.9970669407848997e-05,
+      "loss": 0.9899,
+      "step": 2278
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.279332209312092,
+      "learning_rate": 1.9970610979927096e-05,
+      "loss": 0.9673,
+      "step": 2279
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9251496202591751,
+      "learning_rate": 1.997055249395314e-05,
+      "loss": 0.8439,
+      "step": 2280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8580285376104948,
+      "learning_rate": 1.9970493949927473e-05,
+      "loss": 0.9343,
+      "step": 2281
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6120575065207159,
+      "learning_rate": 1.9970435347850428e-05,
+      "loss": 0.7457,
+      "step": 2282
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.697684733196272,
+      "learning_rate": 1.9970376687722356e-05,
+      "loss": 0.8469,
+      "step": 2283
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.713313074772128,
+      "learning_rate": 1.9970317969543592e-05,
+      "loss": 0.9906,
+      "step": 2284
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6610924889602945,
+      "learning_rate": 1.9970259193314483e-05,
+      "loss": 1.0892,
+      "step": 2285
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2337797719338313,
+      "learning_rate": 1.9970200359035366e-05,
+      "loss": 0.8542,
+      "step": 2286
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.566194240993718,
+      "learning_rate": 1.9970141466706588e-05,
+      "loss": 0.9593,
+      "step": 2287
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8375037969563994,
+      "learning_rate": 1.997008251632849e-05,
+      "loss": 0.8866,
+      "step": 2288
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.72429831909396,
+      "learning_rate": 1.9970023507901416e-05,
+      "loss": 0.733,
+      "step": 2289
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6161200519431234,
+      "learning_rate": 1.9969964441425706e-05,
+      "loss": 0.9838,
+      "step": 2290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.809120494756176,
+      "learning_rate": 1.9969905316901704e-05,
+      "loss": 0.7144,
+      "step": 2291
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.585201149978512,
+      "learning_rate": 1.9969846134329765e-05,
+      "loss": 0.9641,
+      "step": 2292
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7891346294061836,
+      "learning_rate": 1.996978689371022e-05,
+      "loss": 0.9153,
+      "step": 2293
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6004344958439152,
+      "learning_rate": 1.9969727595043417e-05,
+      "loss": 0.8577,
+      "step": 2294
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.492638492141859,
+      "learning_rate": 1.996966823832971e-05,
+      "loss": 0.971,
+      "step": 2295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.830859173668038,
+      "learning_rate": 1.9969608823569435e-05,
+      "loss": 0.8452,
+      "step": 2296
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.829407008589179,
+      "learning_rate": 1.996954935076294e-05,
+      "loss": 0.889,
+      "step": 2297
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6498155412986426,
+      "learning_rate": 1.9969489819910572e-05,
+      "loss": 0.9575,
+      "step": 2298
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3853812345764327,
+      "learning_rate": 1.9969430231012682e-05,
+      "loss": 0.8786,
+      "step": 2299
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9182154407492766,
+      "learning_rate": 1.9969370584069607e-05,
+      "loss": 0.8147,
+      "step": 2300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5691356700285286,
+      "learning_rate": 1.9969310879081705e-05,
+      "loss": 0.9208,
+      "step": 2301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.233576736022847,
+      "learning_rate": 1.996925111604932e-05,
+      "loss": 0.912,
+      "step": 2302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.831721490087331,
+      "learning_rate": 1.9969191294972794e-05,
+      "loss": 0.7961,
+      "step": 2303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6196666213731294,
+      "learning_rate": 1.9969131415852482e-05,
+      "loss": 0.9396,
+      "step": 2304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8134375141042585,
+      "learning_rate": 1.996907147868873e-05,
+      "loss": 0.843,
+      "step": 2305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6986763565335248,
+      "learning_rate": 1.9969011483481888e-05,
+      "loss": 1.0431,
+      "step": 2306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.890069951715086,
+      "learning_rate": 1.9968951430232307e-05,
+      "loss": 0.9605,
+      "step": 2307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.541139717339928,
+      "learning_rate": 1.9968891318940332e-05,
+      "loss": 0.8594,
+      "step": 2308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.877131758892158,
+      "learning_rate": 1.996883114960632e-05,
+      "loss": 0.7752,
+      "step": 2309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.368077636658239,
+      "learning_rate": 1.996877092223061e-05,
+      "loss": 0.9599,
+      "step": 2310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0056829930675675,
+      "learning_rate": 1.9968710636813564e-05,
+      "loss": 0.8466,
+      "step": 2311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.72974417397975,
+      "learning_rate": 1.996865029335553e-05,
+      "loss": 0.9743,
+      "step": 2312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0782237666107695,
+      "learning_rate": 1.9968589891856854e-05,
+      "loss": 0.9588,
+      "step": 2313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7837779057156036,
+      "learning_rate": 1.9968529432317894e-05,
+      "loss": 0.874,
+      "step": 2314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6518134236305726,
+      "learning_rate": 1.9968468914738997e-05,
+      "loss": 0.9853,
+      "step": 2315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.74516503351375,
+      "learning_rate": 1.9968408339120522e-05,
+      "loss": 0.9259,
+      "step": 2316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.255716165534589,
+      "learning_rate": 1.9968347705462817e-05,
+      "loss": 0.8301,
+      "step": 2317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9832737001465823,
+      "learning_rate": 1.9968287013766235e-05,
+      "loss": 1.0554,
+      "step": 2318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6869414600414279,
+      "learning_rate": 1.996822626403113e-05,
+      "loss": 1.0082,
+      "step": 2319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.881250865856896,
+      "learning_rate": 1.9968165456257855e-05,
+      "loss": 0.9647,
+      "step": 2320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0242988902694865,
+      "learning_rate": 1.9968104590446766e-05,
+      "loss": 0.8281,
+      "step": 2321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6303696004378594,
+      "learning_rate": 1.9968043666598216e-05,
+      "loss": 0.8369,
+      "step": 2322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0189012225452947,
+      "learning_rate": 1.996798268471256e-05,
+      "loss": 0.8172,
+      "step": 2323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8527357221430252,
+      "learning_rate": 1.9967921644790153e-05,
+      "loss": 0.9947,
+      "step": 2324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9067377806319732,
+      "learning_rate": 1.9967860546831348e-05,
+      "loss": 0.7746,
+      "step": 2325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8410518031911793,
+      "learning_rate": 1.9967799390836505e-05,
+      "loss": 0.847,
+      "step": 2326
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.725259567005995,
+      "learning_rate": 1.996773817680598e-05,
+      "loss": 0.8753,
+      "step": 2327
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8823482673620322,
+      "learning_rate": 1.9967676904740126e-05,
+      "loss": 0.8624,
+      "step": 2328
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.363236445724382,
+      "learning_rate": 1.99676155746393e-05,
+      "loss": 0.9252,
+      "step": 2329
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7968455221822337,
+      "learning_rate": 1.9967554186503864e-05,
+      "loss": 0.7077,
+      "step": 2330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6067691601196328,
+      "learning_rate": 1.9967492740334168e-05,
+      "loss": 0.9061,
+      "step": 2331
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5325503597140782,
+      "learning_rate": 1.9967431236130576e-05,
+      "loss": 0.9514,
+      "step": 2332
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6925293508302455,
+      "learning_rate": 1.9967369673893443e-05,
+      "loss": 1.0444,
+      "step": 2333
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5071005487804303,
+      "learning_rate": 1.9967308053623128e-05,
+      "loss": 0.8461,
+      "step": 2334
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7531969569983035,
+      "learning_rate": 1.996724637531999e-05,
+      "loss": 1.0151,
+      "step": 2335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6826611722470926,
+      "learning_rate": 1.996718463898439e-05,
+      "loss": 0.8835,
+      "step": 2336
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6400062693916706,
+      "learning_rate": 1.996712284461668e-05,
+      "loss": 0.9553,
+      "step": 2337
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0040683074596393,
+      "learning_rate": 1.996706099221723e-05,
+      "loss": 0.9,
+      "step": 2338
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7610403374675996,
+      "learning_rate": 1.9966999081786396e-05,
+      "loss": 0.8219,
+      "step": 2339
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.978732910350582,
+      "learning_rate": 1.9966937113324536e-05,
+      "loss": 0.9713,
+      "step": 2340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8727450180204737,
+      "learning_rate": 1.9966875086832012e-05,
+      "loss": 0.9686,
+      "step": 2341
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6432408157183431,
+      "learning_rate": 1.9966813002309187e-05,
+      "loss": 0.933,
+      "step": 2342
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8274387085277077,
+      "learning_rate": 1.9966750859756422e-05,
+      "loss": 0.9474,
+      "step": 2343
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.233436549601798,
+      "learning_rate": 1.9966688659174074e-05,
+      "loss": 0.9504,
+      "step": 2344
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9923310490797812,
+      "learning_rate": 1.9966626400562515e-05,
+      "loss": 0.7299,
+      "step": 2345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6325334564401093,
+      "learning_rate": 1.9966564083922098e-05,
+      "loss": 1.0559,
+      "step": 2346
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8227719438387446,
+      "learning_rate": 1.996650170925319e-05,
+      "loss": 0.9386,
+      "step": 2347
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9130099732037023,
+      "learning_rate": 1.9966439276556155e-05,
+      "loss": 0.8895,
+      "step": 2348
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6198771433932977,
+      "learning_rate": 1.9966376785831354e-05,
+      "loss": 0.9298,
+      "step": 2349
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8349272203748241,
+      "learning_rate": 1.996631423707915e-05,
+      "loss": 0.8769,
+      "step": 2350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.688337947260217,
+      "learning_rate": 1.9966251630299915e-05,
+      "loss": 0.9261,
+      "step": 2351
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4777842166891801,
+      "learning_rate": 1.9966188965494e-05,
+      "loss": 0.8301,
+      "step": 2352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5797593370739964,
+      "learning_rate": 1.9966126242661783e-05,
+      "loss": 0.9104,
+      "step": 2353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5959766987964445,
+      "learning_rate": 1.996606346180362e-05,
+      "loss": 0.7174,
+      "step": 2354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8131862691058419,
+      "learning_rate": 1.9966000622919886e-05,
+      "loss": 0.7943,
+      "step": 2355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6619223622779065,
+      "learning_rate": 1.996593772601094e-05,
+      "loss": 0.8331,
+      "step": 2356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8250493057843167,
+      "learning_rate": 1.996587477107715e-05,
+      "loss": 0.9487,
+      "step": 2357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.920149488765991,
+      "learning_rate": 1.9965811758118873e-05,
+      "loss": 0.7104,
+      "step": 2358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.140836409325902,
+      "learning_rate": 1.9965748687136494e-05,
+      "loss": 0.8099,
+      "step": 2359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0107877803388425,
+      "learning_rate": 1.9965685558130368e-05,
+      "loss": 0.9429,
+      "step": 2360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8241099348557543,
+      "learning_rate": 1.9965622371100866e-05,
+      "loss": 0.7821,
+      "step": 2361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8508261968765454,
+      "learning_rate": 1.9965559126048354e-05,
+      "loss": 0.9028,
+      "step": 2362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7236451899256995,
+      "learning_rate": 1.9965495822973204e-05,
+      "loss": 0.9302,
+      "step": 2363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8958566437778726,
+      "learning_rate": 1.996543246187578e-05,
+      "loss": 0.8419,
+      "step": 2364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8046177533802237,
+      "learning_rate": 1.9965369042756453e-05,
+      "loss": 0.9815,
+      "step": 2365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8571226554690272,
+      "learning_rate": 1.99653055656156e-05,
+      "loss": 0.7967,
+      "step": 2366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.142615322125313,
+      "learning_rate": 1.9965242030453577e-05,
+      "loss": 0.9055,
+      "step": 2367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5874817203931908,
+      "learning_rate": 1.9965178437270757e-05,
+      "loss": 0.7773,
+      "step": 2368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7951875404654698,
+      "learning_rate": 1.9965114786067515e-05,
+      "loss": 0.9338,
+      "step": 2369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1469671120890865,
+      "learning_rate": 1.9965051076844224e-05,
+      "loss": 0.8656,
+      "step": 2370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0638339138985913,
+      "learning_rate": 1.9964987309601246e-05,
+      "loss": 0.9443,
+      "step": 2371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6412293844728107,
+      "learning_rate": 1.996492348433896e-05,
+      "loss": 0.8074,
+      "step": 2372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7691226856226283,
+      "learning_rate": 1.9964859601057734e-05,
+      "loss": 0.7166,
+      "step": 2373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.810651379280642,
+      "learning_rate": 1.996479565975794e-05,
+      "loss": 0.758,
+      "step": 2374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6109271751055245,
+      "learning_rate": 1.996473166043995e-05,
+      "loss": 0.8827,
+      "step": 2375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8115097358114423,
+      "learning_rate": 1.996466760310414e-05,
+      "loss": 0.748,
+      "step": 2376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7244882288271657,
+      "learning_rate": 1.996460348775088e-05,
+      "loss": 0.818,
+      "step": 2377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4866047683314776,
+      "learning_rate": 1.9964539314380542e-05,
+      "loss": 0.891,
+      "step": 2378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8331176760327368,
+      "learning_rate": 1.99644750829935e-05,
+      "loss": 0.8762,
+      "step": 2379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8148375022614271,
+      "learning_rate": 1.996441079359013e-05,
+      "loss": 0.8158,
+      "step": 2380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0367275956913216,
+      "learning_rate": 1.996434644617081e-05,
+      "loss": 0.927,
+      "step": 2381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5680132918070149,
+      "learning_rate": 1.9964282040735907e-05,
+      "loss": 0.9231,
+      "step": 2382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7037902538456342,
+      "learning_rate": 1.99642175772858e-05,
+      "loss": 0.9807,
+      "step": 2383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.821271260295339,
+      "learning_rate": 1.9964153055820863e-05,
+      "loss": 0.9331,
+      "step": 2384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6929448965138352,
+      "learning_rate": 1.996408847634147e-05,
+      "loss": 0.9131,
+      "step": 2385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3086501995981568,
+      "learning_rate": 1.9964023838848004e-05,
+      "loss": 0.933,
+      "step": 2386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8531951802216478,
+      "learning_rate": 1.9963959143340833e-05,
+      "loss": 1.0131,
+      "step": 2387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.757805803008038,
+      "learning_rate": 1.996389438982034e-05,
+      "loss": 0.9584,
+      "step": 2388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4141607166671704,
+      "learning_rate": 1.99638295782869e-05,
+      "loss": 0.9355,
+      "step": 2389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7698704292867518,
+      "learning_rate": 1.996376470874089e-05,
+      "loss": 0.8428,
+      "step": 2390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0596685097015746,
+      "learning_rate": 1.9963699781182685e-05,
+      "loss": 0.9017,
+      "step": 2391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9784727004524312,
+      "learning_rate": 1.9963634795612666e-05,
+      "loss": 0.9376,
+      "step": 2392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.712814970915482,
+      "learning_rate": 1.9963569752031207e-05,
+      "loss": 0.9439,
+      "step": 2393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8317550734294135,
+      "learning_rate": 1.9963504650438698e-05,
+      "loss": 0.9241,
+      "step": 2394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6453771255947338,
+      "learning_rate": 1.9963439490835504e-05,
+      "loss": 0.8238,
+      "step": 2395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6032039724474676,
+      "learning_rate": 1.9963374273222015e-05,
+      "loss": 0.9433,
+      "step": 2396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8947982396152725,
+      "learning_rate": 1.9963308997598604e-05,
+      "loss": 0.727,
+      "step": 2397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2541419185896303,
+      "learning_rate": 1.9963243663965658e-05,
+      "loss": 0.7736,
+      "step": 2398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7256114519785286,
+      "learning_rate": 1.9963178272323548e-05,
+      "loss": 0.9378,
+      "step": 2399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.584819593784692,
+      "learning_rate": 1.996311282267266e-05,
+      "loss": 1.0423,
+      "step": 2400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7939708131617969,
+      "learning_rate": 1.9963047315013376e-05,
+      "loss": 1.008,
+      "step": 2401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8280891093731115,
+      "learning_rate": 1.996298174934608e-05,
+      "loss": 0.8014,
+      "step": 2402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3669731704119283,
+      "learning_rate": 1.9962916125671146e-05,
+      "loss": 0.9368,
+      "step": 2403
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9113346791918806,
+      "learning_rate": 1.996285044398896e-05,
+      "loss": 0.9667,
+      "step": 2404
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6217144426504977,
+      "learning_rate": 1.9962784704299907e-05,
+      "loss": 0.9057,
+      "step": 2405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7758361822160721,
+      "learning_rate": 1.9962718906604365e-05,
+      "loss": 0.7738,
+      "step": 2406
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.874945583521751,
+      "learning_rate": 1.9962653050902722e-05,
+      "loss": 0.8677,
+      "step": 2407
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6122926545553338,
+      "learning_rate": 1.996258713719536e-05,
+      "loss": 0.914,
+      "step": 2408
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8204792829112992,
+      "learning_rate": 1.9962521165482657e-05,
+      "loss": 0.986,
+      "step": 2409
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9330901878894298,
+      "learning_rate": 1.9962455135765003e-05,
+      "loss": 0.8706,
+      "step": 2410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7574366206549494,
+      "learning_rate": 1.996238904804278e-05,
+      "loss": 0.7654,
+      "step": 2411
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2367251647319857,
+      "learning_rate": 1.9962322902316376e-05,
+      "loss": 0.8415,
+      "step": 2412
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7896115404104809,
+      "learning_rate": 1.9962256698586174e-05,
+      "loss": 0.7584,
+      "step": 2413
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7082501666543684,
+      "learning_rate": 1.996219043685256e-05,
+      "loss": 1.0187,
+      "step": 2414
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8642542631321326,
+      "learning_rate": 1.9962124117115914e-05,
+      "loss": 1.0169,
+      "step": 2415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.457669244580553,
+      "learning_rate": 1.9962057739376634e-05,
+      "loss": 0.925,
+      "step": 2416
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.606980886179275,
+      "learning_rate": 1.9961991303635096e-05,
+      "loss": 0.8691,
+      "step": 2417
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7929726605279201,
+      "learning_rate": 1.9961924809891693e-05,
+      "loss": 0.9273,
+      "step": 2418
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3256555307178055,
+      "learning_rate": 1.996185825814681e-05,
+      "loss": 0.7908,
+      "step": 2419
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.869424482057943,
+      "learning_rate": 1.9961791648400836e-05,
+      "loss": 0.9227,
+      "step": 2420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7295615025735586,
+      "learning_rate": 1.9961724980654154e-05,
+      "loss": 0.6968,
+      "step": 2421
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7267624907240906,
+      "learning_rate": 1.9961658254907153e-05,
+      "loss": 0.8678,
+      "step": 2422
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8793242862582664,
+      "learning_rate": 1.9961591471160228e-05,
+      "loss": 1.0353,
+      "step": 2423
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7783433269938528,
+      "learning_rate": 1.9961524629413765e-05,
+      "loss": 0.7601,
+      "step": 2424
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8796744458723007,
+      "learning_rate": 1.9961457729668148e-05,
+      "loss": 0.9698,
+      "step": 2425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7223811748526237,
+      "learning_rate": 1.9961390771923772e-05,
+      "loss": 0.9787,
+      "step": 2426
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6315636178424038,
+      "learning_rate": 1.9961323756181024e-05,
+      "loss": 0.8631,
+      "step": 2427
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8255331463522952,
+      "learning_rate": 1.9961256682440298e-05,
+      "loss": 0.8808,
+      "step": 2428
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.757620671322131,
+      "learning_rate": 1.996118955070198e-05,
+      "loss": 0.9605,
+      "step": 2429
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8699178334352693,
+      "learning_rate": 1.9961122360966463e-05,
+      "loss": 0.946,
+      "step": 2430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.719458282323735,
+      "learning_rate": 1.9961055113234137e-05,
+      "loss": 1.0227,
+      "step": 2431
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9983196088280886,
+      "learning_rate": 1.9960987807505398e-05,
+      "loss": 0.9919,
+      "step": 2432
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4099521948831417,
+      "learning_rate": 1.996092044378063e-05,
+      "loss": 0.8076,
+      "step": 2433
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5605320243836034,
+      "learning_rate": 1.996085302206023e-05,
+      "loss": 1.0631,
+      "step": 2434
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8306444300561924,
+      "learning_rate": 1.9960785542344594e-05,
+      "loss": 0.8512,
+      "step": 2435
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3360546960633823,
+      "learning_rate": 1.996071800463411e-05,
+      "loss": 0.9168,
+      "step": 2436
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1319252292309434,
+      "learning_rate": 1.996065040892917e-05,
+      "loss": 0.7923,
+      "step": 2437
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7712221394133163,
+      "learning_rate": 1.9960582755230172e-05,
+      "loss": 0.8631,
+      "step": 2438
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6105511962991443,
+      "learning_rate": 1.9960515043537504e-05,
+      "loss": 0.9548,
+      "step": 2439
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1302045179252405,
+      "learning_rate": 1.9960447273851567e-05,
+      "loss": 0.8955,
+      "step": 2440
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9185871728570891,
+      "learning_rate": 1.9960379446172754e-05,
+      "loss": 0.9376,
+      "step": 2441
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0840257620684763,
+      "learning_rate": 1.9960311560501457e-05,
+      "loss": 0.9448,
+      "step": 2442
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0106396066792103,
+      "learning_rate": 1.996024361683807e-05,
+      "loss": 0.89,
+      "step": 2443
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.734481528787937,
+      "learning_rate": 1.996017561518299e-05,
+      "loss": 0.903,
+      "step": 2444
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7769859771471832,
+      "learning_rate": 1.996010755553662e-05,
+      "loss": 1.0119,
+      "step": 2445
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.949772360622772,
+      "learning_rate": 1.996003943789935e-05,
+      "loss": 0.7892,
+      "step": 2446
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.058262364159942,
+      "learning_rate": 1.9959971262271574e-05,
+      "loss": 0.9144,
+      "step": 2447
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.832498049168593,
+      "learning_rate": 1.9959903028653695e-05,
+      "loss": 0.8315,
+      "step": 2448
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8018474682867602,
+      "learning_rate": 1.9959834737046107e-05,
+      "loss": 0.9557,
+      "step": 2449
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7581431206409859,
+      "learning_rate": 1.9959766387449203e-05,
+      "loss": 0.9392,
+      "step": 2450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8642636010418323,
+      "learning_rate": 1.995969797986339e-05,
+      "loss": 1.0583,
+      "step": 2451
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7820472368726217,
+      "learning_rate": 1.9959629514289065e-05,
+      "loss": 0.8206,
+      "step": 2452
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9474949387543905,
+      "learning_rate": 1.995956099072662e-05,
+      "loss": 0.9263,
+      "step": 2453
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0936147943965095,
+      "learning_rate": 1.995949240917646e-05,
+      "loss": 0.8771,
+      "step": 2454
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.975077556032171,
+      "learning_rate": 1.9959423769638978e-05,
+      "loss": 0.8473,
+      "step": 2455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6999112927925335,
+      "learning_rate": 1.995935507211458e-05,
+      "loss": 1.0443,
+      "step": 2456
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8497150252908716,
+      "learning_rate": 1.9959286316603662e-05,
+      "loss": 0.9469,
+      "step": 2457
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7136119262715184,
+      "learning_rate": 1.995921750310663e-05,
+      "loss": 0.8165,
+      "step": 2458
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7806817480157568,
+      "learning_rate": 1.9959148631623878e-05,
+      "loss": 0.9538,
+      "step": 2459
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7722689132473899,
+      "learning_rate": 1.995907970215581e-05,
+      "loss": 0.9426,
+      "step": 2460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7765447991186243,
+      "learning_rate": 1.9959010714702827e-05,
+      "loss": 1.0341,
+      "step": 2461
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6325002965113737,
+      "learning_rate": 1.995894166926533e-05,
+      "loss": 0.8515,
+      "step": 2462
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.692839139908965,
+      "learning_rate": 1.9958872565843724e-05,
+      "loss": 0.8862,
+      "step": 2463
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5034497494016186,
+      "learning_rate": 1.9958803404438406e-05,
+      "loss": 0.8627,
+      "step": 2464
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8207195140737167,
+      "learning_rate": 1.9958734185049782e-05,
+      "loss": 0.9955,
+      "step": 2465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4529690568288338,
+      "learning_rate": 1.9958664907678257e-05,
+      "loss": 0.9021,
+      "step": 2466
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.779852602247175,
+      "learning_rate": 1.995859557232423e-05,
+      "loss": 0.8517,
+      "step": 2467
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7164158749931793,
+      "learning_rate": 1.9958526178988107e-05,
+      "loss": 1.1089,
+      "step": 2468
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8502401552584489,
+      "learning_rate": 1.9958456727670296e-05,
+      "loss": 0.9493,
+      "step": 2469
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.065839962014,
+      "learning_rate": 1.9958387218371192e-05,
+      "loss": 0.9575,
+      "step": 2470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9476350494377381,
+      "learning_rate": 1.9958317651091206e-05,
+      "loss": 1.0682,
+      "step": 2471
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.853996822309988,
+      "learning_rate": 1.9958248025830743e-05,
+      "loss": 0.7932,
+      "step": 2472
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6869354830490806,
+      "learning_rate": 1.9958178342590205e-05,
+      "loss": 0.9727,
+      "step": 2473
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8918389063526106,
+      "learning_rate": 1.995810860137e-05,
+      "loss": 0.9766,
+      "step": 2474
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6831313976255518,
+      "learning_rate": 1.9958038802170537e-05,
+      "loss": 0.9712,
+      "step": 2475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7271337813270011,
+      "learning_rate": 1.995796894499222e-05,
+      "loss": 0.9902,
+      "step": 2476
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.858441541650961,
+      "learning_rate": 1.9957899029835453e-05,
+      "loss": 0.8721,
+      "step": 2477
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5195923509255702,
+      "learning_rate": 1.9957829056700643e-05,
+      "loss": 1.0194,
+      "step": 2478
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8282456257917514,
+      "learning_rate": 1.9957759025588203e-05,
+      "loss": 0.8758,
+      "step": 2479
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.742554941427461,
+      "learning_rate": 1.9957688936498537e-05,
+      "loss": 0.9161,
+      "step": 2480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9193132997559308,
+      "learning_rate": 1.9957618789432053e-05,
+      "loss": 0.8578,
+      "step": 2481
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.730997007312578,
+      "learning_rate": 1.995754858438916e-05,
+      "loss": 1.0196,
+      "step": 2482
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.605844572176964,
+      "learning_rate": 1.9957478321370266e-05,
+      "loss": 1.0336,
+      "step": 2483
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.899354682428731,
+      "learning_rate": 1.995740800037578e-05,
+      "loss": 0.859,
+      "step": 2484
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5589202878312893,
+      "learning_rate": 1.9957337621406112e-05,
+      "loss": 0.9,
+      "step": 2485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9064030376375722,
+      "learning_rate": 1.9957267184461674e-05,
+      "loss": 0.9174,
+      "step": 2486
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7282887666380826,
+      "learning_rate": 1.9957196689542873e-05,
+      "loss": 0.8615,
+      "step": 2487
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.676219894308753,
+      "learning_rate": 1.995712613665012e-05,
+      "loss": 0.9451,
+      "step": 2488
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.62459673949636,
+      "learning_rate": 1.995705552578383e-05,
+      "loss": 1.0015,
+      "step": 2489
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.690475035868478,
+      "learning_rate": 1.9956984856944403e-05,
+      "loss": 0.9951,
+      "step": 2490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.794852812010284,
+      "learning_rate": 1.9956914130132264e-05,
+      "loss": 0.9531,
+      "step": 2491
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8741019143791684,
+      "learning_rate": 1.9956843345347817e-05,
+      "loss": 0.9664,
+      "step": 2492
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.648098062928289,
+      "learning_rate": 1.9956772502591473e-05,
+      "loss": 0.8351,
+      "step": 2493
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.321556115190333,
+      "learning_rate": 1.9956701601863653e-05,
+      "loss": 1.0213,
+      "step": 2494
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.616001580831195,
+      "learning_rate": 1.9956630643164758e-05,
+      "loss": 1.0277,
+      "step": 2495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7677625796327363,
+      "learning_rate": 1.995655962649521e-05,
+      "loss": 0.9057,
+      "step": 2496
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8095958482620433,
+      "learning_rate": 1.995648855185542e-05,
+      "loss": 1.038,
+      "step": 2497
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8031043285613564,
+      "learning_rate": 1.9956417419245802e-05,
+      "loss": 0.884,
+      "step": 2498
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9290267575306301,
+      "learning_rate": 1.995634622866677e-05,
+      "loss": 0.8433,
+      "step": 2499
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8006420762523431,
+      "learning_rate": 1.9956274980118736e-05,
+      "loss": 0.6646,
+      "step": 2500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9837615936961863,
+      "learning_rate": 1.9956203673602116e-05,
+      "loss": 0.72,
+      "step": 2501
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.655011055660549,
+      "learning_rate": 1.995613230911733e-05,
+      "loss": 0.8504,
+      "step": 2502
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8214588854737206,
+      "learning_rate": 1.9956060886664786e-05,
+      "loss": 0.9254,
+      "step": 2503
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.815119438051416,
+      "learning_rate": 1.9955989406244906e-05,
+      "loss": 0.9606,
+      "step": 2504
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5545463602029463,
+      "learning_rate": 1.9955917867858102e-05,
+      "loss": 0.9366,
+      "step": 2505
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9289340516100613,
+      "learning_rate": 1.9955846271504792e-05,
+      "loss": 0.7773,
+      "step": 2506
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2410324068165877,
+      "learning_rate": 1.9955774617185395e-05,
+      "loss": 0.9469,
+      "step": 2507
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9208124593756668,
+      "learning_rate": 1.9955702904900325e-05,
+      "loss": 0.9644,
+      "step": 2508
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.000979098066601,
+      "learning_rate": 1.995563113465e-05,
+      "loss": 0.9239,
+      "step": 2509
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5024056641136687,
+      "learning_rate": 1.9955559306434843e-05,
+      "loss": 0.8361,
+      "step": 2510
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8631714020869081,
+      "learning_rate": 1.995548742025526e-05,
+      "loss": 1.02,
+      "step": 2511
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.568300828826216,
+      "learning_rate": 1.9955415476111683e-05,
+      "loss": 0.7896,
+      "step": 2512
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1193501971483752,
+      "learning_rate": 1.9955343474004526e-05,
+      "loss": 0.7971,
+      "step": 2513
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6473680195237679,
+      "learning_rate": 1.9955271413934203e-05,
+      "loss": 0.9787,
+      "step": 2514
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4827763585970373,
+      "learning_rate": 1.9955199295901144e-05,
+      "loss": 0.8803,
+      "step": 2515
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.676096120735952,
+      "learning_rate": 1.9955127119905756e-05,
+      "loss": 1.0552,
+      "step": 2516
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9344752927955289,
+      "learning_rate": 1.995505488594847e-05,
+      "loss": 0.9215,
+      "step": 2517
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.846263759763347,
+      "learning_rate": 1.9954982594029703e-05,
+      "loss": 0.8371,
+      "step": 2518
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8165480098355012,
+      "learning_rate": 1.9954910244149873e-05,
+      "loss": 0.7014,
+      "step": 2519
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7087181918748258,
+      "learning_rate": 1.9954837836309407e-05,
+      "loss": 0.9269,
+      "step": 2520
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6539985404690796,
+      "learning_rate": 1.9954765370508723e-05,
+      "loss": 0.8822,
+      "step": 2521
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5320976478816468,
+      "learning_rate": 1.995469284674824e-05,
+      "loss": 0.8673,
+      "step": 2522
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7330035958938232,
+      "learning_rate": 1.995462026502839e-05,
+      "loss": 0.9327,
+      "step": 2523
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5308122136223248,
+      "learning_rate": 1.9954547625349584e-05,
+      "loss": 0.8287,
+      "step": 2524
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7035546342243177,
+      "learning_rate": 1.9954474927712252e-05,
+      "loss": 0.8934,
+      "step": 2525
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5590442897219998,
+      "learning_rate": 1.9954402172116816e-05,
+      "loss": 0.89,
+      "step": 2526
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3462832213834144,
+      "learning_rate": 1.99543293585637e-05,
+      "loss": 0.9116,
+      "step": 2527
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.725006159327294,
+      "learning_rate": 1.9954256487053326e-05,
+      "loss": 0.9788,
+      "step": 2528
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8004672017994197,
+      "learning_rate": 1.995418355758612e-05,
+      "loss": 0.964,
+      "step": 2529
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7429612957406226,
+      "learning_rate": 1.99541105701625e-05,
+      "loss": 0.8129,
+      "step": 2530
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5947141958387623,
+      "learning_rate": 1.99540375247829e-05,
+      "loss": 0.9492,
+      "step": 2531
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7264855065455984,
+      "learning_rate": 1.9953964421447745e-05,
+      "loss": 0.9498,
+      "step": 2532
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1126133401270293,
+      "learning_rate": 1.995389126015746e-05,
+      "loss": 0.9508,
+      "step": 2533
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6964779392411204,
+      "learning_rate": 1.995381804091246e-05,
+      "loss": 1.0164,
+      "step": 2534
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5960920781909511,
+      "learning_rate": 1.9953744763713188e-05,
+      "loss": 0.9679,
+      "step": 2535
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.083560812367723,
+      "learning_rate": 1.995367142856006e-05,
+      "loss": 0.9119,
+      "step": 2536
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9665290356030343,
+      "learning_rate": 1.9953598035453505e-05,
+      "loss": 0.8695,
+      "step": 2537
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9381364814439421,
+      "learning_rate": 1.995352458439395e-05,
+      "loss": 0.8841,
+      "step": 2538
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6665335890638235,
+      "learning_rate": 1.9953451075381826e-05,
+      "loss": 0.9625,
+      "step": 2539
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9157172579496504,
+      "learning_rate": 1.995337750841756e-05,
+      "loss": 1.046,
+      "step": 2540
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.983066115657737,
+      "learning_rate": 1.9953303883501577e-05,
+      "loss": 0.9731,
+      "step": 2541
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4867973236564742,
+      "learning_rate": 1.9953230200634305e-05,
+      "loss": 0.9016,
+      "step": 2542
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3035927684295605,
+      "learning_rate": 1.995315645981618e-05,
+      "loss": 0.9778,
+      "step": 2543
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6958872112743149,
+      "learning_rate": 1.9953082661047624e-05,
+      "loss": 0.9008,
+      "step": 2544
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1657335254179535,
+      "learning_rate": 1.995300880432907e-05,
+      "loss": 0.8824,
+      "step": 2545
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8175346632572504,
+      "learning_rate": 1.9952934889660947e-05,
+      "loss": 0.8199,
+      "step": 2546
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7140116919013881,
+      "learning_rate": 1.9952860917043687e-05,
+      "loss": 1.0338,
+      "step": 2547
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.195309367757774,
+      "learning_rate": 1.9952786886477718e-05,
+      "loss": 0.9933,
+      "step": 2548
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.898589210601413,
+      "learning_rate": 1.9952712797963477e-05,
+      "loss": 1.0261,
+      "step": 2549
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6472070037329614,
+      "learning_rate": 1.995263865150139e-05,
+      "loss": 1.0126,
+      "step": 2550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9926469054173404,
+      "learning_rate": 1.995256444709189e-05,
+      "loss": 0.9981,
+      "step": 2551
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.126338254566385,
+      "learning_rate": 1.9952490184735404e-05,
+      "loss": 0.9897,
+      "step": 2552
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.642539563046559,
+      "learning_rate": 1.9952415864432373e-05,
+      "loss": 1.0238,
+      "step": 2553
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7247687914599235,
+      "learning_rate": 1.9952341486183228e-05,
+      "loss": 0.9747,
+      "step": 2554
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.754659171739046,
+      "learning_rate": 1.9952267049988393e-05,
+      "loss": 0.8921,
+      "step": 2555
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7772217886365823,
+      "learning_rate": 1.9952192555848312e-05,
+      "loss": 0.6806,
+      "step": 2556
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9711922862436375,
+      "learning_rate": 1.995211800376342e-05,
+      "loss": 0.9499,
+      "step": 2557
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5788084163635883,
+      "learning_rate": 1.995204339373414e-05,
+      "loss": 0.9264,
+      "step": 2558
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5881412314284091,
+      "learning_rate": 1.9951968725760916e-05,
+      "loss": 0.9026,
+      "step": 2559
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4969466829684537,
+      "learning_rate": 1.9951893999844178e-05,
+      "loss": 0.933,
+      "step": 2560
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1539246639590792,
+      "learning_rate": 1.995181921598436e-05,
+      "loss": 0.9453,
+      "step": 2561
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.858391766356398,
+      "learning_rate": 1.9951744374181905e-05,
+      "loss": 0.9389,
+      "step": 2562
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5432937952700814,
+      "learning_rate": 1.9951669474437244e-05,
+      "loss": 0.9287,
+      "step": 2563
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7678299777956268,
+      "learning_rate": 1.9951594516750807e-05,
+      "loss": 1.1036,
+      "step": 2564
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.084337267658584,
+      "learning_rate": 1.995151950112304e-05,
+      "loss": 0.92,
+      "step": 2565
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.642903269289873,
+      "learning_rate": 1.9951444427554375e-05,
+      "loss": 0.9012,
+      "step": 2566
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6679725085257735,
+      "learning_rate": 1.9951369296045247e-05,
+      "loss": 0.9276,
+      "step": 2567
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2476553788873743,
+      "learning_rate": 1.9951294106596098e-05,
+      "loss": 0.9598,
+      "step": 2568
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.598768354379379,
+      "learning_rate": 1.9951218859207368e-05,
+      "loss": 0.9135,
+      "step": 2569
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7301086929525138,
+      "learning_rate": 1.995114355387949e-05,
+      "loss": 0.9295,
+      "step": 2570
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7242347174243469,
+      "learning_rate": 1.99510681906129e-05,
+      "loss": 0.8875,
+      "step": 2571
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.523205302249424,
+      "learning_rate": 1.9950992769408043e-05,
+      "loss": 1.0147,
+      "step": 2572
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8165914786208837,
+      "learning_rate": 1.9950917290265356e-05,
+      "loss": 0.9804,
+      "step": 2573
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7627987204049154,
+      "learning_rate": 1.9950841753185276e-05,
+      "loss": 0.9351,
+      "step": 2574
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2425320469535213,
+      "learning_rate": 1.995076615816825e-05,
+      "loss": 0.8689,
+      "step": 2575
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8169921667053277,
+      "learning_rate": 1.9950690505214707e-05,
+      "loss": 0.7811,
+      "step": 2576
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6512653560812374,
+      "learning_rate": 1.9950614794325096e-05,
+      "loss": 0.8547,
+      "step": 2577
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.965706494698456,
+      "learning_rate": 1.9950539025499857e-05,
+      "loss": 0.8211,
+      "step": 2578
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.623416609075085,
+      "learning_rate": 1.995046319873943e-05,
+      "loss": 0.9083,
+      "step": 2579
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.667702566993669,
+      "learning_rate": 1.9950387314044254e-05,
+      "loss": 1.0729,
+      "step": 2580
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7566693514841134,
+      "learning_rate": 1.995031137141477e-05,
+      "loss": 0.8816,
+      "step": 2581
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4493353445866468,
+      "learning_rate": 1.9950235370851428e-05,
+      "loss": 0.9607,
+      "step": 2582
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1675355959404565,
+      "learning_rate": 1.9950159312354664e-05,
+      "loss": 1.0584,
+      "step": 2583
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5254975902369374,
+      "learning_rate": 1.9950083195924922e-05,
+      "loss": 0.9042,
+      "step": 2584
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1038718400678453,
+      "learning_rate": 1.9950007021562643e-05,
+      "loss": 1.0403,
+      "step": 2585
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.629393979289901,
+      "learning_rate": 1.9949930789268274e-05,
+      "loss": 0.9577,
+      "step": 2586
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.958243738412124,
+      "learning_rate": 1.994985449904226e-05,
+      "loss": 0.9488,
+      "step": 2587
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.847150552011069,
+      "learning_rate": 1.994977815088504e-05,
+      "loss": 0.9108,
+      "step": 2588
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7390326854937497,
+      "learning_rate": 1.994970174479707e-05,
+      "loss": 1.0094,
+      "step": 2589
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6182537361206193,
+      "learning_rate": 1.994962528077878e-05,
+      "loss": 1.0125,
+      "step": 2590
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6002773924648468,
+      "learning_rate": 1.9949548758830618e-05,
+      "loss": 0.9383,
+      "step": 2591
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7507349222593867,
+      "learning_rate": 1.9949472178953036e-05,
+      "loss": 0.9919,
+      "step": 2592
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8022854575452565,
+      "learning_rate": 1.9949395541146475e-05,
+      "loss": 0.9388,
+      "step": 2593
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4798887781535197,
+      "learning_rate": 1.994931884541139e-05,
+      "loss": 1.0187,
+      "step": 2594
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5348380203881256,
+      "learning_rate": 1.9949242091748217e-05,
+      "loss": 0.8854,
+      "step": 2595
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.824244492130809,
+      "learning_rate": 1.9949165280157404e-05,
+      "loss": 0.7794,
+      "step": 2596
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9045040098963644,
+      "learning_rate": 1.9949088410639407e-05,
+      "loss": 0.8716,
+      "step": 2597
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7104348951647101,
+      "learning_rate": 1.9949011483194663e-05,
+      "loss": 0.922,
+      "step": 2598
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8356511321654096,
+      "learning_rate": 1.9948934497823626e-05,
+      "loss": 1.011,
+      "step": 2599
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.81872265984676,
+      "learning_rate": 1.994885745452674e-05,
+      "loss": 0.8109,
+      "step": 2600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6884222507635978,
+      "learning_rate": 1.994878035330446e-05,
+      "loss": 0.8234,
+      "step": 2601
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0971093867383157,
+      "learning_rate": 1.9948703194157227e-05,
+      "loss": 0.9084,
+      "step": 2602
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4992697275013092,
+      "learning_rate": 1.9948625977085497e-05,
+      "loss": 0.9007,
+      "step": 2603
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6755346047795279,
+      "learning_rate": 1.9948548702089715e-05,
+      "loss": 0.8155,
+      "step": 2604
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7236229259691018,
+      "learning_rate": 1.9948471369170336e-05,
+      "loss": 1.0363,
+      "step": 2605
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8198663425777524,
+      "learning_rate": 1.9948393978327802e-05,
+      "loss": 0.811,
+      "step": 2606
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7444452166343616,
+      "learning_rate": 1.9948316529562573e-05,
+      "loss": 1.0373,
+      "step": 2607
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.874615667206738,
+      "learning_rate": 1.9948239022875094e-05,
+      "loss": 0.8936,
+      "step": 2608
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8417580268364142,
+      "learning_rate": 1.9948161458265817e-05,
+      "loss": 0.831,
+      "step": 2609
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7422768921782004,
+      "learning_rate": 1.9948083835735198e-05,
+      "loss": 0.9183,
+      "step": 2610
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.015320478496008,
+      "learning_rate": 1.994800615528368e-05,
+      "loss": 0.9734,
+      "step": 2611
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7724389825103521,
+      "learning_rate": 1.9947928416911723e-05,
+      "loss": 0.892,
+      "step": 2612
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6586373544140651,
+      "learning_rate": 1.994785062061978e-05,
+      "loss": 0.9716,
+      "step": 2613
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6819704795749209,
+      "learning_rate": 1.9947772766408298e-05,
+      "loss": 0.6877,
+      "step": 2614
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8847926765746401,
+      "learning_rate": 1.9947694854277734e-05,
+      "loss": 0.9426,
+      "step": 2615
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0424801180472403,
+      "learning_rate": 1.9947616884228545e-05,
+      "loss": 0.7642,
+      "step": 2616
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.84886812415509,
+      "learning_rate": 1.9947538856261177e-05,
+      "loss": 0.9555,
+      "step": 2617
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.074546448649486,
+      "learning_rate": 1.994746077037609e-05,
+      "loss": 0.9534,
+      "step": 2618
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.641505860410998,
+      "learning_rate": 1.9947382626573736e-05,
+      "loss": 0.9797,
+      "step": 2619
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6053176494475687,
+      "learning_rate": 1.994730442485457e-05,
+      "loss": 0.8532,
+      "step": 2620
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8397571981602066,
+      "learning_rate": 1.994722616521905e-05,
+      "loss": 1.0034,
+      "step": 2621
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7848457300941794,
+      "learning_rate": 1.994714784766763e-05,
+      "loss": 0.7016,
+      "step": 2622
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8629634215799449,
+      "learning_rate": 1.9947069472200766e-05,
+      "loss": 0.8786,
+      "step": 2623
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7667516179778379,
+      "learning_rate": 1.9946991038818916e-05,
+      "loss": 0.7314,
+      "step": 2624
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.768337722384772,
+      "learning_rate": 1.994691254752253e-05,
+      "loss": 1.0842,
+      "step": 2625
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7161105828600474,
+      "learning_rate": 1.9946833998312077e-05,
+      "loss": 0.9735,
+      "step": 2626
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.848691435241838,
+      "learning_rate": 1.9946755391188e-05,
+      "loss": 0.9725,
+      "step": 2627
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7386524908206902,
+      "learning_rate": 1.994667672615077e-05,
+      "loss": 0.848,
+      "step": 2628
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5627742804505391,
+      "learning_rate": 1.9946598003200837e-05,
+      "loss": 0.9348,
+      "step": 2629
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8061179242614764,
+      "learning_rate": 1.994651922233866e-05,
+      "loss": 1.0806,
+      "step": 2630
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6705792869222245,
+      "learning_rate": 1.9946440383564703e-05,
+      "loss": 1.0072,
+      "step": 2631
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6928900362992705,
+      "learning_rate": 1.9946361486879416e-05,
+      "loss": 0.9597,
+      "step": 2632
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7922431030841706,
+      "learning_rate": 1.9946282532283266e-05,
+      "loss": 1.0132,
+      "step": 2633
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7427994934219115,
+      "learning_rate": 1.994620351977671e-05,
+      "loss": 0.9005,
+      "step": 2634
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8672322043668397,
+      "learning_rate": 1.9946124449360205e-05,
+      "loss": 0.9521,
+      "step": 2635
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9083867452643155,
+      "learning_rate": 1.9946045321034223e-05,
+      "loss": 0.9362,
+      "step": 2636
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8277793546789798,
+      "learning_rate": 1.9945966134799207e-05,
+      "loss": 0.8925,
+      "step": 2637
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6005595705008826,
+      "learning_rate": 1.9945886890655632e-05,
+      "loss": 0.7314,
+      "step": 2638
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.276783045881591,
+      "learning_rate": 1.9945807588603954e-05,
+      "loss": 1.0035,
+      "step": 2639
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8812889416956966,
+      "learning_rate": 1.9945728228644636e-05,
+      "loss": 0.9718,
+      "step": 2640
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9838986319707197,
+      "learning_rate": 1.994564881077814e-05,
+      "loss": 0.8646,
+      "step": 2641
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7295816665152364,
+      "learning_rate": 1.9945569335004925e-05,
+      "loss": 0.877,
+      "step": 2642
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6227346608022235,
+      "learning_rate": 1.994548980132546e-05,
+      "loss": 0.968,
+      "step": 2643
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.519536749916708,
+      "learning_rate": 1.9945410209740202e-05,
+      "loss": 0.9131,
+      "step": 2644
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5692307413641322,
+      "learning_rate": 1.994533056024962e-05,
+      "loss": 0.8507,
+      "step": 2645
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7640793735600044,
+      "learning_rate": 1.994525085285417e-05,
+      "loss": 0.9626,
+      "step": 2646
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6534559302480853,
+      "learning_rate": 1.9945171087554324e-05,
+      "loss": 0.9435,
+      "step": 2647
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8195256175385355,
+      "learning_rate": 1.9945091264350545e-05,
+      "loss": 0.8103,
+      "step": 2648
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5474633633905903,
+      "learning_rate": 1.9945011383243296e-05,
+      "loss": 0.9813,
+      "step": 2649
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6779365429886939,
+      "learning_rate": 1.994493144423304e-05,
+      "loss": 0.9092,
+      "step": 2650
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.040400933915207,
+      "learning_rate": 1.9944851447320243e-05,
+      "loss": 1.0283,
+      "step": 2651
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8825160669551024,
+      "learning_rate": 1.9944771392505374e-05,
+      "loss": 0.9375,
+      "step": 2652
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.75529590539201,
+      "learning_rate": 1.99446912797889e-05,
+      "loss": 1.1481,
+      "step": 2653
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6399428867615538,
+      "learning_rate": 1.9944611109171284e-05,
+      "loss": 0.8886,
+      "step": 2654
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2303786563470527,
+      "learning_rate": 1.9944530880652993e-05,
+      "loss": 0.9758,
+      "step": 2655
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2252908704238212,
+      "learning_rate": 1.9944450594234492e-05,
+      "loss": 0.8831,
+      "step": 2656
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.704217934287118,
+      "learning_rate": 1.9944370249916254e-05,
+      "loss": 0.9849,
+      "step": 2657
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7161298589240017,
+      "learning_rate": 1.9944289847698744e-05,
+      "loss": 0.9053,
+      "step": 2658
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8883478126492572,
+      "learning_rate": 1.994420938758243e-05,
+      "loss": 0.7982,
+      "step": 2659
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.104055851284933,
+      "learning_rate": 1.9944128869567782e-05,
+      "loss": 0.913,
+      "step": 2660
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5730120382052322,
+      "learning_rate": 1.9944048293655266e-05,
+      "loss": 0.9766,
+      "step": 2661
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.007234025100614,
+      "learning_rate": 1.994396765984535e-05,
+      "loss": 1.0343,
+      "step": 2662
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7448331199629143,
+      "learning_rate": 1.9943886968138512e-05,
+      "loss": 0.8916,
+      "step": 2663
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3800736701458542,
+      "learning_rate": 1.9943806218535212e-05,
+      "loss": 0.8325,
+      "step": 2664
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.727125863304165,
+      "learning_rate": 1.9943725411035926e-05,
+      "loss": 1.0177,
+      "step": 2665
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6282572514335796,
+      "learning_rate": 1.994364454564112e-05,
+      "loss": 0.9465,
+      "step": 2666
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9525760182487142,
+      "learning_rate": 1.9943563622351267e-05,
+      "loss": 0.8268,
+      "step": 2667
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.08126845123687,
+      "learning_rate": 1.994348264116684e-05,
+      "loss": 0.8244,
+      "step": 2668
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1727430930012797,
+      "learning_rate": 1.994340160208831e-05,
+      "loss": 0.9991,
+      "step": 2669
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6848896929075285,
+      "learning_rate": 1.994332050511615e-05,
+      "loss": 0.981,
+      "step": 2670
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8089793698190209,
+      "learning_rate": 1.9943239350250823e-05,
+      "loss": 0.6884,
+      "step": 2671
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8474584510388041,
+      "learning_rate": 1.9943158137492815e-05,
+      "loss": 1.0206,
+      "step": 2672
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.875520362333818,
+      "learning_rate": 1.994307686684259e-05,
+      "loss": 0.9383,
+      "step": 2673
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7055425257113999,
+      "learning_rate": 1.9942995538300624e-05,
+      "loss": 0.8459,
+      "step": 2674
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8230619367620366,
+      "learning_rate": 1.994291415186739e-05,
+      "loss": 0.902,
+      "step": 2675
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8183131421591627,
+      "learning_rate": 1.994283270754336e-05,
+      "loss": 1.0158,
+      "step": 2676
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7877778235326237,
+      "learning_rate": 1.994275120532901e-05,
+      "loss": 0.7357,
+      "step": 2677
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.883137073879971,
+      "learning_rate": 1.994266964522482e-05,
+      "loss": 0.8288,
+      "step": 2678
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.795149056175009,
+      "learning_rate": 1.9942588027231254e-05,
+      "loss": 0.6893,
+      "step": 2679
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.240123974630073,
+      "learning_rate": 1.9942506351348796e-05,
+      "loss": 0.9888,
+      "step": 2680
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.596194764790758,
+      "learning_rate": 1.9942424617577915e-05,
+      "loss": 0.9147,
+      "step": 2681
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6902727925126588,
+      "learning_rate": 1.994234282591909e-05,
+      "loss": 0.9387,
+      "step": 2682
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6811568314907614,
+      "learning_rate": 1.9942260976372802e-05,
+      "loss": 0.8271,
+      "step": 2683
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6867939561790284,
+      "learning_rate": 1.994217906893952e-05,
+      "loss": 0.9044,
+      "step": 2684
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8795059496399413,
+      "learning_rate": 1.9942097103619728e-05,
+      "loss": 0.769,
+      "step": 2685
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5376038327097592,
+      "learning_rate": 1.9942015080413894e-05,
+      "loss": 0.8069,
+      "step": 2686
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7765710493116285,
+      "learning_rate": 1.99419329993225e-05,
+      "loss": 0.8446,
+      "step": 2687
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7543832629001437,
+      "learning_rate": 1.9941850860346024e-05,
+      "loss": 0.8479,
+      "step": 2688
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8437190419463092,
+      "learning_rate": 1.9941768663484948e-05,
+      "loss": 0.935,
+      "step": 2689
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4872195229629845,
+      "learning_rate": 1.9941686408739748e-05,
+      "loss": 1.0152,
+      "step": 2690
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5246904446496736,
+      "learning_rate": 1.99416040961109e-05,
+      "loss": 0.9336,
+      "step": 2691
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5129978890639855,
+      "learning_rate": 1.9941521725598888e-05,
+      "loss": 0.8531,
+      "step": 2692
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6393446048299452,
+      "learning_rate": 1.9941439297204185e-05,
+      "loss": 0.9247,
+      "step": 2693
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7778941834839888,
+      "learning_rate": 1.994135681092728e-05,
+      "loss": 0.8324,
+      "step": 2694
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5447327120728234,
+      "learning_rate": 1.9941274266768643e-05,
+      "loss": 0.8565,
+      "step": 2695
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7016774078675452,
+      "learning_rate": 1.9941191664728763e-05,
+      "loss": 0.9221,
+      "step": 2696
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.753095914067305,
+      "learning_rate": 1.9941109004808117e-05,
+      "loss": 0.9006,
+      "step": 2697
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9654662601764652,
+      "learning_rate": 1.9941026287007186e-05,
+      "loss": 0.9118,
+      "step": 2698
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6605814956969585,
+      "learning_rate": 1.9940943511326453e-05,
+      "loss": 0.8956,
+      "step": 2699
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.688588730940849,
+      "learning_rate": 1.9940860677766404e-05,
+      "loss": 1.0105,
+      "step": 2700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7929454985025923,
+      "learning_rate": 1.9940777786327513e-05,
+      "loss": 1.0268,
+      "step": 2701
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4384008408320175,
+      "learning_rate": 1.9940694837010265e-05,
+      "loss": 0.9145,
+      "step": 2702
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.708853110023607,
+      "learning_rate": 1.994061182981515e-05,
+      "loss": 0.8794,
+      "step": 2703
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6483510161234276,
+      "learning_rate": 1.994052876474264e-05,
+      "loss": 0.9712,
+      "step": 2704
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.846116159521026,
+      "learning_rate": 1.9940445641793228e-05,
+      "loss": 0.75,
+      "step": 2705
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.619254932643776,
+      "learning_rate": 1.994036246096739e-05,
+      "loss": 0.8731,
+      "step": 2706
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4969534585545046,
+      "learning_rate": 1.9940279222265618e-05,
+      "loss": 0.8508,
+      "step": 2707
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.149871325103597,
+      "learning_rate": 1.9940195925688395e-05,
+      "loss": 0.7917,
+      "step": 2708
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8750291146561078,
+      "learning_rate": 1.9940112571236203e-05,
+      "loss": 0.8543,
+      "step": 2709
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.475093834071276,
+      "learning_rate": 1.9940029158909525e-05,
+      "loss": 0.9216,
+      "step": 2710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0171169083913143,
+      "learning_rate": 1.9939945688708854e-05,
+      "loss": 0.876,
+      "step": 2711
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.598850758666792,
+      "learning_rate": 1.9939862160634674e-05,
+      "loss": 0.9301,
+      "step": 2712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.219767059881053,
+      "learning_rate": 1.9939778574687465e-05,
+      "loss": 0.8906,
+      "step": 2713
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7767975840021714,
+      "learning_rate": 1.993969493086772e-05,
+      "loss": 0.7755,
+      "step": 2714
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8516030678209863,
+      "learning_rate": 1.9939611229175926e-05,
+      "loss": 0.7779,
+      "step": 2715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8724549827181393,
+      "learning_rate": 1.9939527469612567e-05,
+      "loss": 0.771,
+      "step": 2716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6744079732720212,
+      "learning_rate": 1.9939443652178136e-05,
+      "loss": 0.9408,
+      "step": 2717
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8691727307319168,
+      "learning_rate": 1.9939359776873112e-05,
+      "loss": 0.9927,
+      "step": 2718
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.781160454008328,
+      "learning_rate": 1.9939275843697988e-05,
+      "loss": 0.8926,
+      "step": 2719
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5219509087940335,
+      "learning_rate": 1.9939191852653257e-05,
+      "loss": 1.0436,
+      "step": 2720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6483350802546028,
+      "learning_rate": 1.9939107803739402e-05,
+      "loss": 0.9344,
+      "step": 2721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7581735214563892,
+      "learning_rate": 1.9939023696956916e-05,
+      "loss": 0.9727,
+      "step": 2722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.415820411449956,
+      "learning_rate": 1.993893953230629e-05,
+      "loss": 0.9251,
+      "step": 2723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5961999538869003,
+      "learning_rate": 1.993885530978801e-05,
+      "loss": 0.9013,
+      "step": 2724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.844633263505003,
+      "learning_rate": 1.9938771029402564e-05,
+      "loss": 0.9263,
+      "step": 2725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7919526690873222,
+      "learning_rate": 1.993868669115045e-05,
+      "loss": 0.9935,
+      "step": 2726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7950873953850401,
+      "learning_rate": 1.993860229503215e-05,
+      "loss": 0.8256,
+      "step": 2727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8711837571016479,
+      "learning_rate": 1.9938517841048167e-05,
+      "loss": 0.9753,
+      "step": 2728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7802617185905311,
+      "learning_rate": 1.9938433329198985e-05,
+      "loss": 0.6686,
+      "step": 2729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.829562995183999,
+      "learning_rate": 1.99383487594851e-05,
+      "loss": 0.932,
+      "step": 2730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6361013871587529,
+      "learning_rate": 1.9938264131906998e-05,
+      "loss": 0.944,
+      "step": 2731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9229081354360382,
+      "learning_rate": 1.993817944646518e-05,
+      "loss": 0.9277,
+      "step": 2732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8258144460839872,
+      "learning_rate": 1.993809470316013e-05,
+      "loss": 0.7702,
+      "step": 2733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.093504959348146,
+      "learning_rate": 1.993800990199235e-05,
+      "loss": 0.9637,
+      "step": 2734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6902879661273988,
+      "learning_rate": 1.993792504296233e-05,
+      "loss": 0.8422,
+      "step": 2735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5925977330501309,
+      "learning_rate": 1.993784012607056e-05,
+      "loss": 0.9057,
+      "step": 2736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8032040351979977,
+      "learning_rate": 1.993775515131754e-05,
+      "loss": 0.7973,
+      "step": 2737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8799495579432275,
+      "learning_rate": 1.9937670118703768e-05,
+      "loss": 0.8493,
+      "step": 2738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7647193428248695,
+      "learning_rate": 1.993758502822973e-05,
+      "loss": 0.73,
+      "step": 2739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.650763716211207,
+      "learning_rate": 1.993749987989593e-05,
+      "loss": 1.0095,
+      "step": 2740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4513633875693808,
+      "learning_rate": 1.9937414673702852e-05,
+      "loss": 0.9317,
+      "step": 2741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4757210494424517,
+      "learning_rate": 1.9937329409651007e-05,
+      "loss": 0.9157,
+      "step": 2742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7042069494757277,
+      "learning_rate": 1.993724408774088e-05,
+      "loss": 0.9668,
+      "step": 2743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9740687161848947,
+      "learning_rate": 1.9937158707972975e-05,
+      "loss": 0.7857,
+      "step": 2744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7624592922068276,
+      "learning_rate": 1.9937073270347782e-05,
+      "loss": 0.825,
+      "step": 2745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.776461429642733,
+      "learning_rate": 1.9936987774865805e-05,
+      "loss": 0.9202,
+      "step": 2746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.775960374432742,
+      "learning_rate": 1.9936902221527536e-05,
+      "loss": 0.6632,
+      "step": 2747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5422471362954004,
+      "learning_rate": 1.993681661033348e-05,
+      "loss": 0.8166,
+      "step": 2748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8133160884299657,
+      "learning_rate": 1.993673094128413e-05,
+      "loss": 0.8752,
+      "step": 2749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.05228855574191,
+      "learning_rate": 1.9936645214379988e-05,
+      "loss": 0.921,
+      "step": 2750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7728242595603851,
+      "learning_rate": 1.993655942962155e-05,
+      "loss": 0.7031,
+      "step": 2751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7798743465364673,
+      "learning_rate": 1.993647358700932e-05,
+      "loss": 0.8891,
+      "step": 2752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7223935185732921,
+      "learning_rate": 1.993638768654379e-05,
+      "loss": 0.931,
+      "step": 2753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7451096582030154,
+      "learning_rate": 1.993630172822547e-05,
+      "loss": 0.9951,
+      "step": 2754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.879535625551704,
+      "learning_rate": 1.9936215712054854e-05,
+      "loss": 0.8759,
+      "step": 2755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9492967285907914,
+      "learning_rate": 1.9936129638032443e-05,
+      "loss": 0.8917,
+      "step": 2756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9756235396883974,
+      "learning_rate": 1.9936043506158742e-05,
+      "loss": 0.8847,
+      "step": 2757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.00737577540969,
+      "learning_rate": 1.9935957316434246e-05,
+      "loss": 0.7965,
+      "step": 2758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6429540437557097,
+      "learning_rate": 1.9935871068859463e-05,
+      "loss": 1.0014,
+      "step": 2759
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.773715443413745,
+      "learning_rate": 1.9935784763434894e-05,
+      "loss": 1.068,
+      "step": 2760
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8469321855518577,
+      "learning_rate": 1.993569840016104e-05,
+      "loss": 0.722,
+      "step": 2761
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.686378402515632,
+      "learning_rate": 1.9935611979038403e-05,
+      "loss": 0.8402,
+      "step": 2762
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.672133374456443,
+      "learning_rate": 1.993552550006749e-05,
+      "loss": 0.8137,
+      "step": 2763
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7294084150133298,
+      "learning_rate": 1.99354389632488e-05,
+      "loss": 0.9278,
+      "step": 2764
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.599977356640584,
+      "learning_rate": 1.9935352368582837e-05,
+      "loss": 0.8848,
+      "step": 2765
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6326546921250231,
+      "learning_rate": 1.9935265716070113e-05,
+      "loss": 0.854,
+      "step": 2766
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5126042371318322,
+      "learning_rate": 1.9935179005711124e-05,
+      "loss": 1.0394,
+      "step": 2767
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6154890077018762,
+      "learning_rate": 1.993509223750637e-05,
+      "loss": 0.9443,
+      "step": 2768
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3668205876403223,
+      "learning_rate": 1.9935005411456373e-05,
+      "loss": 0.8733,
+      "step": 2769
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5820070857613266,
+      "learning_rate": 1.9934918527561627e-05,
+      "loss": 0.9481,
+      "step": 2770
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.811915838641912,
+      "learning_rate": 1.993483158582264e-05,
+      "loss": 0.7494,
+      "step": 2771
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.547142209397669,
+      "learning_rate": 1.9934744586239917e-05,
+      "loss": 1.0006,
+      "step": 2772
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0229019899124174,
+      "learning_rate": 1.9934657528813965e-05,
+      "loss": 0.8802,
+      "step": 2773
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7470935843771787,
+      "learning_rate": 1.993457041354529e-05,
+      "loss": 1.0048,
+      "step": 2774
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.427512148571456,
+      "learning_rate": 1.9934483240434406e-05,
+      "loss": 0.9391,
+      "step": 2775
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7752378428693942,
+      "learning_rate": 1.9934396009481815e-05,
+      "loss": 0.964,
+      "step": 2776
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7132247335390989,
+      "learning_rate": 1.9934308720688024e-05,
+      "loss": 0.9186,
+      "step": 2777
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9628795686927587,
+      "learning_rate": 1.9934221374053538e-05,
+      "loss": 0.8729,
+      "step": 2778
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8012666786487114,
+      "learning_rate": 1.9934133969578875e-05,
+      "loss": 0.7501,
+      "step": 2779
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6110688068780834,
+      "learning_rate": 1.9934046507264537e-05,
+      "loss": 0.9138,
+      "step": 2780
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.00780663473204,
+      "learning_rate": 1.9933958987111035e-05,
+      "loss": 0.7972,
+      "step": 2781
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6269025817300937,
+      "learning_rate": 1.993387140911888e-05,
+      "loss": 0.8953,
+      "step": 2782
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.90703490498457,
+      "learning_rate": 1.993378377328858e-05,
+      "loss": 0.855,
+      "step": 2783
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6874358874328899,
+      "learning_rate": 1.9933696079620647e-05,
+      "loss": 0.9881,
+      "step": 2784
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.304433339834436,
+      "learning_rate": 1.993360832811559e-05,
+      "loss": 0.919,
+      "step": 2785
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.362710405008101,
+      "learning_rate": 1.993352051877392e-05,
+      "loss": 0.9843,
+      "step": 2786
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.768381422709532,
+      "learning_rate": 1.9933432651596146e-05,
+      "loss": 0.8953,
+      "step": 2787
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1351827515256345,
+      "learning_rate": 1.9933344726582787e-05,
+      "loss": 0.9395,
+      "step": 2788
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7351855737386093,
+      "learning_rate": 1.9933256743734346e-05,
+      "loss": 0.8924,
+      "step": 2789
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0004630334680575,
+      "learning_rate": 1.9933168703051343e-05,
+      "loss": 0.9319,
+      "step": 2790
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.610917385936561,
+      "learning_rate": 1.9933080604534285e-05,
+      "loss": 0.8812,
+      "step": 2791
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4938396166430625,
+      "learning_rate": 1.9932992448183688e-05,
+      "loss": 0.8988,
+      "step": 2792
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.987860795760558,
+      "learning_rate": 1.9932904234000063e-05,
+      "loss": 0.9693,
+      "step": 2793
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6739093458337335,
+      "learning_rate": 1.9932815961983925e-05,
+      "loss": 0.8704,
+      "step": 2794
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8027741242591969,
+      "learning_rate": 1.993272763213579e-05,
+      "loss": 0.7195,
+      "step": 2795
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9841572782043286,
+      "learning_rate": 1.9932639244456165e-05,
+      "loss": 0.8252,
+      "step": 2796
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6761592877838718,
+      "learning_rate": 1.9932550798945575e-05,
+      "loss": 0.9328,
+      "step": 2797
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6102347252598566,
+      "learning_rate": 1.9932462295604526e-05,
+      "loss": 0.8002,
+      "step": 2798
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7347101551422364,
+      "learning_rate": 1.993237373443354e-05,
+      "loss": 0.8875,
+      "step": 2799
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2073688236650533,
+      "learning_rate": 1.9932285115433126e-05,
+      "loss": 0.991,
+      "step": 2800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6069486736724825,
+      "learning_rate": 1.9932196438603805e-05,
+      "loss": 0.9003,
+      "step": 2801
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0550734599571023,
+      "learning_rate": 1.993210770394609e-05,
+      "loss": 1.0853,
+      "step": 2802
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.973658125583869,
+      "learning_rate": 1.9932018911460503e-05,
+      "loss": 0.8685,
+      "step": 2803
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8444968523121215,
+      "learning_rate": 1.9931930061147556e-05,
+      "loss": 0.9439,
+      "step": 2804
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.872777816406285,
+      "learning_rate": 1.993184115300777e-05,
+      "loss": 1.0561,
+      "step": 2805
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.819356179067175,
+      "learning_rate": 1.993175218704166e-05,
+      "loss": 0.9634,
+      "step": 2806
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8318945424091273,
+      "learning_rate": 1.9931663163249744e-05,
+      "loss": 0.858,
+      "step": 2807
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7553743386322214,
+      "learning_rate": 1.993157408163254e-05,
+      "loss": 0.7152,
+      "step": 2808
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7534901415845516,
+      "learning_rate": 1.9931484942190567e-05,
+      "loss": 0.8519,
+      "step": 2809
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8108895904416953,
+      "learning_rate": 1.9931395744924345e-05,
+      "loss": 0.973,
+      "step": 2810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5518192615527235,
+      "learning_rate": 1.993130648983439e-05,
+      "loss": 0.97,
+      "step": 2811
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.813747663640297,
+      "learning_rate": 1.9931217176921228e-05,
+      "loss": 1.0233,
+      "step": 2812
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8046420360143054,
+      "learning_rate": 1.9931127806185376e-05,
+      "loss": 0.6838,
+      "step": 2813
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7484236200680445,
+      "learning_rate": 1.993103837762735e-05,
+      "loss": 0.7183,
+      "step": 2814
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.21533066081356,
+      "learning_rate": 1.9930948891247678e-05,
+      "loss": 0.7671,
+      "step": 2815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6585659851414625,
+      "learning_rate": 1.9930859347046875e-05,
+      "loss": 0.9798,
+      "step": 2816
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.724171497189037,
+      "learning_rate": 1.9930769745025468e-05,
+      "loss": 0.9095,
+      "step": 2817
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.200979422995338,
+      "learning_rate": 1.993068008518397e-05,
+      "loss": 1.0584,
+      "step": 2818
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5834889477931147,
+      "learning_rate": 1.993059036752291e-05,
+      "loss": 0.899,
+      "step": 2819
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7231084364001934,
+      "learning_rate": 1.9930500592042815e-05,
+      "loss": 0.975,
+      "step": 2820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7423343419654151,
+      "learning_rate": 1.9930410758744195e-05,
+      "loss": 0.865,
+      "step": 2821
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5489574226723655,
+      "learning_rate": 1.9930320867627584e-05,
+      "loss": 0.8746,
+      "step": 2822
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7135787864876253,
+      "learning_rate": 1.99302309186935e-05,
+      "loss": 0.9558,
+      "step": 2823
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7868948004936374,
+      "learning_rate": 1.9930140911942465e-05,
+      "loss": 0.9229,
+      "step": 2824
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7127812265430735,
+      "learning_rate": 1.993005084737501e-05,
+      "loss": 0.8519,
+      "step": 2825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9953384835095245,
+      "learning_rate": 1.9929960724991652e-05,
+      "loss": 0.896,
+      "step": 2826
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.783666756092239,
+      "learning_rate": 1.9929870544792918e-05,
+      "loss": 0.7544,
+      "step": 2827
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8359004244098514,
+      "learning_rate": 1.9929780306779337e-05,
+      "loss": 0.7297,
+      "step": 2828
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.615605867310219,
+      "learning_rate": 1.9929690010951427e-05,
+      "loss": 0.8458,
+      "step": 2829
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5763780520585187,
+      "learning_rate": 1.992959965730972e-05,
+      "loss": 0.902,
+      "step": 2830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.619468856043281,
+      "learning_rate": 1.9929509245854743e-05,
+      "loss": 0.8954,
+      "step": 2831
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6542181036161296,
+      "learning_rate": 1.9929418776587018e-05,
+      "loss": 0.9478,
+      "step": 2832
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7694369264087098,
+      "learning_rate": 1.9929328249507072e-05,
+      "loss": 0.9778,
+      "step": 2833
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7723840574827034,
+      "learning_rate": 1.9929237664615434e-05,
+      "loss": 0.8165,
+      "step": 2834
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6594329262751053,
+      "learning_rate": 1.992914702191263e-05,
+      "loss": 0.8859,
+      "step": 2835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.05520598683237,
+      "learning_rate": 1.9929056321399193e-05,
+      "loss": 0.9246,
+      "step": 2836
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4184185584021685,
+      "learning_rate": 1.992896556307564e-05,
+      "loss": 0.895,
+      "step": 2837
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9976637079787938,
+      "learning_rate": 1.992887474694251e-05,
+      "loss": 0.8359,
+      "step": 2838
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5675096424503758,
+      "learning_rate": 1.992878387300033e-05,
+      "loss": 0.8799,
+      "step": 2839
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7188235333376622,
+      "learning_rate": 1.992869294124962e-05,
+      "loss": 1.0308,
+      "step": 2840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.716808117117978,
+      "learning_rate": 1.9928601951690922e-05,
+      "loss": 1.0173,
+      "step": 2841
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8516154748821856,
+      "learning_rate": 1.992851090432476e-05,
+      "loss": 0.9066,
+      "step": 2842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.082826933220313,
+      "learning_rate": 1.9928419799151658e-05,
+      "loss": 0.8592,
+      "step": 2843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7737385908516864,
+      "learning_rate": 1.992832863617216e-05,
+      "loss": 0.7244,
+      "step": 2844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.567900310891296,
+      "learning_rate": 1.9928237415386785e-05,
+      "loss": 0.8605,
+      "step": 2845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.843893401478077,
+      "learning_rate": 1.992814613679607e-05,
+      "loss": 1.0477,
+      "step": 2846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8235460446649145,
+      "learning_rate": 1.992805480040055e-05,
+      "loss": 0.8961,
+      "step": 2847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6904694165073915,
+      "learning_rate": 1.9927963406200743e-05,
+      "loss": 1.081,
+      "step": 2848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.647118131297698,
+      "learning_rate": 1.9927871954197194e-05,
+      "loss": 0.9232,
+      "step": 2849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.176905248473648,
+      "learning_rate": 1.9927780444390433e-05,
+      "loss": 0.8918,
+      "step": 2850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8355648466599555,
+      "learning_rate": 1.9927688876780988e-05,
+      "loss": 0.9137,
+      "step": 2851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5928366598896357,
+      "learning_rate": 1.9927597251369398e-05,
+      "loss": 0.9097,
+      "step": 2852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6882507316847335,
+      "learning_rate": 1.992750556815619e-05,
+      "loss": 0.8558,
+      "step": 2853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.651720198284002,
+      "learning_rate": 1.9927413827141904e-05,
+      "loss": 0.8717,
+      "step": 2854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5806750831777487,
+      "learning_rate": 1.9927322028327074e-05,
+      "loss": 0.936,
+      "step": 2855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9223897496480369,
+      "learning_rate": 1.992723017171223e-05,
+      "loss": 0.8615,
+      "step": 2856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.188025075109989,
+      "learning_rate": 1.9927138257297908e-05,
+      "loss": 0.9881,
+      "step": 2857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6998500013350726,
+      "learning_rate": 1.9927046285084642e-05,
+      "loss": 0.9509,
+      "step": 2858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.684353750171474,
+      "learning_rate": 1.9926954255072976e-05,
+      "loss": 0.8485,
+      "step": 2859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7628861680124779,
+      "learning_rate": 1.9926862167263437e-05,
+      "loss": 0.864,
+      "step": 2860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8151692103508261,
+      "learning_rate": 1.9926770021656564e-05,
+      "loss": 0.9377,
+      "step": 2861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7550921485425204,
+      "learning_rate": 1.992667781825289e-05,
+      "loss": 0.8841,
+      "step": 2862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6909389992480683,
+      "learning_rate": 1.9926585557052958e-05,
+      "loss": 0.8395,
+      "step": 2863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5912803490181375,
+      "learning_rate": 1.99264932380573e-05,
+      "loss": 1.029,
+      "step": 2864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7802384174976942,
+      "learning_rate": 1.9926400861266458e-05,
+      "loss": 0.9425,
+      "step": 2865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6232334682522513,
+      "learning_rate": 1.9926308426680966e-05,
+      "loss": 0.9507,
+      "step": 2866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2663707502125856,
+      "learning_rate": 1.9926215934301366e-05,
+      "loss": 0.8963,
+      "step": 2867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.677840082033975,
+      "learning_rate": 1.992612338412819e-05,
+      "loss": 1.0072,
+      "step": 2868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.560317196403694,
+      "learning_rate": 1.9926030776161986e-05,
+      "loss": 0.9345,
+      "step": 2869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.037258813521511,
+      "learning_rate": 1.9925938110403286e-05,
+      "loss": 0.9184,
+      "step": 2870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6575133020321258,
+      "learning_rate": 1.992584538685263e-05,
+      "loss": 0.9706,
+      "step": 2871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7744599904817536,
+      "learning_rate": 1.9925752605510566e-05,
+      "loss": 0.7909,
+      "step": 2872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7466700131629707,
+      "learning_rate": 1.9925659766377622e-05,
+      "loss": 0.9049,
+      "step": 2873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6181984678093135,
+      "learning_rate": 1.9925566869454345e-05,
+      "loss": 0.8412,
+      "step": 2874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.100122525333537,
+      "learning_rate": 1.9925473914741275e-05,
+      "loss": 1.0597,
+      "step": 2875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.701194007032378,
+      "learning_rate": 1.9925380902238956e-05,
+      "loss": 0.8519,
+      "step": 2876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.132297905559653,
+      "learning_rate": 1.9925287831947922e-05,
+      "loss": 0.9366,
+      "step": 2877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0442771657175065,
+      "learning_rate": 1.9925194703868727e-05,
+      "loss": 0.918,
+      "step": 2878
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9470679127828772,
+      "learning_rate": 1.99251015180019e-05,
+      "loss": 0.7434,
+      "step": 2879
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5956726229981724,
+      "learning_rate": 1.9925008274347993e-05,
+      "loss": 0.7745,
+      "step": 2880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2678428646759476,
+      "learning_rate": 1.9924914972907547e-05,
+      "loss": 0.9165,
+      "step": 2881
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6772214685739586,
+      "learning_rate": 1.9924821613681103e-05,
+      "loss": 0.9864,
+      "step": 2882
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6738550502601748,
+      "learning_rate": 1.9924728196669205e-05,
+      "loss": 0.8779,
+      "step": 2883
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4576595663646084,
+      "learning_rate": 1.9924634721872396e-05,
+      "loss": 0.9201,
+      "step": 2884
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.716998332270218,
+      "learning_rate": 1.992454118929122e-05,
+      "loss": 0.8976,
+      "step": 2885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2969346056199758,
+      "learning_rate": 1.9924447598926227e-05,
+      "loss": 0.8773,
+      "step": 2886
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7808020894482386,
+      "learning_rate": 1.9924353950777956e-05,
+      "loss": 1.0995,
+      "step": 2887
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6548714350041729,
+      "learning_rate": 1.9924260244846954e-05,
+      "loss": 0.9982,
+      "step": 2888
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.544824808235642,
+      "learning_rate": 1.9924166481133767e-05,
+      "loss": 0.9645,
+      "step": 2889
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9736092944934702,
+      "learning_rate": 1.9924072659638944e-05,
+      "loss": 0.9296,
+      "step": 2890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8666909880522773,
+      "learning_rate": 1.9923978780363025e-05,
+      "loss": 0.7495,
+      "step": 2891
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7073761488733334,
+      "learning_rate": 1.992388484330656e-05,
+      "loss": 0.9355,
+      "step": 2892
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8331123510962564,
+      "learning_rate": 1.9923790848470098e-05,
+      "loss": 0.9404,
+      "step": 2893
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.134696394295874,
+      "learning_rate": 1.992369679585418e-05,
+      "loss": 0.8551,
+      "step": 2894
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6829084483569487,
+      "learning_rate": 1.992360268545936e-05,
+      "loss": 0.9746,
+      "step": 2895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7062471817139009,
+      "learning_rate": 1.992350851728618e-05,
+      "loss": 0.9307,
+      "step": 2896
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7821984396472783,
+      "learning_rate": 1.9923414291335198e-05,
+      "loss": 0.961,
+      "step": 2897
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8342582566656558,
+      "learning_rate": 1.992332000760695e-05,
+      "loss": 0.8385,
+      "step": 2898
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6524251066110975,
+      "learning_rate": 1.9923225666101994e-05,
+      "loss": 1.018,
+      "step": 2899
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7286313066726455,
+      "learning_rate": 1.9923131266820878e-05,
+      "loss": 0.8752,
+      "step": 2900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5513426471469,
+      "learning_rate": 1.9923036809764145e-05,
+      "loss": 0.9411,
+      "step": 2901
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8482499872196199,
+      "learning_rate": 1.9922942294932356e-05,
+      "loss": 0.9664,
+      "step": 2902
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6250573424534638,
+      "learning_rate": 1.9922847722326052e-05,
+      "loss": 1.0924,
+      "step": 2903
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6542377027501036,
+      "learning_rate": 1.992275309194579e-05,
+      "loss": 1.0262,
+      "step": 2904
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8520972742536532,
+      "learning_rate": 1.9922658403792115e-05,
+      "loss": 0.9372,
+      "step": 2905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6327528638578115,
+      "learning_rate": 1.9922563657865584e-05,
+      "loss": 0.9977,
+      "step": 2906
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8418332547117318,
+      "learning_rate": 1.9922468854166744e-05,
+      "loss": 0.8722,
+      "step": 2907
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6803488567757254,
+      "learning_rate": 1.9922373992696146e-05,
+      "loss": 0.9712,
+      "step": 2908
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6207758296608423,
+      "learning_rate": 1.992227907345435e-05,
+      "loss": 0.9658,
+      "step": 2909
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5693744095270155,
+      "learning_rate": 1.9922184096441904e-05,
+      "loss": 0.8982,
+      "step": 2910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7973354470740794,
+      "learning_rate": 1.992208906165936e-05,
+      "loss": 0.735,
+      "step": 2911
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6969833960393335,
+      "learning_rate": 1.992199396910727e-05,
+      "loss": 0.9546,
+      "step": 2912
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7437976120943564,
+      "learning_rate": 1.992189881878619e-05,
+      "loss": 0.9127,
+      "step": 2913
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8259012344469678,
+      "learning_rate": 1.9921803610696675e-05,
+      "loss": 1.0348,
+      "step": 2914
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6192548742295951,
+      "learning_rate": 1.992170834483928e-05,
+      "loss": 0.8122,
+      "step": 2915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5110517290130767,
+      "learning_rate": 1.9921613021214555e-05,
+      "loss": 0.9393,
+      "step": 2916
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.026007936384917,
+      "learning_rate": 1.992151763982306e-05,
+      "loss": 0.8585,
+      "step": 2917
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6603872542036218,
+      "learning_rate": 1.9921422200665347e-05,
+      "loss": 0.9595,
+      "step": 2918
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.663143095167143,
+      "learning_rate": 1.9921326703741976e-05,
+      "loss": 0.9443,
+      "step": 2919
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2966919892042506,
+      "learning_rate": 1.9921231149053495e-05,
+      "loss": 0.8261,
+      "step": 2920
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.692403064011326,
+      "learning_rate": 1.9921135536600465e-05,
+      "loss": 0.9304,
+      "step": 2921
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6061089214391642,
+      "learning_rate": 1.9921039866383447e-05,
+      "loss": 0.8306,
+      "step": 2922
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5585826423024836,
+      "learning_rate": 1.9920944138402993e-05,
+      "loss": 0.7234,
+      "step": 2923
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6598173737721178,
+      "learning_rate": 1.9920848352659662e-05,
+      "loss": 0.9079,
+      "step": 2924
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9211729418499846,
+      "learning_rate": 1.9920752509154008e-05,
+      "loss": 0.8925,
+      "step": 2925
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9382033545011494,
+      "learning_rate": 1.992065660788659e-05,
+      "loss": 0.8279,
+      "step": 2926
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7150551192964016,
+      "learning_rate": 1.9920560648857976e-05,
+      "loss": 0.9021,
+      "step": 2927
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.440693516270277,
+      "learning_rate": 1.9920464632068713e-05,
+      "loss": 0.9084,
+      "step": 2928
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9518411903336956,
+      "learning_rate": 1.9920368557519362e-05,
+      "loss": 0.8999,
+      "step": 2929
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7347753491188993,
+      "learning_rate": 1.992027242521049e-05,
+      "loss": 0.7946,
+      "step": 2930
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6693617128536826,
+      "learning_rate": 1.9920176235142645e-05,
+      "loss": 0.9117,
+      "step": 2931
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.836335422395093,
+      "learning_rate": 1.9920079987316398e-05,
+      "loss": 0.8668,
+      "step": 2932
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6791803700401957,
+      "learning_rate": 1.9919983681732302e-05,
+      "loss": 1.1103,
+      "step": 2933
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8003374480090761,
+      "learning_rate": 1.9919887318390923e-05,
+      "loss": 1.0023,
+      "step": 2934
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6041689465225508,
+      "learning_rate": 1.9919790897292817e-05,
+      "loss": 0.918,
+      "step": 2935
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7567952195296972,
+      "learning_rate": 1.9919694418438547e-05,
+      "loss": 0.6819,
+      "step": 2936
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6876800074988123,
+      "learning_rate": 1.9919597881828678e-05,
+      "loss": 0.9529,
+      "step": 2937
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.694609224399657,
+      "learning_rate": 1.991950128746377e-05,
+      "loss": 0.9253,
+      "step": 2938
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.509823376940806,
+      "learning_rate": 1.9919404635344383e-05,
+      "loss": 0.8571,
+      "step": 2939
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9943028121294823,
+      "learning_rate": 1.991930792547108e-05,
+      "loss": 0.9965,
+      "step": 2940
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.463619642609269,
+      "learning_rate": 1.991921115784443e-05,
+      "loss": 0.8825,
+      "step": 2941
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.742944003994143,
+      "learning_rate": 1.9919114332464994e-05,
+      "loss": 0.7424,
+      "step": 2942
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7598103446192226,
+      "learning_rate": 1.991901744933333e-05,
+      "loss": 0.8879,
+      "step": 2943
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5268607262823413,
+      "learning_rate": 1.9918920508450003e-05,
+      "loss": 0.9366,
+      "step": 2944
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.536762806946167,
+      "learning_rate": 1.9918823509815585e-05,
+      "loss": 0.9779,
+      "step": 2945
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6833249041187714,
+      "learning_rate": 1.9918726453430635e-05,
+      "loss": 0.9006,
+      "step": 2946
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4776086828273995,
+      "learning_rate": 1.9918629339295722e-05,
+      "loss": 0.9618,
+      "step": 2947
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9047090651916734,
+      "learning_rate": 1.9918532167411407e-05,
+      "loss": 0.7985,
+      "step": 2948
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8055160286834036,
+      "learning_rate": 1.991843493777826e-05,
+      "loss": 0.9842,
+      "step": 2949
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6439363646082463,
+      "learning_rate": 1.991833765039684e-05,
+      "loss": 0.9373,
+      "step": 2950
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.687970237377557,
+      "learning_rate": 1.991824030526772e-05,
+      "loss": 0.8398,
+      "step": 2951
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9582430533830024,
+      "learning_rate": 1.9918142902391463e-05,
+      "loss": 1.0051,
+      "step": 2952
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8628311082702753,
+      "learning_rate": 1.9918045441768642e-05,
+      "loss": 0.8998,
+      "step": 2953
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5504395469509262,
+      "learning_rate": 1.9917947923399818e-05,
+      "loss": 0.9779,
+      "step": 2954
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7550141429046557,
+      "learning_rate": 1.9917850347285563e-05,
+      "loss": 0.9722,
+      "step": 2955
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6932913830738707,
+      "learning_rate": 1.9917752713426442e-05,
+      "loss": 0.8895,
+      "step": 2956
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8284715690465231,
+      "learning_rate": 1.9917655021823022e-05,
+      "loss": 0.7788,
+      "step": 2957
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6692764121447987,
+      "learning_rate": 1.991755727247588e-05,
+      "loss": 0.9361,
+      "step": 2958
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7456833498927427,
+      "learning_rate": 1.9917459465385578e-05,
+      "loss": 0.7659,
+      "step": 2959
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7940618403766911,
+      "learning_rate": 1.9917361600552685e-05,
+      "loss": 0.7062,
+      "step": 2960
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8194986263029382,
+      "learning_rate": 1.9917263677977773e-05,
+      "loss": 1.0282,
+      "step": 2961
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5579424783377644,
+      "learning_rate": 1.9917165697661415e-05,
+      "loss": 0.9102,
+      "step": 2962
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7175099754398981,
+      "learning_rate": 1.9917067659604176e-05,
+      "loss": 1.0654,
+      "step": 2963
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7873311628071522,
+      "learning_rate": 1.991696956380663e-05,
+      "loss": 0.9128,
+      "step": 2964
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4215004531301456,
+      "learning_rate": 1.991687141026935e-05,
+      "loss": 1.0685,
+      "step": 2965
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1052738261030046,
+      "learning_rate": 1.99167731989929e-05,
+      "loss": 0.9596,
+      "step": 2966
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7399069900438775,
+      "learning_rate": 1.991667492997786e-05,
+      "loss": 0.9593,
+      "step": 2967
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5406359454224317,
+      "learning_rate": 1.99165766032248e-05,
+      "loss": 0.9859,
+      "step": 2968
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.605561784281452,
+      "learning_rate": 1.991647821873429e-05,
+      "loss": 0.9319,
+      "step": 2969
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7748217780371696,
+      "learning_rate": 1.9916379776506908e-05,
+      "loss": 0.935,
+      "step": 2970
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.611417495875318,
+      "learning_rate": 1.9916281276543215e-05,
+      "loss": 0.8902,
+      "step": 2971
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7961774003327133,
+      "learning_rate": 1.99161827188438e-05,
+      "loss": 0.9365,
+      "step": 2972
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.713961598639948,
+      "learning_rate": 1.9916084103409228e-05,
+      "loss": 0.8544,
+      "step": 2973
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1173287667628498,
+      "learning_rate": 1.9915985430240075e-05,
+      "loss": 0.8589,
+      "step": 2974
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5846337030485285,
+      "learning_rate": 1.9915886699336915e-05,
+      "loss": 0.8964,
+      "step": 2975
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7593502972803776,
+      "learning_rate": 1.9915787910700322e-05,
+      "loss": 1.0122,
+      "step": 2976
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7727429516830371,
+      "learning_rate": 1.9915689064330875e-05,
+      "loss": 0.774,
+      "step": 2977
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5561856530574254,
+      "learning_rate": 1.991559016022915e-05,
+      "loss": 0.937,
+      "step": 2978
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5614995050266285,
+      "learning_rate": 1.9915491198395714e-05,
+      "loss": 0.9168,
+      "step": 2979
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8093110533844332,
+      "learning_rate": 1.991539217883115e-05,
+      "loss": 0.9984,
+      "step": 2980
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9153321804521537,
+      "learning_rate": 1.991529310153603e-05,
+      "loss": 0.9956,
+      "step": 2981
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5915416311219643,
+      "learning_rate": 1.9915193966510945e-05,
+      "loss": 0.9303,
+      "step": 2982
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.764953939137259,
+      "learning_rate": 1.9915094773756454e-05,
+      "loss": 0.9549,
+      "step": 2983
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.769690324472486,
+      "learning_rate": 1.9914995523273144e-05,
+      "loss": 0.9917,
+      "step": 2984
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9610567728955253,
+      "learning_rate": 1.991489621506159e-05,
+      "loss": 0.9772,
+      "step": 2985
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.033154261243621,
+      "learning_rate": 1.9914796849122374e-05,
+      "loss": 0.9591,
+      "step": 2986
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7527047039472339,
+      "learning_rate": 1.991469742545607e-05,
+      "loss": 1.0073,
+      "step": 2987
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0812257849693325,
+      "learning_rate": 1.9914597944063257e-05,
+      "loss": 0.8309,
+      "step": 2988
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8735976989486023,
+      "learning_rate": 1.991449840494452e-05,
+      "loss": 0.8684,
+      "step": 2989
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9803258738206362,
+      "learning_rate": 1.991439880810043e-05,
+      "loss": 1.1218,
+      "step": 2990
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.743018518992278,
+      "learning_rate": 1.991429915353157e-05,
+      "loss": 0.8669,
+      "step": 2991
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.452310458460812,
+      "learning_rate": 1.9914199441238528e-05,
+      "loss": 0.8891,
+      "step": 2992
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.708723666758284,
+      "learning_rate": 1.9914099671221874e-05,
+      "loss": 0.9363,
+      "step": 2993
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5605942616329684,
+      "learning_rate": 1.9913999843482195e-05,
+      "loss": 0.9064,
+      "step": 2994
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5314673883739123,
+      "learning_rate": 1.9913899958020067e-05,
+      "loss": 0.979,
+      "step": 2995
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6847370197902338,
+      "learning_rate": 1.9913800014836076e-05,
+      "loss": 0.8607,
+      "step": 2996
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.834104371568314,
+      "learning_rate": 1.9913700013930804e-05,
+      "loss": 0.9625,
+      "step": 2997
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6414044947905044,
+      "learning_rate": 1.991359995530483e-05,
+      "loss": 1.0555,
+      "step": 2998
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.21133302777878,
+      "learning_rate": 1.991349983895874e-05,
+      "loss": 0.9998,
+      "step": 2999
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7364607384768909,
+      "learning_rate": 1.9913399664893113e-05,
+      "loss": 0.8773,
+      "step": 3000
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.764371611255586,
+      "learning_rate": 1.991329943310854e-05,
+      "loss": 0.9064,
+      "step": 3001
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6322978536179407,
+      "learning_rate": 1.9913199143605595e-05,
+      "loss": 0.9445,
+      "step": 3002
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7368941953727506,
+      "learning_rate": 1.9913098796384865e-05,
+      "loss": 0.9551,
+      "step": 3003
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7434673639170892,
+      "learning_rate": 1.991299839144694e-05,
+      "loss": 0.9968,
+      "step": 3004
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7763702060570304,
+      "learning_rate": 1.9912897928792395e-05,
+      "loss": 0.7578,
+      "step": 3005
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6027149496777042,
+      "learning_rate": 1.9912797408421823e-05,
+      "loss": 0.9884,
+      "step": 3006
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6702661144484519,
+      "learning_rate": 1.9912696830335806e-05,
+      "loss": 0.9487,
+      "step": 3007
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.84212205056382,
+      "learning_rate": 1.991259619453493e-05,
+      "loss": 1.0303,
+      "step": 3008
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.092119293346054,
+      "learning_rate": 1.9912495501019777e-05,
+      "loss": 0.9704,
+      "step": 3009
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6942762538026808,
+      "learning_rate": 1.9912394749790942e-05,
+      "loss": 0.9025,
+      "step": 3010
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0865852990257916,
+      "learning_rate": 1.9912293940849002e-05,
+      "loss": 1.0052,
+      "step": 3011
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9138946551146914,
+      "learning_rate": 1.991219307419455e-05,
+      "loss": 0.8965,
+      "step": 3012
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5265399455178725,
+      "learning_rate": 1.9912092149828175e-05,
+      "loss": 0.8656,
+      "step": 3013
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8771952953296034,
+      "learning_rate": 1.9911991167750458e-05,
+      "loss": 0.8265,
+      "step": 3014
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7734889257724066,
+      "learning_rate": 1.9911890127961993e-05,
+      "loss": 0.6855,
+      "step": 3015
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8059612394406694,
+      "learning_rate": 1.991178903046336e-05,
+      "loss": 0.9808,
+      "step": 3016
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7033747333545788,
+      "learning_rate": 1.991168787525516e-05,
+      "loss": 0.8467,
+      "step": 3017
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5808871416734411,
+      "learning_rate": 1.991158666233797e-05,
+      "loss": 0.8505,
+      "step": 3018
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.570926569287238,
+      "learning_rate": 1.991148539171239e-05,
+      "loss": 1.0033,
+      "step": 3019
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7957125984881429,
+      "learning_rate": 1.9911384063378996e-05,
+      "loss": 0.8007,
+      "step": 3020
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7359644633170423,
+      "learning_rate": 1.9911282677338395e-05,
+      "loss": 0.9005,
+      "step": 3021
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8282412477448833,
+      "learning_rate": 1.9911181233591163e-05,
+      "loss": 0.8278,
+      "step": 3022
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6905887806155044,
+      "learning_rate": 1.9911079732137897e-05,
+      "loss": 0.9391,
+      "step": 3023
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.785379980534765,
+      "learning_rate": 1.9910978172979187e-05,
+      "loss": 1.0744,
+      "step": 3024
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.588327746317903,
+      "learning_rate": 1.9910876556115626e-05,
+      "loss": 0.9163,
+      "step": 3025
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.325078514728225,
+      "learning_rate": 1.9910774881547803e-05,
+      "loss": 0.9225,
+      "step": 3026
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8322671687394703,
+      "learning_rate": 1.9910673149276308e-05,
+      "loss": 0.9699,
+      "step": 3027
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7113000169846542,
+      "learning_rate": 1.991057135930174e-05,
+      "loss": 0.8887,
+      "step": 3028
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8075930521411957,
+      "learning_rate": 1.991046951162469e-05,
+      "loss": 0.7311,
+      "step": 3029
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8359513410974808,
+      "learning_rate": 1.9910367606245744e-05,
+      "loss": 0.854,
+      "step": 3030
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2806986604899486,
+      "learning_rate": 1.9910265643165502e-05,
+      "loss": 1.0189,
+      "step": 3031
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6932192369111334,
+      "learning_rate": 1.9910163622384556e-05,
+      "loss": 0.8483,
+      "step": 3032
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8899055167603267,
+      "learning_rate": 1.9910061543903502e-05,
+      "loss": 0.9057,
+      "step": 3033
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1902180547718078,
+      "learning_rate": 1.9909959407722926e-05,
+      "loss": 0.9278,
+      "step": 3034
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.856641857708418,
+      "learning_rate": 1.990985721384343e-05,
+      "loss": 0.8786,
+      "step": 3035
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8795656945849015,
+      "learning_rate": 1.9909754962265615e-05,
+      "loss": 1.0026,
+      "step": 3036
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7036957509327604,
+      "learning_rate": 1.9909652652990065e-05,
+      "loss": 0.9992,
+      "step": 3037
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7953465391943703,
+      "learning_rate": 1.990955028601738e-05,
+      "loss": 0.8043,
+      "step": 3038
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.648164911623766,
+      "learning_rate": 1.9909447861348154e-05,
+      "loss": 1.0302,
+      "step": 3039
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8898910216893139,
+      "learning_rate": 1.9909345378982987e-05,
+      "loss": 1.0211,
+      "step": 3040
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.035360030952564,
+      "learning_rate": 1.9909242838922472e-05,
+      "loss": 0.8927,
+      "step": 3041
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1406032537113635,
+      "learning_rate": 1.9909140241167212e-05,
+      "loss": 1.0021,
+      "step": 3042
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8277701660343049,
+      "learning_rate": 1.99090375857178e-05,
+      "loss": 0.8557,
+      "step": 3043
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.584551859585783,
+      "learning_rate": 1.990893487257483e-05,
+      "loss": 0.9287,
+      "step": 3044
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.023784830525438,
+      "learning_rate": 1.990883210173891e-05,
+      "loss": 0.8084,
+      "step": 3045
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8679707182242302,
+      "learning_rate": 1.9908729273210627e-05,
+      "loss": 0.9574,
+      "step": 3046
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7633818573465427,
+      "learning_rate": 1.9908626386990586e-05,
+      "loss": 0.8022,
+      "step": 3047
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7356093077710624,
+      "learning_rate": 1.9908523443079384e-05,
+      "loss": 1.0317,
+      "step": 3048
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7323371004077949,
+      "learning_rate": 1.9908420441477626e-05,
+      "loss": 0.724,
+      "step": 3049
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8208195481288418,
+      "learning_rate": 1.9908317382185904e-05,
+      "loss": 0.9454,
+      "step": 3050
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6321748190132344,
+      "learning_rate": 1.9908214265204825e-05,
+      "loss": 0.8188,
+      "step": 3051
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9670041150249995,
+      "learning_rate": 1.990811109053498e-05,
+      "loss": 0.9133,
+      "step": 3052
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.846208555850161,
+      "learning_rate": 1.990800785817698e-05,
+      "loss": 1.0303,
+      "step": 3053
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2817861418237224,
+      "learning_rate": 1.990790456813142e-05,
+      "loss": 0.7955,
+      "step": 3054
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5158632677753656,
+      "learning_rate": 1.99078012203989e-05,
+      "loss": 0.9339,
+      "step": 3055
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.748576098718806,
+      "learning_rate": 1.9907697814980025e-05,
+      "loss": 0.9295,
+      "step": 3056
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.685530703810208,
+      "learning_rate": 1.9907594351875402e-05,
+      "loss": 0.9064,
+      "step": 3057
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6800525026816666,
+      "learning_rate": 1.9907490831085626e-05,
+      "loss": 1.0204,
+      "step": 3058
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.011602098620982,
+      "learning_rate": 1.99073872526113e-05,
+      "loss": 0.9581,
+      "step": 3059
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.902695795254145,
+      "learning_rate": 1.9907283616453027e-05,
+      "loss": 0.9107,
+      "step": 3060
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.01404972776501,
+      "learning_rate": 1.9907179922611414e-05,
+      "loss": 0.8866,
+      "step": 3061
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8530386478188594,
+      "learning_rate": 1.9907076171087064e-05,
+      "loss": 0.8653,
+      "step": 3062
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.013043326976345,
+      "learning_rate": 1.990697236188058e-05,
+      "loss": 0.9775,
+      "step": 3063
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.086231532708402,
+      "learning_rate": 1.9906868494992566e-05,
+      "loss": 0.9306,
+      "step": 3064
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7960261454175642,
+      "learning_rate": 1.9906764570423624e-05,
+      "loss": 0.7486,
+      "step": 3065
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.785180607232791,
+      "learning_rate": 1.9906660588174367e-05,
+      "loss": 0.8519,
+      "step": 3066
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0102035634382065,
+      "learning_rate": 1.9906556548245395e-05,
+      "loss": 0.9022,
+      "step": 3067
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.704604758095218,
+      "learning_rate": 1.9906452450637312e-05,
+      "loss": 0.9114,
+      "step": 3068
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6835749536360343,
+      "learning_rate": 1.9906348295350727e-05,
+      "loss": 0.9017,
+      "step": 3069
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8230383226783022,
+      "learning_rate": 1.990624408238625e-05,
+      "loss": 1.0158,
+      "step": 3070
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7142978906469284,
+      "learning_rate": 1.990613981174448e-05,
+      "loss": 1.0631,
+      "step": 3071
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8707422307854846,
+      "learning_rate": 1.9906035483426024e-05,
+      "loss": 0.9348,
+      "step": 3072
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0607393687288336,
+      "learning_rate": 1.99059310974315e-05,
+      "loss": 0.8233,
+      "step": 3073
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8265994254413567,
+      "learning_rate": 1.9905826653761506e-05,
+      "loss": 0.7046,
+      "step": 3074
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.505716120146863,
+      "learning_rate": 1.9905722152416652e-05,
+      "loss": 1.0728,
+      "step": 3075
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5352969438809867,
+      "learning_rate": 1.990561759339755e-05,
+      "loss": 0.8514,
+      "step": 3076
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.896395929788609,
+      "learning_rate": 1.9905512976704805e-05,
+      "loss": 0.8805,
+      "step": 3077
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6760545838835723,
+      "learning_rate": 1.9905408302339025e-05,
+      "loss": 0.9317,
+      "step": 3078
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.756886665687722,
+      "learning_rate": 1.9905303570300828e-05,
+      "loss": 0.9911,
+      "step": 3079
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0420591782591555,
+      "learning_rate": 1.990519878059081e-05,
+      "loss": 0.949,
+      "step": 3080
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6148401187545103,
+      "learning_rate": 1.9905093933209593e-05,
+      "loss": 0.9199,
+      "step": 3081
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5941977762849884,
+      "learning_rate": 1.9904989028157784e-05,
+      "loss": 0.9877,
+      "step": 3082
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9105079223149588,
+      "learning_rate": 1.990488406543599e-05,
+      "loss": 0.9675,
+      "step": 3083
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6363085565308442,
+      "learning_rate": 1.9904779045044828e-05,
+      "loss": 0.9423,
+      "step": 3084
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.050462596850154,
+      "learning_rate": 1.9904673966984907e-05,
+      "loss": 0.9415,
+      "step": 3085
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.863424970467947,
+      "learning_rate": 1.9904568831256838e-05,
+      "loss": 0.8999,
+      "step": 3086
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6573500411971767,
+      "learning_rate": 1.990446363786123e-05,
+      "loss": 0.9549,
+      "step": 3087
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5932396578825085,
+      "learning_rate": 1.9904358386798704e-05,
+      "loss": 0.9672,
+      "step": 3088
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8965637509191442,
+      "learning_rate": 1.9904253078069865e-05,
+      "loss": 0.8936,
+      "step": 3089
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3156805487793126,
+      "learning_rate": 1.990414771167533e-05,
+      "loss": 0.9227,
+      "step": 3090
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6980513502793488,
+      "learning_rate": 1.9904042287615712e-05,
+      "loss": 0.8625,
+      "step": 3091
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5158362300944768,
+      "learning_rate": 1.990393680589162e-05,
+      "loss": 0.8227,
+      "step": 3092
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0623719143617483,
+      "learning_rate": 1.9903831266503676e-05,
+      "loss": 0.7948,
+      "step": 3093
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.585004217721968,
+      "learning_rate": 1.9903725669452492e-05,
+      "loss": 1.0032,
+      "step": 3094
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7577513729831777,
+      "learning_rate": 1.990362001473868e-05,
+      "loss": 0.8244,
+      "step": 3095
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8745904122614117,
+      "learning_rate": 1.9903514302362856e-05,
+      "loss": 0.8975,
+      "step": 3096
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.075474187428818,
+      "learning_rate": 1.9903408532325637e-05,
+      "loss": 0.9957,
+      "step": 3097
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7121184914884684,
+      "learning_rate": 1.990330270462764e-05,
+      "loss": 1.0888,
+      "step": 3098
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4811441578297415,
+      "learning_rate": 1.9903196819269476e-05,
+      "loss": 0.7812,
+      "step": 3099
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6488206057736028,
+      "learning_rate": 1.9903090876251767e-05,
+      "loss": 0.926,
+      "step": 3100
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6683691377084506,
+      "learning_rate": 1.9902984875575127e-05,
+      "loss": 0.9134,
+      "step": 3101
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8245407554303287,
+      "learning_rate": 1.9902878817240175e-05,
+      "loss": 0.7367,
+      "step": 3102
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1497155747239316,
+      "learning_rate": 1.9902772701247527e-05,
+      "loss": 1.0278,
+      "step": 3103
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7514885757062479,
+      "learning_rate": 1.9902666527597797e-05,
+      "loss": 0.7262,
+      "step": 3104
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.628361272754276,
+      "learning_rate": 1.9902560296291612e-05,
+      "loss": 0.852,
+      "step": 3105
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7224338716832674,
+      "learning_rate": 1.9902454007329585e-05,
+      "loss": 0.8727,
+      "step": 3106
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7960929128585321,
+      "learning_rate": 1.9902347660712336e-05,
+      "loss": 0.9129,
+      "step": 3107
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8503581007929852,
+      "learning_rate": 1.990224125644048e-05,
+      "loss": 0.9542,
+      "step": 3108
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7066291787306545,
+      "learning_rate": 1.9902134794514648e-05,
+      "loss": 0.873,
+      "step": 3109
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0360275389136686,
+      "learning_rate": 1.9902028274935446e-05,
+      "loss": 1.0468,
+      "step": 3110
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8035684658807265,
+      "learning_rate": 1.9901921697703503e-05,
+      "loss": 0.7847,
+      "step": 3111
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8651796346484744,
+      "learning_rate": 1.9901815062819437e-05,
+      "loss": 0.8265,
+      "step": 3112
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7946370281167876,
+      "learning_rate": 1.9901708370283866e-05,
+      "loss": 0.8685,
+      "step": 3113
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.522191630238739,
+      "learning_rate": 1.9901601620097417e-05,
+      "loss": 0.9871,
+      "step": 3114
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9583654472454806,
+      "learning_rate": 1.990149481226071e-05,
+      "loss": 1.0101,
+      "step": 3115
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8204208787249011,
+      "learning_rate": 1.9901387946774363e-05,
+      "loss": 0.7644,
+      "step": 3116
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1450786906424253,
+      "learning_rate": 1.9901281023639002e-05,
+      "loss": 0.8271,
+      "step": 3117
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.639094557566692,
+      "learning_rate": 1.9901174042855243e-05,
+      "loss": 0.9011,
+      "step": 3118
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9405822962071175,
+      "learning_rate": 1.9901067004423722e-05,
+      "loss": 0.9012,
+      "step": 3119
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.806878626665894,
+      "learning_rate": 1.990095990834505e-05,
+      "loss": 0.7538,
+      "step": 3120
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7275779248640712,
+      "learning_rate": 1.9900852754619853e-05,
+      "loss": 0.9796,
+      "step": 3121
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4931860912334485,
+      "learning_rate": 1.9900745543248762e-05,
+      "loss": 0.9692,
+      "step": 3122
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.980993316571122,
+      "learning_rate": 1.990063827423239e-05,
+      "loss": 0.908,
+      "step": 3123
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5086917342348523,
+      "learning_rate": 1.990053094757137e-05,
+      "loss": 1.0019,
+      "step": 3124
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6142851751090326,
+      "learning_rate": 1.9900423563266325e-05,
+      "loss": 0.9233,
+      "step": 3125
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7638063089778406,
+      "learning_rate": 1.9900316121317883e-05,
+      "loss": 0.9372,
+      "step": 3126
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9582041474910779,
+      "learning_rate": 1.990020862172666e-05,
+      "loss": 1.0348,
+      "step": 3127
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.074906682923206,
+      "learning_rate": 1.990010106449329e-05,
+      "loss": 1.0876,
+      "step": 3128
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6515147201706304,
+      "learning_rate": 1.9899993449618397e-05,
+      "loss": 0.9283,
+      "step": 3129
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9144508878263413,
+      "learning_rate": 1.989988577710261e-05,
+      "loss": 0.9095,
+      "step": 3130
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.240377015195709,
+      "learning_rate": 1.9899778046946552e-05,
+      "loss": 0.8794,
+      "step": 3131
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6340560845226826,
+      "learning_rate": 1.9899670259150855e-05,
+      "loss": 0.9221,
+      "step": 3132
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7510737041069804,
+      "learning_rate": 1.9899562413716137e-05,
+      "loss": 0.7859,
+      "step": 3133
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.855629075128877,
+      "learning_rate": 1.989945451064304e-05,
+      "loss": 0.9537,
+      "step": 3134
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9897404436312067,
+      "learning_rate": 1.989934654993218e-05,
+      "loss": 0.9362,
+      "step": 3135
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.714190834513647,
+      "learning_rate": 1.989923853158419e-05,
+      "loss": 0.982,
+      "step": 3136
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8067046592785377,
+      "learning_rate": 1.9899130455599703e-05,
+      "loss": 0.7444,
+      "step": 3137
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7430357827741798,
+      "learning_rate": 1.989902232197934e-05,
+      "loss": 0.968,
+      "step": 3138
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9398034495523626,
+      "learning_rate": 1.989891413072374e-05,
+      "loss": 0.9904,
+      "step": 3139
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5653435184340887,
+      "learning_rate": 1.9898805881833523e-05,
+      "loss": 0.892,
+      "step": 3140
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0318164235423084,
+      "learning_rate": 1.9898697575309326e-05,
+      "loss": 0.8221,
+      "step": 3141
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.689294709857231,
+      "learning_rate": 1.989858921115178e-05,
+      "loss": 0.892,
+      "step": 3142
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8828361407293082,
+      "learning_rate": 1.989848078936151e-05,
+      "loss": 0.9574,
+      "step": 3143
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7556491319245575,
+      "learning_rate": 1.9898372309939153e-05,
+      "loss": 0.8544,
+      "step": 3144
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.025663232120566,
+      "learning_rate": 1.989826377288534e-05,
+      "loss": 0.9586,
+      "step": 3145
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7726973616484422,
+      "learning_rate": 1.98981551782007e-05,
+      "loss": 0.751,
+      "step": 3146
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7893204666045428,
+      "learning_rate": 1.9898046525885868e-05,
+      "loss": 1.0194,
+      "step": 3147
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7918907176752168,
+      "learning_rate": 1.9897937815941474e-05,
+      "loss": 0.9168,
+      "step": 3148
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.851727868248579,
+      "learning_rate": 1.9897829048368153e-05,
+      "loss": 1.023,
+      "step": 3149
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7910953979680572,
+      "learning_rate": 1.989772022316654e-05,
+      "loss": 0.7313,
+      "step": 3150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.209596067730812,
+      "learning_rate": 1.9897611340337263e-05,
+      "loss": 0.9467,
+      "step": 3151
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.56338968276093,
+      "learning_rate": 1.989750239988096e-05,
+      "loss": 0.8558,
+      "step": 3152
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.69479457325946,
+      "learning_rate": 1.9897393401798265e-05,
+      "loss": 0.9711,
+      "step": 3153
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.876110424109189,
+      "learning_rate": 1.989728434608981e-05,
+      "loss": 0.972,
+      "step": 3154
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0056901429182536,
+      "learning_rate": 1.9897175232756236e-05,
+      "loss": 0.9809,
+      "step": 3155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.754211222078979,
+      "learning_rate": 1.9897066061798172e-05,
+      "loss": 1.0495,
+      "step": 3156
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8534558879051548,
+      "learning_rate": 1.9896956833216257e-05,
+      "loss": 0.7884,
+      "step": 3157
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.771699254324046,
+      "learning_rate": 1.9896847547011124e-05,
+      "loss": 0.8957,
+      "step": 3158
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.82285985888442,
+      "learning_rate": 1.9896738203183413e-05,
+      "loss": 0.8985,
+      "step": 3159
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5479343126692475,
+      "learning_rate": 1.9896628801733763e-05,
+      "loss": 1.0374,
+      "step": 3160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1044486222493095,
+      "learning_rate": 1.9896519342662798e-05,
+      "loss": 1.0733,
+      "step": 3161
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.884260723146339,
+      "learning_rate": 1.989640982597117e-05,
+      "loss": 0.7704,
+      "step": 3162
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.165784294827261,
+      "learning_rate": 1.989630025165951e-05,
+      "loss": 0.873,
+      "step": 3163
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7880255376954909,
+      "learning_rate": 1.9896190619728458e-05,
+      "loss": 0.7065,
+      "step": 3164
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.288885704117888,
+      "learning_rate": 1.9896080930178646e-05,
+      "loss": 0.9161,
+      "step": 3165
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1348050146905715,
+      "learning_rate": 1.9895971183010726e-05,
+      "loss": 0.9201,
+      "step": 3166
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.591216143144748,
+      "learning_rate": 1.9895861378225324e-05,
+      "loss": 0.9434,
+      "step": 3167
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7266590978522456,
+      "learning_rate": 1.989575151582308e-05,
+      "loss": 0.8362,
+      "step": 3168
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7257411582677233,
+      "learning_rate": 1.9895641595804644e-05,
+      "loss": 1.0238,
+      "step": 3169
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.856661713347165,
+      "learning_rate": 1.9895531618170646e-05,
+      "loss": 0.9543,
+      "step": 3170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9628156741532796,
+      "learning_rate": 1.9895421582921735e-05,
+      "loss": 0.8675,
+      "step": 3171
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.920146112557126,
+      "learning_rate": 1.989531149005854e-05,
+      "loss": 0.8308,
+      "step": 3172
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8672541363219998,
+      "learning_rate": 1.9895201339581717e-05,
+      "loss": 0.7558,
+      "step": 3173
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7685875697043845,
+      "learning_rate": 1.989509113149189e-05,
+      "loss": 1.0489,
+      "step": 3174
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9599958595561977,
+      "learning_rate": 1.9894980865789716e-05,
+      "loss": 0.8265,
+      "step": 3175
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9106328197620426,
+      "learning_rate": 1.9894870542475827e-05,
+      "loss": 0.9458,
+      "step": 3176
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5996395422892185,
+      "learning_rate": 1.989476016155087e-05,
+      "loss": 0.9831,
+      "step": 3177
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7853444108824758,
+      "learning_rate": 1.9894649723015487e-05,
+      "loss": 0.7171,
+      "step": 3178
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9719092198631676,
+      "learning_rate": 1.9894539226870325e-05,
+      "loss": 0.9221,
+      "step": 3179
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7691567405038067,
+      "learning_rate": 1.9894428673116016e-05,
+      "loss": 0.9555,
+      "step": 3180
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6942391206098963,
+      "learning_rate": 1.9894318061753214e-05,
+      "loss": 0.9806,
+      "step": 3181
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8289118302431504,
+      "learning_rate": 1.989420739278256e-05,
+      "loss": 0.8578,
+      "step": 3182
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7844831610058698,
+      "learning_rate": 1.98940966662047e-05,
+      "loss": 0.9083,
+      "step": 3183
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.773980285177081,
+      "learning_rate": 1.9893985882020272e-05,
+      "loss": 1.0251,
+      "step": 3184
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9904146645087293,
+      "learning_rate": 1.989387504022993e-05,
+      "loss": 0.9983,
+      "step": 3185
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9292600892481524,
+      "learning_rate": 1.9893764140834312e-05,
+      "loss": 0.8938,
+      "step": 3186
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8598729713841379,
+      "learning_rate": 1.9893653183834068e-05,
+      "loss": 0.7548,
+      "step": 3187
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.79151226660638,
+      "learning_rate": 1.9893542169229844e-05,
+      "loss": 0.9386,
+      "step": 3188
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.327822927266398,
+      "learning_rate": 1.9893431097022284e-05,
+      "loss": 0.7945,
+      "step": 3189
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8061902982892444,
+      "learning_rate": 1.989331996721204e-05,
+      "loss": 0.7253,
+      "step": 3190
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6880981838439628,
+      "learning_rate": 1.9893208779799747e-05,
+      "loss": 0.9443,
+      "step": 3191
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.677623081925577,
+      "learning_rate": 1.989309753478607e-05,
+      "loss": 0.919,
+      "step": 3192
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8286085081229927,
+      "learning_rate": 1.989298623217164e-05,
+      "loss": 0.7234,
+      "step": 3193
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5763326734820464,
+      "learning_rate": 1.989287487195711e-05,
+      "loss": 0.8948,
+      "step": 3194
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9709247067510827,
+      "learning_rate": 1.9892763454143134e-05,
+      "loss": 1.0013,
+      "step": 3195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.711455954691803,
+      "learning_rate": 1.989265197873036e-05,
+      "loss": 0.9522,
+      "step": 3196
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6012073507593574,
+      "learning_rate": 1.989254044571943e-05,
+      "loss": 0.8812,
+      "step": 3197
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8896065842413179,
+      "learning_rate": 1.9892428855111e-05,
+      "loss": 0.9529,
+      "step": 3198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.773292293333319,
+      "learning_rate": 1.989231720690572e-05,
+      "loss": 0.8977,
+      "step": 3199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7836721979175258,
+      "learning_rate": 1.989220550110423e-05,
+      "loss": 0.9174,
+      "step": 3200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7120937456375083,
+      "learning_rate": 1.989209373770719e-05,
+      "loss": 1.035,
+      "step": 3201
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7691711367068508,
+      "learning_rate": 1.9891981916715254e-05,
+      "loss": 0.9155,
+      "step": 3202
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2384115289202766,
+      "learning_rate": 1.9891870038129066e-05,
+      "loss": 0.9335,
+      "step": 3203
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.132495488950891,
+      "learning_rate": 1.9891758101949275e-05,
+      "loss": 0.935,
+      "step": 3204
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0430384293481807,
+      "learning_rate": 1.9891646108176536e-05,
+      "loss": 0.9412,
+      "step": 3205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9072324090261057,
+      "learning_rate": 1.9891534056811507e-05,
+      "loss": 0.9599,
+      "step": 3206
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7843769084508463,
+      "learning_rate": 1.9891421947854833e-05,
+      "loss": 1.0383,
+      "step": 3207
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7788320093697523,
+      "learning_rate": 1.9891309781307167e-05,
+      "loss": 0.9497,
+      "step": 3208
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.677382003526372,
+      "learning_rate": 1.9891197557169164e-05,
+      "loss": 0.9243,
+      "step": 3209
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2878516562418043,
+      "learning_rate": 1.989108527544148e-05,
+      "loss": 0.8651,
+      "step": 3210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8139635415823383,
+      "learning_rate": 1.9890972936124764e-05,
+      "loss": 0.7676,
+      "step": 3211
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0966374456858485,
+      "learning_rate": 1.989086053921967e-05,
+      "loss": 0.9245,
+      "step": 3212
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.818867029254188,
+      "learning_rate": 1.989074808472686e-05,
+      "loss": 0.7506,
+      "step": 3213
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.763742789001338,
+      "learning_rate": 1.989063557264698e-05,
+      "loss": 0.7211,
+      "step": 3214
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7598650850432842,
+      "learning_rate": 1.9890523002980687e-05,
+      "loss": 0.6968,
+      "step": 3215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.091812947934551,
+      "learning_rate": 1.989041037572864e-05,
+      "loss": 0.9093,
+      "step": 3216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7620313936540593,
+      "learning_rate": 1.989029769089149e-05,
+      "loss": 0.8695,
+      "step": 3217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7421291334124231,
+      "learning_rate": 1.98901849484699e-05,
+      "loss": 0.9441,
+      "step": 3218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5929015068358106,
+      "learning_rate": 1.9890072148464518e-05,
+      "loss": 0.9084,
+      "step": 3219
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5627466324901216,
+      "learning_rate": 1.988995929087601e-05,
+      "loss": 0.9943,
+      "step": 3220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6530246715153054,
+      "learning_rate": 1.988984637570502e-05,
+      "loss": 0.9085,
+      "step": 3221
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7547017387250184,
+      "learning_rate": 1.9889733402952222e-05,
+      "loss": 0.7491,
+      "step": 3222
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.892768429249231,
+      "learning_rate": 1.988962037261826e-05,
+      "loss": 0.9962,
+      "step": 3223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2961900708175964,
+      "learning_rate": 1.98895072847038e-05,
+      "loss": 0.7931,
+      "step": 3224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7681824602489783,
+      "learning_rate": 1.9889394139209492e-05,
+      "loss": 0.6863,
+      "step": 3225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6337063690172506,
+      "learning_rate": 1.9889280936136007e-05,
+      "loss": 1.0054,
+      "step": 3226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8687732994486914,
+      "learning_rate": 1.9889167675483994e-05,
+      "loss": 0.9121,
+      "step": 3227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.803241888579486,
+      "learning_rate": 1.9889054357254113e-05,
+      "loss": 0.884,
+      "step": 3228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9969489360482593,
+      "learning_rate": 1.988894098144703e-05,
+      "loss": 0.9073,
+      "step": 3229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6243925237554944,
+      "learning_rate": 1.9888827548063404e-05,
+      "loss": 0.9337,
+      "step": 3230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.599474716489985,
+      "learning_rate": 1.988871405710389e-05,
+      "loss": 0.9428,
+      "step": 3231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5927813712202543,
+      "learning_rate": 1.9888600508569157e-05,
+      "loss": 0.8376,
+      "step": 3232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9191654889207488,
+      "learning_rate": 1.9888486902459852e-05,
+      "loss": 0.9162,
+      "step": 3233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6048416984593346,
+      "learning_rate": 1.9888373238776654e-05,
+      "loss": 0.987,
+      "step": 3234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2123934299501586,
+      "learning_rate": 1.9888259517520212e-05,
+      "loss": 0.8403,
+      "step": 3235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9632342273871692,
+      "learning_rate": 1.9888145738691196e-05,
+      "loss": 0.8038,
+      "step": 3236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.250490381986209,
+      "learning_rate": 1.9888031902290265e-05,
+      "loss": 0.8595,
+      "step": 3237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8669165217367614,
+      "learning_rate": 1.9887918008318076e-05,
+      "loss": 0.9444,
+      "step": 3238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.875231282211587,
+      "learning_rate": 1.98878040567753e-05,
+      "loss": 0.8801,
+      "step": 3239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0604184291877017,
+      "learning_rate": 1.9887690047662604e-05,
+      "loss": 0.8966,
+      "step": 3240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.895399059953481,
+      "learning_rate": 1.988757598098064e-05,
+      "loss": 1.0555,
+      "step": 3241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8060320461027811,
+      "learning_rate": 1.988746185673008e-05,
+      "loss": 0.7511,
+      "step": 3242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6483072881752505,
+      "learning_rate": 1.9887347674911587e-05,
+      "loss": 0.8209,
+      "step": 3243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9867176634677772,
+      "learning_rate": 1.9887233435525827e-05,
+      "loss": 0.9272,
+      "step": 3244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8834852748754356,
+      "learning_rate": 1.9887119138573462e-05,
+      "loss": 0.7575,
+      "step": 3245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0324298156171072,
+      "learning_rate": 1.988700478405516e-05,
+      "loss": 0.8333,
+      "step": 3246
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8964831017406198,
+      "learning_rate": 1.9886890371971585e-05,
+      "loss": 0.9935,
+      "step": 3247
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8630659016717723,
+      "learning_rate": 1.9886775902323405e-05,
+      "loss": 1.0197,
+      "step": 3248
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8592156170480936,
+      "learning_rate": 1.9886661375111286e-05,
+      "loss": 0.9453,
+      "step": 3249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.220722504851943,
+      "learning_rate": 1.9886546790335895e-05,
+      "loss": 0.9377,
+      "step": 3250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6154158726581613,
+      "learning_rate": 1.9886432147997897e-05,
+      "loss": 0.9459,
+      "step": 3251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.628611543357172,
+      "learning_rate": 1.988631744809796e-05,
+      "loss": 0.9483,
+      "step": 3252
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1682123996131426,
+      "learning_rate": 1.9886202690636755e-05,
+      "loss": 0.9803,
+      "step": 3253
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1103099314169214,
+      "learning_rate": 1.9886087875614946e-05,
+      "loss": 0.9275,
+      "step": 3254
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7174730044448263,
+      "learning_rate": 1.9885973003033207e-05,
+      "loss": 1.0718,
+      "step": 3255
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7737841443774423,
+      "learning_rate": 1.98858580728922e-05,
+      "loss": 0.9565,
+      "step": 3256
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.874301532076563,
+      "learning_rate": 1.9885743085192598e-05,
+      "loss": 1.0018,
+      "step": 3257
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7534920163303718,
+      "learning_rate": 1.988562803993507e-05,
+      "loss": 0.737,
+      "step": 3258
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.766422644809778,
+      "learning_rate": 1.988551293712029e-05,
+      "loss": 0.7172,
+      "step": 3259
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3176684158622294,
+      "learning_rate": 1.9885397776748918e-05,
+      "loss": 0.8034,
+      "step": 3260
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6326752474070043,
+      "learning_rate": 1.9885282558821633e-05,
+      "loss": 0.9915,
+      "step": 3261
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7537702765707714,
+      "learning_rate": 1.98851672833391e-05,
+      "loss": 0.7256,
+      "step": 3262
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1555234286886384,
+      "learning_rate": 1.9885051950301998e-05,
+      "loss": 0.9496,
+      "step": 3263
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.256624744905869,
+      "learning_rate": 1.9884936559710995e-05,
+      "loss": 0.7468,
+      "step": 3264
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7464391655542753,
+      "learning_rate": 1.988482111156676e-05,
+      "loss": 0.9389,
+      "step": 3265
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.895337231762071,
+      "learning_rate": 1.9884705605869965e-05,
+      "loss": 0.9519,
+      "step": 3266
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0694430726692103,
+      "learning_rate": 1.9884590042621284e-05,
+      "loss": 1.0076,
+      "step": 3267
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.761511252359846,
+      "learning_rate": 1.9884474421821393e-05,
+      "loss": 1.0483,
+      "step": 3268
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7454420158414232,
+      "learning_rate": 1.9884358743470963e-05,
+      "loss": 0.909,
+      "step": 3269
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6590080671866156,
+      "learning_rate": 1.9884243007570665e-05,
+      "loss": 0.9568,
+      "step": 3270
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.245532404345916,
+      "learning_rate": 1.9884127214121176e-05,
+      "loss": 0.8943,
+      "step": 3271
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6385127227277139,
+      "learning_rate": 1.9884011363123168e-05,
+      "loss": 0.9776,
+      "step": 3272
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6369684901095491,
+      "learning_rate": 1.988389545457732e-05,
+      "loss": 0.9459,
+      "step": 3273
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7018950686213472,
+      "learning_rate": 1.9883779488484297e-05,
+      "loss": 0.9557,
+      "step": 3274
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9697345056524957,
+      "learning_rate": 1.9883663464844786e-05,
+      "loss": 0.9722,
+      "step": 3275
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8205399405530822,
+      "learning_rate": 1.9883547383659456e-05,
+      "loss": 0.8764,
+      "step": 3276
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5068592059229553,
+      "learning_rate": 1.9883431244928984e-05,
+      "loss": 0.8234,
+      "step": 3277
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1163254389263217,
+      "learning_rate": 1.9883315048654048e-05,
+      "loss": 0.8388,
+      "step": 3278
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8479707653095853,
+      "learning_rate": 1.988319879483532e-05,
+      "loss": 0.6992,
+      "step": 3279
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8432446987743147,
+      "learning_rate": 1.988308248347348e-05,
+      "loss": 0.9366,
+      "step": 3280
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7350179429905999,
+      "learning_rate": 1.9882966114569205e-05,
+      "loss": 0.8764,
+      "step": 3281
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9895741277523815,
+      "learning_rate": 1.9882849688123172e-05,
+      "loss": 1.0221,
+      "step": 3282
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8278810416838418,
+      "learning_rate": 1.988273320413606e-05,
+      "loss": 0.9236,
+      "step": 3283
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6755494319885635,
+      "learning_rate": 1.9882616662608547e-05,
+      "loss": 0.8743,
+      "step": 3284
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7809661068738416,
+      "learning_rate": 1.988250006354131e-05,
+      "loss": 0.9152,
+      "step": 3285
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7178543534401653,
+      "learning_rate": 1.988238340693503e-05,
+      "loss": 0.9093,
+      "step": 3286
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3211732014440862,
+      "learning_rate": 1.9882266692790387e-05,
+      "loss": 1.02,
+      "step": 3287
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.863706783148973,
+      "learning_rate": 1.9882149921108053e-05,
+      "loss": 0.9396,
+      "step": 3288
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6879239458187367,
+      "learning_rate": 1.988203309188872e-05,
+      "loss": 0.8841,
+      "step": 3289
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7062233092707024,
+      "learning_rate": 1.9881916205133057e-05,
+      "loss": 0.8577,
+      "step": 3290
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.048843030678155,
+      "learning_rate": 1.988179926084175e-05,
+      "loss": 0.9247,
+      "step": 3291
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.123657809144118,
+      "learning_rate": 1.988168225901548e-05,
+      "loss": 1.0767,
+      "step": 3292
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0432099439356093,
+      "learning_rate": 1.9881565199654926e-05,
+      "loss": 0.834,
+      "step": 3293
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9243292307795388,
+      "learning_rate": 1.9881448082760774e-05,
+      "loss": 0.8174,
+      "step": 3294
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.810513974624993,
+      "learning_rate": 1.9881330908333702e-05,
+      "loss": 0.8629,
+      "step": 3295
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9361627916088087,
+      "learning_rate": 1.988121367637439e-05,
+      "loss": 0.9651,
+      "step": 3296
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9530553207485128,
+      "learning_rate": 1.9881096386883526e-05,
+      "loss": 0.929,
+      "step": 3297
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6700826073104216,
+      "learning_rate": 1.988097903986179e-05,
+      "loss": 1.0643,
+      "step": 3298
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7562371051612329,
+      "learning_rate": 1.9880861635309867e-05,
+      "loss": 0.9151,
+      "step": 3299
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6342126414937281,
+      "learning_rate": 1.9880744173228437e-05,
+      "loss": 0.8796,
+      "step": 3300
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7477164917279207,
+      "learning_rate": 1.988062665361819e-05,
+      "loss": 0.8916,
+      "step": 3301
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9108569142990153,
+      "learning_rate": 1.9880509076479803e-05,
+      "loss": 0.9662,
+      "step": 3302
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8167994135510349,
+      "learning_rate": 1.9880391441813963e-05,
+      "loss": 0.7314,
+      "step": 3303
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7757597626472974,
+      "learning_rate": 1.988027374962136e-05,
+      "loss": 0.9326,
+      "step": 3304
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7083777302418652,
+      "learning_rate": 1.9880155999902672e-05,
+      "loss": 0.9333,
+      "step": 3305
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6915455133584705,
+      "learning_rate": 1.988003819265859e-05,
+      "loss": 0.9226,
+      "step": 3306
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5029007965140437,
+      "learning_rate": 1.98799203278898e-05,
+      "loss": 1.0245,
+      "step": 3307
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7290468962178605,
+      "learning_rate": 1.987980240559698e-05,
+      "loss": 0.9045,
+      "step": 3308
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7518212992822928,
+      "learning_rate": 1.9879684425780825e-05,
+      "loss": 0.7189,
+      "step": 3309
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5953822985434232,
+      "learning_rate": 1.987956638844202e-05,
+      "loss": 0.9655,
+      "step": 3310
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6400326640343121,
+      "learning_rate": 1.9879448293581253e-05,
+      "loss": 0.9648,
+      "step": 3311
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8967651204412959,
+      "learning_rate": 1.9879330141199206e-05,
+      "loss": 0.8991,
+      "step": 3312
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.676154861898504,
+      "learning_rate": 1.9879211931296572e-05,
+      "loss": 0.9706,
+      "step": 3313
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.882217492904402,
+      "learning_rate": 1.9879093663874044e-05,
+      "loss": 1.0096,
+      "step": 3314
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.738953128785975,
+      "learning_rate": 1.98789753389323e-05,
+      "loss": 1.0793,
+      "step": 3315
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0999378548847893,
+      "learning_rate": 1.9878856956472033e-05,
+      "loss": 0.9418,
+      "step": 3316
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9842300444227525,
+      "learning_rate": 1.987873851649394e-05,
+      "loss": 0.9263,
+      "step": 3317
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4495338009690617,
+      "learning_rate": 1.9878620018998696e-05,
+      "loss": 0.8992,
+      "step": 3318
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.119801452943553,
+      "learning_rate": 1.9878501463987002e-05,
+      "loss": 0.9441,
+      "step": 3319
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5361483506664564,
+      "learning_rate": 1.9878382851459545e-05,
+      "loss": 0.8508,
+      "step": 3320
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4022264216307563,
+      "learning_rate": 1.9878264181417016e-05,
+      "loss": 1.0779,
+      "step": 3321
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.653539656350808,
+      "learning_rate": 1.9878145453860106e-05,
+      "loss": 0.9341,
+      "step": 3322
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6600983046313549,
+      "learning_rate": 1.9878026668789503e-05,
+      "loss": 1.085,
+      "step": 3323
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7584015862688489,
+      "learning_rate": 1.9877907826205902e-05,
+      "loss": 0.7852,
+      "step": 3324
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.057153651312899,
+      "learning_rate": 1.9877788926109997e-05,
+      "loss": 0.9928,
+      "step": 3325
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.681432336372676,
+      "learning_rate": 1.9877669968502476e-05,
+      "loss": 0.8964,
+      "step": 3326
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6575677462331238,
+      "learning_rate": 1.9877550953384037e-05,
+      "loss": 0.9763,
+      "step": 3327
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5111012853077064,
+      "learning_rate": 1.9877431880755364e-05,
+      "loss": 0.98,
+      "step": 3328
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6570471351841531,
+      "learning_rate": 1.9877312750617155e-05,
+      "loss": 0.8356,
+      "step": 3329
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9621689837415779,
+      "learning_rate": 1.9877193562970105e-05,
+      "loss": 0.8878,
+      "step": 3330
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9737776126469357,
+      "learning_rate": 1.987707431781491e-05,
+      "loss": 0.9195,
+      "step": 3331
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5991100869819168,
+      "learning_rate": 1.987695501515226e-05,
+      "loss": 0.9504,
+      "step": 3332
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8081404588972352,
+      "learning_rate": 1.9876835654982848e-05,
+      "loss": 0.8916,
+      "step": 3333
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6884218332245773,
+      "learning_rate": 1.9876716237307376e-05,
+      "loss": 0.8863,
+      "step": 3334
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.688707039387109,
+      "learning_rate": 1.9876596762126532e-05,
+      "loss": 0.9482,
+      "step": 3335
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5442974228320203,
+      "learning_rate": 1.9876477229441014e-05,
+      "loss": 0.898,
+      "step": 3336
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.825252258619567,
+      "learning_rate": 1.987635763925152e-05,
+      "loss": 1.1215,
+      "step": 3337
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8555435453102775,
+      "learning_rate": 1.9876237991558746e-05,
+      "loss": 1.0613,
+      "step": 3338
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8348369034650114,
+      "learning_rate": 1.987611828636339e-05,
+      "loss": 1.0087,
+      "step": 3339
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8340255409397637,
+      "learning_rate": 1.987599852366614e-05,
+      "loss": 0.9893,
+      "step": 3340
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8330467175344445,
+      "learning_rate": 1.9875878703467707e-05,
+      "loss": 0.9131,
+      "step": 3341
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6230207766884504,
+      "learning_rate": 1.987575882576878e-05,
+      "loss": 0.879,
+      "step": 3342
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.33161154129389,
+      "learning_rate": 1.987563889057006e-05,
+      "loss": 0.8714,
+      "step": 3343
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7307239664502052,
+      "learning_rate": 1.9875518897872236e-05,
+      "loss": 0.7268,
+      "step": 3344
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7417941286653704,
+      "learning_rate": 1.987539884767602e-05,
+      "loss": 1.0027,
+      "step": 3345
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5335369988977796,
+      "learning_rate": 1.9875278739982104e-05,
+      "loss": 0.9173,
+      "step": 3346
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7211472739317641,
+      "learning_rate": 1.9875158574791192e-05,
+      "loss": 0.9666,
+      "step": 3347
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.647496238967478,
+      "learning_rate": 1.9875038352103975e-05,
+      "loss": 1.0448,
+      "step": 3348
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8794621203545538,
+      "learning_rate": 1.9874918071921162e-05,
+      "loss": 0.7684,
+      "step": 3349
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.002109820903398,
+      "learning_rate": 1.9874797734243446e-05,
+      "loss": 1.01,
+      "step": 3350
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.030171556901663,
+      "learning_rate": 1.9874677339071536e-05,
+      "loss": 0.8229,
+      "step": 3351
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.795812211785222,
+      "learning_rate": 1.9874556886406126e-05,
+      "loss": 0.8629,
+      "step": 3352
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1584174760279002,
+      "learning_rate": 1.987443637624792e-05,
+      "loss": 0.865,
+      "step": 3353
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7766897912998627,
+      "learning_rate": 1.987431580859762e-05,
+      "loss": 0.8809,
+      "step": 3354
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8012771902649604,
+      "learning_rate": 1.9874195183455925e-05,
+      "loss": 0.7371,
+      "step": 3355
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6559565138674017,
+      "learning_rate": 1.987407450082354e-05,
+      "loss": 0.9047,
+      "step": 3356
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.719898315876726,
+      "learning_rate": 1.987395376070117e-05,
+      "loss": 0.9118,
+      "step": 3357
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1938642295951745,
+      "learning_rate": 1.9873832963089512e-05,
+      "loss": 1.081,
+      "step": 3358
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.665453189937007,
+      "learning_rate": 1.9873712107989273e-05,
+      "loss": 0.9722,
+      "step": 3359
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.131692948124317,
+      "learning_rate": 1.9873591195401157e-05,
+      "loss": 0.9405,
+      "step": 3360
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.765987547788053,
+      "learning_rate": 1.9873470225325868e-05,
+      "loss": 0.9691,
+      "step": 3361
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9101458837002614,
+      "learning_rate": 1.987334919776411e-05,
+      "loss": 0.8427,
+      "step": 3362
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5723109019983843,
+      "learning_rate": 1.9873228112716583e-05,
+      "loss": 0.8855,
+      "step": 3363
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7455786107404625,
+      "learning_rate": 1.9873106970184e-05,
+      "loss": 0.943,
+      "step": 3364
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.277476436169869,
+      "learning_rate": 1.987298577016706e-05,
+      "loss": 0.9137,
+      "step": 3365
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7626902721523172,
+      "learning_rate": 1.9872864512666473e-05,
+      "loss": 0.889,
+      "step": 3366
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.736809270233882,
+      "learning_rate": 1.987274319768294e-05,
+      "loss": 0.8254,
+      "step": 3367
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5584516466089746,
+      "learning_rate": 1.9872621825217174e-05,
+      "loss": 0.9061,
+      "step": 3368
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5565395220308456,
+      "learning_rate": 1.9872500395269876e-05,
+      "loss": 0.9837,
+      "step": 3369
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4107865393408745,
+      "learning_rate": 1.9872378907841757e-05,
+      "loss": 0.8162,
+      "step": 3370
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.543437106161686,
+      "learning_rate": 1.987225736293352e-05,
+      "loss": 0.935,
+      "step": 3371
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6320790194712695,
+      "learning_rate": 1.9872135760545877e-05,
+      "loss": 0.9285,
+      "step": 3372
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8017101141966068,
+      "learning_rate": 1.9872014100679533e-05,
+      "loss": 0.9146,
+      "step": 3373
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9047031762956723,
+      "learning_rate": 1.98718923833352e-05,
+      "loss": 0.917,
+      "step": 3374
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.941441876640537,
+      "learning_rate": 1.987177060851358e-05,
+      "loss": 0.9056,
+      "step": 3375
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8989606608198557,
+      "learning_rate": 1.987164877621539e-05,
+      "loss": 0.8399,
+      "step": 3376
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8569115141560437,
+      "learning_rate": 1.9871526886441334e-05,
+      "loss": 0.752,
+      "step": 3377
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8259901066575739,
+      "learning_rate": 1.9871404939192123e-05,
+      "loss": 0.8132,
+      "step": 3378
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.674096432913074,
+      "learning_rate": 1.9871282934468467e-05,
+      "loss": 0.8992,
+      "step": 3379
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.597906159433511,
+      "learning_rate": 1.9871160872271078e-05,
+      "loss": 1.008,
+      "step": 3380
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7953086652384,
+      "learning_rate": 1.9871038752600663e-05,
+      "loss": 0.7456,
+      "step": 3381
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8169952347144537,
+      "learning_rate": 1.987091657545794e-05,
+      "loss": 0.7804,
+      "step": 3382
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.661181963809303,
+      "learning_rate": 1.9870794340843608e-05,
+      "loss": 1.0635,
+      "step": 3383
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.821839515541472,
+      "learning_rate": 1.987067204875839e-05,
+      "loss": 0.7676,
+      "step": 3384
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8331940086679075,
+      "learning_rate": 1.9870549699202994e-05,
+      "loss": 0.9369,
+      "step": 3385
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7436655396276224,
+      "learning_rate": 1.9870427292178136e-05,
+      "loss": 1.0504,
+      "step": 3386
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9260403291275505,
+      "learning_rate": 1.9870304827684522e-05,
+      "loss": 0.9001,
+      "step": 3387
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.527290078504295,
+      "learning_rate": 1.9870182305722867e-05,
+      "loss": 0.8795,
+      "step": 3388
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7426474211492922,
+      "learning_rate": 1.9870059726293892e-05,
+      "loss": 1.0042,
+      "step": 3389
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7163567683457497,
+      "learning_rate": 1.9869937089398296e-05,
+      "loss": 0.9323,
+      "step": 3390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.637572950522375,
+      "learning_rate": 1.986981439503681e-05,
+      "loss": 0.9373,
+      "step": 3391
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7497491179910292,
+      "learning_rate": 1.9869691643210134e-05,
+      "loss": 0.8091,
+      "step": 3392
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7180542059025947,
+      "learning_rate": 1.986956883391899e-05,
+      "loss": 0.8993,
+      "step": 3393
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6628402665812678,
+      "learning_rate": 1.9869445967164093e-05,
+      "loss": 1.0577,
+      "step": 3394
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6866357529839813,
+      "learning_rate": 1.9869323042946155e-05,
+      "loss": 0.9779,
+      "step": 3395
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4340677611532682,
+      "learning_rate": 1.9869200061265893e-05,
+      "loss": 0.7702,
+      "step": 3396
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5376601860807726,
+      "learning_rate": 1.9869077022124026e-05,
+      "loss": 0.9858,
+      "step": 3397
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7579528586988131,
+      "learning_rate": 1.9868953925521267e-05,
+      "loss": 0.8265,
+      "step": 3398
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.418722433414733,
+      "learning_rate": 1.9868830771458337e-05,
+      "loss": 0.774,
+      "step": 3399
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2283903500155935,
+      "learning_rate": 1.9868707559935945e-05,
+      "loss": 0.9623,
+      "step": 3400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8232536006842526,
+      "learning_rate": 1.9868584290954812e-05,
+      "loss": 0.8357,
+      "step": 3401
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.387452550979432,
+      "learning_rate": 1.9868460964515663e-05,
+      "loss": 0.9387,
+      "step": 3402
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.165461609432982,
+      "learning_rate": 1.9868337580619207e-05,
+      "loss": 0.8496,
+      "step": 3403
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7627060751507153,
+      "learning_rate": 1.9868214139266168e-05,
+      "loss": 0.8976,
+      "step": 3404
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.923444752593228,
+      "learning_rate": 1.9868090640457258e-05,
+      "loss": 0.8329,
+      "step": 3405
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9659137595099516,
+      "learning_rate": 1.98679670841932e-05,
+      "loss": 1.0459,
+      "step": 3406
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3091032953517425,
+      "learning_rate": 1.9867843470474714e-05,
+      "loss": 0.9271,
+      "step": 3407
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5963200779558337,
+      "learning_rate": 1.9867719799302524e-05,
+      "loss": 0.9765,
+      "step": 3408
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1800065500336054,
+      "learning_rate": 1.986759607067734e-05,
+      "loss": 0.8702,
+      "step": 3409
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7400851922614138,
+      "learning_rate": 1.9867472284599888e-05,
+      "loss": 0.9994,
+      "step": 3410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8322081626826743,
+      "learning_rate": 1.986734844107089e-05,
+      "loss": 0.7292,
+      "step": 3411
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.049044595446565,
+      "learning_rate": 1.9867224540091064e-05,
+      "loss": 0.9676,
+      "step": 3412
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7964259845618457,
+      "learning_rate": 1.9867100581661135e-05,
+      "loss": 0.7609,
+      "step": 3413
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0096329055905544,
+      "learning_rate": 1.9866976565781823e-05,
+      "loss": 0.9253,
+      "step": 3414
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.804026875278444,
+      "learning_rate": 1.9866852492453846e-05,
+      "loss": 0.907,
+      "step": 3415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8102176303126266,
+      "learning_rate": 1.986672836167793e-05,
+      "loss": 0.7768,
+      "step": 3416
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5942790181036228,
+      "learning_rate": 1.9866604173454802e-05,
+      "loss": 0.833,
+      "step": 3417
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.590233812591058,
+      "learning_rate": 1.986647992778518e-05,
+      "loss": 1.0054,
+      "step": 3418
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8770605917959966,
+      "learning_rate": 1.986635562466978e-05,
+      "loss": 1.031,
+      "step": 3419
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.234636251540259,
+      "learning_rate": 1.9866231264109343e-05,
+      "loss": 0.8784,
+      "step": 3420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.632115863416291,
+      "learning_rate": 1.986610684610458e-05,
+      "loss": 1.0104,
+      "step": 3421
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.886634163590575,
+      "learning_rate": 1.9865982370656224e-05,
+      "loss": 0.9272,
+      "step": 3422
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7685723635952926,
+      "learning_rate": 1.9865857837764992e-05,
+      "loss": 1.014,
+      "step": 3423
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6665084092324975,
+      "learning_rate": 1.986573324743161e-05,
+      "loss": 0.9061,
+      "step": 3424
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4716162564877278,
+      "learning_rate": 1.986560859965681e-05,
+      "loss": 0.851,
+      "step": 3425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6623062593159692,
+      "learning_rate": 1.9865483894441312e-05,
+      "loss": 0.9183,
+      "step": 3426
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6551513190430882,
+      "learning_rate": 1.986535913178584e-05,
+      "loss": 0.9827,
+      "step": 3427
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5168949151755444,
+      "learning_rate": 1.9865234311691127e-05,
+      "loss": 0.8485,
+      "step": 3428
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9921767184839967,
+      "learning_rate": 1.9865109434157896e-05,
+      "loss": 0.8775,
+      "step": 3429
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7679752551969525,
+      "learning_rate": 1.9864984499186873e-05,
+      "loss": 0.8483,
+      "step": 3430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.733193688002764,
+      "learning_rate": 1.9864859506778787e-05,
+      "loss": 0.8558,
+      "step": 3431
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0008890849100838,
+      "learning_rate": 1.9864734456934372e-05,
+      "loss": 0.8404,
+      "step": 3432
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8344741007379333,
+      "learning_rate": 1.986460934965434e-05,
+      "loss": 0.9032,
+      "step": 3433
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.665816027257373,
+      "learning_rate": 1.9864484184939438e-05,
+      "loss": 0.875,
+      "step": 3434
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8193789033991636,
+      "learning_rate": 1.986435896279038e-05,
+      "loss": 0.9027,
+      "step": 3435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9821472390709574,
+      "learning_rate": 1.9864233683207907e-05,
+      "loss": 0.9668,
+      "step": 3436
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.625771324896708,
+      "learning_rate": 1.986410834619274e-05,
+      "loss": 0.9174,
+      "step": 3437
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7698449050987513,
+      "learning_rate": 1.986398295174561e-05,
+      "loss": 0.9073,
+      "step": 3438
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.778231159637329,
+      "learning_rate": 1.9863857499867248e-05,
+      "loss": 0.8766,
+      "step": 3439
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8082393062920465,
+      "learning_rate": 1.9863731990558386e-05,
+      "loss": 0.9248,
+      "step": 3440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6560888860278025,
+      "learning_rate": 1.9863606423819753e-05,
+      "loss": 0.9402,
+      "step": 3441
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6119152787282374,
+      "learning_rate": 1.9863480799652078e-05,
+      "loss": 0.8863,
+      "step": 3442
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4649659640271835,
+      "learning_rate": 1.9863355118056102e-05,
+      "loss": 1.0501,
+      "step": 3443
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2510601363317275,
+      "learning_rate": 1.9863229379032542e-05,
+      "loss": 0.9259,
+      "step": 3444
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.606636830844684,
+      "learning_rate": 1.9863103582582144e-05,
+      "loss": 0.8263,
+      "step": 3445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7023530103359714,
+      "learning_rate": 1.9862977728705633e-05,
+      "loss": 0.9721,
+      "step": 3446
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0745702756197533,
+      "learning_rate": 1.986285181740374e-05,
+      "loss": 0.7891,
+      "step": 3447
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2662554459922823,
+      "learning_rate": 1.98627258486772e-05,
+      "loss": 0.8187,
+      "step": 3448
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9129652354580151,
+      "learning_rate": 1.9862599822526755e-05,
+      "loss": 0.8837,
+      "step": 3449
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7855983457205816,
+      "learning_rate": 1.9862473738953127e-05,
+      "loss": 0.758,
+      "step": 3450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6519312992169821,
+      "learning_rate": 1.9862347597957054e-05,
+      "loss": 1.0108,
+      "step": 3451
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6929424948780163,
+      "learning_rate": 1.9862221399539273e-05,
+      "loss": 0.8807,
+      "step": 3452
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6527553444417131,
+      "learning_rate": 1.9862095143700512e-05,
+      "loss": 1.0571,
+      "step": 3453
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.681742171668804,
+      "learning_rate": 1.986196883044152e-05,
+      "loss": 0.9805,
+      "step": 3454
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7176616255097565,
+      "learning_rate": 1.9861842459763014e-05,
+      "loss": 0.9363,
+      "step": 3455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7139490041637444,
+      "learning_rate": 1.9861716031665744e-05,
+      "loss": 0.9877,
+      "step": 3456
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.869220123629981,
+      "learning_rate": 1.986158954615044e-05,
+      "loss": 0.7632,
+      "step": 3457
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.034452023852105,
+      "learning_rate": 1.986146300321784e-05,
+      "loss": 0.9516,
+      "step": 3458
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7902422233531916,
+      "learning_rate": 1.9861336402868678e-05,
+      "loss": 0.9418,
+      "step": 3459
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6328859997849179,
+      "learning_rate": 1.9861209745103693e-05,
+      "loss": 1.0774,
+      "step": 3460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0559679147445133,
+      "learning_rate": 1.9861083029923623e-05,
+      "loss": 0.9649,
+      "step": 3461
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7956082174118597,
+      "learning_rate": 1.986095625732921e-05,
+      "loss": 0.7706,
+      "step": 3462
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.695468268140954,
+      "learning_rate": 1.9860829427321184e-05,
+      "loss": 0.932,
+      "step": 3463
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0184027886760445,
+      "learning_rate": 1.9860702539900288e-05,
+      "loss": 0.8849,
+      "step": 3464
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.573320961390596,
+      "learning_rate": 1.986057559506726e-05,
+      "loss": 0.9648,
+      "step": 3465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5683035298350858,
+      "learning_rate": 1.9860448592822837e-05,
+      "loss": 0.9162,
+      "step": 3466
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6388181971952342,
+      "learning_rate": 1.9860321533167763e-05,
+      "loss": 0.7709,
+      "step": 3467
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6679334734485072,
+      "learning_rate": 1.9860194416102777e-05,
+      "loss": 0.9304,
+      "step": 3468
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.776304656621465,
+      "learning_rate": 1.9860067241628616e-05,
+      "loss": 0.8236,
+      "step": 3469
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.826077011615013,
+      "learning_rate": 1.985994000974602e-05,
+      "loss": 0.8316,
+      "step": 3470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9486413971086083,
+      "learning_rate": 1.985981272045573e-05,
+      "loss": 0.9959,
+      "step": 3471
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7872319435308874,
+      "learning_rate": 1.9859685373758493e-05,
+      "loss": 0.9568,
+      "step": 3472
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5763918438684845,
+      "learning_rate": 1.9859557969655045e-05,
+      "loss": 0.9829,
+      "step": 3473
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8274134886615883,
+      "learning_rate": 1.9859430508146128e-05,
+      "loss": 0.7274,
+      "step": 3474
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.671926885334101,
+      "learning_rate": 1.9859302989232486e-05,
+      "loss": 0.9946,
+      "step": 3475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.454118349937211,
+      "learning_rate": 1.9859175412914863e-05,
+      "loss": 1.0019,
+      "step": 3476
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7677874438221755,
+      "learning_rate": 1.9859047779194e-05,
+      "loss": 0.9554,
+      "step": 3477
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6437378964950498,
+      "learning_rate": 1.9858920088070636e-05,
+      "loss": 0.7991,
+      "step": 3478
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.904295926781565,
+      "learning_rate": 1.9858792339545515e-05,
+      "loss": 0.9287,
+      "step": 3479
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7959801833577499,
+      "learning_rate": 1.9858664533619388e-05,
+      "loss": 0.6924,
+      "step": 3480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7172925386007127,
+      "learning_rate": 1.9858536670292995e-05,
+      "loss": 0.9842,
+      "step": 3481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1073684137890885,
+      "learning_rate": 1.985840874956708e-05,
+      "loss": 0.954,
+      "step": 3482
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8655382234059121,
+      "learning_rate": 1.9858280771442387e-05,
+      "loss": 1.0405,
+      "step": 3483
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5446576185855037,
+      "learning_rate": 1.985815273591966e-05,
+      "loss": 0.892,
+      "step": 3484
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8345834976937326,
+      "learning_rate": 1.9858024642999655e-05,
+      "loss": 0.7159,
+      "step": 3485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1134596934291774,
+      "learning_rate": 1.98578964926831e-05,
+      "loss": 0.8974,
+      "step": 3486
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.81459575511173,
+      "learning_rate": 1.9857768284970754e-05,
+      "loss": 0.9683,
+      "step": 3487
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.538598629410682,
+      "learning_rate": 1.9857640019863364e-05,
+      "loss": 0.8772,
+      "step": 3488
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7204969278657725,
+      "learning_rate": 1.985751169736167e-05,
+      "loss": 1.0352,
+      "step": 3489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6059978528088636,
+      "learning_rate": 1.985738331746642e-05,
+      "loss": 0.849,
+      "step": 3490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.254182741467438,
+      "learning_rate": 1.9857254880178368e-05,
+      "loss": 0.8811,
+      "step": 3491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4678261602170837,
+      "learning_rate": 1.9857126385498255e-05,
+      "loss": 0.9599,
+      "step": 3492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4414184025524188,
+      "learning_rate": 1.9856997833426828e-05,
+      "loss": 1.0627,
+      "step": 3493
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6479585584446483,
+      "learning_rate": 1.9856869223964844e-05,
+      "loss": 0.9696,
+      "step": 3494
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4090371778209185,
+      "learning_rate": 1.9856740557113046e-05,
+      "loss": 1.0387,
+      "step": 3495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8379844266578662,
+      "learning_rate": 1.9856611832872185e-05,
+      "loss": 0.7719,
+      "step": 3496
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6881166466184496,
+      "learning_rate": 1.9856483051243005e-05,
+      "loss": 0.995,
+      "step": 3497
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7987910262596154,
+      "learning_rate": 1.9856354212226262e-05,
+      "loss": 0.9003,
+      "step": 3498
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9673955928789655,
+      "learning_rate": 1.9856225315822705e-05,
+      "loss": 1.0328,
+      "step": 3499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7922589836091977,
+      "learning_rate": 1.9856096362033083e-05,
+      "loss": 0.7629,
+      "step": 3500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1634334376033513,
+      "learning_rate": 1.9855967350858147e-05,
+      "loss": 0.9183,
+      "step": 3501
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8882458187407416,
+      "learning_rate": 1.985583828229865e-05,
+      "loss": 0.9477,
+      "step": 3502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.135118737806275,
+      "learning_rate": 1.985570915635534e-05,
+      "loss": 0.9901,
+      "step": 3503
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6440544834525384,
+      "learning_rate": 1.985557997302898e-05,
+      "loss": 0.9465,
+      "step": 3504
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.592103434860596,
+      "learning_rate": 1.98554507323203e-05,
+      "loss": 0.9905,
+      "step": 3505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.570372349744076,
+      "learning_rate": 1.9855321434230075e-05,
+      "loss": 0.856,
+      "step": 3506
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5834526434814726,
+      "learning_rate": 1.9855192078759045e-05,
+      "loss": 0.8168,
+      "step": 3507
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3237680349657004,
+      "learning_rate": 1.9855062665907967e-05,
+      "loss": 0.9516,
+      "step": 3508
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7027388749697079,
+      "learning_rate": 1.9854933195677596e-05,
+      "loss": 0.8961,
+      "step": 3509
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6414741420247319,
+      "learning_rate": 1.9854803668068682e-05,
+      "loss": 1.0232,
+      "step": 3510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9377793749329475,
+      "learning_rate": 1.985467408308198e-05,
+      "loss": 0.7947,
+      "step": 3511
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.635009473217777,
+      "learning_rate": 1.9854544440718247e-05,
+      "loss": 0.7989,
+      "step": 3512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.02042600446278,
+      "learning_rate": 1.9854414740978235e-05,
+      "loss": 0.925,
+      "step": 3513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2897952314581613,
+      "learning_rate": 1.98542849838627e-05,
+      "loss": 1.0802,
+      "step": 3514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7721799420835,
+      "learning_rate": 1.98541551693724e-05,
+      "loss": 0.8191,
+      "step": 3515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6234337466699538,
+      "learning_rate": 1.985402529750809e-05,
+      "loss": 1.003,
+      "step": 3516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7009368359722064,
+      "learning_rate": 1.9853895368270524e-05,
+      "loss": 1.0043,
+      "step": 3517
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6327747672393413,
+      "learning_rate": 1.9853765381660456e-05,
+      "loss": 0.9955,
+      "step": 3518
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8825875159985668,
+      "learning_rate": 1.985363533767865e-05,
+      "loss": 0.9173,
+      "step": 3519
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7457912356725753,
+      "learning_rate": 1.9853505236325857e-05,
+      "loss": 1.0018,
+      "step": 3520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6194625450598241,
+      "learning_rate": 1.985337507760284e-05,
+      "loss": 0.9933,
+      "step": 3521
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5818887166950841,
+      "learning_rate": 1.985324486151035e-05,
+      "loss": 1.0269,
+      "step": 3522
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.206467649563391,
+      "learning_rate": 1.9853114588049154e-05,
+      "loss": 0.8538,
+      "step": 3523
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.220013052919128,
+      "learning_rate": 1.9852984257220004e-05,
+      "loss": 0.7805,
+      "step": 3524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6783036622445147,
+      "learning_rate": 1.9852853869023656e-05,
+      "loss": 0.9421,
+      "step": 3525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1342126473343512,
+      "learning_rate": 1.9852723423460878e-05,
+      "loss": 1.0167,
+      "step": 3526
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5192369696529995,
+      "learning_rate": 1.9852592920532424e-05,
+      "loss": 0.9627,
+      "step": 3527
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7689744067472792,
+      "learning_rate": 1.9852462360239053e-05,
+      "loss": 0.7567,
+      "step": 3528
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5664951362304025,
+      "learning_rate": 1.9852331742581528e-05,
+      "loss": 0.9829,
+      "step": 3529
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8250006018677525,
+      "learning_rate": 1.9852201067560607e-05,
+      "loss": 1.0532,
+      "step": 3530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.636124685549085,
+      "learning_rate": 1.9852070335177057e-05,
+      "loss": 0.9572,
+      "step": 3531
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6493257337269227,
+      "learning_rate": 1.985193954543163e-05,
+      "loss": 0.806,
+      "step": 3532
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5974479025515795,
+      "learning_rate": 1.985180869832509e-05,
+      "loss": 0.9285,
+      "step": 3533
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.42555859721322,
+      "learning_rate": 1.9851677793858206e-05,
+      "loss": 0.8973,
+      "step": 3534
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7498138944580244,
+      "learning_rate": 1.985154683203173e-05,
+      "loss": 1.0258,
+      "step": 3535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7060175073414847,
+      "learning_rate": 1.9851415812846432e-05,
+      "loss": 0.9421,
+      "step": 3536
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5469884174228044,
+      "learning_rate": 1.985128473630307e-05,
+      "loss": 0.9274,
+      "step": 3537
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6650664513606221,
+      "learning_rate": 1.9851153602402414e-05,
+      "loss": 0.9775,
+      "step": 3538
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7295924757176098,
+      "learning_rate": 1.9851022411145222e-05,
+      "loss": 0.9495,
+      "step": 3539
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6017675033428402,
+      "learning_rate": 1.9850891162532255e-05,
+      "loss": 0.9876,
+      "step": 3540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8308688130574627,
+      "learning_rate": 1.9850759856564285e-05,
+      "loss": 0.8026,
+      "step": 3541
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.575559850128685,
+      "learning_rate": 1.9850628493242067e-05,
+      "loss": 0.7867,
+      "step": 3542
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4248106685418767,
+      "learning_rate": 1.9850497072566377e-05,
+      "loss": 0.9839,
+      "step": 3543
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5437533431473045,
+      "learning_rate": 1.9850365594537972e-05,
+      "loss": 0.9389,
+      "step": 3544
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7586550592036312,
+      "learning_rate": 1.985023405915762e-05,
+      "loss": 0.7404,
+      "step": 3545
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8071536160896559,
+      "learning_rate": 1.9850102466426086e-05,
+      "loss": 0.763,
+      "step": 3546
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.541145445851392,
+      "learning_rate": 1.984997081634414e-05,
+      "loss": 0.8627,
+      "step": 3547
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6062703242225305,
+      "learning_rate": 1.9849839108912544e-05,
+      "loss": 0.8374,
+      "step": 3548
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5906124090369,
+      "learning_rate": 1.9849707344132065e-05,
+      "loss": 0.9046,
+      "step": 3549
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3993939027144644,
+      "learning_rate": 1.984957552200347e-05,
+      "loss": 0.8156,
+      "step": 3550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5794653048018048,
+      "learning_rate": 1.984944364252753e-05,
+      "loss": 0.8764,
+      "step": 3551
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7530150092732923,
+      "learning_rate": 1.984931170570501e-05,
+      "loss": 0.8487,
+      "step": 3552
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5631589694284174,
+      "learning_rate": 1.984917971153668e-05,
+      "loss": 0.9207,
+      "step": 3553
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7066015308691411,
+      "learning_rate": 1.984904766002331e-05,
+      "loss": 0.8564,
+      "step": 3554
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6171739304201869,
+      "learning_rate": 1.984891555116566e-05,
+      "loss": 0.935,
+      "step": 3555
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7510049087598124,
+      "learning_rate": 1.9848783384964513e-05,
+      "loss": 0.7802,
+      "step": 3556
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8520260386356748,
+      "learning_rate": 1.9848651161420625e-05,
+      "loss": 0.8606,
+      "step": 3557
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7439646998599985,
+      "learning_rate": 1.9848518880534775e-05,
+      "loss": 0.9653,
+      "step": 3558
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0721545577424276,
+      "learning_rate": 1.9848386542307728e-05,
+      "loss": 0.7827,
+      "step": 3559
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.854255456200187,
+      "learning_rate": 1.9848254146740257e-05,
+      "loss": 1.0365,
+      "step": 3560
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.106000642023113,
+      "learning_rate": 1.9848121693833135e-05,
+      "loss": 0.9537,
+      "step": 3561
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2593147450651725,
+      "learning_rate": 1.9847989183587127e-05,
+      "loss": 0.9291,
+      "step": 3562
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8011124685504792,
+      "learning_rate": 1.9847856616003013e-05,
+      "loss": 0.7884,
+      "step": 3563
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9643252277014038,
+      "learning_rate": 1.9847723991081555e-05,
+      "loss": 0.9468,
+      "step": 3564
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2773353982800084,
+      "learning_rate": 1.9847591308823535e-05,
+      "loss": 0.9747,
+      "step": 3565
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.165492651122838,
+      "learning_rate": 1.9847458569229716e-05,
+      "loss": 0.8995,
+      "step": 3566
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.63539105443488,
+      "learning_rate": 1.9847325772300878e-05,
+      "loss": 1.0032,
+      "step": 3567
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7350856052637098,
+      "learning_rate": 1.984719291803779e-05,
+      "loss": 0.9739,
+      "step": 3568
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5590539653486326,
+      "learning_rate": 1.984706000644123e-05,
+      "loss": 0.9011,
+      "step": 3569
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7428609538245563,
+      "learning_rate": 1.984692703751197e-05,
+      "loss": 0.8841,
+      "step": 3570
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7884127499918341,
+      "learning_rate": 1.984679401125078e-05,
+      "loss": 0.7326,
+      "step": 3571
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.006915165127363,
+      "learning_rate": 1.9846660927658437e-05,
+      "loss": 0.9555,
+      "step": 3572
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6748689190660397,
+      "learning_rate": 1.9846527786735722e-05,
+      "loss": 0.9943,
+      "step": 3573
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6257991685232454,
+      "learning_rate": 1.98463945884834e-05,
+      "loss": 1.0522,
+      "step": 3574
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7073994585932242,
+      "learning_rate": 1.984626133290225e-05,
+      "loss": 0.9604,
+      "step": 3575
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8736657579486113,
+      "learning_rate": 1.9846128019993054e-05,
+      "loss": 0.9447,
+      "step": 3576
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7037301499741817,
+      "learning_rate": 1.984599464975658e-05,
+      "loss": 0.9103,
+      "step": 3577
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.606569134891912,
+      "learning_rate": 1.984586122219361e-05,
+      "loss": 0.7613,
+      "step": 3578
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.178016755333668,
+      "learning_rate": 1.9845727737304917e-05,
+      "loss": 0.8144,
+      "step": 3579
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.897905837420245,
+      "learning_rate": 1.9845594195091283e-05,
+      "loss": 0.8131,
+      "step": 3580
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5199293719651892,
+      "learning_rate": 1.9845460595553482e-05,
+      "loss": 0.8438,
+      "step": 3581
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9389640977536728,
+      "learning_rate": 1.9845326938692292e-05,
+      "loss": 0.8367,
+      "step": 3582
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5389851640560037,
+      "learning_rate": 1.984519322450849e-05,
+      "loss": 0.9839,
+      "step": 3583
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8906008379667862,
+      "learning_rate": 1.9845059453002858e-05,
+      "loss": 0.892,
+      "step": 3584
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7868038264156751,
+      "learning_rate": 1.984492562417617e-05,
+      "loss": 0.7677,
+      "step": 3585
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.866570498525933,
+      "learning_rate": 1.984479173802921e-05,
+      "loss": 0.9444,
+      "step": 3586
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6112478902918412,
+      "learning_rate": 1.9844657794562755e-05,
+      "loss": 0.9563,
+      "step": 3587
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9177263418926764,
+      "learning_rate": 1.984452379377759e-05,
+      "loss": 1.0026,
+      "step": 3588
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.871749652002769,
+      "learning_rate": 1.9844389735674484e-05,
+      "loss": 0.9172,
+      "step": 3589
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.705771144090008,
+      "learning_rate": 1.9844255620254227e-05,
+      "loss": 0.8674,
+      "step": 3590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8130538428943057,
+      "learning_rate": 1.98441214475176e-05,
+      "loss": 0.8661,
+      "step": 3591
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8055332296518902,
+      "learning_rate": 1.9843987217465378e-05,
+      "loss": 0.7926,
+      "step": 3592
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6550494350860305,
+      "learning_rate": 1.9843852930098345e-05,
+      "loss": 0.9827,
+      "step": 3593
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.75188949341092,
+      "learning_rate": 1.9843718585417284e-05,
+      "loss": 0.8711,
+      "step": 3594
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6261462972516758,
+      "learning_rate": 1.9843584183422983e-05,
+      "loss": 0.9444,
+      "step": 3595
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1246212922895555,
+      "learning_rate": 1.984344972411621e-05,
+      "loss": 0.9243,
+      "step": 3596
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5155070870757643,
+      "learning_rate": 1.984331520749776e-05,
+      "loss": 0.9814,
+      "step": 3597
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7424982849098158,
+      "learning_rate": 1.9843180633568412e-05,
+      "loss": 1.0012,
+      "step": 3598
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9133782962071784,
+      "learning_rate": 1.9843046002328953e-05,
+      "loss": 0.9512,
+      "step": 3599
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7532410503839364,
+      "learning_rate": 1.984291131378016e-05,
+      "loss": 0.9414,
+      "step": 3600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0276538655561445,
+      "learning_rate": 1.9842776567922825e-05,
+      "loss": 0.8959,
+      "step": 3601
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9881522218550036,
+      "learning_rate": 1.9842641764757724e-05,
+      "loss": 0.9052,
+      "step": 3602
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.060358744901419,
+      "learning_rate": 1.984250690428565e-05,
+      "loss": 0.8971,
+      "step": 3603
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7582033573981979,
+      "learning_rate": 1.9842371986507384e-05,
+      "loss": 0.7122,
+      "step": 3604
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7054991618070259,
+      "learning_rate": 1.9842237011423712e-05,
+      "loss": 0.9067,
+      "step": 3605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.92543400505898,
+      "learning_rate": 1.9842101979035416e-05,
+      "loss": 0.7908,
+      "step": 3606
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7328916100034641,
+      "learning_rate": 1.984196688934329e-05,
+      "loss": 0.9139,
+      "step": 3607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9464125297699273,
+      "learning_rate": 1.984183174234812e-05,
+      "loss": 1.0196,
+      "step": 3608
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7317890329259125,
+      "learning_rate": 1.9841696538050686e-05,
+      "loss": 0.9016,
+      "step": 3609
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.495363630947435,
+      "learning_rate": 1.984156127645178e-05,
+      "loss": 0.8519,
+      "step": 3610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8016339842797249,
+      "learning_rate": 1.9841425957552188e-05,
+      "loss": 1.0458,
+      "step": 3611
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.849749261512491,
+      "learning_rate": 1.98412905813527e-05,
+      "loss": 0.9219,
+      "step": 3612
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7591711504119616,
+      "learning_rate": 1.98411551478541e-05,
+      "loss": 0.8853,
+      "step": 3613
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6874912555827302,
+      "learning_rate": 1.984101965705718e-05,
+      "loss": 1.0294,
+      "step": 3614
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.072644691001653,
+      "learning_rate": 1.984088410896273e-05,
+      "loss": 0.8483,
+      "step": 3615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7726802697610953,
+      "learning_rate": 1.9840748503571535e-05,
+      "loss": 0.9679,
+      "step": 3616
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8056911475332169,
+      "learning_rate": 1.9840612840884386e-05,
+      "loss": 0.7614,
+      "step": 3617
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8002117730741813,
+      "learning_rate": 1.9840477120902076e-05,
+      "loss": 0.9104,
+      "step": 3618
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5593636499517538,
+      "learning_rate": 1.984034134362539e-05,
+      "loss": 0.8476,
+      "step": 3619
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.680321719808709,
+      "learning_rate": 1.9840205509055123e-05,
+      "loss": 0.7837,
+      "step": 3620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7633954064909876,
+      "learning_rate": 1.9840069617192063e-05,
+      "loss": 0.7318,
+      "step": 3621
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6854771932953074,
+      "learning_rate": 1.9839933668037003e-05,
+      "loss": 0.8555,
+      "step": 3622
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6851971110819695,
+      "learning_rate": 1.9839797661590734e-05,
+      "loss": 0.8612,
+      "step": 3623
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7992861908756279,
+      "learning_rate": 1.9839661597854048e-05,
+      "loss": 0.7383,
+      "step": 3624
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6919808321021015,
+      "learning_rate": 1.983952547682774e-05,
+      "loss": 0.8645,
+      "step": 3625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4936165286256773,
+      "learning_rate": 1.9839389298512593e-05,
+      "loss": 0.93,
+      "step": 3626
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5179159510726792,
+      "learning_rate": 1.9839253062909408e-05,
+      "loss": 1.0145,
+      "step": 3627
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8300787843881914,
+      "learning_rate": 1.983911677001898e-05,
+      "loss": 0.9995,
+      "step": 3628
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6559197516496031,
+      "learning_rate": 1.98389804198421e-05,
+      "loss": 0.9111,
+      "step": 3629
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8861237233925439,
+      "learning_rate": 1.9838844012379557e-05,
+      "loss": 0.9374,
+      "step": 3630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7717956858542321,
+      "learning_rate": 1.983870754763215e-05,
+      "loss": 0.7334,
+      "step": 3631
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7708612082662821,
+      "learning_rate": 1.9838571025600674e-05,
+      "loss": 0.7528,
+      "step": 3632
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.528368631329786,
+      "learning_rate": 1.983843444628592e-05,
+      "loss": 1.0593,
+      "step": 3633
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2464063927759423,
+      "learning_rate": 1.983829780968869e-05,
+      "loss": 0.946,
+      "step": 3634
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.690035947714621,
+      "learning_rate": 1.9838161115809773e-05,
+      "loss": 0.871,
+      "step": 3635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7680470699431812,
+      "learning_rate": 1.983802436464997e-05,
+      "loss": 0.865,
+      "step": 3636
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7986137614694451,
+      "learning_rate": 1.983788755621007e-05,
+      "loss": 0.7553,
+      "step": 3637
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.010709226440249,
+      "learning_rate": 1.9837750690490876e-05,
+      "loss": 0.9517,
+      "step": 3638
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8991284809392088,
+      "learning_rate": 1.9837613767493186e-05,
+      "loss": 0.9628,
+      "step": 3639
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.531529894385908,
+      "learning_rate": 1.983747678721779e-05,
+      "loss": 0.8927,
+      "step": 3640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8424350052373855,
+      "learning_rate": 1.983733974966549e-05,
+      "loss": 1.0308,
+      "step": 3641
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.089665546155778,
+      "learning_rate": 1.9837202654837085e-05,
+      "loss": 0.9848,
+      "step": 3642
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5238720219158401,
+      "learning_rate": 1.983706550273337e-05,
+      "loss": 0.9916,
+      "step": 3643
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9380393967773921,
+      "learning_rate": 1.9836928293355146e-05,
+      "loss": 0.9122,
+      "step": 3644
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7871015623896425,
+      "learning_rate": 1.9836791026703213e-05,
+      "loss": 0.7957,
+      "step": 3645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.47527701119078,
+      "learning_rate": 1.983665370277837e-05,
+      "loss": 0.8662,
+      "step": 3646
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.075691867660224,
+      "learning_rate": 1.9836516321581408e-05,
+      "loss": 0.9252,
+      "step": 3647
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7945532473097955,
+      "learning_rate": 1.9836378883113137e-05,
+      "loss": 0.7157,
+      "step": 3648
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6328451127727852,
+      "learning_rate": 1.9836241387374354e-05,
+      "loss": 0.9604,
+      "step": 3649
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.736362544983498,
+      "learning_rate": 1.9836103834365864e-05,
+      "loss": 0.8864,
+      "step": 3650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7104363341272253,
+      "learning_rate": 1.983596622408846e-05,
+      "loss": 0.9534,
+      "step": 3651
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.555825335572532,
+      "learning_rate": 1.9835828556542944e-05,
+      "loss": 0.8324,
+      "step": 3652
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.855123099581955,
+      "learning_rate": 1.9835690831730125e-05,
+      "loss": 0.9717,
+      "step": 3653
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7676723902782134,
+      "learning_rate": 1.9835553049650796e-05,
+      "loss": 0.915,
+      "step": 3654
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9223931904827356,
+      "learning_rate": 1.9835415210305763e-05,
+      "loss": 0.9447,
+      "step": 3655
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7574160026501678,
+      "learning_rate": 1.983527731369583e-05,
+      "loss": 0.7474,
+      "step": 3656
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.494691910341888,
+      "learning_rate": 1.98351393598218e-05,
+      "loss": 0.8906,
+      "step": 3657
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6749202251897721,
+      "learning_rate": 1.9835001348684476e-05,
+      "loss": 0.8928,
+      "step": 3658
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6858522590835125,
+      "learning_rate": 1.9834863280284657e-05,
+      "loss": 0.8908,
+      "step": 3659
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9551025645975015,
+      "learning_rate": 1.983472515462315e-05,
+      "loss": 0.8718,
+      "step": 3660
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.9019927765866,
+      "learning_rate": 1.983458697170076e-05,
+      "loss": 0.8236,
+      "step": 3661
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5727819506422596,
+      "learning_rate": 1.9834448731518296e-05,
+      "loss": 0.9656,
+      "step": 3662
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.764210656511387,
+      "learning_rate": 1.9834310434076553e-05,
+      "loss": 0.7341,
+      "step": 3663
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7629378980052064,
+      "learning_rate": 1.9834172079376342e-05,
+      "loss": 0.9108,
+      "step": 3664
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8681835931855284,
+      "learning_rate": 1.983403366741847e-05,
+      "loss": 0.9395,
+      "step": 3665
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9352368495710879,
+      "learning_rate": 1.9833895198203737e-05,
+      "loss": 0.8805,
+      "step": 3666
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6023179005495551,
+      "learning_rate": 1.9833756671732956e-05,
+      "loss": 0.8574,
+      "step": 3667
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7564573666415533,
+      "learning_rate": 1.983361808800693e-05,
+      "loss": 0.873,
+      "step": 3668
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5006425578179416,
+      "learning_rate": 1.9833479447026464e-05,
+      "loss": 0.9527,
+      "step": 3669
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7405261971325658,
+      "learning_rate": 1.983334074879237e-05,
+      "loss": 0.9775,
+      "step": 3670
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7530963105462396,
+      "learning_rate": 1.9833201993305452e-05,
+      "loss": 0.6668,
+      "step": 3671
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6419770445280812,
+      "learning_rate": 1.9833063180566517e-05,
+      "loss": 1.1111,
+      "step": 3672
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6444980562312086,
+      "learning_rate": 1.9832924310576377e-05,
+      "loss": 1.003,
+      "step": 3673
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8453964389232309,
+      "learning_rate": 1.983278538333584e-05,
+      "loss": 0.8043,
+      "step": 3674
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7591639507538428,
+      "learning_rate": 1.983264639884571e-05,
+      "loss": 1.0059,
+      "step": 3675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6897312382602434,
+      "learning_rate": 1.98325073571068e-05,
+      "loss": 0.8942,
+      "step": 3676
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.597144235283309,
+      "learning_rate": 1.983236825811992e-05,
+      "loss": 0.8472,
+      "step": 3677
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.210335198206724,
+      "learning_rate": 1.9832229101885882e-05,
+      "loss": 0.8827,
+      "step": 3678
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7558508127134782,
+      "learning_rate": 1.983208988840549e-05,
+      "loss": 0.6665,
+      "step": 3679
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5661484291247771,
+      "learning_rate": 1.9831950617679562e-05,
+      "loss": 0.9516,
+      "step": 3680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8648972129013839,
+      "learning_rate": 1.98318112897089e-05,
+      "loss": 0.8872,
+      "step": 3681
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0892667855161053,
+      "learning_rate": 1.983167190449432e-05,
+      "loss": 0.9284,
+      "step": 3682
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.304375724621369,
+      "learning_rate": 1.9831532462036634e-05,
+      "loss": 0.8985,
+      "step": 3683
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5702289046700024,
+      "learning_rate": 1.983139296233666e-05,
+      "loss": 1.0458,
+      "step": 3684
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9561764161857984,
+      "learning_rate": 1.9831253405395196e-05,
+      "loss": 0.9866,
+      "step": 3685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6245680875883521,
+      "learning_rate": 1.9831113791213062e-05,
+      "loss": 0.7821,
+      "step": 3686
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7757510330980977,
+      "learning_rate": 1.9830974119791075e-05,
+      "loss": 0.87,
+      "step": 3687
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7988592727683326,
+      "learning_rate": 1.9830834391130043e-05,
+      "loss": 0.8597,
+      "step": 3688
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.733048817608301,
+      "learning_rate": 1.983069460523078e-05,
+      "loss": 0.9352,
+      "step": 3689
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2850798958373595,
+      "learning_rate": 1.98305547620941e-05,
+      "loss": 0.9207,
+      "step": 3690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.66299590501206,
+      "learning_rate": 1.9830414861720818e-05,
+      "loss": 0.914,
+      "step": 3691
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7826387912957895,
+      "learning_rate": 1.9830274904111754e-05,
+      "loss": 0.7848,
+      "step": 3692
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7448728052205567,
+      "learning_rate": 1.983013488926771e-05,
+      "loss": 0.8637,
+      "step": 3693
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0227281478758763,
+      "learning_rate": 1.9829994817189512e-05,
+      "loss": 0.803,
+      "step": 3694
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.730152616685844,
+      "learning_rate": 1.982985468787797e-05,
+      "loss": 0.9254,
+      "step": 3695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6283359492276839,
+      "learning_rate": 1.9829714501333905e-05,
+      "loss": 0.9625,
+      "step": 3696
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.307792803519718,
+      "learning_rate": 1.9829574257558128e-05,
+      "loss": 0.8799,
+      "step": 3697
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6290618635228016,
+      "learning_rate": 1.982943395655146e-05,
+      "loss": 1.0441,
+      "step": 3698
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.340415368749226,
+      "learning_rate": 1.982929359831471e-05,
+      "loss": 0.8501,
+      "step": 3699
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.061647720647425,
+      "learning_rate": 1.9829153182848705e-05,
+      "loss": 0.9887,
+      "step": 3700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8962624378385213,
+      "learning_rate": 1.9829012710154258e-05,
+      "loss": 1.0156,
+      "step": 3701
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.511772466373624,
+      "learning_rate": 1.9828872180232185e-05,
+      "loss": 0.8455,
+      "step": 3702
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9817995388254876,
+      "learning_rate": 1.982873159308331e-05,
+      "loss": 1.0056,
+      "step": 3703
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.626461395820741,
+      "learning_rate": 1.9828590948708446e-05,
+      "loss": 1.0415,
+      "step": 3704
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7404879662255415,
+      "learning_rate": 1.9828450247108413e-05,
+      "loss": 0.7702,
+      "step": 3705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5373228994137564,
+      "learning_rate": 1.982830948828403e-05,
+      "loss": 0.836,
+      "step": 3706
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6262483064361621,
+      "learning_rate": 1.9828168672236125e-05,
+      "loss": 0.9251,
+      "step": 3707
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0763146239737966,
+      "learning_rate": 1.9828027798965502e-05,
+      "loss": 0.9781,
+      "step": 3708
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.823263155540023,
+      "learning_rate": 1.9827886868472993e-05,
+      "loss": 0.9679,
+      "step": 3709
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.754842306911729,
+      "learning_rate": 1.9827745880759414e-05,
+      "loss": 0.8781,
+      "step": 3710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8097727823145283,
+      "learning_rate": 1.982760483582559e-05,
+      "loss": 0.7605,
+      "step": 3711
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.801739053914763,
+      "learning_rate": 1.9827463733672334e-05,
+      "loss": 0.7855,
+      "step": 3712
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6437731544432181,
+      "learning_rate": 1.9827322574300477e-05,
+      "loss": 0.9316,
+      "step": 3713
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6677052872888065,
+      "learning_rate": 1.9827181357710837e-05,
+      "loss": 0.9399,
+      "step": 3714
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7862564767469866,
+      "learning_rate": 1.9827040083904236e-05,
+      "loss": 0.7144,
+      "step": 3715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5374469942203386,
+      "learning_rate": 1.9826898752881494e-05,
+      "loss": 1.0456,
+      "step": 3716
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7042762335501735,
+      "learning_rate": 1.9826757364643435e-05,
+      "loss": 0.951,
+      "step": 3717
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8325970875260242,
+      "learning_rate": 1.9826615919190886e-05,
+      "loss": 0.7887,
+      "step": 3718
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7874744973006447,
+      "learning_rate": 1.9826474416524667e-05,
+      "loss": 0.9632,
+      "step": 3719
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7870927385466382,
+      "learning_rate": 1.9826332856645603e-05,
+      "loss": 0.6847,
+      "step": 3720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6596544733548155,
+      "learning_rate": 1.9826191239554523e-05,
+      "loss": 0.9317,
+      "step": 3721
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3030048118082007,
+      "learning_rate": 1.9826049565252242e-05,
+      "loss": 0.8057,
+      "step": 3722
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5023828472161214,
+      "learning_rate": 1.982590783373959e-05,
+      "loss": 0.8534,
+      "step": 3723
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.932577875427455,
+      "learning_rate": 1.9825766045017395e-05,
+      "loss": 0.944,
+      "step": 3724
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8893716013919044,
+      "learning_rate": 1.9825624199086473e-05,
+      "loss": 0.8479,
+      "step": 3725
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7934322848679225,
+      "learning_rate": 1.982548229594766e-05,
+      "loss": 0.718,
+      "step": 3726
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.804153919022296,
+      "learning_rate": 1.9825340335601776e-05,
+      "loss": 0.8971,
+      "step": 3727
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6213034114709903,
+      "learning_rate": 1.9825198318049656e-05,
+      "loss": 0.8859,
+      "step": 3728
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6216679564903334,
+      "learning_rate": 1.9825056243292112e-05,
+      "loss": 0.8636,
+      "step": 3729
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6430242154151633,
+      "learning_rate": 1.9824914111329988e-05,
+      "loss": 0.8416,
+      "step": 3730
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6617923857625263,
+      "learning_rate": 1.98247719221641e-05,
+      "loss": 0.9079,
+      "step": 3731
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6851016718993574,
+      "learning_rate": 1.982462967579528e-05,
+      "loss": 1.0621,
+      "step": 3732
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.803626794476327,
+      "learning_rate": 1.9824487372224354e-05,
+      "loss": 0.6958,
+      "step": 3733
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7130886535532754,
+      "learning_rate": 1.9824345011452152e-05,
+      "loss": 0.9381,
+      "step": 3734
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5561052835723101,
+      "learning_rate": 1.9824202593479507e-05,
+      "loss": 0.897,
+      "step": 3735
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6268110244159328,
+      "learning_rate": 1.982406011830724e-05,
+      "loss": 1.0133,
+      "step": 3736
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8354670538648409,
+      "learning_rate": 1.9823917585936185e-05,
+      "loss": 0.7221,
+      "step": 3737
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6153590257550965,
+      "learning_rate": 1.9823774996367177e-05,
+      "loss": 0.8853,
+      "step": 3738
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7721595565074706,
+      "learning_rate": 1.9823632349601033e-05,
+      "loss": 0.9269,
+      "step": 3739
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7599796127353998,
+      "learning_rate": 1.9823489645638596e-05,
+      "loss": 0.7098,
+      "step": 3740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8916806402710034,
+      "learning_rate": 1.9823346884480694e-05,
+      "loss": 0.9672,
+      "step": 3741
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6818514389354073,
+      "learning_rate": 1.982320406612815e-05,
+      "loss": 0.8117,
+      "step": 3742
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9384840624013184,
+      "learning_rate": 1.9823061190581807e-05,
+      "loss": 0.8878,
+      "step": 3743
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8063339342821059,
+      "learning_rate": 1.982291825784249e-05,
+      "loss": 0.6676,
+      "step": 3744
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7471978468753884,
+      "learning_rate": 1.9822775267911033e-05,
+      "loss": 0.8825,
+      "step": 3745
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7025415444459926,
+      "learning_rate": 1.982263222078827e-05,
+      "loss": 0.8415,
+      "step": 3746
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9690259439137205,
+      "learning_rate": 1.982248911647503e-05,
+      "loss": 0.9418,
+      "step": 3747
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8339808679075904,
+      "learning_rate": 1.982234595497215e-05,
+      "loss": 0.9773,
+      "step": 3748
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.232139930837333,
+      "learning_rate": 1.9822202736280462e-05,
+      "loss": 0.8711,
+      "step": 3749
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.654687114275166,
+      "learning_rate": 1.9822059460400804e-05,
+      "loss": 0.95,
+      "step": 3750
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8887422581754296,
+      "learning_rate": 1.9821916127334e-05,
+      "loss": 0.8474,
+      "step": 3751
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6871134404031107,
+      "learning_rate": 1.9821772737080892e-05,
+      "loss": 0.9369,
+      "step": 3752
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0934967880497353,
+      "learning_rate": 1.9821629289642316e-05,
+      "loss": 0.8394,
+      "step": 3753
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6928971099030359,
+      "learning_rate": 1.9821485785019103e-05,
+      "loss": 0.8928,
+      "step": 3754
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6420085931473734,
+      "learning_rate": 1.982134222321209e-05,
+      "loss": 0.7725,
+      "step": 3755
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9320915580132643,
+      "learning_rate": 1.9821198604222114e-05,
+      "loss": 0.8327,
+      "step": 3756
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.159177756858594,
+      "learning_rate": 1.9821054928050012e-05,
+      "loss": 0.8971,
+      "step": 3757
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7572695492117085,
+      "learning_rate": 1.9820911194696613e-05,
+      "loss": 0.9644,
+      "step": 3758
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.241763530408293,
+      "learning_rate": 1.9820767404162765e-05,
+      "loss": 0.9195,
+      "step": 3759
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9348333145750538,
+      "learning_rate": 1.98206235564493e-05,
+      "loss": 0.8538,
+      "step": 3760
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7266859859232866,
+      "learning_rate": 1.9820479651557056e-05,
+      "loss": 0.867,
+      "step": 3761
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9380753688126207,
+      "learning_rate": 1.982033568948687e-05,
+      "loss": 0.985,
+      "step": 3762
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7005669541560096,
+      "learning_rate": 1.982019167023958e-05,
+      "loss": 0.9743,
+      "step": 3763
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.071572832922676,
+      "learning_rate": 1.9820047593816024e-05,
+      "loss": 0.8498,
+      "step": 3764
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6058318597893402,
+      "learning_rate": 1.981990346021704e-05,
+      "loss": 0.9539,
+      "step": 3765
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.662042877929629,
+      "learning_rate": 1.9819759269443475e-05,
+      "loss": 1.0079,
+      "step": 3766
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6070406464470133,
+      "learning_rate": 1.9819615021496158e-05,
+      "loss": 0.9799,
+      "step": 3767
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7389433837335964,
+      "learning_rate": 1.9819470716375933e-05,
+      "loss": 0.8604,
+      "step": 3768
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7261759652300308,
+      "learning_rate": 1.9819326354083644e-05,
+      "loss": 0.8027,
+      "step": 3769
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0712948984356614,
+      "learning_rate": 1.9819181934620128e-05,
+      "loss": 0.8935,
+      "step": 3770
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6293300340683203,
+      "learning_rate": 1.9819037457986224e-05,
+      "loss": 1.0134,
+      "step": 3771
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.53040610227442,
+      "learning_rate": 1.9818892924182774e-05,
+      "loss": 0.8803,
+      "step": 3772
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7828939176351406,
+      "learning_rate": 1.9818748333210624e-05,
+      "loss": 0.7468,
+      "step": 3773
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.079875363318771,
+      "learning_rate": 1.981860368507061e-05,
+      "loss": 0.8776,
+      "step": 3774
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6256582563946675,
+      "learning_rate": 1.981845897976358e-05,
+      "loss": 1.0808,
+      "step": 3775
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5463559251067467,
+      "learning_rate": 1.9818314217290373e-05,
+      "loss": 0.9781,
+      "step": 3776
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.874753415885691,
+      "learning_rate": 1.981816939765183e-05,
+      "loss": 0.926,
+      "step": 3777
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7236698618027724,
+      "learning_rate": 1.9818024520848797e-05,
+      "loss": 0.9351,
+      "step": 3778
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.108900943939093,
+      "learning_rate": 1.9817879586882117e-05,
+      "loss": 0.9807,
+      "step": 3779
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6671593591123102,
+      "learning_rate": 1.9817734595752632e-05,
+      "loss": 0.8275,
+      "step": 3780
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8089572380970459,
+      "learning_rate": 1.9817589547461193e-05,
+      "loss": 0.7581,
+      "step": 3781
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9084997912571033,
+      "learning_rate": 1.9817444442008634e-05,
+      "loss": 0.9061,
+      "step": 3782
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7803811975563777,
+      "learning_rate": 1.981729927939581e-05,
+      "loss": 0.7967,
+      "step": 3783
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6920208511563377,
+      "learning_rate": 1.981715405962356e-05,
+      "loss": 0.9684,
+      "step": 3784
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0646325675013233,
+      "learning_rate": 1.981700878269273e-05,
+      "loss": 0.9079,
+      "step": 3785
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.697474109825748,
+      "learning_rate": 1.9816863448604165e-05,
+      "loss": 1.0157,
+      "step": 3786
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2981926787794396,
+      "learning_rate": 1.9816718057358715e-05,
+      "loss": 0.9235,
+      "step": 3787
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8141913561695422,
+      "learning_rate": 1.9816572608957227e-05,
+      "loss": 0.9452,
+      "step": 3788
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.776261506500827,
+      "learning_rate": 1.981642710340054e-05,
+      "loss": 0.8602,
+      "step": 3789
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.883862601068315,
+      "learning_rate": 1.981628154068951e-05,
+      "loss": 1.1352,
+      "step": 3790
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.705565440818296,
+      "learning_rate": 1.981613592082498e-05,
+      "loss": 0.8745,
+      "step": 3791
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8880480856579718,
+      "learning_rate": 1.98159902438078e-05,
+      "loss": 0.8167,
+      "step": 3792
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.793367228721765,
+      "learning_rate": 1.9815844509638817e-05,
+      "loss": 0.8504,
+      "step": 3793
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6106893229022727,
+      "learning_rate": 1.981569871831888e-05,
+      "loss": 0.9392,
+      "step": 3794
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1585649272858793,
+      "learning_rate": 1.9815552869848836e-05,
+      "loss": 0.9675,
+      "step": 3795
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8333111997645388,
+      "learning_rate": 1.9815406964229537e-05,
+      "loss": 0.7636,
+      "step": 3796
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.84137502443352,
+      "learning_rate": 1.981526100146183e-05,
+      "loss": 0.9376,
+      "step": 3797
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8413329164572227,
+      "learning_rate": 1.9815114981546564e-05,
+      "loss": 0.9465,
+      "step": 3798
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9110984433669869,
+      "learning_rate": 1.9814968904484595e-05,
+      "loss": 0.9282,
+      "step": 3799
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.792759253990416,
+      "learning_rate": 1.981482277027677e-05,
+      "loss": 0.8695,
+      "step": 3800
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8286644399092369,
+      "learning_rate": 1.9814676578923936e-05,
+      "loss": 0.9728,
+      "step": 3801
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6933951174726862,
+      "learning_rate": 1.981453033042695e-05,
+      "loss": 1.0051,
+      "step": 3802
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.652163623223841,
+      "learning_rate": 1.981438402478666e-05,
+      "loss": 0.9347,
+      "step": 3803
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6470631624885594,
+      "learning_rate": 1.981423766200392e-05,
+      "loss": 0.858,
+      "step": 3804
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.853415102385258,
+      "learning_rate": 1.9814091242079583e-05,
+      "loss": 0.9706,
+      "step": 3805
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8890943338281063,
+      "learning_rate": 1.98139447650145e-05,
+      "loss": 1.0061,
+      "step": 3806
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7429090855014846,
+      "learning_rate": 1.981379823080952e-05,
+      "loss": 0.9394,
+      "step": 3807
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6032178119945353,
+      "learning_rate": 1.98136516394655e-05,
+      "loss": 0.963,
+      "step": 3808
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8606715042043906,
+      "learning_rate": 1.9813504990983293e-05,
+      "loss": 0.9325,
+      "step": 3809
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.183450907101736,
+      "learning_rate": 1.9813358285363754e-05,
+      "loss": 0.9627,
+      "step": 3810
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.981215521668806,
+      "learning_rate": 1.981321152260774e-05,
+      "loss": 0.8495,
+      "step": 3811
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8251136490568505,
+      "learning_rate": 1.9813064702716094e-05,
+      "loss": 0.8989,
+      "step": 3812
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8292171531226409,
+      "learning_rate": 1.9812917825689683e-05,
+      "loss": 0.7037,
+      "step": 3813
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.703386818124439,
+      "learning_rate": 1.981277089152936e-05,
+      "loss": 0.9584,
+      "step": 3814
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7067901121641582,
+      "learning_rate": 1.981262390023597e-05,
+      "loss": 0.9915,
+      "step": 3815
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7879534361671137,
+      "learning_rate": 1.9812476851810384e-05,
+      "loss": 0.6923,
+      "step": 3816
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7678733309501917,
+      "learning_rate": 1.981232974625345e-05,
+      "loss": 0.7854,
+      "step": 3817
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.584129947951871,
+      "learning_rate": 1.9812182583566026e-05,
+      "loss": 0.9561,
+      "step": 3818
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5892123102209628,
+      "learning_rate": 1.981203536374897e-05,
+      "loss": 0.9647,
+      "step": 3819
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.007839614016216,
+      "learning_rate": 1.9811888086803137e-05,
+      "loss": 0.8854,
+      "step": 3820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8736658630455372,
+      "learning_rate": 1.981174075272938e-05,
+      "loss": 0.9365,
+      "step": 3821
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.564215941872792,
+      "learning_rate": 1.9811593361528566e-05,
+      "loss": 0.9886,
+      "step": 3822
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7266424878333064,
+      "learning_rate": 1.981144591320155e-05,
+      "loss": 0.7723,
+      "step": 3823
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6362467379089884,
+      "learning_rate": 1.9811298407749188e-05,
+      "loss": 0.8544,
+      "step": 3824
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8156568573362113,
+      "learning_rate": 1.9811150845172342e-05,
+      "loss": 0.8921,
+      "step": 3825
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.167850035078355,
+      "learning_rate": 1.981100322547187e-05,
+      "loss": 0.9302,
+      "step": 3826
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4943997467552412,
+      "learning_rate": 1.9810855548648628e-05,
+      "loss": 0.8773,
+      "step": 3827
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.819860331161937,
+      "learning_rate": 1.981070781470348e-05,
+      "loss": 1.106,
+      "step": 3828
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7310409204092037,
+      "learning_rate": 1.9810560023637285e-05,
+      "loss": 0.9553,
+      "step": 3829
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2343807861107945,
+      "learning_rate": 1.9810412175450907e-05,
+      "loss": 0.9937,
+      "step": 3830
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7828479355128912,
+      "learning_rate": 1.9810264270145198e-05,
+      "loss": 1.0105,
+      "step": 3831
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6202970399219312,
+      "learning_rate": 1.9810116307721028e-05,
+      "loss": 0.9423,
+      "step": 3832
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1058939069536287,
+      "learning_rate": 1.9809968288179254e-05,
+      "loss": 1.0086,
+      "step": 3833
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9279992923623042,
+      "learning_rate": 1.980982021152074e-05,
+      "loss": 0.8182,
+      "step": 3834
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5270064020185061,
+      "learning_rate": 1.9809672077746347e-05,
+      "loss": 0.8785,
+      "step": 3835
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8877670510930387,
+      "learning_rate": 1.9809523886856937e-05,
+      "loss": 1.0664,
+      "step": 3836
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.116724888905511,
+      "learning_rate": 1.9809375638853373e-05,
+      "loss": 0.902,
+      "step": 3837
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.937635206736642,
+      "learning_rate": 1.9809227333736518e-05,
+      "loss": 0.8817,
+      "step": 3838
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6518168868806198,
+      "learning_rate": 1.9809078971507234e-05,
+      "loss": 0.9076,
+      "step": 3839
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5742629243902253,
+      "learning_rate": 1.980893055216639e-05,
+      "loss": 0.8717,
+      "step": 3840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9481229908769777,
+      "learning_rate": 1.980878207571485e-05,
+      "loss": 0.975,
+      "step": 3841
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6898396867860244,
+      "learning_rate": 1.980863354215347e-05,
+      "loss": 0.962,
+      "step": 3842
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9967697887142002,
+      "learning_rate": 1.980848495148312e-05,
+      "loss": 0.9181,
+      "step": 3843
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.896503450569554,
+      "learning_rate": 1.9808336303704666e-05,
+      "loss": 0.9331,
+      "step": 3844
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.130174343270586,
+      "learning_rate": 1.9808187598818974e-05,
+      "loss": 0.9468,
+      "step": 3845
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.728511125617987,
+      "learning_rate": 1.9808038836826906e-05,
+      "loss": 0.792,
+      "step": 3846
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6910867598532877,
+      "learning_rate": 1.980789001772933e-05,
+      "loss": 0.8822,
+      "step": 3847
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7787205813306315,
+      "learning_rate": 1.980774114152712e-05,
+      "loss": 0.9359,
+      "step": 3848
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8083239313725973,
+      "learning_rate": 1.980759220822113e-05,
+      "loss": 0.749,
+      "step": 3849
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7331099815858555,
+      "learning_rate": 1.9807443217812236e-05,
+      "loss": 0.8901,
+      "step": 3850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6056337984882119,
+      "learning_rate": 1.9807294170301303e-05,
+      "loss": 0.8966,
+      "step": 3851
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6114664380539636,
+      "learning_rate": 1.9807145065689196e-05,
+      "loss": 1.0028,
+      "step": 3852
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.930396221981986,
+      "learning_rate": 1.9806995903976785e-05,
+      "loss": 0.8165,
+      "step": 3853
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9821127460108472,
+      "learning_rate": 1.980684668516494e-05,
+      "loss": 0.8689,
+      "step": 3854
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9301590104083783,
+      "learning_rate": 1.9806697409254532e-05,
+      "loss": 0.8225,
+      "step": 3855
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5277268962681494,
+      "learning_rate": 1.9806548076246425e-05,
+      "loss": 0.9249,
+      "step": 3856
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5073554525539132,
+      "learning_rate": 1.9806398686141487e-05,
+      "loss": 0.911,
+      "step": 3857
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.763516045914178,
+      "learning_rate": 1.9806249238940596e-05,
+      "loss": 0.9572,
+      "step": 3858
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6124163356253285,
+      "learning_rate": 1.9806099734644615e-05,
+      "loss": 0.9452,
+      "step": 3859
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6618369997292741,
+      "learning_rate": 1.980595017325442e-05,
+      "loss": 0.9072,
+      "step": 3860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.760997242442872,
+      "learning_rate": 1.9805800554770872e-05,
+      "loss": 0.8839,
+      "step": 3861
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7985864596934324,
+      "learning_rate": 1.980565087919485e-05,
+      "loss": 0.8896,
+      "step": 3862
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6438078843941926,
+      "learning_rate": 1.9805501146527228e-05,
+      "loss": 0.9444,
+      "step": 3863
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7210813662003488,
+      "learning_rate": 1.9805351356768872e-05,
+      "loss": 0.9644,
+      "step": 3864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7759988157996098,
+      "learning_rate": 1.980520150992066e-05,
+      "loss": 0.8307,
+      "step": 3865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5739812118213694,
+      "learning_rate": 1.9805051605983454e-05,
+      "loss": 0.8755,
+      "step": 3866
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.803512400233742,
+      "learning_rate": 1.9804901644958137e-05,
+      "loss": 0.7141,
+      "step": 3867
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8603384290316705,
+      "learning_rate": 1.9804751626845577e-05,
+      "loss": 0.8271,
+      "step": 3868
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.915710944043393,
+      "learning_rate": 1.9804601551646652e-05,
+      "loss": 0.9841,
+      "step": 3869
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8283908517247334,
+      "learning_rate": 1.980445141936223e-05,
+      "loss": 0.9924,
+      "step": 3870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5933878149395118,
+      "learning_rate": 1.9804301229993185e-05,
+      "loss": 1.0466,
+      "step": 3871
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.924205209536407,
+      "learning_rate": 1.98041509835404e-05,
+      "loss": 0.9749,
+      "step": 3872
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.937470659286298,
+      "learning_rate": 1.9804000680004744e-05,
+      "loss": 0.9757,
+      "step": 3873
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8202406168819956,
+      "learning_rate": 1.9803850319387087e-05,
+      "loss": 0.9665,
+      "step": 3874
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2366272686403565,
+      "learning_rate": 1.9803699901688313e-05,
+      "loss": 0.9243,
+      "step": 3875
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4863850802196912,
+      "learning_rate": 1.9803549426909293e-05,
+      "loss": 0.9293,
+      "step": 3876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.708728310389718,
+      "learning_rate": 1.9803398895050905e-05,
+      "loss": 0.8704,
+      "step": 3877
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7622459165510047,
+      "learning_rate": 1.9803248306114026e-05,
+      "loss": 1.096,
+      "step": 3878
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8138914701997852,
+      "learning_rate": 1.980309766009953e-05,
+      "loss": 0.9505,
+      "step": 3879
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9011244761612367,
+      "learning_rate": 1.9802946957008296e-05,
+      "loss": 0.8428,
+      "step": 3880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7959153169541703,
+      "learning_rate": 1.98027961968412e-05,
+      "loss": 0.7617,
+      "step": 3881
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5404131848934475,
+      "learning_rate": 1.9802645379599122e-05,
+      "loss": 0.9751,
+      "step": 3882
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6586832205135549,
+      "learning_rate": 1.980249450528294e-05,
+      "loss": 0.9569,
+      "step": 3883
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.025338418374752,
+      "learning_rate": 1.9802343573893536e-05,
+      "loss": 0.9568,
+      "step": 3884
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1688016946340247,
+      "learning_rate": 1.9802192585431778e-05,
+      "loss": 0.9395,
+      "step": 3885
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.787586208816512,
+      "learning_rate": 1.9802041539898553e-05,
+      "loss": 0.8327,
+      "step": 3886
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.578351190166966,
+      "learning_rate": 1.980189043729474e-05,
+      "loss": 0.7359,
+      "step": 3887
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8180937716138126,
+      "learning_rate": 1.9801739277621215e-05,
+      "loss": 0.9364,
+      "step": 3888
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7654812884901774,
+      "learning_rate": 1.980158806087886e-05,
+      "loss": 0.8141,
+      "step": 3889
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2124147278661783,
+      "learning_rate": 1.9801436787068563e-05,
+      "loss": 0.9147,
+      "step": 3890
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9293924147792243,
+      "learning_rate": 1.980128545619119e-05,
+      "loss": 0.9632,
+      "step": 3891
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0572205346151,
+      "learning_rate": 1.9801134068247636e-05,
+      "loss": 0.9379,
+      "step": 3892
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9665258434986793,
+      "learning_rate": 1.9800982623238775e-05,
+      "loss": 0.7424,
+      "step": 3893
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7632206067234792,
+      "learning_rate": 1.9800831121165488e-05,
+      "loss": 0.8449,
+      "step": 3894
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8233514700913185,
+      "learning_rate": 1.9800679562028662e-05,
+      "loss": 0.8513,
+      "step": 3895
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7757382561734444,
+      "learning_rate": 1.9800527945829175e-05,
+      "loss": 0.681,
+      "step": 3896
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7810048871982305,
+      "learning_rate": 1.9800376272567913e-05,
+      "loss": 0.7813,
+      "step": 3897
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.070706556404286,
+      "learning_rate": 1.9800224542245756e-05,
+      "loss": 0.9307,
+      "step": 3898
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6034887428257727,
+      "learning_rate": 1.980007275486359e-05,
+      "loss": 0.9386,
+      "step": 3899
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6853848799382654,
+      "learning_rate": 1.9799920910422296e-05,
+      "loss": 0.9672,
+      "step": 3900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8382817079175579,
+      "learning_rate": 1.9799769008922762e-05,
+      "loss": 0.7727,
+      "step": 3901
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5748314786314552,
+      "learning_rate": 1.979961705036587e-05,
+      "loss": 0.8666,
+      "step": 3902
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7649550669273937,
+      "learning_rate": 1.9799465034752505e-05,
+      "loss": 0.9236,
+      "step": 3903
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6098864052816628,
+      "learning_rate": 1.979931296208355e-05,
+      "loss": 0.8769,
+      "step": 3904
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7073062785584017,
+      "learning_rate": 1.9799160832359896e-05,
+      "loss": 0.9991,
+      "step": 3905
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6072087343947337,
+      "learning_rate": 1.9799008645582424e-05,
+      "loss": 0.9265,
+      "step": 3906
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6007258279066383,
+      "learning_rate": 1.9798856401752017e-05,
+      "loss": 0.8164,
+      "step": 3907
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.057708643038147,
+      "learning_rate": 1.979870410086957e-05,
+      "loss": 0.9073,
+      "step": 3908
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7526673385120914,
+      "learning_rate": 1.979855174293597e-05,
+      "loss": 0.9783,
+      "step": 3909
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8655256785958384,
+      "learning_rate": 1.9798399327952093e-05,
+      "loss": 0.7234,
+      "step": 3910
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.846036176349655,
+      "learning_rate": 1.9798246855918833e-05,
+      "loss": 0.8376,
+      "step": 3911
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8551541170838355,
+      "learning_rate": 1.979809432683708e-05,
+      "loss": 0.8769,
+      "step": 3912
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5322425480860271,
+      "learning_rate": 1.9797941740707718e-05,
+      "loss": 1.0846,
+      "step": 3913
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8940937984295978,
+      "learning_rate": 1.9797789097531638e-05,
+      "loss": 0.9688,
+      "step": 3914
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8458198778113591,
+      "learning_rate": 1.979763639730973e-05,
+      "loss": 0.6914,
+      "step": 3915
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7337271359576203,
+      "learning_rate": 1.979748364004288e-05,
+      "loss": 0.7788,
+      "step": 3916
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7223545143912016,
+      "learning_rate": 1.9797330825731976e-05,
+      "loss": 0.8659,
+      "step": 3917
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.111310165991768,
+      "learning_rate": 1.979717795437791e-05,
+      "loss": 0.9418,
+      "step": 3918
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6689596323585623,
+      "learning_rate": 1.9797025025981576e-05,
+      "loss": 1.0303,
+      "step": 3919
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4139669556974823,
+      "learning_rate": 1.979687204054386e-05,
+      "loss": 0.8579,
+      "step": 3920
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.709891203494449,
+      "learning_rate": 1.9796718998065653e-05,
+      "loss": 0.8914,
+      "step": 3921
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.031406753995376,
+      "learning_rate": 1.9796565898547845e-05,
+      "loss": 0.9051,
+      "step": 3922
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.264380359728782,
+      "learning_rate": 1.979641274199133e-05,
+      "loss": 1.0161,
+      "step": 3923
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.73791772543447,
+      "learning_rate": 1.9796259528396998e-05,
+      "loss": 1.012,
+      "step": 3924
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8385639496017956,
+      "learning_rate": 1.979610625776574e-05,
+      "loss": 1.0884,
+      "step": 3925
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.657527092988495,
+      "learning_rate": 1.9795952930098453e-05,
+      "loss": 0.9624,
+      "step": 3926
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.638257196075443,
+      "learning_rate": 1.9795799545396022e-05,
+      "loss": 0.8647,
+      "step": 3927
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.336824227905678,
+      "learning_rate": 1.9795646103659346e-05,
+      "loss": 0.8923,
+      "step": 3928
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.019024468133063,
+      "learning_rate": 1.9795492604889323e-05,
+      "loss": 0.8537,
+      "step": 3929
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6923881991795133,
+      "learning_rate": 1.9795339049086837e-05,
+      "loss": 0.9219,
+      "step": 3930
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.515911669068663,
+      "learning_rate": 1.9795185436252785e-05,
+      "loss": 0.9908,
+      "step": 3931
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7392694083425106,
+      "learning_rate": 1.9795031766388063e-05,
+      "loss": 0.7749,
+      "step": 3932
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.795736181676135,
+      "learning_rate": 1.9794878039493564e-05,
+      "loss": 1.042,
+      "step": 3933
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7652968347411837,
+      "learning_rate": 1.9794724255570187e-05,
+      "loss": 0.9355,
+      "step": 3934
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6785152943107071,
+      "learning_rate": 1.9794570414618823e-05,
+      "loss": 0.8641,
+      "step": 3935
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9614415257158895,
+      "learning_rate": 1.9794416516640365e-05,
+      "loss": 0.8824,
+      "step": 3936
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4877354087789,
+      "learning_rate": 1.9794262561635716e-05,
+      "loss": 1.008,
+      "step": 3937
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3123995936702526,
+      "learning_rate": 1.9794108549605774e-05,
+      "loss": 0.9798,
+      "step": 3938
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8342262294962708,
+      "learning_rate": 1.979395448055143e-05,
+      "loss": 0.909,
+      "step": 3939
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6551978293033047,
+      "learning_rate": 1.9793800354473575e-05,
+      "loss": 0.909,
+      "step": 3940
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6623678258604033,
+      "learning_rate": 1.979364617137312e-05,
+      "loss": 1.0007,
+      "step": 3941
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.686510967088842,
+      "learning_rate": 1.9793491931250955e-05,
+      "loss": 1.0173,
+      "step": 3942
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.054677766643272,
+      "learning_rate": 1.9793337634107977e-05,
+      "loss": 0.8962,
+      "step": 3943
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6926296851871878,
+      "learning_rate": 1.9793183279945093e-05,
+      "loss": 0.9304,
+      "step": 3944
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.803231352882931,
+      "learning_rate": 1.979302886876319e-05,
+      "loss": 0.7464,
+      "step": 3945
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6443144190671966,
+      "learning_rate": 1.9792874400563172e-05,
+      "loss": 0.9676,
+      "step": 3946
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5945501224615788,
+      "learning_rate": 1.979271987534594e-05,
+      "loss": 0.9445,
+      "step": 3947
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.020930095368236,
+      "learning_rate": 1.9792565293112396e-05,
+      "loss": 0.9793,
+      "step": 3948
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8206987711568285,
+      "learning_rate": 1.9792410653863436e-05,
+      "loss": 0.9527,
+      "step": 3949
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.98770030072312,
+      "learning_rate": 1.9792255957599956e-05,
+      "loss": 0.8849,
+      "step": 3950
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7502711853938169,
+      "learning_rate": 1.9792101204322866e-05,
+      "loss": 0.9078,
+      "step": 3951
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.656339222446456,
+      "learning_rate": 1.979194639403306e-05,
+      "loss": 0.9871,
+      "step": 3952
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8528545496342335,
+      "learning_rate": 1.9791791526731444e-05,
+      "loss": 0.8498,
+      "step": 3953
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9090779172499914,
+      "learning_rate": 1.9791636602418916e-05,
+      "loss": 1.064,
+      "step": 3954
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5591945119282307,
+      "learning_rate": 1.979148162109638e-05,
+      "loss": 0.7574,
+      "step": 3955
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6769523183905441,
+      "learning_rate": 1.9791326582764742e-05,
+      "loss": 0.8429,
+      "step": 3956
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.746396555342519,
+      "learning_rate": 1.9791171487424895e-05,
+      "loss": 0.9608,
+      "step": 3957
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9597041789164642,
+      "learning_rate": 1.979101633507775e-05,
+      "loss": 0.9868,
+      "step": 3958
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.829194129926128,
+      "learning_rate": 1.979086112572421e-05,
+      "loss": 0.9835,
+      "step": 3959
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.709363563261955,
+      "learning_rate": 1.9790705859365178e-05,
+      "loss": 1.0895,
+      "step": 3960
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.841209408887883,
+      "learning_rate": 1.9790550536001554e-05,
+      "loss": 0.7282,
+      "step": 3961
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.910360830037553,
+      "learning_rate": 1.9790395155634243e-05,
+      "loss": 0.9931,
+      "step": 3962
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0613251553582383,
+      "learning_rate": 1.9790239718264155e-05,
+      "loss": 0.7817,
+      "step": 3963
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.796558146684219,
+      "learning_rate": 1.9790084223892192e-05,
+      "loss": 0.8978,
+      "step": 3964
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.641573107311541,
+      "learning_rate": 1.978992867251926e-05,
+      "loss": 1.0065,
+      "step": 3965
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6602328230224477,
+      "learning_rate": 1.9789773064146257e-05,
+      "loss": 0.9588,
+      "step": 3966
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1310633358432685,
+      "learning_rate": 1.97896173987741e-05,
+      "loss": 0.8875,
+      "step": 3967
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.544861327569847,
+      "learning_rate": 1.9789461676403694e-05,
+      "loss": 0.9081,
+      "step": 3968
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7013591452577532,
+      "learning_rate": 1.9789305897035942e-05,
+      "loss": 0.8898,
+      "step": 3969
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9901512628057827,
+      "learning_rate": 1.978915006067175e-05,
+      "loss": 0.9197,
+      "step": 3970
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.180440133248528,
+      "learning_rate": 1.978899416731203e-05,
+      "loss": 0.8555,
+      "step": 3971
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4244417777028522,
+      "learning_rate": 1.978883821695768e-05,
+      "loss": 0.7896,
+      "step": 3972
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.830392788828692,
+      "learning_rate": 1.978868220960962e-05,
+      "loss": 0.9525,
+      "step": 3973
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7709549324600726,
+      "learning_rate": 1.9788526145268752e-05,
+      "loss": 0.8583,
+      "step": 3974
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9741385501675466,
+      "learning_rate": 1.978837002393599e-05,
+      "loss": 0.979,
+      "step": 3975
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7218487402865916,
+      "learning_rate": 1.9788213845612233e-05,
+      "loss": 0.8606,
+      "step": 3976
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.781829680643822,
+      "learning_rate": 1.9788057610298396e-05,
+      "loss": 0.9509,
+      "step": 3977
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.809881066702976,
+      "learning_rate": 1.9787901317995393e-05,
+      "loss": 0.8704,
+      "step": 3978
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.203600841951371,
+      "learning_rate": 1.9787744968704125e-05,
+      "loss": 0.999,
+      "step": 3979
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6581326712694935,
+      "learning_rate": 1.9787588562425514e-05,
+      "loss": 1.0055,
+      "step": 3980
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6351184215401222,
+      "learning_rate": 1.9787432099160457e-05,
+      "loss": 1.0008,
+      "step": 3981
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8015561510447302,
+      "learning_rate": 1.9787275578909873e-05,
+      "loss": 0.7916,
+      "step": 3982
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9304330233038507,
+      "learning_rate": 1.9787119001674677e-05,
+      "loss": 0.9259,
+      "step": 3983
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7068961950243513,
+      "learning_rate": 1.978696236745577e-05,
+      "loss": 1.0679,
+      "step": 3984
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.926860818143291,
+      "learning_rate": 1.978680567625407e-05,
+      "loss": 0.8604,
+      "step": 3985
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8348183633911098,
+      "learning_rate": 1.9786648928070493e-05,
+      "loss": 0.7114,
+      "step": 3986
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7132481008088862,
+      "learning_rate": 1.978649212290594e-05,
+      "loss": 0.8312,
+      "step": 3987
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4455284407831925,
+      "learning_rate": 1.978633526076134e-05,
+      "loss": 0.9749,
+      "step": 3988
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7884550934126916,
+      "learning_rate": 1.9786178341637594e-05,
+      "loss": 0.7362,
+      "step": 3989
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8075934683345098,
+      "learning_rate": 1.9786021365535616e-05,
+      "loss": 0.8275,
+      "step": 3990
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8355454692444002,
+      "learning_rate": 1.9785864332456327e-05,
+      "loss": 0.841,
+      "step": 3991
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5146829282515482,
+      "learning_rate": 1.9785707242400638e-05,
+      "loss": 0.8649,
+      "step": 3992
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.674054719253842,
+      "learning_rate": 1.9785550095369465e-05,
+      "loss": 0.9315,
+      "step": 3993
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.481881000388384,
+      "learning_rate": 1.9785392891363718e-05,
+      "loss": 0.8895,
+      "step": 3994
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0096625747318173,
+      "learning_rate": 1.9785235630384316e-05,
+      "loss": 0.8506,
+      "step": 3995
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.132639318823953,
+      "learning_rate": 1.9785078312432176e-05,
+      "loss": 0.7753,
+      "step": 3996
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3347259076730627,
+      "learning_rate": 1.9784920937508208e-05,
+      "loss": 0.922,
+      "step": 3997
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.696057424904861,
+      "learning_rate": 1.978476350561334e-05,
+      "loss": 1.0335,
+      "step": 3998
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8415906836599731,
+      "learning_rate": 1.9784606016748474e-05,
+      "loss": 0.9765,
+      "step": 3999
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.443852472406114,
+      "learning_rate": 1.9784448470914538e-05,
+      "loss": 0.836,
+      "step": 4000
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9004069468011968,
+      "learning_rate": 1.9784290868112442e-05,
+      "loss": 0.9138,
+      "step": 4001
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5786722968018712,
+      "learning_rate": 1.978413320834311e-05,
+      "loss": 0.9541,
+      "step": 4002
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7670834361814247,
+      "learning_rate": 1.9783975491607457e-05,
+      "loss": 0.9196,
+      "step": 4003
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9136347650237695,
+      "learning_rate": 1.9783817717906398e-05,
+      "loss": 0.8165,
+      "step": 4004
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7958656551809582,
+      "learning_rate": 1.9783659887240855e-05,
+      "loss": 0.9123,
+      "step": 4005
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7144699927456255,
+      "learning_rate": 1.978350199961175e-05,
+      "loss": 1.0152,
+      "step": 4006
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8654289126993675,
+      "learning_rate": 1.978334405502e-05,
+      "loss": 0.9311,
+      "step": 4007
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.934616475129806,
+      "learning_rate": 1.9783186053466518e-05,
+      "loss": 0.9488,
+      "step": 4008
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7993893231539548,
+      "learning_rate": 1.9783027994952233e-05,
+      "loss": 0.9771,
+      "step": 4009
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.255138676709329,
+      "learning_rate": 1.9782869879478058e-05,
+      "loss": 0.9549,
+      "step": 4010
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6933050424665697,
+      "learning_rate": 1.9782711707044923e-05,
+      "loss": 0.7827,
+      "step": 4011
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.929029065360081,
+      "learning_rate": 1.9782553477653738e-05,
+      "loss": 0.8984,
+      "step": 4012
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7940788332540929,
+      "learning_rate": 1.9782395191305436e-05,
+      "loss": 0.7312,
+      "step": 4013
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9549297643031702,
+      "learning_rate": 1.9782236848000927e-05,
+      "loss": 0.9378,
+      "step": 4014
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9092783560951885,
+      "learning_rate": 1.9782078447741136e-05,
+      "loss": 1.1542,
+      "step": 4015
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7159451357926667,
+      "learning_rate": 1.9781919990526993e-05,
+      "loss": 0.9816,
+      "step": 4016
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7846254652773358,
+      "learning_rate": 1.9781761476359414e-05,
+      "loss": 0.7502,
+      "step": 4017
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9119070983893696,
+      "learning_rate": 1.9781602905239323e-05,
+      "loss": 1.0127,
+      "step": 4018
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6319621859906523,
+      "learning_rate": 1.978144427716764e-05,
+      "loss": 0.9001,
+      "step": 4019
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7869002676856207,
+      "learning_rate": 1.9781285592145292e-05,
+      "loss": 0.9561,
+      "step": 4020
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7721848949404566,
+      "learning_rate": 1.9781126850173204e-05,
+      "loss": 1.0499,
+      "step": 4021
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7671685146599444,
+      "learning_rate": 1.97809680512523e-05,
+      "loss": 0.7641,
+      "step": 4022
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7456537823726854,
+      "learning_rate": 1.9780809195383498e-05,
+      "loss": 0.8626,
+      "step": 4023
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6011132769482748,
+      "learning_rate": 1.9780650282567737e-05,
+      "loss": 0.9866,
+      "step": 4024
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6572412350298145,
+      "learning_rate": 1.9780491312805927e-05,
+      "loss": 1.0902,
+      "step": 4025
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8798191374980648,
+      "learning_rate": 1.9780332286099004e-05,
+      "loss": 0.9995,
+      "step": 4026
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5304862894645517,
+      "learning_rate": 1.9780173202447885e-05,
+      "loss": 0.9865,
+      "step": 4027
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.831294564905965,
+      "learning_rate": 1.9780014061853506e-05,
+      "loss": 0.96,
+      "step": 4028
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1835421977980176,
+      "learning_rate": 1.9779854864316784e-05,
+      "loss": 1.0224,
+      "step": 4029
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6863362218276605,
+      "learning_rate": 1.9779695609838656e-05,
+      "loss": 0.9336,
+      "step": 4030
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8049418090697449,
+      "learning_rate": 1.9779536298420042e-05,
+      "loss": 0.9807,
+      "step": 4031
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.648111718501788,
+      "learning_rate": 1.977937693006187e-05,
+      "loss": 0.9424,
+      "step": 4032
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.603243240764367,
+      "learning_rate": 1.9779217504765072e-05,
+      "loss": 1.0846,
+      "step": 4033
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0359402381385623,
+      "learning_rate": 1.9779058022530572e-05,
+      "loss": 0.9167,
+      "step": 4034
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.281915931386886,
+      "learning_rate": 1.97788984833593e-05,
+      "loss": 0.9221,
+      "step": 4035
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6311182071357888,
+      "learning_rate": 1.9778738887252187e-05,
+      "loss": 0.8311,
+      "step": 4036
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8571263973315192,
+      "learning_rate": 1.977857923421016e-05,
+      "loss": 0.7784,
+      "step": 4037
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5938834675270883,
+      "learning_rate": 1.9778419524234148e-05,
+      "loss": 0.9038,
+      "step": 4038
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7839794188610911,
+      "learning_rate": 1.977825975732508e-05,
+      "loss": 0.6876,
+      "step": 4039
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7159021359885518,
+      "learning_rate": 1.977809993348389e-05,
+      "loss": 0.8486,
+      "step": 4040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7024541949120484,
+      "learning_rate": 1.9777940052711506e-05,
+      "loss": 0.9487,
+      "step": 4041
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5225534017737488,
+      "learning_rate": 1.977778011500886e-05,
+      "loss": 0.9855,
+      "step": 4042
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9018686491543193,
+      "learning_rate": 1.977762012037688e-05,
+      "loss": 0.7803,
+      "step": 4043
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6138045509523589,
+      "learning_rate": 1.9777460068816503e-05,
+      "loss": 0.8116,
+      "step": 4044
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8679762481395772,
+      "learning_rate": 1.9777299960328657e-05,
+      "loss": 0.9449,
+      "step": 4045
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7050317105795891,
+      "learning_rate": 1.9777139794914276e-05,
+      "loss": 0.9075,
+      "step": 4046
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7777976914441861,
+      "learning_rate": 1.9776979572574292e-05,
+      "loss": 0.7616,
+      "step": 4047
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7098525944561327,
+      "learning_rate": 1.9776819293309636e-05,
+      "loss": 0.9699,
+      "step": 4048
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.333350602458244,
+      "learning_rate": 1.9776658957121242e-05,
+      "loss": 0.9529,
+      "step": 4049
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8227804406140806,
+      "learning_rate": 1.9776498564010045e-05,
+      "loss": 0.9936,
+      "step": 4050
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6413270517198137,
+      "learning_rate": 1.977633811397698e-05,
+      "loss": 0.8076,
+      "step": 4051
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5998701616703674,
+      "learning_rate": 1.9776177607022976e-05,
+      "loss": 0.881,
+      "step": 4052
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8259717400360562,
+      "learning_rate": 1.9776017043148975e-05,
+      "loss": 0.7456,
+      "step": 4053
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9009414534747262,
+      "learning_rate": 1.9775856422355904e-05,
+      "loss": 0.9896,
+      "step": 4054
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.606682118086946,
+      "learning_rate": 1.9775695744644706e-05,
+      "loss": 0.7808,
+      "step": 4055
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7344199929188653,
+      "learning_rate": 1.977553501001631e-05,
+      "loss": 1.0229,
+      "step": 4056
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5010876239876403,
+      "learning_rate": 1.9775374218471653e-05,
+      "loss": 0.8975,
+      "step": 4057
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9837998303952062,
+      "learning_rate": 1.9775213370011675e-05,
+      "loss": 1.0449,
+      "step": 4058
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5178684850683952,
+      "learning_rate": 1.9775052464637306e-05,
+      "loss": 0.8604,
+      "step": 4059
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7735193622347014,
+      "learning_rate": 1.9774891502349493e-05,
+      "loss": 0.8647,
+      "step": 4060
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8560475149155415,
+      "learning_rate": 1.9774730483149166e-05,
+      "loss": 0.8262,
+      "step": 4061
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8068094188381234,
+      "learning_rate": 1.977456940703726e-05,
+      "loss": 0.9883,
+      "step": 4062
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.16388520513654,
+      "learning_rate": 1.9774408274014715e-05,
+      "loss": 0.915,
+      "step": 4063
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5107571900918317,
+      "learning_rate": 1.9774247084082474e-05,
+      "loss": 0.7985,
+      "step": 4064
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7379651552913526,
+      "learning_rate": 1.977408583724147e-05,
+      "loss": 0.8805,
+      "step": 4065
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7957189945652521,
+      "learning_rate": 1.9773924533492646e-05,
+      "loss": 0.7336,
+      "step": 4066
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.763053696089848,
+      "learning_rate": 1.9773763172836936e-05,
+      "loss": 0.7782,
+      "step": 4067
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.594459402524619,
+      "learning_rate": 1.9773601755275287e-05,
+      "loss": 0.8097,
+      "step": 4068
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6363857187665751,
+      "learning_rate": 1.9773440280808633e-05,
+      "loss": 1.0425,
+      "step": 4069
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.779081841178094,
+      "learning_rate": 1.9773278749437912e-05,
+      "loss": 0.7896,
+      "step": 4070
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7034809228244043,
+      "learning_rate": 1.977311716116407e-05,
+      "loss": 0.9221,
+      "step": 4071
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0376071603334207,
+      "learning_rate": 1.9772955515988045e-05,
+      "loss": 0.8171,
+      "step": 4072
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7819535869125762,
+      "learning_rate": 1.9772793813910783e-05,
+      "loss": 0.7495,
+      "step": 4073
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7152641701298639,
+      "learning_rate": 1.9772632054933216e-05,
+      "loss": 0.9803,
+      "step": 4074
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5356984408950012,
+      "learning_rate": 1.9772470239056295e-05,
+      "loss": 0.9362,
+      "step": 4075
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7921243298219882,
+      "learning_rate": 1.9772308366280955e-05,
+      "loss": 0.7286,
+      "step": 4076
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7701622290310732,
+      "learning_rate": 1.9772146436608143e-05,
+      "loss": 0.9069,
+      "step": 4077
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6457739951465267,
+      "learning_rate": 1.9771984450038802e-05,
+      "loss": 0.8948,
+      "step": 4078
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.857132364485072,
+      "learning_rate": 1.9771822406573876e-05,
+      "loss": 0.844,
+      "step": 4079
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4781436045450043,
+      "learning_rate": 1.97716603062143e-05,
+      "loss": 0.7875,
+      "step": 4080
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.671480194781321,
+      "learning_rate": 1.9771498148961027e-05,
+      "loss": 0.9972,
+      "step": 4081
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6701060726821753,
+      "learning_rate": 1.9771335934815e-05,
+      "loss": 1.0177,
+      "step": 4082
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8405203289470449,
+      "learning_rate": 1.977117366377716e-05,
+      "loss": 0.9135,
+      "step": 4083
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5057721660800154,
+      "learning_rate": 1.9771011335848453e-05,
+      "loss": 0.9701,
+      "step": 4084
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5700304963353071,
+      "learning_rate": 1.9770848951029825e-05,
+      "loss": 0.9421,
+      "step": 4085
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7915888778575586,
+      "learning_rate": 1.977068650932222e-05,
+      "loss": 0.9256,
+      "step": 4086
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6756151629789005,
+      "learning_rate": 1.9770524010726587e-05,
+      "loss": 0.9051,
+      "step": 4087
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.352207503416674,
+      "learning_rate": 1.977036145524387e-05,
+      "loss": 0.8812,
+      "step": 4088
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7284671522670585,
+      "learning_rate": 1.9770198842875015e-05,
+      "loss": 0.7309,
+      "step": 4089
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3765536425242426,
+      "learning_rate": 1.9770036173620965e-05,
+      "loss": 0.9139,
+      "step": 4090
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6588766044218366,
+      "learning_rate": 1.9769873447482676e-05,
+      "loss": 0.9271,
+      "step": 4091
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7851774325303844,
+      "learning_rate": 1.976971066446109e-05,
+      "loss": 0.7106,
+      "step": 4092
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2148674042056093,
+      "learning_rate": 1.9769547824557155e-05,
+      "loss": 1.0168,
+      "step": 4093
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9635024425879386,
+      "learning_rate": 1.976938492777182e-05,
+      "loss": 0.9581,
+      "step": 4094
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4399751021201905,
+      "learning_rate": 1.976922197410603e-05,
+      "loss": 0.9857,
+      "step": 4095
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7994728573867886,
+      "learning_rate": 1.976905896356074e-05,
+      "loss": 0.9654,
+      "step": 4096
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.568616040213193,
+      "learning_rate": 1.9768895896136893e-05,
+      "loss": 1.0221,
+      "step": 4097
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9084334994110328,
+      "learning_rate": 1.9768732771835446e-05,
+      "loss": 0.8808,
+      "step": 4098
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9543478333290638,
+      "learning_rate": 1.976856959065734e-05,
+      "loss": 0.9195,
+      "step": 4099
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.761976693196312,
+      "learning_rate": 1.976840635260353e-05,
+      "loss": 0.9322,
+      "step": 4100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6499091290695391,
+      "learning_rate": 1.9768243057674967e-05,
+      "loss": 0.8525,
+      "step": 4101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7489817983842786,
+      "learning_rate": 1.97680797058726e-05,
+      "loss": 0.8358,
+      "step": 4102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7723219709747361,
+      "learning_rate": 1.9767916297197378e-05,
+      "loss": 0.7307,
+      "step": 4103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9298980581786778,
+      "learning_rate": 1.976775283165026e-05,
+      "loss": 0.8657,
+      "step": 4104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2557465744406104,
+      "learning_rate": 1.976758930923219e-05,
+      "loss": 0.9024,
+      "step": 4105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7095779556696595,
+      "learning_rate": 1.976742572994412e-05,
+      "loss": 0.9463,
+      "step": 4106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.766330609714557,
+      "learning_rate": 1.9767262093787007e-05,
+      "loss": 0.9625,
+      "step": 4107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6994921161896246,
+      "learning_rate": 1.97670984007618e-05,
+      "loss": 0.9547,
+      "step": 4108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2383940279755743,
+      "learning_rate": 1.976693465086946e-05,
+      "loss": 0.9167,
+      "step": 4109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7165896327716623,
+      "learning_rate": 1.976677084411093e-05,
+      "loss": 0.9306,
+      "step": 4110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0563179013920925,
+      "learning_rate": 1.9766606980487167e-05,
+      "loss": 0.9427,
+      "step": 4111
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7077379439817526,
+      "learning_rate": 1.9766443059999127e-05,
+      "loss": 0.9341,
+      "step": 4112
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6534145784486929,
+      "learning_rate": 1.9766279082647763e-05,
+      "loss": 0.9277,
+      "step": 4113
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.336599555847774,
+      "learning_rate": 1.976611504843403e-05,
+      "loss": 1.0054,
+      "step": 4114
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.569357325147629,
+      "learning_rate": 1.9765950957358887e-05,
+      "loss": 0.8647,
+      "step": 4115
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8411369540886435,
+      "learning_rate": 1.9765786809423283e-05,
+      "loss": 0.9974,
+      "step": 4116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0055039222275948,
+      "learning_rate": 1.9765622604628177e-05,
+      "loss": 0.8949,
+      "step": 4117
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.74671702950343,
+      "learning_rate": 1.976545834297452e-05,
+      "loss": 0.933,
+      "step": 4118
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8062975013795477,
+      "learning_rate": 1.9765294024463276e-05,
+      "loss": 0.9848,
+      "step": 4119
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4955244137524266,
+      "learning_rate": 1.97651296490954e-05,
+      "loss": 0.8937,
+      "step": 4120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7301042389442471,
+      "learning_rate": 1.9764965216871848e-05,
+      "loss": 1.0551,
+      "step": 4121
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7472415174372875,
+      "learning_rate": 1.9764800727793576e-05,
+      "loss": 0.9106,
+      "step": 4122
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.105311185791734,
+      "learning_rate": 1.9764636181861544e-05,
+      "loss": 0.9448,
+      "step": 4123
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6192201343677253,
+      "learning_rate": 1.9764471579076705e-05,
+      "loss": 0.8449,
+      "step": 4124
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9304199314334531,
+      "learning_rate": 1.9764306919440023e-05,
+      "loss": 0.9747,
+      "step": 4125
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7477219370524654,
+      "learning_rate": 1.9764142202952456e-05,
+      "loss": 0.7313,
+      "step": 4126
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7101297603266574,
+      "learning_rate": 1.976397742961496e-05,
+      "loss": 1.0141,
+      "step": 4127
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.047220400234825,
+      "learning_rate": 1.9763812599428498e-05,
+      "loss": 0.9534,
+      "step": 4128
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5952807869262657,
+      "learning_rate": 1.9763647712394028e-05,
+      "loss": 0.783,
+      "step": 4129
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5407470147503806,
+      "learning_rate": 1.9763482768512507e-05,
+      "loss": 0.9811,
+      "step": 4130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.782960797047389,
+      "learning_rate": 1.97633177677849e-05,
+      "loss": 0.7678,
+      "step": 4131
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5997611800264442,
+      "learning_rate": 1.976315271021217e-05,
+      "loss": 0.9091,
+      "step": 4132
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7168294691413688,
+      "learning_rate": 1.976298759579527e-05,
+      "loss": 0.8506,
+      "step": 4133
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.850879006169308,
+      "learning_rate": 1.9762822424535163e-05,
+      "loss": 0.8627,
+      "step": 4134
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9021655629475565,
+      "learning_rate": 1.9762657196432814e-05,
+      "loss": 0.8312,
+      "step": 4135
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0261025383635642,
+      "learning_rate": 1.9762491911489184e-05,
+      "loss": 0.8476,
+      "step": 4136
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8113492025959269,
+      "learning_rate": 1.976232656970524e-05,
+      "loss": 0.7784,
+      "step": 4137
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.192405268361903,
+      "learning_rate": 1.9762161171081934e-05,
+      "loss": 0.8308,
+      "step": 4138
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9846630588709238,
+      "learning_rate": 1.976199571562024e-05,
+      "loss": 0.8147,
+      "step": 4139
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7566646320338045,
+      "learning_rate": 1.9761830203321114e-05,
+      "loss": 0.7112,
+      "step": 4140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8085032194627155,
+      "learning_rate": 1.9761664634185523e-05,
+      "loss": 1.0063,
+      "step": 4141
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1300933522085552,
+      "learning_rate": 1.976149900821443e-05,
+      "loss": 1.1067,
+      "step": 4142
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9366099352835182,
+      "learning_rate": 1.9761333325408795e-05,
+      "loss": 0.935,
+      "step": 4143
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.857645226875001,
+      "learning_rate": 1.9761167585769588e-05,
+      "loss": 0.8735,
+      "step": 4144
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7154588361658625,
+      "learning_rate": 1.9761001789297778e-05,
+      "loss": 1.0275,
+      "step": 4145
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8349326403335722,
+      "learning_rate": 1.9760835935994322e-05,
+      "loss": 0.8884,
+      "step": 4146
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8823676107250245,
+      "learning_rate": 1.976067002586019e-05,
+      "loss": 0.8445,
+      "step": 4147
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5310353206285745,
+      "learning_rate": 1.9760504058896348e-05,
+      "loss": 0.9608,
+      "step": 4148
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7012037766328052,
+      "learning_rate": 1.9760338035103758e-05,
+      "loss": 0.8526,
+      "step": 4149
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9501953201408557,
+      "learning_rate": 1.976017195448339e-05,
+      "loss": 0.8977,
+      "step": 4150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.230909000340651,
+      "learning_rate": 1.9760005817036215e-05,
+      "loss": 0.9477,
+      "step": 4151
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8216374340709958,
+      "learning_rate": 1.9759839622763194e-05,
+      "loss": 0.9195,
+      "step": 4152
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5352505568558246,
+      "learning_rate": 1.9759673371665297e-05,
+      "loss": 0.9164,
+      "step": 4153
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7739860293991834,
+      "learning_rate": 1.975950706374349e-05,
+      "loss": 0.9597,
+      "step": 4154
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.931006078223674,
+      "learning_rate": 1.9759340698998746e-05,
+      "loss": 1.0064,
+      "step": 4155
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8124783601572507,
+      "learning_rate": 1.9759174277432026e-05,
+      "loss": 0.9334,
+      "step": 4156
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.730615488294838,
+      "learning_rate": 1.9759007799044307e-05,
+      "loss": 0.8528,
+      "step": 4157
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6295021672887993,
+      "learning_rate": 1.9758841263836553e-05,
+      "loss": 0.9006,
+      "step": 4158
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7181016517523395,
+      "learning_rate": 1.9758674671809742e-05,
+      "loss": 1.0134,
+      "step": 4159
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6966899301461393,
+      "learning_rate": 1.975850802296483e-05,
+      "loss": 0.7987,
+      "step": 4160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.835102624680902,
+      "learning_rate": 1.97583413173028e-05,
+      "loss": 0.709,
+      "step": 4161
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5743280140936464,
+      "learning_rate": 1.9758174554824614e-05,
+      "loss": 0.8503,
+      "step": 4162
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5749424957949436,
+      "learning_rate": 1.9758007735531247e-05,
+      "loss": 0.9338,
+      "step": 4163
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.914513770603447,
+      "learning_rate": 1.975784085942367e-05,
+      "loss": 0.9642,
+      "step": 4164
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4886092233067614,
+      "learning_rate": 1.9757673926502854e-05,
+      "loss": 0.8502,
+      "step": 4165
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8872607442175224,
+      "learning_rate": 1.975750693676977e-05,
+      "loss": 0.8915,
+      "step": 4166
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7088558093144488,
+      "learning_rate": 1.9757339890225395e-05,
+      "loss": 0.9696,
+      "step": 4167
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7526815014593353,
+      "learning_rate": 1.9757172786870695e-05,
+      "loss": 0.9421,
+      "step": 4168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0256834135516137,
+      "learning_rate": 1.9757005626706648e-05,
+      "loss": 1.0083,
+      "step": 4169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0124121056441324,
+      "learning_rate": 1.9756838409734222e-05,
+      "loss": 1.0475,
+      "step": 4170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8779936221258215,
+      "learning_rate": 1.9756671135954396e-05,
+      "loss": 0.9205,
+      "step": 4171
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6380836416528382,
+      "learning_rate": 1.975650380536814e-05,
+      "loss": 0.9127,
+      "step": 4172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7031198871851874,
+      "learning_rate": 1.9756336417976435e-05,
+      "loss": 0.9243,
+      "step": 4173
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.047540464218416,
+      "learning_rate": 1.9756168973780245e-05,
+      "loss": 0.9724,
+      "step": 4174
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.264851445419456,
+      "learning_rate": 1.975600147278055e-05,
+      "loss": 0.8708,
+      "step": 4175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7608855937193724,
+      "learning_rate": 1.975583391497833e-05,
+      "loss": 0.701,
+      "step": 4176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.382999149204823,
+      "learning_rate": 1.9755666300374554e-05,
+      "loss": 0.9411,
+      "step": 4177
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7333529954145301,
+      "learning_rate": 1.9755498628970202e-05,
+      "loss": 0.8448,
+      "step": 4178
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.791722267595404,
+      "learning_rate": 1.9755330900766247e-05,
+      "loss": 1.0524,
+      "step": 4179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8730619287327195,
+      "learning_rate": 1.975516311576367e-05,
+      "loss": 0.9793,
+      "step": 4180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8525289298208623,
+      "learning_rate": 1.975499527396344e-05,
+      "loss": 0.7594,
+      "step": 4181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.476821954239126,
+      "learning_rate": 1.975482737536654e-05,
+      "loss": 0.9226,
+      "step": 4182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.784415633269061,
+      "learning_rate": 1.975465941997395e-05,
+      "loss": 0.708,
+      "step": 4183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.943878566310865,
+      "learning_rate": 1.9754491407786642e-05,
+      "loss": 0.9449,
+      "step": 4184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7616674833288253,
+      "learning_rate": 1.9754323338805597e-05,
+      "loss": 0.8747,
+      "step": 4185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6425078749943052,
+      "learning_rate": 1.9754155213031793e-05,
+      "loss": 0.875,
+      "step": 4186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6863707841487847,
+      "learning_rate": 1.9753987030466212e-05,
+      "loss": 0.9598,
+      "step": 4187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4835496005970201,
+      "learning_rate": 1.975381879110983e-05,
+      "loss": 0.8376,
+      "step": 4188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.843694137952165,
+      "learning_rate": 1.975365049496362e-05,
+      "loss": 0.8579,
+      "step": 4189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.027974135512824,
+      "learning_rate": 1.9753482142028578e-05,
+      "loss": 0.9591,
+      "step": 4190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.065312689850435,
+      "learning_rate": 1.975331373230567e-05,
+      "loss": 1.0485,
+      "step": 4191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8469022821721862,
+      "learning_rate": 1.9753145265795885e-05,
+      "loss": 0.7519,
+      "step": 4192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.771226132722556,
+      "learning_rate": 1.9752976742500198e-05,
+      "loss": 0.8349,
+      "step": 4193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7919626315701491,
+      "learning_rate": 1.975280816241959e-05,
+      "loss": 0.7405,
+      "step": 4194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.790236935846118,
+      "learning_rate": 1.9752639525555048e-05,
+      "loss": 0.8741,
+      "step": 4195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2933774311585826,
+      "learning_rate": 1.9752470831907553e-05,
+      "loss": 0.9453,
+      "step": 4196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8121456097789432,
+      "learning_rate": 1.9752302081478084e-05,
+      "loss": 0.7228,
+      "step": 4197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.192023361436427,
+      "learning_rate": 1.9752133274267626e-05,
+      "loss": 0.9526,
+      "step": 4198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8890406475758479,
+      "learning_rate": 1.9751964410277157e-05,
+      "loss": 0.9991,
+      "step": 4199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7457483357687575,
+      "learning_rate": 1.9751795489507665e-05,
+      "loss": 0.767,
+      "step": 4200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.940999659269498,
+      "learning_rate": 1.9751626511960134e-05,
+      "loss": 0.8961,
+      "step": 4201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7836823342466184,
+      "learning_rate": 1.9751457477635544e-05,
+      "loss": 0.9174,
+      "step": 4202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.403667541751709,
+      "learning_rate": 1.9751288386534885e-05,
+      "loss": 0.9703,
+      "step": 4203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8424712697506487,
+      "learning_rate": 1.9751119238659132e-05,
+      "loss": 0.959,
+      "step": 4204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7937651736766362,
+      "learning_rate": 1.9750950034009282e-05,
+      "loss": 0.9034,
+      "step": 4205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7890701388736435,
+      "learning_rate": 1.975078077258631e-05,
+      "loss": 0.723,
+      "step": 4206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8458107218557442,
+      "learning_rate": 1.9750611454391208e-05,
+      "loss": 1.0935,
+      "step": 4207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8112517298664927,
+      "learning_rate": 1.9750442079424953e-05,
+      "loss": 0.7677,
+      "step": 4208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5685189356803748,
+      "learning_rate": 1.9750272647688544e-05,
+      "loss": 0.9955,
+      "step": 4209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7075835595849675,
+      "learning_rate": 1.9750103159182957e-05,
+      "loss": 1.016,
+      "step": 4210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1391413742147325,
+      "learning_rate": 1.974993361390918e-05,
+      "loss": 0.8825,
+      "step": 4211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6077669663584784,
+      "learning_rate": 1.9749764011868212e-05,
+      "loss": 0.9876,
+      "step": 4212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.802655077638821,
+      "learning_rate": 1.9749594353061024e-05,
+      "loss": 0.8658,
+      "step": 4213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7713418476605134,
+      "learning_rate": 1.974942463748861e-05,
+      "loss": 0.8336,
+      "step": 4214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.649618857821199,
+      "learning_rate": 1.9749254865151966e-05,
+      "loss": 0.9109,
+      "step": 4215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.630633810642605,
+      "learning_rate": 1.9749085036052066e-05,
+      "loss": 0.8789,
+      "step": 4216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0723242528947097,
+      "learning_rate": 1.974891515018991e-05,
+      "loss": 0.9082,
+      "step": 4217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8759251902952824,
+      "learning_rate": 1.9748745207566483e-05,
+      "loss": 0.8401,
+      "step": 4218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8215667121949688,
+      "learning_rate": 1.9748575208182776e-05,
+      "loss": 1.0212,
+      "step": 4219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.028055182909521,
+      "learning_rate": 1.9748405152039777e-05,
+      "loss": 0.8547,
+      "step": 4220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6809739155761583,
+      "learning_rate": 1.974823503913848e-05,
+      "loss": 1.0747,
+      "step": 4221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8042223685423637,
+      "learning_rate": 1.9748064869479866e-05,
+      "loss": 0.7927,
+      "step": 4222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1423432231497372,
+      "learning_rate": 1.9747894643064938e-05,
+      "loss": 0.9185,
+      "step": 4223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8974775017762946,
+      "learning_rate": 1.9747724359894677e-05,
+      "loss": 0.9662,
+      "step": 4224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7260707199285137,
+      "learning_rate": 1.974755401997008e-05,
+      "loss": 0.9266,
+      "step": 4225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7299104807144603,
+      "learning_rate": 1.974738362329214e-05,
+      "loss": 0.8295,
+      "step": 4226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5800967560201369,
+      "learning_rate": 1.9747213169861844e-05,
+      "loss": 0.8774,
+      "step": 4227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1758479035852387,
+      "learning_rate": 1.974704265968019e-05,
+      "loss": 1.0347,
+      "step": 4228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7586720321246957,
+      "learning_rate": 1.9746872092748162e-05,
+      "loss": 0.7824,
+      "step": 4229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.500822079127389,
+      "learning_rate": 1.9746701469066766e-05,
+      "loss": 0.9028,
+      "step": 4230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7902505796848738,
+      "learning_rate": 1.974653078863698e-05,
+      "loss": 0.9972,
+      "step": 4231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6733163690604036,
+      "learning_rate": 1.974636005145981e-05,
+      "loss": 0.9164,
+      "step": 4232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7418500748813583,
+      "learning_rate": 1.9746189257536252e-05,
+      "loss": 0.9838,
+      "step": 4233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0215173111354017,
+      "learning_rate": 1.9746018406867285e-05,
+      "loss": 0.8718,
+      "step": 4234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8571969664149433,
+      "learning_rate": 1.9745847499453918e-05,
+      "loss": 0.9316,
+      "step": 4235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4407208917897067,
+      "learning_rate": 1.974567653529714e-05,
+      "loss": 0.8192,
+      "step": 4236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.639466284462648,
+      "learning_rate": 1.9745505514397952e-05,
+      "loss": 0.9254,
+      "step": 4237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2664583432167293,
+      "learning_rate": 1.9745334436757342e-05,
+      "loss": 1.0357,
+      "step": 4238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.761322664956236,
+      "learning_rate": 1.9745163302376306e-05,
+      "loss": 0.8898,
+      "step": 4239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7943825692899068,
+      "learning_rate": 1.974499211125585e-05,
+      "loss": 0.6786,
+      "step": 4240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7800127412420879,
+      "learning_rate": 1.9744820863396962e-05,
+      "loss": 0.8758,
+      "step": 4241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9154128943990578,
+      "learning_rate": 1.9744649558800644e-05,
+      "loss": 0.9884,
+      "step": 4242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7455029901854555,
+      "learning_rate": 1.974447819746789e-05,
+      "loss": 0.9384,
+      "step": 4243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7015000486408718,
+      "learning_rate": 1.9744306779399697e-05,
+      "loss": 0.8376,
+      "step": 4244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.257237294757785,
+      "learning_rate": 1.9744135304597068e-05,
+      "loss": 0.8464,
+      "step": 4245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7763126771674198,
+      "learning_rate": 1.9743963773060996e-05,
+      "loss": 0.9568,
+      "step": 4246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.114415189444847,
+      "learning_rate": 1.974379218479248e-05,
+      "loss": 0.9683,
+      "step": 4247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0051949621268617,
+      "learning_rate": 1.9743620539792525e-05,
+      "loss": 0.8998,
+      "step": 4248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6129071586172128,
+      "learning_rate": 1.9743448838062126e-05,
+      "loss": 0.8861,
+      "step": 4249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8011019019075674,
+      "learning_rate": 1.974327707960228e-05,
+      "loss": 0.8649,
+      "step": 4250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5090844305351987,
+      "learning_rate": 1.9743105264413994e-05,
+      "loss": 0.8137,
+      "step": 4251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5643570949516938,
+      "learning_rate": 1.9742933392498263e-05,
+      "loss": 0.8764,
+      "step": 4252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8256746870025387,
+      "learning_rate": 1.974276146385609e-05,
+      "loss": 0.8653,
+      "step": 4253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8637087431714319,
+      "learning_rate": 1.9742589478488474e-05,
+      "loss": 0.9141,
+      "step": 4254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6966168900136016,
+      "learning_rate": 1.974241743639642e-05,
+      "loss": 0.9719,
+      "step": 4255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1539574529290433,
+      "learning_rate": 1.9742245337580924e-05,
+      "loss": 0.902,
+      "step": 4256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.15139211883127,
+      "learning_rate": 1.974207318204299e-05,
+      "loss": 0.8954,
+      "step": 4257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6891118970330004,
+      "learning_rate": 1.9741900969783625e-05,
+      "loss": 0.8183,
+      "step": 4258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7592667759837595,
+      "learning_rate": 1.974172870080383e-05,
+      "loss": 0.7301,
+      "step": 4259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8452159686197434,
+      "learning_rate": 1.97415563751046e-05,
+      "loss": 0.9407,
+      "step": 4260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2827178547834235,
+      "learning_rate": 1.974138399268695e-05,
+      "loss": 0.9214,
+      "step": 4261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7595711215433453,
+      "learning_rate": 1.9741211553551877e-05,
+      "loss": 0.7318,
+      "step": 4262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3635976355959776,
+      "learning_rate": 1.9741039057700385e-05,
+      "loss": 0.9188,
+      "step": 4263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.620892191670292,
+      "learning_rate": 1.974086650513348e-05,
+      "loss": 0.9188,
+      "step": 4264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.630146622199892,
+      "learning_rate": 1.974069389585217e-05,
+      "loss": 0.9514,
+      "step": 4265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9033032694862624,
+      "learning_rate": 1.9740521229857453e-05,
+      "loss": 0.9473,
+      "step": 4266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6602549152997785,
+      "learning_rate": 1.9740348507150338e-05,
+      "loss": 0.8657,
+      "step": 4267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7809781272207672,
+      "learning_rate": 1.9740175727731832e-05,
+      "loss": 0.7255,
+      "step": 4268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.747170001172922,
+      "learning_rate": 1.9740002891602936e-05,
+      "loss": 0.9603,
+      "step": 4269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6280698471697572,
+      "learning_rate": 1.9739829998764662e-05,
+      "loss": 1.0286,
+      "step": 4270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7906331558059254,
+      "learning_rate": 1.9739657049218014e-05,
+      "loss": 0.7279,
+      "step": 4271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.888030846746654,
+      "learning_rate": 1.9739484042963997e-05,
+      "loss": 1.0803,
+      "step": 4272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7441734859559004,
+      "learning_rate": 1.973931098000362e-05,
+      "loss": 0.9272,
+      "step": 4273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.092773892439429,
+      "learning_rate": 1.9739137860337893e-05,
+      "loss": 0.847,
+      "step": 4274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7435441904968807,
+      "learning_rate": 1.9738964683967823e-05,
+      "loss": 0.8517,
+      "step": 4275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6959443926550064,
+      "learning_rate": 1.9738791450894415e-05,
+      "loss": 1.0051,
+      "step": 4276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7371429476942709,
+      "learning_rate": 1.9738618161118684e-05,
+      "loss": 0.8035,
+      "step": 4277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.11079617264472,
+      "learning_rate": 1.973844481464163e-05,
+      "loss": 0.8531,
+      "step": 4278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7776482859598082,
+      "learning_rate": 1.973827141146427e-05,
+      "loss": 0.883,
+      "step": 4279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.524198101903834,
+      "learning_rate": 1.973809795158761e-05,
+      "loss": 0.8522,
+      "step": 4280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.103293349057757,
+      "learning_rate": 1.973792443501266e-05,
+      "loss": 0.8857,
+      "step": 4281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7385250737937656,
+      "learning_rate": 1.9737750861740434e-05,
+      "loss": 0.8339,
+      "step": 4282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8603227534894035,
+      "learning_rate": 1.9737577231771934e-05,
+      "loss": 0.7644,
+      "step": 4283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7927638426521607,
+      "learning_rate": 1.973740354510818e-05,
+      "loss": 0.7194,
+      "step": 4284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7021822507970576,
+      "learning_rate": 1.973722980175018e-05,
+      "loss": 1.0224,
+      "step": 4285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2112556267505727,
+      "learning_rate": 1.9737056001698946e-05,
+      "loss": 1.0587,
+      "step": 4286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8505517586229117,
+      "learning_rate": 1.9736882144955487e-05,
+      "loss": 0.8455,
+      "step": 4287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4547586395214416,
+      "learning_rate": 1.973670823152082e-05,
+      "loss": 0.8469,
+      "step": 4288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7964305433849981,
+      "learning_rate": 1.9736534261395953e-05,
+      "loss": 0.8085,
+      "step": 4289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6659923075591843,
+      "learning_rate": 1.9736360234581902e-05,
+      "loss": 0.8004,
+      "step": 4290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6274123731640688,
+      "learning_rate": 1.9736186151079677e-05,
+      "loss": 0.8977,
+      "step": 4291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8844639499338351,
+      "learning_rate": 1.9736012010890294e-05,
+      "loss": 1.0409,
+      "step": 4292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5713525204020284,
+      "learning_rate": 1.9735837814014772e-05,
+      "loss": 0.8958,
+      "step": 4293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7318556268679381,
+      "learning_rate": 1.9735663560454115e-05,
+      "loss": 0.9138,
+      "step": 4294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1827763146732835,
+      "learning_rate": 1.9735489250209342e-05,
+      "loss": 0.9526,
+      "step": 4295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4668086411160413,
+      "learning_rate": 1.973531488328147e-05,
+      "loss": 0.8739,
+      "step": 4296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9301503018377344,
+      "learning_rate": 1.9735140459671512e-05,
+      "loss": 0.967,
+      "step": 4297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.682587373411129,
+      "learning_rate": 1.9734965979380487e-05,
+      "loss": 0.8903,
+      "step": 4298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6443097084336034,
+      "learning_rate": 1.9734791442409403e-05,
+      "loss": 0.9573,
+      "step": 4299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9286888917133822,
+      "learning_rate": 1.9734616848759288e-05,
+      "loss": 0.8114,
+      "step": 4300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5584772968654137,
+      "learning_rate": 1.9734442198431146e-05,
+      "loss": 0.9172,
+      "step": 4301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7261483983485302,
+      "learning_rate": 1.9734267491426003e-05,
+      "loss": 0.9422,
+      "step": 4302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.634793641328585,
+      "learning_rate": 1.973409272774487e-05,
+      "loss": 1.1932,
+      "step": 4303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7755206627534224,
+      "learning_rate": 1.9733917907388772e-05,
+      "loss": 1.0706,
+      "step": 4304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7849272678670546,
+      "learning_rate": 1.9733743030358716e-05,
+      "loss": 0.7492,
+      "step": 4305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.892053429437752,
+      "learning_rate": 1.973356809665573e-05,
+      "loss": 0.7887,
+      "step": 4306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3367228500382016,
+      "learning_rate": 1.9733393106280826e-05,
+      "loss": 0.7571,
+      "step": 4307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5626619217277895,
+      "learning_rate": 1.9733218059235032e-05,
+      "loss": 0.8507,
+      "step": 4308
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6177547456539012,
+      "learning_rate": 1.9733042955519356e-05,
+      "loss": 0.8908,
+      "step": 4309
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.450934240315167,
+      "learning_rate": 1.9732867795134824e-05,
+      "loss": 0.8928,
+      "step": 4310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5904209718863023,
+      "learning_rate": 1.9732692578082453e-05,
+      "loss": 0.9081,
+      "step": 4311
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.273623048549037,
+      "learning_rate": 1.9732517304363266e-05,
+      "loss": 1.0116,
+      "step": 4312
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5720138518886768,
+      "learning_rate": 1.9732341973978283e-05,
+      "loss": 0.9516,
+      "step": 4313
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6416355869323196,
+      "learning_rate": 1.973216658692852e-05,
+      "loss": 1.002,
+      "step": 4314
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6933288651389937,
+      "learning_rate": 1.9731991143215005e-05,
+      "loss": 0.9865,
+      "step": 4315
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.136936471079121,
+      "learning_rate": 1.9731815642838756e-05,
+      "loss": 0.9231,
+      "step": 4316
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6889610075086823,
+      "learning_rate": 1.9731640085800794e-05,
+      "loss": 1.0491,
+      "step": 4317
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.721297080468653,
+      "learning_rate": 1.9731464472102145e-05,
+      "loss": 0.997,
+      "step": 4318
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.066664952273591,
+      "learning_rate": 1.9731288801743826e-05,
+      "loss": 0.8768,
+      "step": 4319
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.811389416607597,
+      "learning_rate": 1.9731113074726863e-05,
+      "loss": 1.0009,
+      "step": 4320
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.086309297800912,
+      "learning_rate": 1.973093729105228e-05,
+      "loss": 0.8495,
+      "step": 4321
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8337168153200688,
+      "learning_rate": 1.9730761450721103e-05,
+      "loss": 0.742,
+      "step": 4322
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6061903448663326,
+      "learning_rate": 1.9730585553734345e-05,
+      "loss": 0.8714,
+      "step": 4323
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4950378958807398,
+      "learning_rate": 1.9730409600093042e-05,
+      "loss": 0.9435,
+      "step": 4324
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5426544870058003,
+      "learning_rate": 1.9730233589798213e-05,
+      "loss": 0.897,
+      "step": 4325
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7200704287457718,
+      "learning_rate": 1.9730057522850882e-05,
+      "loss": 0.872,
+      "step": 4326
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8879920782442623,
+      "learning_rate": 1.9729881399252077e-05,
+      "loss": 0.8757,
+      "step": 4327
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8386511963716965,
+      "learning_rate": 1.972970521900282e-05,
+      "loss": 0.9379,
+      "step": 4328
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6266239212149156,
+      "learning_rate": 1.972952898210414e-05,
+      "loss": 0.7839,
+      "step": 4329
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.088815354627164,
+      "learning_rate": 1.9729352688557064e-05,
+      "loss": 0.8556,
+      "step": 4330
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.02333897033219,
+      "learning_rate": 1.9729176338362613e-05,
+      "loss": 0.9735,
+      "step": 4331
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6378706336080764,
+      "learning_rate": 1.9728999931521818e-05,
+      "loss": 1.0055,
+      "step": 4332
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5515540520660376,
+      "learning_rate": 1.9728823468035708e-05,
+      "loss": 0.9857,
+      "step": 4333
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4003735876623304,
+      "learning_rate": 1.9728646947905307e-05,
+      "loss": 0.974,
+      "step": 4334
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2374428518477045,
+      "learning_rate": 1.972847037113164e-05,
+      "loss": 0.838,
+      "step": 4335
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6124696469196491,
+      "learning_rate": 1.9728293737715742e-05,
+      "loss": 0.9091,
+      "step": 4336
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5001012100065632,
+      "learning_rate": 1.9728117047658636e-05,
+      "loss": 0.7673,
+      "step": 4337
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8696345094985116,
+      "learning_rate": 1.9727940300961354e-05,
+      "loss": 0.874,
+      "step": 4338
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6715073477377305,
+      "learning_rate": 1.9727763497624924e-05,
+      "loss": 0.8934,
+      "step": 4339
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7164513185393893,
+      "learning_rate": 1.9727586637650373e-05,
+      "loss": 0.9003,
+      "step": 4340
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8332082231250395,
+      "learning_rate": 1.9727409721038736e-05,
+      "loss": 0.7793,
+      "step": 4341
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4592213338709028,
+      "learning_rate": 1.9727232747791037e-05,
+      "loss": 0.7325,
+      "step": 4342
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5822519696245534,
+      "learning_rate": 1.972705571790831e-05,
+      "loss": 0.9249,
+      "step": 4343
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.684834726165629,
+      "learning_rate": 1.9726878631391585e-05,
+      "loss": 1.0532,
+      "step": 4344
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9012129190330027,
+      "learning_rate": 1.9726701488241895e-05,
+      "loss": 0.8856,
+      "step": 4345
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6171254444460614,
+      "learning_rate": 1.9726524288460266e-05,
+      "loss": 0.9793,
+      "step": 4346
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8384340370126215,
+      "learning_rate": 1.9726347032047735e-05,
+      "loss": 0.9021,
+      "step": 4347
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8578328628791132,
+      "learning_rate": 1.9726169719005336e-05,
+      "loss": 1.2462,
+      "step": 4348
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6528481811244373,
+      "learning_rate": 1.9725992349334092e-05,
+      "loss": 0.8116,
+      "step": 4349
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9609000879608116,
+      "learning_rate": 1.9725814923035044e-05,
+      "loss": 0.8845,
+      "step": 4350
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.026942473438204,
+      "learning_rate": 1.9725637440109223e-05,
+      "loss": 0.9331,
+      "step": 4351
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9564337091954167,
+      "learning_rate": 1.972545990055766e-05,
+      "loss": 0.9664,
+      "step": 4352
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1791573311065178,
+      "learning_rate": 1.972528230438139e-05,
+      "loss": 0.9228,
+      "step": 4353
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3627371804349413,
+      "learning_rate": 1.972510465158145e-05,
+      "loss": 1.0797,
+      "step": 4354
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4931085400267445,
+      "learning_rate": 1.972492694215887e-05,
+      "loss": 0.8501,
+      "step": 4355
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5717106244984187,
+      "learning_rate": 1.972474917611469e-05,
+      "loss": 0.8912,
+      "step": 4356
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0850405796214955,
+      "learning_rate": 1.9724571353449933e-05,
+      "loss": 0.919,
+      "step": 4357
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1705867865627053,
+      "learning_rate": 1.972439347416565e-05,
+      "loss": 0.884,
+      "step": 4358
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8118242819779298,
+      "learning_rate": 1.972421553826287e-05,
+      "loss": 0.7182,
+      "step": 4359
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.875731451598332,
+      "learning_rate": 1.9724037545742624e-05,
+      "loss": 0.8771,
+      "step": 4360
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8061419935244418,
+      "learning_rate": 1.9723859496605958e-05,
+      "loss": 1.0405,
+      "step": 4361
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.242660522331164,
+      "learning_rate": 1.97236813908539e-05,
+      "loss": 0.8348,
+      "step": 4362
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6778157334715404,
+      "learning_rate": 1.972350322848749e-05,
+      "loss": 0.9245,
+      "step": 4363
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8107522220604069,
+      "learning_rate": 1.972332500950777e-05,
+      "loss": 0.9976,
+      "step": 4364
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6736197429192414,
+      "learning_rate": 1.9723146733915766e-05,
+      "loss": 0.9427,
+      "step": 4365
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8029527356287576,
+      "learning_rate": 1.9722968401712527e-05,
+      "loss": 1.0966,
+      "step": 4366
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7510063531476734,
+      "learning_rate": 1.972279001289909e-05,
+      "loss": 0.715,
+      "step": 4367
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7358583484540542,
+      "learning_rate": 1.972261156747649e-05,
+      "loss": 0.8772,
+      "step": 4368
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7752471968086019,
+      "learning_rate": 1.9722433065445767e-05,
+      "loss": 0.8704,
+      "step": 4369
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6636978686526238,
+      "learning_rate": 1.972225450680796e-05,
+      "loss": 0.8734,
+      "step": 4370
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7709753917106557,
+      "learning_rate": 1.972207589156411e-05,
+      "loss": 0.737,
+      "step": 4371
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8048038653024572,
+      "learning_rate": 1.9721897219715254e-05,
+      "loss": 1.014,
+      "step": 4372
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.786685375657261,
+      "learning_rate": 1.972171849126244e-05,
+      "loss": 0.8512,
+      "step": 4373
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.754638310382371,
+      "learning_rate": 1.9721539706206697e-05,
+      "loss": 1.0308,
+      "step": 4374
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7747806393875971,
+      "learning_rate": 1.9721360864549078e-05,
+      "loss": 0.8361,
+      "step": 4375
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1606348783201983,
+      "learning_rate": 1.9721181966290614e-05,
+      "loss": 0.9403,
+      "step": 4376
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.160204542701167,
+      "learning_rate": 1.972100301143235e-05,
+      "loss": 0.9424,
+      "step": 4377
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7142636598937904,
+      "learning_rate": 1.9720823999975333e-05,
+      "loss": 0.9152,
+      "step": 4378
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7188449385686377,
+      "learning_rate": 1.97206449319206e-05,
+      "loss": 0.857,
+      "step": 4379
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7734759173703798,
+      "learning_rate": 1.9720465807269192e-05,
+      "loss": 0.9164,
+      "step": 4380
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9419578798526491,
+      "learning_rate": 1.972028662602216e-05,
+      "loss": 0.8977,
+      "step": 4381
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7910047910825534,
+      "learning_rate": 1.9720107388180536e-05,
+      "loss": 0.7407,
+      "step": 4382
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.531502169111113,
+      "learning_rate": 1.9719928093745372e-05,
+      "loss": 0.9042,
+      "step": 4383
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.876722348872026,
+      "learning_rate": 1.971974874271771e-05,
+      "loss": 0.8959,
+      "step": 4384
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6126065573621078,
+      "learning_rate": 1.9719569335098592e-05,
+      "loss": 0.9922,
+      "step": 4385
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.59303627649931,
+      "learning_rate": 1.9719389870889064e-05,
+      "loss": 0.9152,
+      "step": 4386
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7395531464622853,
+      "learning_rate": 1.9719210350090172e-05,
+      "loss": 0.8501,
+      "step": 4387
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8516994882548898,
+      "learning_rate": 1.971903077270296e-05,
+      "loss": 0.8468,
+      "step": 4388
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2147570196279025,
+      "learning_rate": 1.9718851138728474e-05,
+      "loss": 0.9226,
+      "step": 4389
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7479982158712009,
+      "learning_rate": 1.971867144816776e-05,
+      "loss": 0.845,
+      "step": 4390
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8093116624329133,
+      "learning_rate": 1.9718491701021865e-05,
+      "loss": 0.9711,
+      "step": 4391
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5938000188522832,
+      "learning_rate": 1.971831189729183e-05,
+      "loss": 0.8899,
+      "step": 4392
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5695257991146176,
+      "learning_rate": 1.9718132036978712e-05,
+      "loss": 0.9324,
+      "step": 4393
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7777778441382042,
+      "learning_rate": 1.971795212008355e-05,
+      "loss": 0.7073,
+      "step": 4394
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0981057787180113,
+      "learning_rate": 1.9717772146607396e-05,
+      "loss": 1.0321,
+      "step": 4395
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.143431839675845,
+      "learning_rate": 1.971759211655129e-05,
+      "loss": 0.9038,
+      "step": 4396
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6810847753424074,
+      "learning_rate": 1.971741202991629e-05,
+      "loss": 1.0301,
+      "step": 4397
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.810179274537569,
+      "learning_rate": 1.9717231886703444e-05,
+      "loss": 0.7529,
+      "step": 4398
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.910970652067455,
+      "learning_rate": 1.9717051686913793e-05,
+      "loss": 0.9017,
+      "step": 4399
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.82088120735044,
+      "learning_rate": 1.9716871430548394e-05,
+      "loss": 1.0204,
+      "step": 4400
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5527396185504498,
+      "learning_rate": 1.971669111760829e-05,
+      "loss": 0.9059,
+      "step": 4401
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3385429985267097,
+      "learning_rate": 1.9716510748094536e-05,
+      "loss": 1.0008,
+      "step": 4402
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6850539002792546,
+      "learning_rate": 1.971633032200818e-05,
+      "loss": 0.9754,
+      "step": 4403
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6214890933871342,
+      "learning_rate": 1.971614983935027e-05,
+      "loss": 0.9748,
+      "step": 4404
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6336392919358933,
+      "learning_rate": 1.9715969300121863e-05,
+      "loss": 0.8623,
+      "step": 4405
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7646987641472431,
+      "learning_rate": 1.971578870432401e-05,
+      "loss": 0.9301,
+      "step": 4406
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.729186828598191,
+      "learning_rate": 1.9715608051957753e-05,
+      "loss": 0.8865,
+      "step": 4407
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1695540268038176,
+      "learning_rate": 1.971542734302415e-05,
+      "loss": 1.0034,
+      "step": 4408
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6017179620845476,
+      "learning_rate": 1.9715246577524258e-05,
+      "loss": 0.8968,
+      "step": 4409
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9200584899427366,
+      "learning_rate": 1.9715065755459122e-05,
+      "loss": 0.9143,
+      "step": 4410
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.690982010567964,
+      "learning_rate": 1.9714884876829797e-05,
+      "loss": 0.9588,
+      "step": 4411
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.095251749882023,
+      "learning_rate": 1.9714703941637333e-05,
+      "loss": 0.8379,
+      "step": 4412
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.099306528988774,
+      "learning_rate": 1.9714522949882794e-05,
+      "loss": 0.9948,
+      "step": 4413
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0551290691863735,
+      "learning_rate": 1.9714341901567223e-05,
+      "loss": 1.0125,
+      "step": 4414
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8948205879939266,
+      "learning_rate": 1.971416079669168e-05,
+      "loss": 0.7463,
+      "step": 4415
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8839503306560772,
+      "learning_rate": 1.9713979635257215e-05,
+      "loss": 0.9135,
+      "step": 4416
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.654514103816669,
+      "learning_rate": 1.9713798417264885e-05,
+      "loss": 1.0193,
+      "step": 4417
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6012696332223952,
+      "learning_rate": 1.9713617142715748e-05,
+      "loss": 0.9636,
+      "step": 4418
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5059046066891901,
+      "learning_rate": 1.9713435811610855e-05,
+      "loss": 0.9727,
+      "step": 4419
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.652964806573785,
+      "learning_rate": 1.9713254423951263e-05,
+      "loss": 0.9103,
+      "step": 4420
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8345199915108004,
+      "learning_rate": 1.971307297973803e-05,
+      "loss": 0.7049,
+      "step": 4421
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7019373468037755,
+      "learning_rate": 1.971289147897221e-05,
+      "loss": 0.8737,
+      "step": 4422
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6224972223161618,
+      "learning_rate": 1.971270992165486e-05,
+      "loss": 0.9527,
+      "step": 4423
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6615561697293308,
+      "learning_rate": 1.971252830778704e-05,
+      "loss": 0.9011,
+      "step": 4424
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7239278570192496,
+      "learning_rate": 1.9712346637369803e-05,
+      "loss": 0.8758,
+      "step": 4425
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1378263166087708,
+      "learning_rate": 1.971216491040421e-05,
+      "loss": 0.9979,
+      "step": 4426
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.769112062559253,
+      "learning_rate": 1.971198312689132e-05,
+      "loss": 0.9304,
+      "step": 4427
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.876851590207032,
+      "learning_rate": 1.9711801286832186e-05,
+      "loss": 1.0211,
+      "step": 4428
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.502278925203113,
+      "learning_rate": 1.971161939022787e-05,
+      "loss": 0.9597,
+      "step": 4429
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5262270329844183,
+      "learning_rate": 1.9711437437079436e-05,
+      "loss": 0.8549,
+      "step": 4430
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.635270781147331,
+      "learning_rate": 1.9711255427387935e-05,
+      "loss": 0.9218,
+      "step": 4431
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.692385886246487,
+      "learning_rate": 1.9711073361154433e-05,
+      "loss": 0.9838,
+      "step": 4432
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7144125519539606,
+      "learning_rate": 1.9710891238379984e-05,
+      "loss": 0.9086,
+      "step": 4433
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7607096687006005,
+      "learning_rate": 1.9710709059065653e-05,
+      "loss": 0.8617,
+      "step": 4434
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.719052308728783,
+      "learning_rate": 1.97105268232125e-05,
+      "loss": 0.8974,
+      "step": 4435
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7740162785986178,
+      "learning_rate": 1.9710344530821586e-05,
+      "loss": 0.7107,
+      "step": 4436
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6134495386352805,
+      "learning_rate": 1.9710162181893972e-05,
+      "loss": 1.0992,
+      "step": 4437
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.768010147484293,
+      "learning_rate": 1.9709979776430717e-05,
+      "loss": 0.7243,
+      "step": 4438
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6201939346742953,
+      "learning_rate": 1.9709797314432885e-05,
+      "loss": 0.9401,
+      "step": 4439
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7065402728272292,
+      "learning_rate": 1.9709614795901543e-05,
+      "loss": 0.9449,
+      "step": 4440
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8474557239583624,
+      "learning_rate": 1.9709432220837748e-05,
+      "loss": 0.9485,
+      "step": 4441
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5376521951605344,
+      "learning_rate": 1.9709249589242564e-05,
+      "loss": 1.0804,
+      "step": 4442
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9646879512792803,
+      "learning_rate": 1.970906690111706e-05,
+      "loss": 0.7924,
+      "step": 4443
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9453214113228132,
+      "learning_rate": 1.9708884156462286e-05,
+      "loss": 1.076,
+      "step": 4444
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.023685108003052,
+      "learning_rate": 1.970870135527932e-05,
+      "loss": 0.927,
+      "step": 4445
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6448452295775995,
+      "learning_rate": 1.970851849756922e-05,
+      "loss": 0.9669,
+      "step": 4446
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7832730807931134,
+      "learning_rate": 1.9708335583333052e-05,
+      "loss": 1.0053,
+      "step": 4447
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7272395969265342,
+      "learning_rate": 1.970815261257188e-05,
+      "loss": 1.0059,
+      "step": 4448
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7990096788432831,
+      "learning_rate": 1.970796958528677e-05,
+      "loss": 0.7893,
+      "step": 4449
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.76783003964226,
+      "learning_rate": 1.970778650147879e-05,
+      "loss": 0.7248,
+      "step": 4450
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7341175440339232,
+      "learning_rate": 1.9707603361149e-05,
+      "loss": 0.97,
+      "step": 4451
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5568840761102407,
+      "learning_rate": 1.970742016429847e-05,
+      "loss": 0.8696,
+      "step": 4452
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7306368927422952,
+      "learning_rate": 1.970723691092827e-05,
+      "loss": 0.7591,
+      "step": 4453
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9748437214279693,
+      "learning_rate": 1.970705360103946e-05,
+      "loss": 0.8951,
+      "step": 4454
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8063075387649139,
+      "learning_rate": 1.9706870234633114e-05,
+      "loss": 0.753,
+      "step": 4455
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2791348705472902,
+      "learning_rate": 1.970668681171029e-05,
+      "loss": 0.9805,
+      "step": 4456
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9136419915138554,
+      "learning_rate": 1.970650333227207e-05,
+      "loss": 0.9476,
+      "step": 4457
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9149579488909905,
+      "learning_rate": 1.970631979631951e-05,
+      "loss": 0.9456,
+      "step": 4458
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5685012880552083,
+      "learning_rate": 1.9706136203853684e-05,
+      "loss": 0.9846,
+      "step": 4459
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.498406839221745,
+      "learning_rate": 1.970595255487566e-05,
+      "loss": 0.8582,
+      "step": 4460
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7672990553340338,
+      "learning_rate": 1.970576884938651e-05,
+      "loss": 0.7254,
+      "step": 4461
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7755522695198187,
+      "learning_rate": 1.9705585087387297e-05,
+      "loss": 0.8834,
+      "step": 4462
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4488771719121742,
+      "learning_rate": 1.9705401268879095e-05,
+      "loss": 0.9362,
+      "step": 4463
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6538408171465613,
+      "learning_rate": 1.9705217393862977e-05,
+      "loss": 1.0115,
+      "step": 4464
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8246164573185605,
+      "learning_rate": 1.970503346234001e-05,
+      "loss": 0.7716,
+      "step": 4465
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.032189022197064,
+      "learning_rate": 1.9704849474311266e-05,
+      "loss": 0.976,
+      "step": 4466
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.698564651071886,
+      "learning_rate": 1.9704665429777815e-05,
+      "loss": 1.0395,
+      "step": 4467
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9024735126272536,
+      "learning_rate": 1.9704481328740726e-05,
+      "loss": 0.9689,
+      "step": 4468
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7794453220639045,
+      "learning_rate": 1.9704297171201082e-05,
+      "loss": 0.7297,
+      "step": 4469
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.139414625709916,
+      "learning_rate": 1.970411295715994e-05,
+      "loss": 0.9487,
+      "step": 4470
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.999292658326995,
+      "learning_rate": 1.9703928686618384e-05,
+      "loss": 0.9591,
+      "step": 4471
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.681003640013536,
+      "learning_rate": 1.9703744359577483e-05,
+      "loss": 1.0441,
+      "step": 4472
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.712209742578257,
+      "learning_rate": 1.9703559976038308e-05,
+      "loss": 0.793,
+      "step": 4473
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8029355321135908,
+      "learning_rate": 1.9703375536001937e-05,
+      "loss": 0.7518,
+      "step": 4474
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7961791713920996,
+      "learning_rate": 1.970319103946944e-05,
+      "loss": 0.7595,
+      "step": 4475
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0983856833188677,
+      "learning_rate": 1.9703006486441895e-05,
+      "loss": 0.8108,
+      "step": 4476
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5462543160152122,
+      "learning_rate": 1.9702821876920374e-05,
+      "loss": 0.8955,
+      "step": 4477
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.746726510082591,
+      "learning_rate": 1.970263721090595e-05,
+      "loss": 0.8659,
+      "step": 4478
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2927271488403633,
+      "learning_rate": 1.97024524883997e-05,
+      "loss": 0.9322,
+      "step": 4479
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6142216693424245,
+      "learning_rate": 1.9702267709402703e-05,
+      "loss": 0.9618,
+      "step": 4480
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.711191937256012,
+      "learning_rate": 1.970208287391603e-05,
+      "loss": 0.8988,
+      "step": 4481
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8213923587204061,
+      "learning_rate": 1.970189798194076e-05,
+      "loss": 0.7964,
+      "step": 4482
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.659219201102557,
+      "learning_rate": 1.9701713033477965e-05,
+      "loss": 1.024,
+      "step": 4483
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7677045871415844,
+      "learning_rate": 1.9701528028528728e-05,
+      "loss": 0.9345,
+      "step": 4484
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5435025691679494,
+      "learning_rate": 1.970134296709412e-05,
+      "loss": 0.8422,
+      "step": 4485
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7167798741996654,
+      "learning_rate": 1.970115784917523e-05,
+      "loss": 0.9976,
+      "step": 4486
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8461675388258343,
+      "learning_rate": 1.970097267477312e-05,
+      "loss": 0.9249,
+      "step": 4487
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0783953651029328,
+      "learning_rate": 1.9700787443888877e-05,
+      "loss": 0.9106,
+      "step": 4488
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7247543334792195,
+      "learning_rate": 1.9700602156523582e-05,
+      "loss": 0.9012,
+      "step": 4489
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.750907782574964,
+      "learning_rate": 1.9700416812678303e-05,
+      "loss": 1.0446,
+      "step": 4490
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8407309693883362,
+      "learning_rate": 1.9700231412354134e-05,
+      "loss": 0.9168,
+      "step": 4491
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7080181853806569,
+      "learning_rate": 1.9700045955552143e-05,
+      "loss": 0.9525,
+      "step": 4492
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.634929604660027,
+      "learning_rate": 1.9699860442273417e-05,
+      "loss": 0.9129,
+      "step": 4493
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8019399337724772,
+      "learning_rate": 1.9699674872519026e-05,
+      "loss": 0.765,
+      "step": 4494
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8734489302208455,
+      "learning_rate": 1.969948924629006e-05,
+      "loss": 0.9445,
+      "step": 4495
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6379933790088237,
+      "learning_rate": 1.9699303563587596e-05,
+      "loss": 0.9609,
+      "step": 4496
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7373715418020117,
+      "learning_rate": 1.9699117824412718e-05,
+      "loss": 0.9508,
+      "step": 4497
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9033131033825461,
+      "learning_rate": 1.9698932028766506e-05,
+      "loss": 0.9132,
+      "step": 4498
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.426388237119813,
+      "learning_rate": 1.969874617665004e-05,
+      "loss": 0.9906,
+      "step": 4499
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9205777860804698,
+      "learning_rate": 1.96985602680644e-05,
+      "loss": 0.9706,
+      "step": 4500
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9761028738839161,
+      "learning_rate": 1.969837430301068e-05,
+      "loss": 0.9128,
+      "step": 4501
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.647434696419593,
+      "learning_rate": 1.9698188281489944e-05,
+      "loss": 0.8482,
+      "step": 4502
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.792656099836318,
+      "learning_rate": 1.969800220350329e-05,
+      "loss": 1.0597,
+      "step": 4503
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9042788032291675,
+      "learning_rate": 1.96978160690518e-05,
+      "loss": 0.9295,
+      "step": 4504
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7999726371380159,
+      "learning_rate": 1.969762987813655e-05,
+      "loss": 0.7721,
+      "step": 4505
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9298109500300236,
+      "learning_rate": 1.969744363075863e-05,
+      "loss": 0.8706,
+      "step": 4506
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.475856100800338,
+      "learning_rate": 1.9697257326919123e-05,
+      "loss": 0.8822,
+      "step": 4507
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7714053169882339,
+      "learning_rate": 1.9697070966619115e-05,
+      "loss": 0.7678,
+      "step": 4508
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7351280134374112,
+      "learning_rate": 1.969688454985969e-05,
+      "loss": 0.7471,
+      "step": 4509
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.728966912006734,
+      "learning_rate": 1.969669807664193e-05,
+      "loss": 0.9808,
+      "step": 4510
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6980332120514798,
+      "learning_rate": 1.969651154696693e-05,
+      "loss": 0.893,
+      "step": 4511
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9298909667257296,
+      "learning_rate": 1.9696324960835763e-05,
+      "loss": 0.8563,
+      "step": 4512
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7164002941908738,
+      "learning_rate": 1.9696138318249526e-05,
+      "loss": 0.8557,
+      "step": 4513
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7418626125155636,
+      "learning_rate": 1.96959516192093e-05,
+      "loss": 0.9126,
+      "step": 4514
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6697986836892924,
+      "learning_rate": 1.9695764863716173e-05,
+      "loss": 0.9532,
+      "step": 4515
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7530102348350747,
+      "learning_rate": 1.9695578051771238e-05,
+      "loss": 0.8873,
+      "step": 4516
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8094660265223976,
+      "learning_rate": 1.9695391183375575e-05,
+      "loss": 0.7317,
+      "step": 4517
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7402826042227242,
+      "learning_rate": 1.9695204258530273e-05,
+      "loss": 0.963,
+      "step": 4518
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9267653044388449,
+      "learning_rate": 1.9695017277236425e-05,
+      "loss": 0.8581,
+      "step": 4519
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6603684542526114,
+      "learning_rate": 1.969483023949512e-05,
+      "loss": 0.827,
+      "step": 4520
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.740994646606985,
+      "learning_rate": 1.969464314530744e-05,
+      "loss": 0.8595,
+      "step": 4521
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7121317385431818,
+      "learning_rate": 1.9694455994674476e-05,
+      "loss": 0.9537,
+      "step": 4522
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.83800426941244,
+      "learning_rate": 1.9694268787597328e-05,
+      "loss": 0.9746,
+      "step": 4523
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7312248676818385,
+      "learning_rate": 1.9694081524077072e-05,
+      "loss": 0.898,
+      "step": 4524
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7429293999312032,
+      "learning_rate": 1.9693894204114803e-05,
+      "loss": 1.0381,
+      "step": 4525
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8493147165335613,
+      "learning_rate": 1.9693706827711615e-05,
+      "loss": 0.9136,
+      "step": 4526
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5361497395195773,
+      "learning_rate": 1.9693519394868595e-05,
+      "loss": 0.8649,
+      "step": 4527
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6480348228013957,
+      "learning_rate": 1.9693331905586842e-05,
+      "loss": 0.9681,
+      "step": 4528
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6554300619711348,
+      "learning_rate": 1.9693144359867436e-05,
+      "loss": 1.0341,
+      "step": 4529
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9104211434099394,
+      "learning_rate": 1.969295675771148e-05,
+      "loss": 0.8793,
+      "step": 4530
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7195662163979113,
+      "learning_rate": 1.9692769099120054e-05,
+      "loss": 0.765,
+      "step": 4531
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7076694700110546,
+      "learning_rate": 1.9692581384094262e-05,
+      "loss": 0.8185,
+      "step": 4532
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1170292680452336,
+      "learning_rate": 1.9692393612635192e-05,
+      "loss": 0.8015,
+      "step": 4533
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.619196368367749,
+      "learning_rate": 1.9692205784743938e-05,
+      "loss": 0.8716,
+      "step": 4534
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.62458439283311,
+      "learning_rate": 1.9692017900421594e-05,
+      "loss": 0.983,
+      "step": 4535
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4695051593609025,
+      "learning_rate": 1.969182995966925e-05,
+      "loss": 0.8876,
+      "step": 4536
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7432112220754734,
+      "learning_rate": 1.969164196248801e-05,
+      "loss": 0.8806,
+      "step": 4537
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.786782936751102,
+      "learning_rate": 1.9691453908878957e-05,
+      "loss": 0.7085,
+      "step": 4538
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7583199840865997,
+      "learning_rate": 1.969126579884319e-05,
+      "loss": 0.8487,
+      "step": 4539
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9201709875780886,
+      "learning_rate": 1.969107763238181e-05,
+      "loss": 0.9758,
+      "step": 4540
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9950434311800125,
+      "learning_rate": 1.9690889409495905e-05,
+      "loss": 0.937,
+      "step": 4541
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7555286936582715,
+      "learning_rate": 1.9690701130186576e-05,
+      "loss": 0.982,
+      "step": 4542
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6836632099670303,
+      "learning_rate": 1.9690512794454915e-05,
+      "loss": 0.9374,
+      "step": 4543
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6613296026780904,
+      "learning_rate": 1.969032440230202e-05,
+      "loss": 1.0113,
+      "step": 4544
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2084489437851036,
+      "learning_rate": 1.969013595372899e-05,
+      "loss": 0.8521,
+      "step": 4545
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7430976505077065,
+      "learning_rate": 1.9689947448736918e-05,
+      "loss": 0.9932,
+      "step": 4546
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6576833311331574,
+      "learning_rate": 1.9689758887326905e-05,
+      "loss": 0.9221,
+      "step": 4547
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8864809420821094,
+      "learning_rate": 1.9689570269500053e-05,
+      "loss": 0.9224,
+      "step": 4548
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5413253295852145,
+      "learning_rate": 1.968938159525745e-05,
+      "loss": 0.8993,
+      "step": 4549
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7135295924468443,
+      "learning_rate": 1.9689192864600203e-05,
+      "loss": 1.0195,
+      "step": 4550
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6537688211530275,
+      "learning_rate": 1.9689004077529405e-05,
+      "loss": 0.7888,
+      "step": 4551
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.142300480365961,
+      "learning_rate": 1.968881523404616e-05,
+      "loss": 0.8802,
+      "step": 4552
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.304535136386689,
+      "learning_rate": 1.968862633415156e-05,
+      "loss": 1.0037,
+      "step": 4553
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6935470978794263,
+      "learning_rate": 1.9688437377846714e-05,
+      "loss": 0.8334,
+      "step": 4554
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5090739442831977,
+      "learning_rate": 1.968824836513272e-05,
+      "loss": 0.846,
+      "step": 4555
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6496532366577514,
+      "learning_rate": 1.9688059296010676e-05,
+      "loss": 0.8506,
+      "step": 4556
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6231303718048962,
+      "learning_rate": 1.968787017048168e-05,
+      "loss": 0.827,
+      "step": 4557
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8019129767388589,
+      "learning_rate": 1.968768098854684e-05,
+      "loss": 0.8441,
+      "step": 4558
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6195121700929065,
+      "learning_rate": 1.9687491750207255e-05,
+      "loss": 0.8625,
+      "step": 4559
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5092451177141344,
+      "learning_rate": 1.9687302455464025e-05,
+      "loss": 1.1036,
+      "step": 4560
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7808103072129279,
+      "learning_rate": 1.9687113104318252e-05,
+      "loss": 0.6962,
+      "step": 4561
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6835990969153685,
+      "learning_rate": 1.968692369677104e-05,
+      "loss": 0.8882,
+      "step": 4562
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.192368329955546,
+      "learning_rate": 1.968673423282349e-05,
+      "loss": 0.8951,
+      "step": 4563
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5113188261145414,
+      "learning_rate": 1.968654471247671e-05,
+      "loss": 0.9052,
+      "step": 4564
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.530391320295916,
+      "learning_rate": 1.9686355135731797e-05,
+      "loss": 1.0183,
+      "step": 4565
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.657237527540704,
+      "learning_rate": 1.968616550258986e-05,
+      "loss": 0.914,
+      "step": 4566
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8112596490561494,
+      "learning_rate": 1.9685975813052e-05,
+      "loss": 0.9232,
+      "step": 4567
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4587166024233122,
+      "learning_rate": 1.968578606711932e-05,
+      "loss": 0.9548,
+      "step": 4568
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.082119999172934,
+      "learning_rate": 1.968559626479293e-05,
+      "loss": 0.9467,
+      "step": 4569
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8516013759669399,
+      "learning_rate": 1.9685406406073933e-05,
+      "loss": 0.8623,
+      "step": 4570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6490712633454059,
+      "learning_rate": 1.968521649096343e-05,
+      "loss": 0.9681,
+      "step": 4571
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4944776048873047,
+      "learning_rate": 1.9685026519462535e-05,
+      "loss": 0.8316,
+      "step": 4572
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7979885086327665,
+      "learning_rate": 1.9684836491572342e-05,
+      "loss": 0.7912,
+      "step": 4573
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6995418926775048,
+      "learning_rate": 1.9684646407293974e-05,
+      "loss": 0.8949,
+      "step": 4574
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7840287081698343,
+      "learning_rate": 1.9684456266628524e-05,
+      "loss": 0.7454,
+      "step": 4575
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9388649117888836,
+      "learning_rate": 1.96842660695771e-05,
+      "loss": 0.9397,
+      "step": 4576
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1248343973537684,
+      "learning_rate": 1.9684075816140817e-05,
+      "loss": 0.9176,
+      "step": 4577
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6377255167774671,
+      "learning_rate": 1.968388550632078e-05,
+      "loss": 0.8709,
+      "step": 4578
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7950959815492742,
+      "learning_rate": 1.9683695140118093e-05,
+      "loss": 0.8012,
+      "step": 4579
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5419064422679254,
+      "learning_rate": 1.9683504717533868e-05,
+      "loss": 0.8806,
+      "step": 4580
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4936154227854794,
+      "learning_rate": 1.9683314238569215e-05,
+      "loss": 0.8323,
+      "step": 4581
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6680598635929338,
+      "learning_rate": 1.9683123703225236e-05,
+      "loss": 0.918,
+      "step": 4582
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7573024489916387,
+      "learning_rate": 1.968293311150305e-05,
+      "loss": 1.0154,
+      "step": 4583
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0075309410213067,
+      "learning_rate": 1.9682742463403755e-05,
+      "loss": 0.8262,
+      "step": 4584
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8051440678179618,
+      "learning_rate": 1.9682551758928473e-05,
+      "loss": 0.9383,
+      "step": 4585
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3530999871964204,
+      "learning_rate": 1.968236099807831e-05,
+      "loss": 0.8906,
+      "step": 4586
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6287616904759656,
+      "learning_rate": 1.9682170180854373e-05,
+      "loss": 0.9313,
+      "step": 4587
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7666386989008482,
+      "learning_rate": 1.9681979307257778e-05,
+      "loss": 0.7541,
+      "step": 4588
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5146797311524258,
+      "learning_rate": 1.9681788377289633e-05,
+      "loss": 1.0285,
+      "step": 4589
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9153260093729088,
+      "learning_rate": 1.9681597390951052e-05,
+      "loss": 0.965,
+      "step": 4590
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0376922759481646,
+      "learning_rate": 1.9681406348243147e-05,
+      "loss": 0.8084,
+      "step": 4591
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7817448798590858,
+      "learning_rate": 1.968121524916703e-05,
+      "loss": 0.9022,
+      "step": 4592
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6319364872216218,
+      "learning_rate": 1.9681024093723805e-05,
+      "loss": 0.9015,
+      "step": 4593
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.82114101302273,
+      "learning_rate": 1.9680832881914598e-05,
+      "loss": 0.9005,
+      "step": 4594
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8380622872280292,
+      "learning_rate": 1.9680641613740515e-05,
+      "loss": 0.833,
+      "step": 4595
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.881638055682309,
+      "learning_rate": 1.968045028920268e-05,
+      "loss": 0.9591,
+      "step": 4596
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1726802222767114,
+      "learning_rate": 1.968025890830219e-05,
+      "loss": 0.9491,
+      "step": 4597
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8644801555547952,
+      "learning_rate": 1.968006747104017e-05,
+      "loss": 0.9193,
+      "step": 4598
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6497054533269773,
+      "learning_rate": 1.9679875977417734e-05,
+      "loss": 0.934,
+      "step": 4599
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0636268220593514,
+      "learning_rate": 1.9679684427435993e-05,
+      "loss": 0.8342,
+      "step": 4600
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4680208755350006,
+      "learning_rate": 1.9679492821096065e-05,
+      "loss": 0.9746,
+      "step": 4601
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.137636263916981,
+      "learning_rate": 1.9679301158399067e-05,
+      "loss": 1.021,
+      "step": 4602
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7468036854594207,
+      "learning_rate": 1.9679109439346114e-05,
+      "loss": 0.9808,
+      "step": 4603
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0145725040276123,
+      "learning_rate": 1.967891766393832e-05,
+      "loss": 0.9112,
+      "step": 4604
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3103969895230416,
+      "learning_rate": 1.9678725832176803e-05,
+      "loss": 0.9471,
+      "step": 4605
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.572807184479499,
+      "learning_rate": 1.967853394406268e-05,
+      "loss": 0.8978,
+      "step": 4606
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.815499753059093,
+      "learning_rate": 1.9678341999597066e-05,
+      "loss": 0.6975,
+      "step": 4607
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7567048209771629,
+      "learning_rate": 1.9678149998781083e-05,
+      "loss": 0.7682,
+      "step": 4608
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9971488435876033,
+      "learning_rate": 1.9677957941615847e-05,
+      "loss": 0.9132,
+      "step": 4609
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8442746146652627,
+      "learning_rate": 1.9677765828102477e-05,
+      "loss": 0.9773,
+      "step": 4610
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8808522682777373,
+      "learning_rate": 1.967757365824209e-05,
+      "loss": 0.8778,
+      "step": 4611
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7367773839393075,
+      "learning_rate": 1.9677381432035803e-05,
+      "loss": 0.979,
+      "step": 4612
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7509308124286889,
+      "learning_rate": 1.967718914948474e-05,
+      "loss": 0.9947,
+      "step": 4613
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.896332109558975,
+      "learning_rate": 1.9676996810590016e-05,
+      "loss": 0.8723,
+      "step": 4614
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6465666118452718,
+      "learning_rate": 1.9676804415352752e-05,
+      "loss": 0.9025,
+      "step": 4615
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5387800874403679,
+      "learning_rate": 1.967661196377407e-05,
+      "loss": 0.937,
+      "step": 4616
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0657857613049107,
+      "learning_rate": 1.9676419455855093e-05,
+      "loss": 0.8999,
+      "step": 4617
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5569770706214718,
+      "learning_rate": 1.9676226891596934e-05,
+      "loss": 0.7729,
+      "step": 4618
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.147949476571576,
+      "learning_rate": 1.967603427100072e-05,
+      "loss": 1.0316,
+      "step": 4619
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7801497985092907,
+      "learning_rate": 1.967584159406757e-05,
+      "loss": 0.874,
+      "step": 4620
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6565049189369414,
+      "learning_rate": 1.967564886079861e-05,
+      "loss": 1.006,
+      "step": 4621
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6796652780285461,
+      "learning_rate": 1.9675456071194957e-05,
+      "loss": 0.9736,
+      "step": 4622
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.746160573962089,
+      "learning_rate": 1.9675263225257736e-05,
+      "loss": 0.8872,
+      "step": 4623
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1665028071169656,
+      "learning_rate": 1.967507032298807e-05,
+      "loss": 1.0179,
+      "step": 4624
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7108084877545622,
+      "learning_rate": 1.9674877364387083e-05,
+      "loss": 1.0003,
+      "step": 4625
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6472957356683642,
+      "learning_rate": 1.9674684349455893e-05,
+      "loss": 0.873,
+      "step": 4626
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.499667316433825,
+      "learning_rate": 1.967449127819563e-05,
+      "loss": 0.9687,
+      "step": 4627
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4823253531886427,
+      "learning_rate": 1.9674298150607415e-05,
+      "loss": 0.9024,
+      "step": 4628
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7553464097925675,
+      "learning_rate": 1.9674104966692376e-05,
+      "loss": 0.9199,
+      "step": 4629
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6464394382824632,
+      "learning_rate": 1.9673911726451632e-05,
+      "loss": 0.847,
+      "step": 4630
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.290105384345352,
+      "learning_rate": 1.967371842988631e-05,
+      "loss": 0.9142,
+      "step": 4631
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7218604817327094,
+      "learning_rate": 1.967352507699754e-05,
+      "loss": 0.9641,
+      "step": 4632
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7831937623620364,
+      "learning_rate": 1.9673331667786445e-05,
+      "loss": 1.0137,
+      "step": 4633
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7586389192902172,
+      "learning_rate": 1.9673138202254148e-05,
+      "loss": 0.7297,
+      "step": 4634
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6188413289412122,
+      "learning_rate": 1.9672944680401776e-05,
+      "loss": 0.8799,
+      "step": 4635
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8300836607173505,
+      "learning_rate": 1.967275110223046e-05,
+      "loss": 0.9876,
+      "step": 4636
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8747439358466573,
+      "learning_rate": 1.9672557467741325e-05,
+      "loss": 0.8625,
+      "step": 4637
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7422778139646022,
+      "learning_rate": 1.9672363776935503e-05,
+      "loss": 0.9247,
+      "step": 4638
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.13070715362344,
+      "learning_rate": 1.9672170029814112e-05,
+      "loss": 1.136,
+      "step": 4639
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7664363970299772,
+      "learning_rate": 1.967197622637828e-05,
+      "loss": 0.7689,
+      "step": 4640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.538865236267219,
+      "learning_rate": 1.967178236662915e-05,
+      "loss": 0.9736,
+      "step": 4641
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7513923269421607,
+      "learning_rate": 1.9671588450567833e-05,
+      "loss": 0.8146,
+      "step": 4642
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6235883640590927,
+      "learning_rate": 1.9671394478195472e-05,
+      "loss": 0.8522,
+      "step": 4643
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5559427657284766,
+      "learning_rate": 1.9671200449513188e-05,
+      "loss": 0.8717,
+      "step": 4644
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.033766187724597,
+      "learning_rate": 1.9671006364522115e-05,
+      "loss": 0.9285,
+      "step": 4645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7251020802822552,
+      "learning_rate": 1.9670812223223378e-05,
+      "loss": 1.01,
+      "step": 4646
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1145753782209122,
+      "learning_rate": 1.9670618025618112e-05,
+      "loss": 0.9141,
+      "step": 4647
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7454091979439132,
+      "learning_rate": 1.9670423771707442e-05,
+      "loss": 0.8376,
+      "step": 4648
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7679897732628878,
+      "learning_rate": 1.967022946149251e-05,
+      "loss": 0.8295,
+      "step": 4649
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7874812957464326,
+      "learning_rate": 1.9670035094974435e-05,
+      "loss": 0.9785,
+      "step": 4650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2586062370643396,
+      "learning_rate": 1.9669840672154358e-05,
+      "loss": 0.9536,
+      "step": 4651
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.828093507031996,
+      "learning_rate": 1.9669646193033406e-05,
+      "loss": 0.9365,
+      "step": 4652
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5133720139004188,
+      "learning_rate": 1.9669451657612712e-05,
+      "loss": 0.9339,
+      "step": 4653
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.594675481030571,
+      "learning_rate": 1.966925706589341e-05,
+      "loss": 0.76,
+      "step": 4654
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6059363145788148,
+      "learning_rate": 1.966906241787663e-05,
+      "loss": 0.7944,
+      "step": 4655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.792246600335743,
+      "learning_rate": 1.966886771356351e-05,
+      "loss": 0.7463,
+      "step": 4656
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9767232290072059,
+      "learning_rate": 1.966867295295518e-05,
+      "loss": 0.9103,
+      "step": 4657
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1277922714203723,
+      "learning_rate": 1.9668478136052776e-05,
+      "loss": 1.0294,
+      "step": 4658
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.01124972213023,
+      "learning_rate": 1.966828326285743e-05,
+      "loss": 0.8945,
+      "step": 4659
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.201492311315359,
+      "learning_rate": 1.966808833337028e-05,
+      "loss": 0.9411,
+      "step": 4660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1407492939925037,
+      "learning_rate": 1.9667893347592458e-05,
+      "loss": 0.9454,
+      "step": 4661
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9988752084141526,
+      "learning_rate": 1.96676983055251e-05,
+      "loss": 0.9151,
+      "step": 4662
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8312533553271193,
+      "learning_rate": 1.9667503207169337e-05,
+      "loss": 0.8952,
+      "step": 4663
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7004535899872,
+      "learning_rate": 1.9667308052526318e-05,
+      "loss": 1.1224,
+      "step": 4664
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2029030631063655,
+      "learning_rate": 1.9667112841597164e-05,
+      "loss": 0.84,
+      "step": 4665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6443442793510046,
+      "learning_rate": 1.9666917574383026e-05,
+      "loss": 0.976,
+      "step": 4666
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.310980332071054,
+      "learning_rate": 1.9666722250885026e-05,
+      "loss": 0.9221,
+      "step": 4667
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8272056518887245,
+      "learning_rate": 1.9666526871104312e-05,
+      "loss": 0.845,
+      "step": 4668
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6998784166644063,
+      "learning_rate": 1.9666331435042017e-05,
+      "loss": 0.8625,
+      "step": 4669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0342065995201573,
+      "learning_rate": 1.9666135942699278e-05,
+      "loss": 0.8755,
+      "step": 4670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6288262843038748,
+      "learning_rate": 1.966594039407724e-05,
+      "loss": 1.0477,
+      "step": 4671
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1331652498298035,
+      "learning_rate": 1.9665744789177033e-05,
+      "loss": 0.849,
+      "step": 4672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9653061405395698,
+      "learning_rate": 1.96655491279998e-05,
+      "loss": 0.9313,
+      "step": 4673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6565240712508358,
+      "learning_rate": 1.9665353410546683e-05,
+      "loss": 0.9232,
+      "step": 4674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.536002677724345,
+      "learning_rate": 1.9665157636818816e-05,
+      "loss": 0.8009,
+      "step": 4675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4291562823156196,
+      "learning_rate": 1.9664961806817343e-05,
+      "loss": 0.9182,
+      "step": 4676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.388149611560935,
+      "learning_rate": 1.96647659205434e-05,
+      "loss": 0.8667,
+      "step": 4677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.227253236676382,
+      "learning_rate": 1.9664569977998132e-05,
+      "loss": 1.0166,
+      "step": 4678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6287736891290254,
+      "learning_rate": 1.9664373979182678e-05,
+      "loss": 1.067,
+      "step": 4679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4971696938421788,
+      "learning_rate": 1.9664177924098178e-05,
+      "loss": 0.9395,
+      "step": 4680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5858861870213137,
+      "learning_rate": 1.9663981812745777e-05,
+      "loss": 0.9422,
+      "step": 4681
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2422911092480122,
+      "learning_rate": 1.966378564512661e-05,
+      "loss": 0.9761,
+      "step": 4682
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0211675040678365,
+      "learning_rate": 1.9663589421241827e-05,
+      "loss": 0.9972,
+      "step": 4683
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.426899902023638,
+      "learning_rate": 1.9663393141092565e-05,
+      "loss": 0.8633,
+      "step": 4684
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8260705464919984,
+      "learning_rate": 1.966319680467997e-05,
+      "loss": 0.8779,
+      "step": 4685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.762105704255822,
+      "learning_rate": 1.9663000412005182e-05,
+      "loss": 0.9948,
+      "step": 4686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7686520424909664,
+      "learning_rate": 1.966280396306935e-05,
+      "loss": 1.0064,
+      "step": 4687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0790876410592944,
+      "learning_rate": 1.966260745787361e-05,
+      "loss": 0.8331,
+      "step": 4688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.657699961509795,
+      "learning_rate": 1.966241089641911e-05,
+      "loss": 0.9357,
+      "step": 4689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7709362135762144,
+      "learning_rate": 1.9662214278707e-05,
+      "loss": 0.7302,
+      "step": 4690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6942217203636176,
+      "learning_rate": 1.9662017604738416e-05,
+      "loss": 0.9224,
+      "step": 4691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6078638118938275,
+      "learning_rate": 1.9661820874514504e-05,
+      "loss": 0.893,
+      "step": 4692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6202725168898697,
+      "learning_rate": 1.9661624088036414e-05,
+      "loss": 0.8448,
+      "step": 4693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6393269812147153,
+      "learning_rate": 1.966142724530529e-05,
+      "loss": 0.9289,
+      "step": 4694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7164191845790167,
+      "learning_rate": 1.966123034632228e-05,
+      "loss": 0.9038,
+      "step": 4695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7648733706913361,
+      "learning_rate": 1.9661033391088526e-05,
+      "loss": 0.7329,
+      "step": 4696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8243910102204306,
+      "learning_rate": 1.9660836379605176e-05,
+      "loss": 0.7241,
+      "step": 4697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8058382774730102,
+      "learning_rate": 1.966063931187338e-05,
+      "loss": 0.8074,
+      "step": 4698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7679217574402961,
+      "learning_rate": 1.9660442187894283e-05,
+      "loss": 0.9772,
+      "step": 4699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8819375959352025,
+      "learning_rate": 1.9660245007669035e-05,
+      "loss": 0.8533,
+      "step": 4700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.95751364848122,
+      "learning_rate": 1.966004777119878e-05,
+      "loss": 0.8936,
+      "step": 4701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8676404303466781,
+      "learning_rate": 1.9659850478484668e-05,
+      "loss": 0.808,
+      "step": 4702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.089905053828856,
+      "learning_rate": 1.965965312952785e-05,
+      "loss": 0.8409,
+      "step": 4703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8926773881814378,
+      "learning_rate": 1.965945572432947e-05,
+      "loss": 0.8473,
+      "step": 4704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9595247479971223,
+      "learning_rate": 1.9659258262890683e-05,
+      "loss": 1.0071,
+      "step": 4705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8009499364481185,
+      "learning_rate": 1.9659060745212638e-05,
+      "loss": 0.8838,
+      "step": 4706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9226504239011406,
+      "learning_rate": 1.9658863171296484e-05,
+      "loss": 0.8567,
+      "step": 4707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7710814029816613,
+      "learning_rate": 1.9658665541143366e-05,
+      "loss": 0.716,
+      "step": 4708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6732315358398637,
+      "learning_rate": 1.965846785475444e-05,
+      "loss": 0.9189,
+      "step": 4709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7901637406794549,
+      "learning_rate": 1.965827011213086e-05,
+      "loss": 0.7691,
+      "step": 4710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9144910843749918,
+      "learning_rate": 1.965807231327377e-05,
+      "loss": 0.9747,
+      "step": 4711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3984843077309104,
+      "learning_rate": 1.965787445818433e-05,
+      "loss": 0.8372,
+      "step": 4712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8621138063606224,
+      "learning_rate": 1.9657676546863682e-05,
+      "loss": 0.9587,
+      "step": 4713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9617715866346666,
+      "learning_rate": 1.9657478579312984e-05,
+      "loss": 0.9666,
+      "step": 4714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7766011942416786,
+      "learning_rate": 1.965728055553339e-05,
+      "loss": 0.8701,
+      "step": 4715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8219889667201556,
+      "learning_rate": 1.965708247552605e-05,
+      "loss": 0.9628,
+      "step": 4716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.02198802178253,
+      "learning_rate": 1.9656884339292123e-05,
+      "loss": 0.9104,
+      "step": 4717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6776273084058537,
+      "learning_rate": 1.9656686146832754e-05,
+      "loss": 0.801,
+      "step": 4718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3966553752565387,
+      "learning_rate": 1.96564878981491e-05,
+      "loss": 0.9791,
+      "step": 4719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8876611436045834,
+      "learning_rate": 1.965628959324232e-05,
+      "loss": 1.0203,
+      "step": 4720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7703110029363869,
+      "learning_rate": 1.9656091232113562e-05,
+      "loss": 0.9597,
+      "step": 4721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7841038848490511,
+      "learning_rate": 1.9655892814763984e-05,
+      "loss": 0.9165,
+      "step": 4722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6184036364802397,
+      "learning_rate": 1.9655694341194743e-05,
+      "loss": 0.9784,
+      "step": 4723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.161781606572563,
+      "learning_rate": 1.965549581140699e-05,
+      "loss": 1.0807,
+      "step": 4724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0076314889141176,
+      "learning_rate": 1.9655297225401884e-05,
+      "loss": 0.9226,
+      "step": 4725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.823516323195565,
+      "learning_rate": 1.965509858318058e-05,
+      "loss": 0.9914,
+      "step": 4726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.590853350180954,
+      "learning_rate": 1.9654899884744238e-05,
+      "loss": 0.8398,
+      "step": 4727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7886950214718371,
+      "learning_rate": 1.965470113009401e-05,
+      "loss": 0.825,
+      "step": 4728
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.776417010140368,
+      "learning_rate": 1.9654502319231056e-05,
+      "loss": 0.9181,
+      "step": 4729
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6160653483805014,
+      "learning_rate": 1.9654303452156535e-05,
+      "loss": 0.9322,
+      "step": 4730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7374946981406678,
+      "learning_rate": 1.96541045288716e-05,
+      "loss": 0.9188,
+      "step": 4731
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5804200726141397,
+      "learning_rate": 1.9653905549377413e-05,
+      "loss": 1.0638,
+      "step": 4732
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9482663628146657,
+      "learning_rate": 1.965370651367513e-05,
+      "loss": 0.9613,
+      "step": 4733
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9853733591070268,
+      "learning_rate": 1.9653507421765914e-05,
+      "loss": 0.8456,
+      "step": 4734
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6792810836733831,
+      "learning_rate": 1.965330827365092e-05,
+      "loss": 1.0048,
+      "step": 4735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6279332844414702,
+      "learning_rate": 1.965310906933131e-05,
+      "loss": 0.9092,
+      "step": 4736
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.006812830923065,
+      "learning_rate": 1.9652909808808242e-05,
+      "loss": 0.8128,
+      "step": 4737
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6881472234749282,
+      "learning_rate": 1.965271049208288e-05,
+      "loss": 0.894,
+      "step": 4738
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7755694765634855,
+      "learning_rate": 1.9652511119156375e-05,
+      "loss": 0.819,
+      "step": 4739
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.465876204499296,
+      "learning_rate": 1.96523116900299e-05,
+      "loss": 0.9335,
+      "step": 4740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4077346713087717,
+      "learning_rate": 1.9652112204704608e-05,
+      "loss": 0.9807,
+      "step": 4741
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6973292378081575,
+      "learning_rate": 1.9651912663181665e-05,
+      "loss": 1.0676,
+      "step": 4742
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.720457455612874,
+      "learning_rate": 1.965171306546223e-05,
+      "loss": 0.8669,
+      "step": 4743
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6633000832715332,
+      "learning_rate": 1.9651513411547466e-05,
+      "loss": 1.0105,
+      "step": 4744
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7417666277755612,
+      "learning_rate": 1.9651313701438533e-05,
+      "loss": 0.8745,
+      "step": 4745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8028596033646398,
+      "learning_rate": 1.9651113935136596e-05,
+      "loss": 0.7958,
+      "step": 4746
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8019688888669627,
+      "learning_rate": 1.965091411264282e-05,
+      "loss": 0.764,
+      "step": 4747
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6813420952219575,
+      "learning_rate": 1.9650714233958367e-05,
+      "loss": 0.964,
+      "step": 4748
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.964205338247612,
+      "learning_rate": 1.96505142990844e-05,
+      "loss": 0.8104,
+      "step": 4749
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8751069666574125,
+      "learning_rate": 1.9650314308022083e-05,
+      "loss": 0.969,
+      "step": 4750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6916292523753438,
+      "learning_rate": 1.9650114260772583e-05,
+      "loss": 0.9216,
+      "step": 4751
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7940256636994352,
+      "learning_rate": 1.964991415733706e-05,
+      "loss": 0.996,
+      "step": 4752
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.565478363249466,
+      "learning_rate": 1.964971399771668e-05,
+      "loss": 0.9241,
+      "step": 4753
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8053854573424832,
+      "learning_rate": 1.964951378191261e-05,
+      "loss": 1.0073,
+      "step": 4754
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.865346800919814,
+      "learning_rate": 1.9649313509926022e-05,
+      "loss": 0.9199,
+      "step": 4755
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8149136953468006,
+      "learning_rate": 1.964911318175807e-05,
+      "loss": 0.9231,
+      "step": 4756
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7281433949975682,
+      "learning_rate": 1.964891279740993e-05,
+      "loss": 0.9828,
+      "step": 4757
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.552810933198308,
+      "learning_rate": 1.9648712356882763e-05,
+      "loss": 0.8794,
+      "step": 4758
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5665676182166846,
+      "learning_rate": 1.9648511860177737e-05,
+      "loss": 1.0128,
+      "step": 4759
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9293406071309716,
+      "learning_rate": 1.964831130729602e-05,
+      "loss": 0.846,
+      "step": 4760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5365520681838292,
+      "learning_rate": 1.9648110698238782e-05,
+      "loss": 0.9518,
+      "step": 4761
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.524721967289196,
+      "learning_rate": 1.964791003300719e-05,
+      "loss": 0.8045,
+      "step": 4762
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5492449062270348,
+      "learning_rate": 1.9647709311602408e-05,
+      "loss": 0.8862,
+      "step": 4763
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4890783785770718,
+      "learning_rate": 1.964750853402561e-05,
+      "loss": 0.7901,
+      "step": 4764
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6253112449369371,
+      "learning_rate": 1.9647307700277963e-05,
+      "loss": 0.9908,
+      "step": 4765
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7633040305024763,
+      "learning_rate": 1.9647106810360636e-05,
+      "loss": 0.9066,
+      "step": 4766
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7788103957956356,
+      "learning_rate": 1.9646905864274798e-05,
+      "loss": 0.9791,
+      "step": 4767
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5866646519131886,
+      "learning_rate": 1.964670486202162e-05,
+      "loss": 0.9572,
+      "step": 4768
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6997208658752314,
+      "learning_rate": 1.9646503803602272e-05,
+      "loss": 1.1081,
+      "step": 4769
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8604274100388567,
+      "learning_rate": 1.9646302689017925e-05,
+      "loss": 0.922,
+      "step": 4770
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6692172059048922,
+      "learning_rate": 1.964610151826975e-05,
+      "loss": 0.965,
+      "step": 4771
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2904050673451315,
+      "learning_rate": 1.9645900291358918e-05,
+      "loss": 1.0152,
+      "step": 4772
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9943116355505888,
+      "learning_rate": 1.9645699008286602e-05,
+      "loss": 0.9537,
+      "step": 4773
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6741175200028608,
+      "learning_rate": 1.9645497669053973e-05,
+      "loss": 0.8381,
+      "step": 4774
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4570385213490238,
+      "learning_rate": 1.9645296273662202e-05,
+      "loss": 0.9764,
+      "step": 4775
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4979033458124844,
+      "learning_rate": 1.964509482211246e-05,
+      "loss": 0.7793,
+      "step": 4776
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8432118370083508,
+      "learning_rate": 1.9644893314405924e-05,
+      "loss": 0.9471,
+      "step": 4777
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1268206801903413,
+      "learning_rate": 1.964469175054377e-05,
+      "loss": 0.8928,
+      "step": 4778
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8883816904392552,
+      "learning_rate": 1.964449013052716e-05,
+      "loss": 0.7846,
+      "step": 4779
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5742330096071675,
+      "learning_rate": 1.964428845435728e-05,
+      "loss": 1.0763,
+      "step": 4780
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.703179469332593,
+      "learning_rate": 1.9644086722035295e-05,
+      "loss": 0.8714,
+      "step": 4781
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.829324912289471,
+      "learning_rate": 1.9643884933562387e-05,
+      "loss": 0.8655,
+      "step": 4782
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.78214307091216,
+      "learning_rate": 1.9643683088939726e-05,
+      "loss": 0.7292,
+      "step": 4783
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8075614299565095,
+      "learning_rate": 1.9643481188168492e-05,
+      "loss": 0.8173,
+      "step": 4784
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.819344164178443,
+      "learning_rate": 1.9643279231249855e-05,
+      "loss": 0.7488,
+      "step": 4785
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6885165428504978,
+      "learning_rate": 1.9643077218184993e-05,
+      "loss": 0.9092,
+      "step": 4786
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7463953389748992,
+      "learning_rate": 1.9642875148975084e-05,
+      "loss": 0.893,
+      "step": 4787
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.644578251395674,
+      "learning_rate": 1.9642673023621303e-05,
+      "loss": 1.0691,
+      "step": 4788
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6548727999474007,
+      "learning_rate": 1.9642470842124823e-05,
+      "loss": 0.9652,
+      "step": 4789
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8607560134576544,
+      "learning_rate": 1.964226860448683e-05,
+      "loss": 0.82,
+      "step": 4790
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8858085488659238,
+      "learning_rate": 1.9642066310708493e-05,
+      "loss": 0.9533,
+      "step": 4791
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2017694166123523,
+      "learning_rate": 1.9641863960790998e-05,
+      "loss": 0.9,
+      "step": 4792
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8561617587142478,
+      "learning_rate": 1.9641661554735513e-05,
+      "loss": 0.9158,
+      "step": 4793
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.236516687658466,
+      "learning_rate": 1.9641459092543224e-05,
+      "loss": 1.0422,
+      "step": 4794
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1543549701290083,
+      "learning_rate": 1.9641256574215308e-05,
+      "loss": 0.9138,
+      "step": 4795
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7877972415426155,
+      "learning_rate": 1.9641053999752942e-05,
+      "loss": 0.8493,
+      "step": 4796
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4518372900261847,
+      "learning_rate": 1.964085136915731e-05,
+      "loss": 0.9797,
+      "step": 4797
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7221894297697027,
+      "learning_rate": 1.964064868242959e-05,
+      "loss": 0.778,
+      "step": 4798
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.527282490478921,
+      "learning_rate": 1.964044593957096e-05,
+      "loss": 0.8477,
+      "step": 4799
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.845013973649404,
+      "learning_rate": 1.96402431405826e-05,
+      "loss": 0.9323,
+      "step": 4800
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7408019061110345,
+      "learning_rate": 1.9640040285465693e-05,
+      "loss": 0.9548,
+      "step": 4801
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7413234484296838,
+      "learning_rate": 1.963983737422142e-05,
+      "loss": 0.8028,
+      "step": 4802
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9448438576106115,
+      "learning_rate": 1.9639634406850962e-05,
+      "loss": 0.9069,
+      "step": 4803
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5059299114250544,
+      "learning_rate": 1.96394313833555e-05,
+      "loss": 0.8808,
+      "step": 4804
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8293924891421214,
+      "learning_rate": 1.9639228303736216e-05,
+      "loss": 0.9122,
+      "step": 4805
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7709950279653859,
+      "learning_rate": 1.9639025167994295e-05,
+      "loss": 0.8006,
+      "step": 4806
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6382759788062966,
+      "learning_rate": 1.9638821976130916e-05,
+      "loss": 0.9303,
+      "step": 4807
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6518604574909028,
+      "learning_rate": 1.9638618728147266e-05,
+      "loss": 0.7407,
+      "step": 4808
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8192152753011934,
+      "learning_rate": 1.9638415424044527e-05,
+      "loss": 0.8639,
+      "step": 4809
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.753692069784189,
+      "learning_rate": 1.963821206382388e-05,
+      "loss": 0.9584,
+      "step": 4810
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.83621098278794,
+      "learning_rate": 1.963800864748651e-05,
+      "loss": 0.7905,
+      "step": 4811
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.765007708974011,
+      "learning_rate": 1.9637805175033603e-05,
+      "loss": 0.8896,
+      "step": 4812
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6994653596649132,
+      "learning_rate": 1.9637601646466343e-05,
+      "loss": 0.8003,
+      "step": 4813
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9466340406125464,
+      "learning_rate": 1.9637398061785914e-05,
+      "loss": 1.1192,
+      "step": 4814
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7355245839109903,
+      "learning_rate": 1.9637194420993504e-05,
+      "loss": 0.9974,
+      "step": 4815
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1342629017179817,
+      "learning_rate": 1.9636990724090297e-05,
+      "loss": 0.8775,
+      "step": 4816
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5709117649205009,
+      "learning_rate": 1.963678697107748e-05,
+      "loss": 1.0225,
+      "step": 4817
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8106285890644853,
+      "learning_rate": 1.9636583161956235e-05,
+      "loss": 0.7266,
+      "step": 4818
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4092760878911204,
+      "learning_rate": 1.9636379296727754e-05,
+      "loss": 0.8199,
+      "step": 4819
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6627993553234388,
+      "learning_rate": 1.9636175375393223e-05,
+      "loss": 0.9585,
+      "step": 4820
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4789278763772917,
+      "learning_rate": 1.9635971397953826e-05,
+      "loss": 0.8439,
+      "step": 4821
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9076581198039912,
+      "learning_rate": 1.963576736441075e-05,
+      "loss": 0.8617,
+      "step": 4822
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7818370569758283,
+      "learning_rate": 1.9635563274765194e-05,
+      "loss": 0.6731,
+      "step": 4823
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9429571880180199,
+      "learning_rate": 1.9635359129018334e-05,
+      "loss": 0.9211,
+      "step": 4824
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7362390833340628,
+      "learning_rate": 1.963515492717136e-05,
+      "loss": 0.8197,
+      "step": 4825
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1581835449741655,
+      "learning_rate": 1.9634950669225465e-05,
+      "loss": 0.9432,
+      "step": 4826
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.191095092898845,
+      "learning_rate": 1.9634746355181836e-05,
+      "loss": 0.8918,
+      "step": 4827
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8892914143822186,
+      "learning_rate": 1.9634541985041667e-05,
+      "loss": 0.9161,
+      "step": 4828
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7182787128495924,
+      "learning_rate": 1.9634337558806142e-05,
+      "loss": 0.7708,
+      "step": 4829
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7948168288900224,
+      "learning_rate": 1.9634133076476453e-05,
+      "loss": 1.0412,
+      "step": 4830
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6347534471809408,
+      "learning_rate": 1.9633928538053792e-05,
+      "loss": 0.6887,
+      "step": 4831
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7611566326632628,
+      "learning_rate": 1.9633723943539352e-05,
+      "loss": 0.9322,
+      "step": 4832
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.066668699288202,
+      "learning_rate": 1.9633519292934317e-05,
+      "loss": 0.9258,
+      "step": 4833
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.623173442931025,
+      "learning_rate": 1.9633314586239884e-05,
+      "loss": 0.9271,
+      "step": 4834
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.754686817884286,
+      "learning_rate": 1.9633109823457243e-05,
+      "loss": 0.9329,
+      "step": 4835
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6600696925079996,
+      "learning_rate": 1.9632905004587588e-05,
+      "loss": 0.8099,
+      "step": 4836
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1081805754775895,
+      "learning_rate": 1.963270012963211e-05,
+      "loss": 0.979,
+      "step": 4837
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.75679177639708,
+      "learning_rate": 1.9632495198592003e-05,
+      "loss": 1.06,
+      "step": 4838
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5559453696235765,
+      "learning_rate": 1.9632290211468458e-05,
+      "loss": 1.02,
+      "step": 4839
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8940758835655906,
+      "learning_rate": 1.963208516826267e-05,
+      "loss": 0.951,
+      "step": 4840
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7673765056968462,
+      "learning_rate": 1.9631880068975834e-05,
+      "loss": 0.8838,
+      "step": 4841
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.804913426940321,
+      "learning_rate": 1.9631674913609137e-05,
+      "loss": 0.9075,
+      "step": 4842
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.18178018071669,
+      "learning_rate": 1.9631469702163786e-05,
+      "loss": 0.9142,
+      "step": 4843
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.891465667229099,
+      "learning_rate": 1.9631264434640965e-05,
+      "loss": 0.8563,
+      "step": 4844
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0357299805282234,
+      "learning_rate": 1.9631059111041876e-05,
+      "loss": 1.0487,
+      "step": 4845
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8374853453776315,
+      "learning_rate": 1.9630853731367715e-05,
+      "loss": 0.775,
+      "step": 4846
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.295355850438296,
+      "learning_rate": 1.9630648295619668e-05,
+      "loss": 0.7138,
+      "step": 4847
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7900288046673272,
+      "learning_rate": 1.963044280379894e-05,
+      "loss": 0.9813,
+      "step": 4848
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7753216747119256,
+      "learning_rate": 1.9630237255906724e-05,
+      "loss": 0.7881,
+      "step": 4849
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.794571442114409,
+      "learning_rate": 1.963003165194422e-05,
+      "loss": 0.9631,
+      "step": 4850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6162451280192824,
+      "learning_rate": 1.9629825991912622e-05,
+      "loss": 0.8851,
+      "step": 4851
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6817418183370156,
+      "learning_rate": 1.962962027581313e-05,
+      "loss": 0.9557,
+      "step": 4852
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.334951613604829,
+      "learning_rate": 1.9629414503646936e-05,
+      "loss": 1.0138,
+      "step": 4853
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7461582948702157,
+      "learning_rate": 1.9629208675415246e-05,
+      "loss": 1.0595,
+      "step": 4854
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7263267311246827,
+      "learning_rate": 1.9629002791119252e-05,
+      "loss": 0.7009,
+      "step": 4855
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7911252090444212,
+      "learning_rate": 1.9628796850760156e-05,
+      "loss": 0.9969,
+      "step": 4856
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4356828939175268,
+      "learning_rate": 1.962859085433916e-05,
+      "loss": 0.8144,
+      "step": 4857
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7649684663243836,
+      "learning_rate": 1.9628384801857454e-05,
+      "loss": 0.9316,
+      "step": 4858
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.72512594072494,
+      "learning_rate": 1.9628178693316243e-05,
+      "loss": 1.0281,
+      "step": 4859
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.294067309040135,
+      "learning_rate": 1.962797252871673e-05,
+      "loss": 0.9852,
+      "step": 4860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.672451683443658,
+      "learning_rate": 1.9627766308060115e-05,
+      "loss": 0.8768,
+      "step": 4861
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6586274650078014,
+      "learning_rate": 1.9627560031347594e-05,
+      "loss": 0.9936,
+      "step": 4862
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8442636062485598,
+      "learning_rate": 1.962735369858037e-05,
+      "loss": 0.9132,
+      "step": 4863
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.47260457214293,
+      "learning_rate": 1.9627147309759646e-05,
+      "loss": 0.9733,
+      "step": 4864
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7063881700015846,
+      "learning_rate": 1.9626940864886622e-05,
+      "loss": 0.8412,
+      "step": 4865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.582116679401087,
+      "learning_rate": 1.96267343639625e-05,
+      "loss": 0.8987,
+      "step": 4866
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7221878486178899,
+      "learning_rate": 1.9626527806988485e-05,
+      "loss": 1.0684,
+      "step": 4867
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6839184385899764,
+      "learning_rate": 1.9626321193965774e-05,
+      "loss": 0.9172,
+      "step": 4868
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0702863692150544,
+      "learning_rate": 1.9626114524895574e-05,
+      "loss": 0.8367,
+      "step": 4869
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.544500719615065,
+      "learning_rate": 1.962590779977909e-05,
+      "loss": 1.022,
+      "step": 4870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5613475736164761,
+      "learning_rate": 1.9625701018617524e-05,
+      "loss": 0.9297,
+      "step": 4871
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0601323427021554,
+      "learning_rate": 1.9625494181412076e-05,
+      "loss": 0.9931,
+      "step": 4872
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0762891527697316,
+      "learning_rate": 1.9625287288163954e-05,
+      "loss": 0.8938,
+      "step": 4873
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0841372251709447,
+      "learning_rate": 1.9625080338874363e-05,
+      "loss": 0.9146,
+      "step": 4874
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6454980352575366,
+      "learning_rate": 1.9624873333544507e-05,
+      "loss": 0.9509,
+      "step": 4875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6644955205325864,
+      "learning_rate": 1.9624666272175588e-05,
+      "loss": 0.931,
+      "step": 4876
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6835467097651493,
+      "learning_rate": 1.962445915476882e-05,
+      "loss": 0.8409,
+      "step": 4877
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6303969024640186,
+      "learning_rate": 1.9624251981325403e-05,
+      "loss": 1.0364,
+      "step": 4878
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.688751826071363,
+      "learning_rate": 1.9624044751846543e-05,
+      "loss": 0.7934,
+      "step": 4879
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7631717216070346,
+      "learning_rate": 1.9623837466333445e-05,
+      "loss": 0.7729,
+      "step": 4880
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0463354250605796,
+      "learning_rate": 1.962363012478732e-05,
+      "loss": 0.8859,
+      "step": 4881
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6994047228772506,
+      "learning_rate": 1.9623422727209376e-05,
+      "loss": 0.9105,
+      "step": 4882
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5971051298956351,
+      "learning_rate": 1.9623215273600817e-05,
+      "loss": 1.0603,
+      "step": 4883
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.568113049890623,
+      "learning_rate": 1.962300776396285e-05,
+      "loss": 0.8627,
+      "step": 4884
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8971384828704345,
+      "learning_rate": 1.9622800198296686e-05,
+      "loss": 0.9211,
+      "step": 4885
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7311329447896775,
+      "learning_rate": 1.9622592576603536e-05,
+      "loss": 0.8734,
+      "step": 4886
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7897670944769688,
+      "learning_rate": 1.96223848988846e-05,
+      "loss": 0.9563,
+      "step": 4887
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7616961546902798,
+      "learning_rate": 1.9622177165141098e-05,
+      "loss": 0.7127,
+      "step": 4888
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6149921987590823,
+      "learning_rate": 1.962196937537423e-05,
+      "loss": 0.9856,
+      "step": 4889
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.676035262184356,
+      "learning_rate": 1.962176152958521e-05,
+      "loss": 0.8839,
+      "step": 4890
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7125231266118859,
+      "learning_rate": 1.962155362777525e-05,
+      "loss": 0.8483,
+      "step": 4891
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7916738934752208,
+      "learning_rate": 1.9621345669945562e-05,
+      "loss": 0.9544,
+      "step": 4892
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0283194355051335,
+      "learning_rate": 1.9621137656097347e-05,
+      "loss": 0.981,
+      "step": 4893
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7413658803517196,
+      "learning_rate": 1.9620929586231828e-05,
+      "loss": 0.9867,
+      "step": 4894
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.624862754240939,
+      "learning_rate": 1.9620721460350207e-05,
+      "loss": 0.9447,
+      "step": 4895
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.602339965951357,
+      "learning_rate": 1.9620513278453704e-05,
+      "loss": 1.0228,
+      "step": 4896
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.938651194366426,
+      "learning_rate": 1.9620305040543523e-05,
+      "loss": 0.9623,
+      "step": 4897
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6689269947307346,
+      "learning_rate": 1.9620096746620882e-05,
+      "loss": 0.8369,
+      "step": 4898
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6800462550871125,
+      "learning_rate": 1.9619888396686992e-05,
+      "loss": 0.9422,
+      "step": 4899
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7000466431317103,
+      "learning_rate": 1.9619679990743065e-05,
+      "loss": 0.894,
+      "step": 4900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5987012652030397,
+      "learning_rate": 1.9619471528790317e-05,
+      "loss": 0.8707,
+      "step": 4901
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8052248866353242,
+      "learning_rate": 1.961926301082996e-05,
+      "loss": 0.8461,
+      "step": 4902
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9334389171537634,
+      "learning_rate": 1.9619054436863206e-05,
+      "loss": 0.837,
+      "step": 4903
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8113945606259549,
+      "learning_rate": 1.9618845806891276e-05,
+      "loss": 0.7627,
+      "step": 4904
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7728990122473955,
+      "learning_rate": 1.9618637120915376e-05,
+      "loss": 1.0191,
+      "step": 4905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.736615370904323,
+      "learning_rate": 1.9618428378936728e-05,
+      "loss": 0.9905,
+      "step": 4906
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.465086173660602,
+      "learning_rate": 1.9618219580956544e-05,
+      "loss": 0.876,
+      "step": 4907
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6008287966928012,
+      "learning_rate": 1.9618010726976044e-05,
+      "loss": 0.9519,
+      "step": 4908
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9528481250125753,
+      "learning_rate": 1.9617801816996437e-05,
+      "loss": 0.9198,
+      "step": 4909
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.534500992339961,
+      "learning_rate": 1.961759285101894e-05,
+      "loss": 0.8473,
+      "step": 4910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.790319153594747,
+      "learning_rate": 1.961738382904478e-05,
+      "loss": 0.897,
+      "step": 4911
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5659040633911376,
+      "learning_rate": 1.961717475107516e-05,
+      "loss": 0.9777,
+      "step": 4912
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7325479358835976,
+      "learning_rate": 1.9616965617111305e-05,
+      "loss": 0.9339,
+      "step": 4913
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9069727714381322,
+      "learning_rate": 1.9616756427154434e-05,
+      "loss": 1.0416,
+      "step": 4914
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0589867367374244,
+      "learning_rate": 1.961654718120576e-05,
+      "loss": 0.9201,
+      "step": 4915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6448321476887466,
+      "learning_rate": 1.9616337879266507e-05,
+      "loss": 0.8651,
+      "step": 4916
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7919375162475317,
+      "learning_rate": 1.9616128521337886e-05,
+      "loss": 0.7308,
+      "step": 4917
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.54082516146962,
+      "learning_rate": 1.9615919107421123e-05,
+      "loss": 0.8339,
+      "step": 4918
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7911666436947717,
+      "learning_rate": 1.9615709637517437e-05,
+      "loss": 0.8402,
+      "step": 4919
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6262265975403205,
+      "learning_rate": 1.961550011162804e-05,
+      "loss": 0.8895,
+      "step": 4920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5014379193636638,
+      "learning_rate": 1.9615290529754162e-05,
+      "loss": 0.9882,
+      "step": 4921
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7764186801205197,
+      "learning_rate": 1.961508089189702e-05,
+      "loss": 1.0499,
+      "step": 4922
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6429208477752433,
+      "learning_rate": 1.9614871198057826e-05,
+      "loss": 1.0219,
+      "step": 4923
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1019289416160123,
+      "learning_rate": 1.961466144823781e-05,
+      "loss": 0.9165,
+      "step": 4924
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4525200946858678,
+      "learning_rate": 1.9614451642438193e-05,
+      "loss": 0.7576,
+      "step": 4925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6931774347660948,
+      "learning_rate": 1.9614241780660196e-05,
+      "loss": 1.0538,
+      "step": 4926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7972628153234596,
+      "learning_rate": 1.9614031862905034e-05,
+      "loss": 0.9639,
+      "step": 4927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6865910916945217,
+      "learning_rate": 1.961382188917394e-05,
+      "loss": 1.0024,
+      "step": 4928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.067282967916779,
+      "learning_rate": 1.9613611859468128e-05,
+      "loss": 1.0402,
+      "step": 4929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8279532334408607,
+      "learning_rate": 1.9613401773788826e-05,
+      "loss": 0.7528,
+      "step": 4930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4979884416280709,
+      "learning_rate": 1.9613191632137256e-05,
+      "loss": 1.0526,
+      "step": 4931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9058794072012124,
+      "learning_rate": 1.961298143451464e-05,
+      "loss": 0.812,
+      "step": 4932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5861547897758872,
+      "learning_rate": 1.9612771180922203e-05,
+      "loss": 1.0009,
+      "step": 4933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.654771119686301,
+      "learning_rate": 1.9612560871361163e-05,
+      "loss": 0.9927,
+      "step": 4934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7824441923896408,
+      "learning_rate": 1.9612350505832757e-05,
+      "loss": 0.7387,
+      "step": 4935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7737117175711112,
+      "learning_rate": 1.9612140084338204e-05,
+      "loss": 0.915,
+      "step": 4936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7865037344198362,
+      "learning_rate": 1.9611929606878724e-05,
+      "loss": 0.6592,
+      "step": 4937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8035326200838204,
+      "learning_rate": 1.9611719073455547e-05,
+      "loss": 0.9243,
+      "step": 4938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6753596329169722,
+      "learning_rate": 1.9611508484069902e-05,
+      "loss": 0.9512,
+      "step": 4939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.87288528622405,
+      "learning_rate": 1.961129783872301e-05,
+      "loss": 0.9907,
+      "step": 4940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5581286271584933,
+      "learning_rate": 1.9611087137416097e-05,
+      "loss": 0.8803,
+      "step": 4941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.707373994691376,
+      "learning_rate": 1.9610876380150396e-05,
+      "loss": 0.8167,
+      "step": 4942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9715262102089315,
+      "learning_rate": 1.9610665566927127e-05,
+      "loss": 1.0806,
+      "step": 4943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.743335069855392,
+      "learning_rate": 1.961045469774752e-05,
+      "loss": 0.8718,
+      "step": 4944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.993775511739409,
+      "learning_rate": 1.9610243772612804e-05,
+      "loss": 0.8643,
+      "step": 4945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9751040717328423,
+      "learning_rate": 1.9610032791524206e-05,
+      "loss": 0.8471,
+      "step": 4946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5370690698410687,
+      "learning_rate": 1.9609821754482957e-05,
+      "loss": 0.9389,
+      "step": 4947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.514341116478535,
+      "learning_rate": 1.960961066149028e-05,
+      "loss": 1.0106,
+      "step": 4948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6027220719439597,
+      "learning_rate": 1.960939951254741e-05,
+      "loss": 0.8776,
+      "step": 4949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6152805807678674,
+      "learning_rate": 1.9609188307655572e-05,
+      "loss": 0.8938,
+      "step": 4950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8882824212396717,
+      "learning_rate": 1.9608977046815997e-05,
+      "loss": 0.7789,
+      "step": 4951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7965565917647324,
+      "learning_rate": 1.9608765730029917e-05,
+      "loss": 1.014,
+      "step": 4952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7025548848196121,
+      "learning_rate": 1.960855435729856e-05,
+      "loss": 0.9608,
+      "step": 4953
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9160120604748772,
+      "learning_rate": 1.9608342928623157e-05,
+      "loss": 0.793,
+      "step": 4954
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.067374013926294,
+      "learning_rate": 1.960813144400494e-05,
+      "loss": 0.9061,
+      "step": 4955
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8556940797868167,
+      "learning_rate": 1.960791990344514e-05,
+      "loss": 0.9951,
+      "step": 4956
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7653486635554705,
+      "learning_rate": 1.960770830694499e-05,
+      "loss": 0.9351,
+      "step": 4957
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3516068294352257,
+      "learning_rate": 1.960749665450572e-05,
+      "loss": 0.9322,
+      "step": 4958
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.135953620328196,
+      "learning_rate": 1.9607284946128562e-05,
+      "loss": 1.0038,
+      "step": 4959
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6687426307463102,
+      "learning_rate": 1.9607073181814748e-05,
+      "loss": 0.9868,
+      "step": 4960
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.825967476524879,
+      "learning_rate": 1.9606861361565514e-05,
+      "loss": 0.8761,
+      "step": 4961
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0401559872129553,
+      "learning_rate": 1.960664948538209e-05,
+      "loss": 0.9701,
+      "step": 4962
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7121990684418924,
+      "learning_rate": 1.9606437553265713e-05,
+      "loss": 0.9377,
+      "step": 4963
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0634451926585835,
+      "learning_rate": 1.9606225565217615e-05,
+      "loss": 0.7706,
+      "step": 4964
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.913641175328537,
+      "learning_rate": 1.960601352123903e-05,
+      "loss": 0.8874,
+      "step": 4965
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7634252030696285,
+      "learning_rate": 1.9605801421331194e-05,
+      "loss": 0.8918,
+      "step": 4966
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0885613373129615,
+      "learning_rate": 1.960558926549534e-05,
+      "loss": 0.8348,
+      "step": 4967
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5680716312311538,
+      "learning_rate": 1.9605377053732705e-05,
+      "loss": 1.0215,
+      "step": 4968
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6488381575336453,
+      "learning_rate": 1.9605164786044523e-05,
+      "loss": 0.8916,
+      "step": 4969
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.060953451065899,
+      "learning_rate": 1.9604952462432032e-05,
+      "loss": 0.8242,
+      "step": 4970
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8394324503273634,
+      "learning_rate": 1.9604740082896467e-05,
+      "loss": 0.8597,
+      "step": 4971
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8076000853805468,
+      "learning_rate": 1.960452764743906e-05,
+      "loss": 0.7552,
+      "step": 4972
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5231863703444066,
+      "learning_rate": 1.960431515606106e-05,
+      "loss": 0.9653,
+      "step": 4973
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6094946064892204,
+      "learning_rate": 1.960410260876369e-05,
+      "loss": 0.99,
+      "step": 4974
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5087318978622064,
+      "learning_rate": 1.9603890005548198e-05,
+      "loss": 0.9688,
+      "step": 4975
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7563029623791104,
+      "learning_rate": 1.9603677346415815e-05,
+      "loss": 0.887,
+      "step": 4976
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2610933002084614,
+      "learning_rate": 1.9603464631367782e-05,
+      "loss": 0.8877,
+      "step": 4977
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7851776157081112,
+      "learning_rate": 1.960325186040534e-05,
+      "loss": 1.0196,
+      "step": 4978
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6241782734521466,
+      "learning_rate": 1.9603039033529723e-05,
+      "loss": 0.9786,
+      "step": 4979
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.018984836216597,
+      "learning_rate": 1.9602826150742174e-05,
+      "loss": 0.8456,
+      "step": 4980
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.788059082154867,
+      "learning_rate": 1.960261321204393e-05,
+      "loss": 1.0225,
+      "step": 4981
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9263892128579505,
+      "learning_rate": 1.960240021743623e-05,
+      "loss": 1.0318,
+      "step": 4982
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8054324759928715,
+      "learning_rate": 1.960218716692032e-05,
+      "loss": 0.871,
+      "step": 4983
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.91899355185884,
+      "learning_rate": 1.9601974060497433e-05,
+      "loss": 0.931,
+      "step": 4984
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.783052756771213,
+      "learning_rate": 1.9601760898168814e-05,
+      "loss": 0.873,
+      "step": 4985
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5777686904397321,
+      "learning_rate": 1.9601547679935706e-05,
+      "loss": 0.9024,
+      "step": 4986
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.412240705643697,
+      "learning_rate": 1.9601334405799344e-05,
+      "loss": 1.0099,
+      "step": 4987
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7730769689472243,
+      "learning_rate": 1.9601121075760977e-05,
+      "loss": 0.9401,
+      "step": 4988
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6885179897709413,
+      "learning_rate": 1.960090768982184e-05,
+      "loss": 0.914,
+      "step": 4989
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0666037217947215,
+      "learning_rate": 1.960069424798318e-05,
+      "loss": 0.8011,
+      "step": 4990
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.746150811614304,
+      "learning_rate": 1.9600480750246235e-05,
+      "loss": 0.8855,
+      "step": 4991
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.68598956848924,
+      "learning_rate": 1.9600267196612253e-05,
+      "loss": 0.8638,
+      "step": 4992
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8175665156435944,
+      "learning_rate": 1.960005358708248e-05,
+      "loss": 0.7633,
+      "step": 4993
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7739679279926321,
+      "learning_rate": 1.9599839921658154e-05,
+      "loss": 0.9235,
+      "step": 4994
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.260164133691529,
+      "learning_rate": 1.9599626200340523e-05,
+      "loss": 0.9181,
+      "step": 4995
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6288679120290155,
+      "learning_rate": 1.9599412423130823e-05,
+      "loss": 0.9833,
+      "step": 4996
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9332120249616445,
+      "learning_rate": 1.9599198590030308e-05,
+      "loss": 0.911,
+      "step": 4997
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8462524094222208,
+      "learning_rate": 1.959898470104022e-05,
+      "loss": 0.7217,
+      "step": 4998
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7171412468296556,
+      "learning_rate": 1.9598770756161807e-05,
+      "loss": 1.0875,
+      "step": 4999
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.829939473905139,
+      "learning_rate": 1.9598556755396307e-05,
+      "loss": 0.9167,
+      "step": 5000
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.948637647546077,
+      "learning_rate": 1.9598342698744973e-05,
+      "loss": 0.9199,
+      "step": 5001
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7003925028266602,
+      "learning_rate": 1.9598128586209047e-05,
+      "loss": 0.9228,
+      "step": 5002
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.15722641161309,
+      "learning_rate": 1.959791441778978e-05,
+      "loss": 1.0442,
+      "step": 5003
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8060194232082312,
+      "learning_rate": 1.9597700193488418e-05,
+      "loss": 0.8192,
+      "step": 5004
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4751819574392537,
+      "learning_rate": 1.9597485913306203e-05,
+      "loss": 0.9053,
+      "step": 5005
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.794575258537518,
+      "learning_rate": 1.959727157724439e-05,
+      "loss": 0.9785,
+      "step": 5006
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6415165167018966,
+      "learning_rate": 1.959705718530422e-05,
+      "loss": 0.9325,
+      "step": 5007
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.371514521764121,
+      "learning_rate": 1.9596842737486945e-05,
+      "loss": 1.0153,
+      "step": 5008
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4831843154522413,
+      "learning_rate": 1.9596628233793814e-05,
+      "loss": 0.9794,
+      "step": 5009
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5961765322600494,
+      "learning_rate": 1.9596413674226076e-05,
+      "loss": 0.9133,
+      "step": 5010
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6142853849116487,
+      "learning_rate": 1.959619905878498e-05,
+      "loss": 0.9327,
+      "step": 5011
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.161780938530338,
+      "learning_rate": 1.9595984387471772e-05,
+      "loss": 0.8099,
+      "step": 5012
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.000766348187498,
+      "learning_rate": 1.959576966028771e-05,
+      "loss": 0.8992,
+      "step": 5013
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8116388444000193,
+      "learning_rate": 1.9595554877234036e-05,
+      "loss": 0.7998,
+      "step": 5014
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.715593958158269,
+      "learning_rate": 1.9595340038312e-05,
+      "loss": 0.9308,
+      "step": 5015
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.613310859237584,
+      "learning_rate": 1.959512514352286e-05,
+      "loss": 0.9722,
+      "step": 5016
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7300429636211767,
+      "learning_rate": 1.9594910192867866e-05,
+      "loss": 0.8723,
+      "step": 5017
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2861680271417724,
+      "learning_rate": 1.9594695186348265e-05,
+      "loss": 0.9093,
+      "step": 5018
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0728534439116433,
+      "learning_rate": 1.959448012396531e-05,
+      "loss": 0.8983,
+      "step": 5019
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7090726912532728,
+      "learning_rate": 1.9594265005720255e-05,
+      "loss": 0.9557,
+      "step": 5020
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5914354302872673,
+      "learning_rate": 1.9594049831614352e-05,
+      "loss": 1.0015,
+      "step": 5021
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8527819842065836,
+      "learning_rate": 1.9593834601648853e-05,
+      "loss": 0.7051,
+      "step": 5022
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5889035110052505,
+      "learning_rate": 1.9593619315825013e-05,
+      "loss": 0.9654,
+      "step": 5023
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.278095840492453,
+      "learning_rate": 1.9593403974144084e-05,
+      "loss": 0.94,
+      "step": 5024
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.789716699959214,
+      "learning_rate": 1.9593188576607316e-05,
+      "loss": 0.7518,
+      "step": 5025
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1194107425445896,
+      "learning_rate": 1.959297312321597e-05,
+      "loss": 0.937,
+      "step": 5026
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.956614556694495,
+      "learning_rate": 1.95927576139713e-05,
+      "loss": 0.8124,
+      "step": 5027
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0984608192867493,
+      "learning_rate": 1.959254204887455e-05,
+      "loss": 0.8863,
+      "step": 5028
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9222032853092381,
+      "learning_rate": 1.959232642792699e-05,
+      "loss": 0.9265,
+      "step": 5029
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.931277395143879,
+      "learning_rate": 1.959211075112987e-05,
+      "loss": 0.8899,
+      "step": 5030
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.082483587047054,
+      "learning_rate": 1.959189501848444e-05,
+      "loss": 0.9523,
+      "step": 5031
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5036853081483248,
+      "learning_rate": 1.9591679229991963e-05,
+      "loss": 0.9172,
+      "step": 5032
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8132692503498706,
+      "learning_rate": 1.959146338565369e-05,
+      "loss": 0.9907,
+      "step": 5033
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7780079024503823,
+      "learning_rate": 1.959124748547088e-05,
+      "loss": 0.9312,
+      "step": 5034
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.306924242145549,
+      "learning_rate": 1.9591031529444795e-05,
+      "loss": 0.9365,
+      "step": 5035
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6241590057243658,
+      "learning_rate": 1.9590815517576684e-05,
+      "loss": 0.9862,
+      "step": 5036
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.047579725931465,
+      "learning_rate": 1.9590599449867813e-05,
+      "loss": 0.9574,
+      "step": 5037
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8157943419190906,
+      "learning_rate": 1.959038332631943e-05,
+      "loss": 0.778,
+      "step": 5038
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3504106988467877,
+      "learning_rate": 1.9590167146932804e-05,
+      "loss": 0.9711,
+      "step": 5039
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5281593385752938,
+      "learning_rate": 1.958995091170919e-05,
+      "loss": 0.8202,
+      "step": 5040
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6365132501929642,
+      "learning_rate": 1.958973462064984e-05,
+      "loss": 0.855,
+      "step": 5041
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.719939225442151,
+      "learning_rate": 1.9589518273756023e-05,
+      "loss": 0.9892,
+      "step": 5042
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6809333813062157,
+      "learning_rate": 1.9589301871028993e-05,
+      "loss": 0.81,
+      "step": 5043
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8670996533281619,
+      "learning_rate": 1.958908541247001e-05,
+      "loss": 0.9077,
+      "step": 5044
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.25215504013579,
+      "learning_rate": 1.9588868898080337e-05,
+      "loss": 0.9628,
+      "step": 5045
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5626918302066801,
+      "learning_rate": 1.9588652327861236e-05,
+      "loss": 0.9574,
+      "step": 5046
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6598552146575227,
+      "learning_rate": 1.9588435701813962e-05,
+      "loss": 0.8789,
+      "step": 5047
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0000851905528134,
+      "learning_rate": 1.958821901993978e-05,
+      "loss": 0.8747,
+      "step": 5048
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5733792148492982,
+      "learning_rate": 1.9588002282239955e-05,
+      "loss": 0.8359,
+      "step": 5049
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8608390961728716,
+      "learning_rate": 1.9587785488715742e-05,
+      "loss": 0.7767,
+      "step": 5050
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7224954080749706,
+      "learning_rate": 1.9587568639368405e-05,
+      "loss": 0.9414,
+      "step": 5051
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6406309088496966,
+      "learning_rate": 1.958735173419921e-05,
+      "loss": 1.0214,
+      "step": 5052
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7264010712589628,
+      "learning_rate": 1.958713477320942e-05,
+      "loss": 0.9351,
+      "step": 5053
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6395860633898593,
+      "learning_rate": 1.9586917756400293e-05,
+      "loss": 0.8807,
+      "step": 5054
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8781790689721047,
+      "learning_rate": 1.95867006837731e-05,
+      "loss": 0.8846,
+      "step": 5055
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7452021288690114,
+      "learning_rate": 1.9586483555329098e-05,
+      "loss": 0.9025,
+      "step": 5056
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8306491727018934,
+      "learning_rate": 1.958626637106955e-05,
+      "loss": 0.8976,
+      "step": 5057
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7942664195686245,
+      "learning_rate": 1.9586049130995728e-05,
+      "loss": 0.7734,
+      "step": 5058
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5112430945209494,
+      "learning_rate": 1.958583183510889e-05,
+      "loss": 1.0556,
+      "step": 5059
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.589601612038538,
+      "learning_rate": 1.9585614483410308e-05,
+      "loss": 0.8776,
+      "step": 5060
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3862318547435866,
+      "learning_rate": 1.9585397075901242e-05,
+      "loss": 0.9749,
+      "step": 5061
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7270279451727113,
+      "learning_rate": 1.958517961258296e-05,
+      "loss": 1.0432,
+      "step": 5062
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5519025849036834,
+      "learning_rate": 1.958496209345673e-05,
+      "loss": 0.9482,
+      "step": 5063
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7830913906015908,
+      "learning_rate": 1.9584744518523812e-05,
+      "loss": 0.8676,
+      "step": 5064
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0250442875454864,
+      "learning_rate": 1.958452688778548e-05,
+      "loss": 0.8655,
+      "step": 5065
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8088266542528152,
+      "learning_rate": 1.9584309201242993e-05,
+      "loss": 0.7304,
+      "step": 5066
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6608463671257776,
+      "learning_rate": 1.958409145889763e-05,
+      "loss": 1.0235,
+      "step": 5067
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9276697853707112,
+      "learning_rate": 1.9583873660750646e-05,
+      "loss": 0.9084,
+      "step": 5068
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7125692396696193,
+      "learning_rate": 1.958365580680332e-05,
+      "loss": 1.0493,
+      "step": 5069
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6810438599728768,
+      "learning_rate": 1.9583437897056915e-05,
+      "loss": 0.9451,
+      "step": 5070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.406942540795005,
+      "learning_rate": 1.95832199315127e-05,
+      "loss": 0.9266,
+      "step": 5071
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1416326760208695,
+      "learning_rate": 1.9583001910171947e-05,
+      "loss": 0.8823,
+      "step": 5072
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7417331923309367,
+      "learning_rate": 1.958278383303592e-05,
+      "loss": 0.9355,
+      "step": 5073
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6735619740565644,
+      "learning_rate": 1.9582565700105893e-05,
+      "loss": 0.9196,
+      "step": 5074
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6339639207009775,
+      "learning_rate": 1.958234751138313e-05,
+      "loss": 0.9895,
+      "step": 5075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6729747409148863,
+      "learning_rate": 1.9582129266868914e-05,
+      "loss": 0.9427,
+      "step": 5076
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8431684163618318,
+      "learning_rate": 1.9581910966564502e-05,
+      "loss": 0.9164,
+      "step": 5077
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.728307529897513,
+      "learning_rate": 1.9581692610471174e-05,
+      "loss": 0.9968,
+      "step": 5078
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9883306297564625,
+      "learning_rate": 1.9581474198590196e-05,
+      "loss": 0.9517,
+      "step": 5079
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4202425251478032,
+      "learning_rate": 1.9581255730922842e-05,
+      "loss": 0.9809,
+      "step": 5080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8434447705877912,
+      "learning_rate": 1.9581037207470382e-05,
+      "loss": 0.9413,
+      "step": 5081
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6845999581209343,
+      "learning_rate": 1.9580818628234095e-05,
+      "loss": 0.9461,
+      "step": 5082
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5987264556815757,
+      "learning_rate": 1.9580599993215243e-05,
+      "loss": 0.7866,
+      "step": 5083
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6990059882455373,
+      "learning_rate": 1.9580381302415107e-05,
+      "loss": 0.8581,
+      "step": 5084
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.870111066479485,
+      "learning_rate": 1.9580162555834962e-05,
+      "loss": 0.7872,
+      "step": 5085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7855812497210963,
+      "learning_rate": 1.957994375347607e-05,
+      "loss": 0.6807,
+      "step": 5086
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7156578527227042,
+      "learning_rate": 1.957972489533972e-05,
+      "loss": 0.8867,
+      "step": 5087
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.913797385754987,
+      "learning_rate": 1.9579505981427177e-05,
+      "loss": 1.0086,
+      "step": 5088
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.791056100516419,
+      "learning_rate": 1.9579287011739712e-05,
+      "loss": 0.9581,
+      "step": 5089
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7861538371279335,
+      "learning_rate": 1.957906798627861e-05,
+      "loss": 0.769,
+      "step": 5090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6814956913066992,
+      "learning_rate": 1.957884890504514e-05,
+      "loss": 0.7988,
+      "step": 5091
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.904072841187664,
+      "learning_rate": 1.957862976804058e-05,
+      "loss": 0.8789,
+      "step": 5092
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7194836275082128,
+      "learning_rate": 1.9578410575266203e-05,
+      "loss": 0.9119,
+      "step": 5093
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.851303562869575,
+      "learning_rate": 1.9578191326723292e-05,
+      "loss": 0.9296,
+      "step": 5094
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7086366187566135,
+      "learning_rate": 1.9577972022413113e-05,
+      "loss": 0.8155,
+      "step": 5095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5014215193871245,
+      "learning_rate": 1.9577752662336952e-05,
+      "loss": 0.8282,
+      "step": 5096
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9164158047027933,
+      "learning_rate": 1.957753324649608e-05,
+      "loss": 0.8775,
+      "step": 5097
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.754128173827367,
+      "learning_rate": 1.957731377489178e-05,
+      "loss": 0.9024,
+      "step": 5098
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6601825575812776,
+      "learning_rate": 1.9577094247525324e-05,
+      "loss": 0.9683,
+      "step": 5099
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.743228201561033,
+      "learning_rate": 1.9576874664397998e-05,
+      "loss": 0.8501,
+      "step": 5100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5981810596110246,
+      "learning_rate": 1.9576655025511072e-05,
+      "loss": 0.9158,
+      "step": 5101
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7817092737819569,
+      "learning_rate": 1.957643533086583e-05,
+      "loss": 0.7409,
+      "step": 5102
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6034864816938732,
+      "learning_rate": 1.957621558046355e-05,
+      "loss": 0.98,
+      "step": 5103
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7501919393963586,
+      "learning_rate": 1.9575995774305506e-05,
+      "loss": 0.7247,
+      "step": 5104
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.761884900859941,
+      "learning_rate": 1.9575775912392992e-05,
+      "loss": 0.7195,
+      "step": 5105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6201072958872547,
+      "learning_rate": 1.9575555994727273e-05,
+      "loss": 0.9019,
+      "step": 5106
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5286720496676605,
+      "learning_rate": 1.9575336021309635e-05,
+      "loss": 0.9131,
+      "step": 5107
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7671839344647344,
+      "learning_rate": 1.957511599214136e-05,
+      "loss": 0.9553,
+      "step": 5108
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7399476225132007,
+      "learning_rate": 1.957489590722373e-05,
+      "loss": 0.7405,
+      "step": 5109
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5287623884487171,
+      "learning_rate": 1.9574675766558023e-05,
+      "loss": 0.8933,
+      "step": 5110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.028562878138543,
+      "learning_rate": 1.9574455570145524e-05,
+      "loss": 0.9243,
+      "step": 5111
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9509880707964935,
+      "learning_rate": 1.9574235317987514e-05,
+      "loss": 0.865,
+      "step": 5112
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7988250004292614,
+      "learning_rate": 1.9574015010085273e-05,
+      "loss": 0.8039,
+      "step": 5113
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.681481607346226,
+      "learning_rate": 1.9573794646440084e-05,
+      "loss": 0.9374,
+      "step": 5114
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.070437696572177,
+      "learning_rate": 1.9573574227053233e-05,
+      "loss": 0.8444,
+      "step": 5115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.780961720865504,
+      "learning_rate": 1.9573353751926004e-05,
+      "loss": 0.9333,
+      "step": 5116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.759361221135415,
+      "learning_rate": 1.9573133221059676e-05,
+      "loss": 0.8539,
+      "step": 5117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7570988173874897,
+      "learning_rate": 1.9572912634455538e-05,
+      "loss": 0.9881,
+      "step": 5118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2423654368434014,
+      "learning_rate": 1.9572691992114867e-05,
+      "loss": 0.9315,
+      "step": 5119
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7126885906856586,
+      "learning_rate": 1.9572471294038956e-05,
+      "loss": 0.9792,
+      "step": 5120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6337422583243775,
+      "learning_rate": 1.9572250540229088e-05,
+      "loss": 0.8892,
+      "step": 5121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2330124770025015,
+      "learning_rate": 1.9572029730686546e-05,
+      "loss": 0.9015,
+      "step": 5122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7869841576376602,
+      "learning_rate": 1.9571808865412617e-05,
+      "loss": 0.8235,
+      "step": 5123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.487313891692777,
+      "learning_rate": 1.9571587944408582e-05,
+      "loss": 0.8405,
+      "step": 5124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8254536197157796,
+      "learning_rate": 1.957136696767574e-05,
+      "loss": 0.9361,
+      "step": 5125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.2458725381653455,
+      "learning_rate": 1.9571145935215358e-05,
+      "loss": 1.0213,
+      "step": 5126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6317317919675158,
+      "learning_rate": 1.9570924847028744e-05,
+      "loss": 0.9958,
+      "step": 5127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5660429701085232,
+      "learning_rate": 1.957070370311717e-05,
+      "loss": 0.9151,
+      "step": 5128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.599175440614603,
+      "learning_rate": 1.9570482503481933e-05,
+      "loss": 0.7692,
+      "step": 5129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.595871952864152,
+      "learning_rate": 1.957026124812431e-05,
+      "loss": 0.9531,
+      "step": 5130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8603979699567945,
+      "learning_rate": 1.9570039937045604e-05,
+      "loss": 0.9377,
+      "step": 5131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9340791800211177,
+      "learning_rate": 1.9569818570247093e-05,
+      "loss": 0.81,
+      "step": 5132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1898669852706716,
+      "learning_rate": 1.9569597147730066e-05,
+      "loss": 1.0279,
+      "step": 5133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6183432545309557,
+      "learning_rate": 1.9569375669495815e-05,
+      "loss": 1.0245,
+      "step": 5134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7497779835777567,
+      "learning_rate": 1.9569154135545634e-05,
+      "loss": 0.8276,
+      "step": 5135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.654881064009871,
+      "learning_rate": 1.9568932545880804e-05,
+      "loss": 1.0061,
+      "step": 5136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6402510325711248,
+      "learning_rate": 1.9568710900502622e-05,
+      "loss": 0.9113,
+      "step": 5137
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9255831468869828,
+      "learning_rate": 1.9568489199412373e-05,
+      "loss": 0.9282,
+      "step": 5138
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.616116227770459,
+      "learning_rate": 1.956826744261135e-05,
+      "loss": 0.9543,
+      "step": 5139
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9077216078099407,
+      "learning_rate": 1.9568045630100846e-05,
+      "loss": 0.829,
+      "step": 5140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9090757405418268,
+      "learning_rate": 1.9567823761882154e-05,
+      "loss": 0.7547,
+      "step": 5141
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8243368734040427,
+      "learning_rate": 1.9567601837956557e-05,
+      "loss": 0.7493,
+      "step": 5142
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.656930794538969,
+      "learning_rate": 1.9567379858325358e-05,
+      "loss": 0.88,
+      "step": 5143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7471573633647464,
+      "learning_rate": 1.9567157822989843e-05,
+      "loss": 0.906,
+      "step": 5144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.542560289511785,
+      "learning_rate": 1.9566935731951307e-05,
+      "loss": 0.9313,
+      "step": 5145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9415765441246897,
+      "learning_rate": 1.956671358521104e-05,
+      "loss": 0.9208,
+      "step": 5146
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6268742785769514,
+      "learning_rate": 1.9566491382770342e-05,
+      "loss": 1.1056,
+      "step": 5147
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.823564604381398,
+      "learning_rate": 1.9566269124630498e-05,
+      "loss": 0.8676,
+      "step": 5148
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0137661573392776,
+      "learning_rate": 1.9566046810792806e-05,
+      "loss": 0.875,
+      "step": 5149
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8051028946965524,
+      "learning_rate": 1.9565824441258562e-05,
+      "loss": 0.9604,
+      "step": 5150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8086395465138199,
+      "learning_rate": 1.9565602016029062e-05,
+      "loss": 0.9403,
+      "step": 5151
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8739732297003798,
+      "learning_rate": 1.95653795351056e-05,
+      "loss": 0.7479,
+      "step": 5152
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7171436644166524,
+      "learning_rate": 1.9565156998489465e-05,
+      "loss": 1.0446,
+      "step": 5153
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6895713981121072,
+      "learning_rate": 1.956493440618196e-05,
+      "loss": 1.0138,
+      "step": 5154
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5781394823315402,
+      "learning_rate": 1.9564711758184378e-05,
+      "loss": 0.8213,
+      "step": 5155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.569840016051968,
+      "learning_rate": 1.956448905449802e-05,
+      "loss": 0.9981,
+      "step": 5156
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7611763024878601,
+      "learning_rate": 1.9564266295124174e-05,
+      "loss": 0.7912,
+      "step": 5157
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6268438327067363,
+      "learning_rate": 1.956404348006414e-05,
+      "loss": 0.8503,
+      "step": 5158
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3431620514045384,
+      "learning_rate": 1.956382060931922e-05,
+      "loss": 0.9552,
+      "step": 5159
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7581565354781219,
+      "learning_rate": 1.956359768289071e-05,
+      "loss": 0.7449,
+      "step": 5160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9437670031291914,
+      "learning_rate": 1.9563374700779904e-05,
+      "loss": 0.9526,
+      "step": 5161
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.183980043575955,
+      "learning_rate": 1.9563151662988106e-05,
+      "loss": 0.898,
+      "step": 5162
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7333017905921884,
+      "learning_rate": 1.9562928569516607e-05,
+      "loss": 0.8542,
+      "step": 5163
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4627195001426299,
+      "learning_rate": 1.9562705420366712e-05,
+      "loss": 0.9797,
+      "step": 5164
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7370338897475865,
+      "learning_rate": 1.9562482215539722e-05,
+      "loss": 1.0767,
+      "step": 5165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.868719881585108,
+      "learning_rate": 1.9562258955036928e-05,
+      "loss": 0.8498,
+      "step": 5166
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6795310353715844,
+      "learning_rate": 1.9562035638859638e-05,
+      "loss": 1.034,
+      "step": 5167
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9672267670344505,
+      "learning_rate": 1.9561812267009148e-05,
+      "loss": 0.8332,
+      "step": 5168
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8800921925141394,
+      "learning_rate": 1.956158883948676e-05,
+      "loss": 1.0258,
+      "step": 5169
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6758555266826967,
+      "learning_rate": 1.9561365356293774e-05,
+      "loss": 0.765,
+      "step": 5170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.431781516291709,
+      "learning_rate": 1.9561141817431494e-05,
+      "loss": 1.0027,
+      "step": 5171
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.998836619864577,
+      "learning_rate": 1.9560918222901216e-05,
+      "loss": 0.8659,
+      "step": 5172
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.182099855881105,
+      "learning_rate": 1.9560694572704245e-05,
+      "loss": 1.0394,
+      "step": 5173
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4817428419644068,
+      "learning_rate": 1.9560470866841888e-05,
+      "loss": 0.9563,
+      "step": 5174
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9569750759800955,
+      "learning_rate": 1.9560247105315443e-05,
+      "loss": 0.9768,
+      "step": 5175
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8040985503747937,
+      "learning_rate": 1.9560023288126208e-05,
+      "loss": 0.8163,
+      "step": 5176
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7239451890897353,
+      "learning_rate": 1.955979941527549e-05,
+      "loss": 0.9362,
+      "step": 5177
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9365526253860272,
+      "learning_rate": 1.9559575486764596e-05,
+      "loss": 0.9674,
+      "step": 5178
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.498922650197344,
+      "learning_rate": 1.9559351502594826e-05,
+      "loss": 0.9043,
+      "step": 5179
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5403861849825167,
+      "learning_rate": 1.9559127462767487e-05,
+      "loss": 0.7736,
+      "step": 5180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8277983463338731,
+      "learning_rate": 1.9558903367283882e-05,
+      "loss": 0.7866,
+      "step": 5181
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.775475622369038,
+      "learning_rate": 1.9558679216145314e-05,
+      "loss": 0.9904,
+      "step": 5182
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5281658117365184,
+      "learning_rate": 1.9558455009353086e-05,
+      "loss": 0.9569,
+      "step": 5183
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7349184059975096,
+      "learning_rate": 1.955823074690851e-05,
+      "loss": 0.9192,
+      "step": 5184
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.805916884512184,
+      "learning_rate": 1.955800642881289e-05,
+      "loss": 0.9081,
+      "step": 5185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8227918550331125,
+      "learning_rate": 1.9557782055067528e-05,
+      "loss": 0.7195,
+      "step": 5186
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6699386705807375,
+      "learning_rate": 1.9557557625673734e-05,
+      "loss": 0.9156,
+      "step": 5187
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3825057289545253,
+      "learning_rate": 1.9557333140632816e-05,
+      "loss": 0.9503,
+      "step": 5188
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8288755585009255,
+      "learning_rate": 1.9557108599946076e-05,
+      "loss": 0.8041,
+      "step": 5189
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7392976926203847,
+      "learning_rate": 1.9556884003614824e-05,
+      "loss": 0.9296,
+      "step": 5190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6118039976563674,
+      "learning_rate": 1.9556659351640368e-05,
+      "loss": 1.0328,
+      "step": 5191
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7126083555008071,
+      "learning_rate": 1.9556434644024013e-05,
+      "loss": 0.8459,
+      "step": 5192
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.62323216326804,
+      "learning_rate": 1.9556209880767073e-05,
+      "loss": 0.9482,
+      "step": 5193
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6831771531781243,
+      "learning_rate": 1.9555985061870854e-05,
+      "loss": 0.9488,
+      "step": 5194
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8521975749468866,
+      "learning_rate": 1.9555760187336665e-05,
+      "loss": 1.0456,
+      "step": 5195
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4789162964570624,
+      "learning_rate": 1.955553525716581e-05,
+      "loss": 0.8295,
+      "step": 5196
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.789599931134345,
+      "learning_rate": 1.9555310271359608e-05,
+      "loss": 0.9306,
+      "step": 5197
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8894074774913878,
+      "learning_rate": 1.9555085229919366e-05,
+      "loss": 0.95,
+      "step": 5198
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2246169096894053,
+      "learning_rate": 1.955486013284639e-05,
+      "loss": 0.9409,
+      "step": 5199
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7118887555850368,
+      "learning_rate": 1.9554634980141993e-05,
+      "loss": 0.8501,
+      "step": 5200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7686160012843388,
+      "learning_rate": 1.955440977180749e-05,
+      "loss": 1.0259,
+      "step": 5201
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6192052692967118,
+      "learning_rate": 1.9554184507844183e-05,
+      "loss": 0.812,
+      "step": 5202
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6970799347477619,
+      "learning_rate": 1.9553959188253388e-05,
+      "loss": 1.0167,
+      "step": 5203
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7911316009739089,
+      "learning_rate": 1.9553733813036425e-05,
+      "loss": 0.7687,
+      "step": 5204
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.747120728108359,
+      "learning_rate": 1.955350838219459e-05,
+      "loss": 0.7092,
+      "step": 5205
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5943973118734653,
+      "learning_rate": 1.9553282895729212e-05,
+      "loss": 0.8388,
+      "step": 5206
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7624672493474765,
+      "learning_rate": 1.9553057353641595e-05,
+      "loss": 0.7476,
+      "step": 5207
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3300769167315996,
+      "learning_rate": 1.955283175593305e-05,
+      "loss": 0.9138,
+      "step": 5208
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.75347097299342,
+      "learning_rate": 1.95526061026049e-05,
+      "loss": 0.9296,
+      "step": 5209
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.761056924788106,
+      "learning_rate": 1.955238039365845e-05,
+      "loss": 0.9572,
+      "step": 5210
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7319372294278643,
+      "learning_rate": 1.9552154629095015e-05,
+      "loss": 0.9977,
+      "step": 5211
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.850853360538751,
+      "learning_rate": 1.9551928808915914e-05,
+      "loss": 0.9405,
+      "step": 5212
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.011167974607973,
+      "learning_rate": 1.9551702933122462e-05,
+      "loss": 0.8994,
+      "step": 5213
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9730022645693248,
+      "learning_rate": 1.9551477001715965e-05,
+      "loss": 0.8179,
+      "step": 5214
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6457791656698342,
+      "learning_rate": 1.955125101469775e-05,
+      "loss": 0.9987,
+      "step": 5215
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7742249304213965,
+      "learning_rate": 1.9551024972069127e-05,
+      "loss": 0.7053,
+      "step": 5216
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6092577017052383,
+      "learning_rate": 1.955079887383141e-05,
+      "loss": 0.8514,
+      "step": 5217
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7331064820394124,
+      "learning_rate": 1.9550572719985922e-05,
+      "loss": 1.0243,
+      "step": 5218
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8005164392934576,
+      "learning_rate": 1.955034651053397e-05,
+      "loss": 0.7461,
+      "step": 5219
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7135695656179375,
+      "learning_rate": 1.9550120245476886e-05,
+      "loss": 0.9585,
+      "step": 5220
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6003811189809345,
+      "learning_rate": 1.9549893924815974e-05,
+      "loss": 0.9058,
+      "step": 5221
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.991097439596687,
+      "learning_rate": 1.9549667548552557e-05,
+      "loss": 0.8725,
+      "step": 5222
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.803085741961407,
+      "learning_rate": 1.9549441116687945e-05,
+      "loss": 0.9487,
+      "step": 5223
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.742193094625305,
+      "learning_rate": 1.954921462922347e-05,
+      "loss": 0.9015,
+      "step": 5224
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5835775061983637,
+      "learning_rate": 1.9548988086160445e-05,
+      "loss": 0.886,
+      "step": 5225
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8574871565725657,
+      "learning_rate": 1.954876148750019e-05,
+      "loss": 0.7706,
+      "step": 5226
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.568782393483165,
+      "learning_rate": 1.9548534833244017e-05,
+      "loss": 0.7828,
+      "step": 5227
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8615733860825465,
+      "learning_rate": 1.9548308123393255e-05,
+      "loss": 0.9028,
+      "step": 5228
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9257569157210799,
+      "learning_rate": 1.9548081357949217e-05,
+      "loss": 0.9737,
+      "step": 5229
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6945220326318713,
+      "learning_rate": 1.954785453691323e-05,
+      "loss": 0.8842,
+      "step": 5230
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8406968086871316,
+      "learning_rate": 1.9547627660286607e-05,
+      "loss": 1.001,
+      "step": 5231
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6075468368751937,
+      "learning_rate": 1.9547400728070675e-05,
+      "loss": 0.8245,
+      "step": 5232
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6644124362151287,
+      "learning_rate": 1.9547173740266753e-05,
+      "loss": 0.8883,
+      "step": 5233
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.848731521390837,
+      "learning_rate": 1.9546946696876164e-05,
+      "loss": 0.969,
+      "step": 5234
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.824710012867441,
+      "learning_rate": 1.954671959790023e-05,
+      "loss": 0.8695,
+      "step": 5235
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.811540013106361,
+      "learning_rate": 1.9546492443340268e-05,
+      "loss": 0.9926,
+      "step": 5236
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6632943885490021,
+      "learning_rate": 1.9546265233197607e-05,
+      "loss": 1.023,
+      "step": 5237
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.843505139351382,
+      "learning_rate": 1.954603796747357e-05,
+      "loss": 0.9745,
+      "step": 5238
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.935411118675836,
+      "learning_rate": 1.9545810646169474e-05,
+      "loss": 0.8343,
+      "step": 5239
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7564852885708075,
+      "learning_rate": 1.9545583269286646e-05,
+      "loss": 0.9433,
+      "step": 5240
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7740020589536223,
+      "learning_rate": 1.9545355836826415e-05,
+      "loss": 0.967,
+      "step": 5241
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7577583694567447,
+      "learning_rate": 1.9545128348790096e-05,
+      "loss": 1.0008,
+      "step": 5242
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9160962691619225,
+      "learning_rate": 1.9544900805179016e-05,
+      "loss": 0.9379,
+      "step": 5243
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6167185394684545,
+      "learning_rate": 1.9544673205994505e-05,
+      "loss": 0.9878,
+      "step": 5244
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6601549347353382,
+      "learning_rate": 1.9544445551237883e-05,
+      "loss": 0.9329,
+      "step": 5245
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8250134750403395,
+      "learning_rate": 1.9544217840910483e-05,
+      "loss": 0.9996,
+      "step": 5246
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.889372892198248,
+      "learning_rate": 1.9543990075013618e-05,
+      "loss": 0.8811,
+      "step": 5247
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7223598460129503,
+      "learning_rate": 1.9543762253548623e-05,
+      "loss": 0.8301,
+      "step": 5248
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6680452673478048,
+      "learning_rate": 1.9543534376516824e-05,
+      "loss": 0.8977,
+      "step": 5249
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.081064749514489,
+      "learning_rate": 1.954330644391955e-05,
+      "loss": 0.9847,
+      "step": 5250
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7490776218754434,
+      "learning_rate": 1.9543078455758118e-05,
+      "loss": 0.944,
+      "step": 5251
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5825088543135961,
+      "learning_rate": 1.9542850412033863e-05,
+      "loss": 0.9214,
+      "step": 5252
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5525621221735695,
+      "learning_rate": 1.9542622312748113e-05,
+      "loss": 0.9911,
+      "step": 5253
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7163822339038213,
+      "learning_rate": 1.9542394157902196e-05,
+      "loss": 0.9124,
+      "step": 5254
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.14857471306957,
+      "learning_rate": 1.9542165947497436e-05,
+      "loss": 0.8499,
+      "step": 5255
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6979089537707854,
+      "learning_rate": 1.954193768153517e-05,
+      "loss": 0.936,
+      "step": 5256
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1311829145659478,
+      "learning_rate": 1.9541709360016715e-05,
+      "loss": 0.9338,
+      "step": 5257
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5470508469618902,
+      "learning_rate": 1.9541480982943413e-05,
+      "loss": 0.9786,
+      "step": 5258
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6050198742079682,
+      "learning_rate": 1.9541252550316585e-05,
+      "loss": 0.8798,
+      "step": 5259
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9765001174098276,
+      "learning_rate": 1.9541024062137565e-05,
+      "loss": 0.9632,
+      "step": 5260
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7455505068669437,
+      "learning_rate": 1.954079551840768e-05,
+      "loss": 0.9523,
+      "step": 5261
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6413220081242488,
+      "learning_rate": 1.954056691912827e-05,
+      "loss": 0.9148,
+      "step": 5262
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8190424811899725,
+      "learning_rate": 1.954033826430065e-05,
+      "loss": 0.7775,
+      "step": 5263
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8860928238754523,
+      "learning_rate": 1.9540109553926165e-05,
+      "loss": 0.7876,
+      "step": 5264
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6828697933986898,
+      "learning_rate": 1.953988078800614e-05,
+      "loss": 0.9816,
+      "step": 5265
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.502738451607436,
+      "learning_rate": 1.953965196654191e-05,
+      "loss": 0.8861,
+      "step": 5266
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1234694625002097,
+      "learning_rate": 1.9539423089534805e-05,
+      "loss": 0.9096,
+      "step": 5267
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1258967852239303,
+      "learning_rate": 1.953919415698616e-05,
+      "loss": 0.9751,
+      "step": 5268
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2800108098377243,
+      "learning_rate": 1.9538965168897307e-05,
+      "loss": 0.8691,
+      "step": 5269
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2019227908517034,
+      "learning_rate": 1.9538736125269578e-05,
+      "loss": 0.9111,
+      "step": 5270
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9966027878000487,
+      "learning_rate": 1.9538507026104304e-05,
+      "loss": 0.9194,
+      "step": 5271
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.782059150335296,
+      "learning_rate": 1.953827787140283e-05,
+      "loss": 0.7048,
+      "step": 5272
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0419580918905127,
+      "learning_rate": 1.9538048661166474e-05,
+      "loss": 0.8588,
+      "step": 5273
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6203121078608795,
+      "learning_rate": 1.9537819395396586e-05,
+      "loss": 0.9046,
+      "step": 5274
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7425926167497144,
+      "learning_rate": 1.9537590074094492e-05,
+      "loss": 0.731,
+      "step": 5275
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5530179318516024,
+      "learning_rate": 1.9537360697261524e-05,
+      "loss": 0.9763,
+      "step": 5276
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5016150626384914,
+      "learning_rate": 1.953713126489903e-05,
+      "loss": 0.9013,
+      "step": 5277
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7084329323903422,
+      "learning_rate": 1.9536901777008332e-05,
+      "loss": 0.9803,
+      "step": 5278
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7758531654116538,
+      "learning_rate": 1.9536672233590774e-05,
+      "loss": 1.0705,
+      "step": 5279
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.415796926644435,
+      "learning_rate": 1.9536442634647696e-05,
+      "loss": 0.9231,
+      "step": 5280
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.015142005950909,
+      "learning_rate": 1.9536212980180422e-05,
+      "loss": 0.8394,
+      "step": 5281
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.015459115136848,
+      "learning_rate": 1.95359832701903e-05,
+      "loss": 1.115,
+      "step": 5282
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.864039272647281,
+      "learning_rate": 1.9535753504678667e-05,
+      "loss": 0.9418,
+      "step": 5283
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7394483609058522,
+      "learning_rate": 1.9535523683646857e-05,
+      "loss": 0.9197,
+      "step": 5284
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5854275271184786,
+      "learning_rate": 1.9535293807096208e-05,
+      "loss": 0.8105,
+      "step": 5285
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6454234627086821,
+      "learning_rate": 1.953506387502806e-05,
+      "loss": 0.8542,
+      "step": 5286
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.511235091097577,
+      "learning_rate": 1.9534833887443752e-05,
+      "loss": 0.828,
+      "step": 5287
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9457081266974134,
+      "learning_rate": 1.9534603844344623e-05,
+      "loss": 0.9956,
+      "step": 5288
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8779753431420496,
+      "learning_rate": 1.9534373745732008e-05,
+      "loss": 0.8959,
+      "step": 5289
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9178576374111194,
+      "learning_rate": 1.9534143591607256e-05,
+      "loss": 1.0322,
+      "step": 5290
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8662910880282699,
+      "learning_rate": 1.9533913381971698e-05,
+      "loss": 0.7302,
+      "step": 5291
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7826955747669873,
+      "learning_rate": 1.953368311682668e-05,
+      "loss": 0.8565,
+      "step": 5292
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7964542650510167,
+      "learning_rate": 1.953345279617354e-05,
+      "loss": 0.7567,
+      "step": 5293
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6137696848875023,
+      "learning_rate": 1.9533222420013618e-05,
+      "loss": 0.888,
+      "step": 5294
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5825224445369757,
+      "learning_rate": 1.9532991988348254e-05,
+      "loss": 0.9575,
+      "step": 5295
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6442944124012893,
+      "learning_rate": 1.9532761501178797e-05,
+      "loss": 0.9075,
+      "step": 5296
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6556931684690634,
+      "learning_rate": 1.9532530958506585e-05,
+      "loss": 1.0076,
+      "step": 5297
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.289226200986102,
+      "learning_rate": 1.953230036033296e-05,
+      "loss": 0.899,
+      "step": 5298
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4692785097934908,
+      "learning_rate": 1.9532069706659263e-05,
+      "loss": 0.9117,
+      "step": 5299
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8339465299606417,
+      "learning_rate": 1.9531838997486838e-05,
+      "loss": 0.9307,
+      "step": 5300
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9959655903880873,
+      "learning_rate": 1.953160823281703e-05,
+      "loss": 1.0878,
+      "step": 5301
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7320270739959593,
+      "learning_rate": 1.953137741265118e-05,
+      "loss": 0.971,
+      "step": 5302
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9900132527644658,
+      "learning_rate": 1.9531146536990634e-05,
+      "loss": 0.864,
+      "step": 5303
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7147721897195314,
+      "learning_rate": 1.9530915605836735e-05,
+      "loss": 0.8889,
+      "step": 5304
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8750180149361566,
+      "learning_rate": 1.9530684619190827e-05,
+      "loss": 0.8374,
+      "step": 5305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1209383971258116,
+      "learning_rate": 1.9530453577054257e-05,
+      "loss": 1.0068,
+      "step": 5306
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7464257217641674,
+      "learning_rate": 1.9530222479428368e-05,
+      "loss": 0.8364,
+      "step": 5307
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7333656003521416,
+      "learning_rate": 1.9529991326314506e-05,
+      "loss": 0.9595,
+      "step": 5308
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7801685027680443,
+      "learning_rate": 1.952976011771402e-05,
+      "loss": 0.8096,
+      "step": 5309
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8767441832743956,
+      "learning_rate": 1.9529528853628254e-05,
+      "loss": 0.9618,
+      "step": 5310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7144882880485373,
+      "learning_rate": 1.952929753405855e-05,
+      "loss": 0.9782,
+      "step": 5311
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.540971120729845,
+      "learning_rate": 1.952906615900626e-05,
+      "loss": 0.717,
+      "step": 5312
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7828562394473262,
+      "learning_rate": 1.952883472847273e-05,
+      "loss": 0.6946,
+      "step": 5313
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7862598825895826,
+      "learning_rate": 1.9528603242459308e-05,
+      "loss": 0.9163,
+      "step": 5314
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6619926931137672,
+      "learning_rate": 1.952837170096734e-05,
+      "loss": 0.9451,
+      "step": 5315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2643240194151644,
+      "learning_rate": 1.952814010399818e-05,
+      "loss": 0.9042,
+      "step": 5316
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8076685306014773,
+      "learning_rate": 1.9527908451553168e-05,
+      "loss": 0.888,
+      "step": 5317
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.525848383782918,
+      "learning_rate": 1.9527676743633654e-05,
+      "loss": 0.9653,
+      "step": 5318
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7173733359311778,
+      "learning_rate": 1.9527444980240993e-05,
+      "loss": 0.9321,
+      "step": 5319
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.060848093812822,
+      "learning_rate": 1.952721316137653e-05,
+      "loss": 0.9152,
+      "step": 5320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8442536854592997,
+      "learning_rate": 1.952698128704162e-05,
+      "loss": 1.0763,
+      "step": 5321
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6908074358193346,
+      "learning_rate": 1.9526749357237604e-05,
+      "loss": 0.802,
+      "step": 5322
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9328394768197559,
+      "learning_rate": 1.952651737196584e-05,
+      "loss": 0.9057,
+      "step": 5323
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3365099720973257,
+      "learning_rate": 1.9526285331227672e-05,
+      "loss": 0.9047,
+      "step": 5324
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7165172930924288,
+      "learning_rate": 1.9526053235024457e-05,
+      "loss": 0.9731,
+      "step": 5325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.009341374210776,
+      "learning_rate": 1.9525821083357545e-05,
+      "loss": 0.9569,
+      "step": 5326
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6809503739426892,
+      "learning_rate": 1.9525588876228284e-05,
+      "loss": 0.9731,
+      "step": 5327
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6161414360354214,
+      "learning_rate": 1.952535661363803e-05,
+      "loss": 0.8903,
+      "step": 5328
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6634254779029691,
+      "learning_rate": 1.9525124295588135e-05,
+      "loss": 1.0164,
+      "step": 5329
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6974300224070469,
+      "learning_rate": 1.952489192207995e-05,
+      "loss": 0.9562,
+      "step": 5330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.407340473179403,
+      "learning_rate": 1.952465949311483e-05,
+      "loss": 0.9165,
+      "step": 5331
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6821148501876082,
+      "learning_rate": 1.9524427008694125e-05,
+      "loss": 0.691,
+      "step": 5332
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.741861485830763,
+      "learning_rate": 1.952419446881919e-05,
+      "loss": 0.9632,
+      "step": 5333
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1462468119351144,
+      "learning_rate": 1.9523961873491385e-05,
+      "loss": 1.0495,
+      "step": 5334
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.585538739003118,
+      "learning_rate": 1.9523729222712053e-05,
+      "loss": 0.8468,
+      "step": 5335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.462976415691202,
+      "learning_rate": 1.9523496516482557e-05,
+      "loss": 0.9731,
+      "step": 5336
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9604192542788503,
+      "learning_rate": 1.9523263754804248e-05,
+      "loss": 0.8574,
+      "step": 5337
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.09102078785821,
+      "learning_rate": 1.9523030937678484e-05,
+      "loss": 0.8945,
+      "step": 5338
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8905214986580736,
+      "learning_rate": 1.9522798065106615e-05,
+      "loss": 0.9284,
+      "step": 5339
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7220881404118291,
+      "learning_rate": 1.9522565137090005e-05,
+      "loss": 1.0757,
+      "step": 5340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8048904677853965,
+      "learning_rate": 1.9522332153630006e-05,
+      "loss": 0.9473,
+      "step": 5341
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7469465665094186,
+      "learning_rate": 1.952209911472797e-05,
+      "loss": 0.8947,
+      "step": 5342
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8509410438029643,
+      "learning_rate": 1.952186602038526e-05,
+      "loss": 0.9013,
+      "step": 5343
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7406840296649513,
+      "learning_rate": 1.9521632870603235e-05,
+      "loss": 0.7015,
+      "step": 5344
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.881791113469076,
+      "learning_rate": 1.952139966538325e-05,
+      "loss": 0.9571,
+      "step": 5345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3344371337308365,
+      "learning_rate": 1.9521166404726656e-05,
+      "loss": 0.8,
+      "step": 5346
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8005053595197159,
+      "learning_rate": 1.952093308863482e-05,
+      "loss": 0.7785,
+      "step": 5347
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2308430368483414,
+      "learning_rate": 1.9520699717109096e-05,
+      "loss": 0.963,
+      "step": 5348
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.141862657695238,
+      "learning_rate": 1.9520466290150842e-05,
+      "loss": 0.9635,
+      "step": 5349
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7690635340096987,
+      "learning_rate": 1.9520232807761424e-05,
+      "loss": 0.9612,
+      "step": 5350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6086560707543935,
+      "learning_rate": 1.9519999269942194e-05,
+      "loss": 0.9082,
+      "step": 5351
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.148215861267133,
+      "learning_rate": 1.9519765676694513e-05,
+      "loss": 0.9579,
+      "step": 5352
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6263565406697633,
+      "learning_rate": 1.9519532028019744e-05,
+      "loss": 1.1081,
+      "step": 5353
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6711783138491898,
+      "learning_rate": 1.9519298323919245e-05,
+      "loss": 0.7825,
+      "step": 5354
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7585137333288545,
+      "learning_rate": 1.9519064564394383e-05,
+      "loss": 0.7283,
+      "step": 5355
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6859913151432588,
+      "learning_rate": 1.9518830749446507e-05,
+      "loss": 0.881,
+      "step": 5356
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9769859974620596,
+      "learning_rate": 1.9518596879076982e-05,
+      "loss": 1.0287,
+      "step": 5357
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6344124652595886,
+      "learning_rate": 1.9518362953287178e-05,
+      "loss": 0.8539,
+      "step": 5358
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.321037428429909,
+      "learning_rate": 1.9518128972078448e-05,
+      "loss": 0.9184,
+      "step": 5359
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.126235497209555,
+      "learning_rate": 1.951789493545216e-05,
+      "loss": 0.7913,
+      "step": 5360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.802766408502546,
+      "learning_rate": 1.9517660843409678e-05,
+      "loss": 0.9803,
+      "step": 5361
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7969476759637737,
+      "learning_rate": 1.9517426695952358e-05,
+      "loss": 0.9796,
+      "step": 5362
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7614613642630681,
+      "learning_rate": 1.9517192493081563e-05,
+      "loss": 0.746,
+      "step": 5363
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9344862934305203,
+      "learning_rate": 1.9516958234798663e-05,
+      "loss": 1.0259,
+      "step": 5364
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5940066182959411,
+      "learning_rate": 1.951672392110502e-05,
+      "loss": 0.994,
+      "step": 5365
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.037261670871271,
+      "learning_rate": 1.9516489552002e-05,
+      "loss": 0.9577,
+      "step": 5366
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8079767287737025,
+      "learning_rate": 1.951625512749096e-05,
+      "loss": 0.9531,
+      "step": 5367
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5643013264454202,
+      "learning_rate": 1.9516020647573272e-05,
+      "loss": 1.0211,
+      "step": 5368
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7939932365251154,
+      "learning_rate": 1.95157861122503e-05,
+      "loss": 0.788,
+      "step": 5369
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.617211919111542,
+      "learning_rate": 1.9515551521523406e-05,
+      "loss": 0.8465,
+      "step": 5370
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5316581278255934,
+      "learning_rate": 1.9515316875393958e-05,
+      "loss": 0.9709,
+      "step": 5371
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8396077799473223,
+      "learning_rate": 1.9515082173863327e-05,
+      "loss": 1.0722,
+      "step": 5372
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.759523013306363,
+      "learning_rate": 1.9514847416932873e-05,
+      "loss": 0.7269,
+      "step": 5373
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5334566191658405,
+      "learning_rate": 1.9514612604603963e-05,
+      "loss": 0.9835,
+      "step": 5374
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.006439312783685,
+      "learning_rate": 1.9514377736877967e-05,
+      "loss": 0.9162,
+      "step": 5375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8454950772641976,
+      "learning_rate": 1.9514142813756253e-05,
+      "loss": 0.9533,
+      "step": 5376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8500522432142463,
+      "learning_rate": 1.9513907835240187e-05,
+      "loss": 0.8693,
+      "step": 5377
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6990182442796222,
+      "learning_rate": 1.9513672801331135e-05,
+      "loss": 0.7939,
+      "step": 5378
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7972617933078932,
+      "learning_rate": 1.951343771203047e-05,
+      "loss": 0.7572,
+      "step": 5379
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8249004907013517,
+      "learning_rate": 1.951320256733956e-05,
+      "loss": 0.9,
+      "step": 5380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6310457845830766,
+      "learning_rate": 1.951296736725977e-05,
+      "loss": 0.8368,
+      "step": 5381
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7553806299864445,
+      "learning_rate": 1.9512732111792477e-05,
+      "loss": 0.9626,
+      "step": 5382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.058543774517998,
+      "learning_rate": 1.9512496800939043e-05,
+      "loss": 0.9241,
+      "step": 5383
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8649875400827219,
+      "learning_rate": 1.9512261434700842e-05,
+      "loss": 0.7468,
+      "step": 5384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6635490954387002,
+      "learning_rate": 1.951202601307924e-05,
+      "loss": 1.0096,
+      "step": 5385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.919000734723904,
+      "learning_rate": 1.9511790536075613e-05,
+      "loss": 0.8563,
+      "step": 5386
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7892176436917233,
+      "learning_rate": 1.951155500369133e-05,
+      "loss": 0.9227,
+      "step": 5387
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9020564435855092,
+      "learning_rate": 1.9511319415927766e-05,
+      "loss": 0.8915,
+      "step": 5388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8179972257705157,
+      "learning_rate": 1.9511083772786285e-05,
+      "loss": 0.8943,
+      "step": 5389
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.021895058749251,
+      "learning_rate": 1.9510848074268263e-05,
+      "loss": 0.7737,
+      "step": 5390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6891199887726087,
+      "learning_rate": 1.9510612320375073e-05,
+      "loss": 0.9041,
+      "step": 5391
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.658799981338792,
+      "learning_rate": 1.951037651110809e-05,
+      "loss": 0.8776,
+      "step": 5392
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.070308890738805,
+      "learning_rate": 1.9510140646468682e-05,
+      "loss": 0.9247,
+      "step": 5393
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.71083629856045,
+      "learning_rate": 1.9509904726458223e-05,
+      "loss": 0.988,
+      "step": 5394
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7228799266418808,
+      "learning_rate": 1.9509668751078087e-05,
+      "loss": 0.7754,
+      "step": 5395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5574001126053683,
+      "learning_rate": 1.9509432720329653e-05,
+      "loss": 0.9811,
+      "step": 5396
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0767273571643208,
+      "learning_rate": 1.9509196634214288e-05,
+      "loss": 0.9355,
+      "step": 5397
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7131219082339386,
+      "learning_rate": 1.950896049273337e-05,
+      "loss": 0.8516,
+      "step": 5398
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4912028452985295,
+      "learning_rate": 1.9508724295888277e-05,
+      "loss": 0.8892,
+      "step": 5399
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7892981931898446,
+      "learning_rate": 1.9508488043680376e-05,
+      "loss": 0.9332,
+      "step": 5400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7925425576756322,
+      "learning_rate": 1.950825173611105e-05,
+      "loss": 0.9925,
+      "step": 5401
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5555795503219432,
+      "learning_rate": 1.950801537318167e-05,
+      "loss": 0.7728,
+      "step": 5402
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7787813645102626,
+      "learning_rate": 1.9507778954893614e-05,
+      "loss": 0.7983,
+      "step": 5403
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.688929958336824,
+      "learning_rate": 1.950754248124826e-05,
+      "loss": 0.9675,
+      "step": 5404
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.740479027460686,
+      "learning_rate": 1.9507305952246988e-05,
+      "loss": 1.0137,
+      "step": 5405
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6490363103791827,
+      "learning_rate": 1.9507069367891166e-05,
+      "loss": 0.9066,
+      "step": 5406
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.126309609276336,
+      "learning_rate": 1.950683272818218e-05,
+      "loss": 0.9525,
+      "step": 5407
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6297375098091107,
+      "learning_rate": 1.9506596033121397e-05,
+      "loss": 0.9031,
+      "step": 5408
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.063698273492525,
+      "learning_rate": 1.9506359282710207e-05,
+      "loss": 0.9538,
+      "step": 5409
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6487620943937056,
+      "learning_rate": 1.950612247694998e-05,
+      "loss": 0.9825,
+      "step": 5410
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.253286770522573,
+      "learning_rate": 1.9505885615842102e-05,
+      "loss": 1.0461,
+      "step": 5411
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1452991197044327,
+      "learning_rate": 1.950564869938795e-05,
+      "loss": 0.9489,
+      "step": 5412
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9251266503977102,
+      "learning_rate": 1.95054117275889e-05,
+      "loss": 0.8763,
+      "step": 5413
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5367244480995392,
+      "learning_rate": 1.9505174700446333e-05,
+      "loss": 1.0075,
+      "step": 5414
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.971817922121483,
+      "learning_rate": 1.950493761796163e-05,
+      "loss": 0.9632,
+      "step": 5415
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1129032725239543,
+      "learning_rate": 1.950470048013617e-05,
+      "loss": 1.0063,
+      "step": 5416
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8562801834062778,
+      "learning_rate": 1.9504463286971332e-05,
+      "loss": 0.7769,
+      "step": 5417
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.552064536995138,
+      "learning_rate": 1.9504226038468506e-05,
+      "loss": 0.8209,
+      "step": 5418
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6870117425573543,
+      "learning_rate": 1.950398873462906e-05,
+      "loss": 1.0762,
+      "step": 5419
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7988538529160336,
+      "learning_rate": 1.9503751375454388e-05,
+      "loss": 0.74,
+      "step": 5420
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.701815132796825,
+      "learning_rate": 1.9503513960945864e-05,
+      "loss": 0.9148,
+      "step": 5421
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.756972830951306,
+      "learning_rate": 1.9503276491104876e-05,
+      "loss": 0.6982,
+      "step": 5422
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0538067971749254,
+      "learning_rate": 1.9503038965932802e-05,
+      "loss": 0.8383,
+      "step": 5423
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.844709763064207,
+      "learning_rate": 1.9502801385431028e-05,
+      "loss": 0.9535,
+      "step": 5424
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8946759010794472,
+      "learning_rate": 1.9502563749600933e-05,
+      "loss": 0.985,
+      "step": 5425
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8652318161177932,
+      "learning_rate": 1.9502326058443904e-05,
+      "loss": 0.9584,
+      "step": 5426
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8396521840812694,
+      "learning_rate": 1.9502088311961324e-05,
+      "loss": 0.9944,
+      "step": 5427
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7691331826195228,
+      "learning_rate": 1.9501850510154578e-05,
+      "loss": 0.7662,
+      "step": 5428
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6652388990437186,
+      "learning_rate": 1.950161265302505e-05,
+      "loss": 0.9347,
+      "step": 5429
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8631228847719945,
+      "learning_rate": 1.9501374740574128e-05,
+      "loss": 0.8657,
+      "step": 5430
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6528099967538668,
+      "learning_rate": 1.9501136772803192e-05,
+      "loss": 0.876,
+      "step": 5431
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7560890038429193,
+      "learning_rate": 1.9500898749713628e-05,
+      "loss": 0.6628,
+      "step": 5432
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.152034716498684,
+      "learning_rate": 1.9500660671306826e-05,
+      "loss": 0.9732,
+      "step": 5433
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1234127573049784,
+      "learning_rate": 1.950042253758417e-05,
+      "loss": 0.9302,
+      "step": 5434
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7717484127707819,
+      "learning_rate": 1.950018434854704e-05,
+      "loss": 0.8698,
+      "step": 5435
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7338308793602197,
+      "learning_rate": 1.9499946104196835e-05,
+      "loss": 1.0222,
+      "step": 5436
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.057595168458925,
+      "learning_rate": 1.9499707804534934e-05,
+      "loss": 0.9138,
+      "step": 5437
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1499471962308236,
+      "learning_rate": 1.949946944956273e-05,
+      "loss": 0.9338,
+      "step": 5438
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.065412651013168,
+      "learning_rate": 1.9499231039281603e-05,
+      "loss": 0.9354,
+      "step": 5439
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8011553720864757,
+      "learning_rate": 1.9498992573692948e-05,
+      "loss": 1.0134,
+      "step": 5440
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7966379494059974,
+      "learning_rate": 1.949875405279815e-05,
+      "loss": 1.0361,
+      "step": 5441
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9450399198885355,
+      "learning_rate": 1.94985154765986e-05,
+      "loss": 0.8227,
+      "step": 5442
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.656569668380468,
+      "learning_rate": 1.949827684509568e-05,
+      "loss": 0.9984,
+      "step": 5443
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7359466578439349,
+      "learning_rate": 1.9498038158290792e-05,
+      "loss": 0.7281,
+      "step": 5444
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7432831728369,
+      "learning_rate": 1.9497799416185318e-05,
+      "loss": 0.9733,
+      "step": 5445
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8000075112438245,
+      "learning_rate": 1.949756061878065e-05,
+      "loss": 1.0169,
+      "step": 5446
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6147732201781544,
+      "learning_rate": 1.949732176607817e-05,
+      "loss": 1.0034,
+      "step": 5447
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6660794450492264,
+      "learning_rate": 1.9497082858079284e-05,
+      "loss": 0.9985,
+      "step": 5448
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.806771203805873,
+      "learning_rate": 1.9496843894785372e-05,
+      "loss": 0.746,
+      "step": 5449
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.394120431628494,
+      "learning_rate": 1.9496604876197826e-05,
+      "loss": 0.9757,
+      "step": 5450
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8928078748579586,
+      "learning_rate": 1.949636580231804e-05,
+      "loss": 0.916,
+      "step": 5451
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8060035112975872,
+      "learning_rate": 1.949612667314741e-05,
+      "loss": 0.9461,
+      "step": 5452
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5616168716297434,
+      "learning_rate": 1.949588748868732e-05,
+      "loss": 0.8879,
+      "step": 5453
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0112121818345052,
+      "learning_rate": 1.9495648248939166e-05,
+      "loss": 0.9419,
+      "step": 5454
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6258149183567963,
+      "learning_rate": 1.9495408953904346e-05,
+      "loss": 0.8213,
+      "step": 5455
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8130786310000964,
+      "learning_rate": 1.9495169603584244e-05,
+      "loss": 0.7201,
+      "step": 5456
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0362332613441816,
+      "learning_rate": 1.949493019798026e-05,
+      "loss": 1.0616,
+      "step": 5457
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1998179866325174,
+      "learning_rate": 1.9494690737093787e-05,
+      "loss": 0.8715,
+      "step": 5458
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.643748949691601,
+      "learning_rate": 1.9494451220926224e-05,
+      "loss": 0.8486,
+      "step": 5459
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8693358138230047,
+      "learning_rate": 1.9494211649478952e-05,
+      "loss": 0.845,
+      "step": 5460
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6615876639481437,
+      "learning_rate": 1.949397202275338e-05,
+      "loss": 0.8308,
+      "step": 5461
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7038472695997218,
+      "learning_rate": 1.9493732340750892e-05,
+      "loss": 0.9864,
+      "step": 5462
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5342087393103714,
+      "learning_rate": 1.949349260347289e-05,
+      "loss": 0.921,
+      "step": 5463
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.949388914788814,
+      "learning_rate": 1.949325281092077e-05,
+      "loss": 0.8769,
+      "step": 5464
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6618532808698103,
+      "learning_rate": 1.9493012963095923e-05,
+      "loss": 0.8967,
+      "step": 5465
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3319521944430703,
+      "learning_rate": 1.9492773059999752e-05,
+      "loss": 0.9886,
+      "step": 5466
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7000484236713922,
+      "learning_rate": 1.949253310163365e-05,
+      "loss": 0.9773,
+      "step": 5467
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8156734736388234,
+      "learning_rate": 1.9492293087999016e-05,
+      "loss": 0.9553,
+      "step": 5468
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7895244232453892,
+      "learning_rate": 1.9492053019097246e-05,
+      "loss": 0.716,
+      "step": 5469
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7265252986143085,
+      "learning_rate": 1.949181289492974e-05,
+      "loss": 0.8623,
+      "step": 5470
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9501513402818642,
+      "learning_rate": 1.9491572715497894e-05,
+      "loss": 0.9471,
+      "step": 5471
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.92058764943706,
+      "learning_rate": 1.9491332480803106e-05,
+      "loss": 0.9317,
+      "step": 5472
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.881500210806186,
+      "learning_rate": 1.949109219084677e-05,
+      "loss": 0.8743,
+      "step": 5473
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6435518791024006,
+      "learning_rate": 1.9490851845630297e-05,
+      "loss": 0.9855,
+      "step": 5474
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.770910364814021,
+      "learning_rate": 1.9490611445155078e-05,
+      "loss": 0.8039,
+      "step": 5475
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7323131820333424,
+      "learning_rate": 1.9490370989422516e-05,
+      "loss": 0.9953,
+      "step": 5476
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5640199755673532,
+      "learning_rate": 1.9490130478434008e-05,
+      "loss": 0.9735,
+      "step": 5477
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6660940313357722,
+      "learning_rate": 1.9489889912190954e-05,
+      "loss": 0.9979,
+      "step": 5478
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5570356053079941,
+      "learning_rate": 1.948964929069476e-05,
+      "loss": 0.8227,
+      "step": 5479
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.659572134883831,
+      "learning_rate": 1.948940861394682e-05,
+      "loss": 0.8634,
+      "step": 5480
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7980927103608756,
+      "learning_rate": 1.948916788194854e-05,
+      "loss": 0.739,
+      "step": 5481
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7764351743730191,
+      "learning_rate": 1.9488927094701322e-05,
+      "loss": 0.9523,
+      "step": 5482
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8038485752345754,
+      "learning_rate": 1.9488686252206564e-05,
+      "loss": 0.8473,
+      "step": 5483
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1546739969468405,
+      "learning_rate": 1.948844535446567e-05,
+      "loss": 0.884,
+      "step": 5484
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8958724539361402,
+      "learning_rate": 1.948820440148005e-05,
+      "loss": 0.8848,
+      "step": 5485
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7459953402249777,
+      "learning_rate": 1.948796339325109e-05,
+      "loss": 0.8528,
+      "step": 5486
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6983155513100296,
+      "learning_rate": 1.948772232978021e-05,
+      "loss": 0.9267,
+      "step": 5487
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.998839089126332,
+      "learning_rate": 1.9487481211068807e-05,
+      "loss": 1.0813,
+      "step": 5488
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9213334747811353,
+      "learning_rate": 1.9487240037118283e-05,
+      "loss": 0.7958,
+      "step": 5489
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0358702066168286,
+      "learning_rate": 1.9486998807930043e-05,
+      "loss": 0.9884,
+      "step": 5490
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7557295206674894,
+      "learning_rate": 1.9486757523505495e-05,
+      "loss": 0.8322,
+      "step": 5491
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9797670687411517,
+      "learning_rate": 1.9486516183846036e-05,
+      "loss": 0.8813,
+      "step": 5492
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8934882602935197,
+      "learning_rate": 1.9486274788953083e-05,
+      "loss": 1.0586,
+      "step": 5493
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7782557966229107,
+      "learning_rate": 1.948603333882803e-05,
+      "loss": 0.7351,
+      "step": 5494
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7331199489087407,
+      "learning_rate": 1.9485791833472286e-05,
+      "loss": 0.7738,
+      "step": 5495
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9380555908901185,
+      "learning_rate": 1.9485550272887265e-05,
+      "loss": 0.9725,
+      "step": 5496
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7395072836710586,
+      "learning_rate": 1.9485308657074364e-05,
+      "loss": 0.8705,
+      "step": 5497
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8202465360805494,
+      "learning_rate": 1.9485066986034992e-05,
+      "loss": 0.7393,
+      "step": 5498
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0307778116348683,
+      "learning_rate": 1.948482525977056e-05,
+      "loss": 0.9277,
+      "step": 5499
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.934145372073317,
+      "learning_rate": 1.948458347828247e-05,
+      "loss": 0.9204,
+      "step": 5500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5937175827602337,
+      "learning_rate": 1.9484341641572134e-05,
+      "loss": 0.8536,
+      "step": 5501
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7460415128724187,
+      "learning_rate": 1.9484099749640956e-05,
+      "loss": 0.7178,
+      "step": 5502
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.828206030989378,
+      "learning_rate": 1.9483857802490348e-05,
+      "loss": 0.9546,
+      "step": 5503
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8287121873839793,
+      "learning_rate": 1.9483615800121717e-05,
+      "loss": 0.8492,
+      "step": 5504
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7673060225927533,
+      "learning_rate": 1.9483373742536475e-05,
+      "loss": 0.7291,
+      "step": 5505
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8103437855375273,
+      "learning_rate": 1.9483131629736024e-05,
+      "loss": 0.9071,
+      "step": 5506
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.686276597223403,
+      "learning_rate": 1.9482889461721782e-05,
+      "loss": 0.8711,
+      "step": 5507
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.759712858105561,
+      "learning_rate": 1.948264723849515e-05,
+      "loss": 0.8255,
+      "step": 5508
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8912560989126042,
+      "learning_rate": 1.948240496005755e-05,
+      "loss": 1.0085,
+      "step": 5509
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6002566309286257,
+      "learning_rate": 1.9482162626410383e-05,
+      "loss": 0.9348,
+      "step": 5510
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7817963883324806,
+      "learning_rate": 1.9481920237555064e-05,
+      "loss": 0.7982,
+      "step": 5511
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7694163284637636,
+      "learning_rate": 1.9481677793493005e-05,
+      "loss": 0.9742,
+      "step": 5512
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.784088921052071,
+      "learning_rate": 1.9481435294225614e-05,
+      "loss": 0.9277,
+      "step": 5513
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5436639660810532,
+      "learning_rate": 1.9481192739754307e-05,
+      "loss": 0.906,
+      "step": 5514
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6876157433010102,
+      "learning_rate": 1.948095013008049e-05,
+      "loss": 0.8847,
+      "step": 5515
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6757464144396605,
+      "learning_rate": 1.9480707465205582e-05,
+      "loss": 0.9008,
+      "step": 5516
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6723534615082183,
+      "learning_rate": 1.9480464745130997e-05,
+      "loss": 0.89,
+      "step": 5517
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9564593978170635,
+      "learning_rate": 1.948022196985814e-05,
+      "loss": 0.9555,
+      "step": 5518
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6628376535812934,
+      "learning_rate": 1.9479979139388432e-05,
+      "loss": 0.9198,
+      "step": 5519
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.834261410477382,
+      "learning_rate": 1.9479736253723282e-05,
+      "loss": 0.9604,
+      "step": 5520
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7168222318085167,
+      "learning_rate": 1.9479493312864107e-05,
+      "loss": 1.0602,
+      "step": 5521
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.783896884242137,
+      "learning_rate": 1.9479250316812323e-05,
+      "loss": 0.8808,
+      "step": 5522
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6703196661948931,
+      "learning_rate": 1.9479007265569342e-05,
+      "loss": 0.9258,
+      "step": 5523
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7288938933473699,
+      "learning_rate": 1.9478764159136578e-05,
+      "loss": 0.9752,
+      "step": 5524
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6634119400150003,
+      "learning_rate": 1.9478520997515446e-05,
+      "loss": 0.9461,
+      "step": 5525
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.998343307016111,
+      "learning_rate": 1.9478277780707367e-05,
+      "loss": 0.9064,
+      "step": 5526
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.304445854370152,
+      "learning_rate": 1.9478034508713753e-05,
+      "loss": 0.8896,
+      "step": 5527
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6065060935761744,
+      "learning_rate": 1.9477791181536024e-05,
+      "loss": 0.9889,
+      "step": 5528
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4838389895876092,
+      "learning_rate": 1.9477547799175592e-05,
+      "loss": 0.894,
+      "step": 5529
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8989183985666438,
+      "learning_rate": 1.9477304361633878e-05,
+      "loss": 0.8752,
+      "step": 5530
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9301597032196237,
+      "learning_rate": 1.9477060868912294e-05,
+      "loss": 0.7919,
+      "step": 5531
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2591478288962463,
+      "learning_rate": 1.9476817321012263e-05,
+      "loss": 0.9874,
+      "step": 5532
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2143841450883697,
+      "learning_rate": 1.94765737179352e-05,
+      "loss": 0.9008,
+      "step": 5533
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7060207088293349,
+      "learning_rate": 1.9476330059682528e-05,
+      "loss": 0.833,
+      "step": 5534
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5210214848679866,
+      "learning_rate": 1.9476086346255657e-05,
+      "loss": 1.0003,
+      "step": 5535
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7846014269305795,
+      "learning_rate": 1.9475842577656018e-05,
+      "loss": 0.7887,
+      "step": 5536
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5725471819794203,
+      "learning_rate": 1.947559875388502e-05,
+      "loss": 0.9155,
+      "step": 5537
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7264169170016823,
+      "learning_rate": 1.9475354874944083e-05,
+      "loss": 0.8656,
+      "step": 5538
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.753291797228115,
+      "learning_rate": 1.9475110940834633e-05,
+      "loss": 1.0163,
+      "step": 5539
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.320664688958106,
+      "learning_rate": 1.947486695155809e-05,
+      "loss": 0.9628,
+      "step": 5540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3330419275640573,
+      "learning_rate": 1.947462290711587e-05,
+      "loss": 0.9788,
+      "step": 5541
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8880903946477006,
+      "learning_rate": 1.9474378807509395e-05,
+      "loss": 0.817,
+      "step": 5542
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.95718896563869,
+      "learning_rate": 1.9474134652740083e-05,
+      "loss": 0.959,
+      "step": 5543
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5600956600484042,
+      "learning_rate": 1.9473890442809366e-05,
+      "loss": 1.0581,
+      "step": 5544
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2265161450375093,
+      "learning_rate": 1.9473646177718658e-05,
+      "loss": 0.9954,
+      "step": 5545
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.101908858672013,
+      "learning_rate": 1.947340185746938e-05,
+      "loss": 0.9042,
+      "step": 5546
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9571247372658185,
+      "learning_rate": 1.9473157482062958e-05,
+      "loss": 1.1105,
+      "step": 5547
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7770571510684299,
+      "learning_rate": 1.9472913051500816e-05,
+      "loss": 0.9545,
+      "step": 5548
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.120507007770714,
+      "learning_rate": 1.9472668565784375e-05,
+      "loss": 0.8927,
+      "step": 5549
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.948059368608613,
+      "learning_rate": 1.9472424024915056e-05,
+      "loss": 0.9075,
+      "step": 5550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7527353797511043,
+      "learning_rate": 1.9472179428894288e-05,
+      "loss": 0.9706,
+      "step": 5551
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.473640740963955,
+      "learning_rate": 1.9471934777723492e-05,
+      "loss": 0.9232,
+      "step": 5552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5618962017925164,
+      "learning_rate": 1.9471690071404096e-05,
+      "loss": 0.9476,
+      "step": 5553
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8725771282937445,
+      "learning_rate": 1.9471445309937515e-05,
+      "loss": 1.0202,
+      "step": 5554
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.681901321324112,
+      "learning_rate": 1.9471200493325186e-05,
+      "loss": 0.8725,
+      "step": 5555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8780934895573493,
+      "learning_rate": 1.9470955621568525e-05,
+      "loss": 0.9783,
+      "step": 5556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.753527258684672,
+      "learning_rate": 1.9470710694668966e-05,
+      "loss": 0.8807,
+      "step": 5557
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8153533441599508,
+      "learning_rate": 1.9470465712627928e-05,
+      "loss": 0.9196,
+      "step": 5558
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6368043341642968,
+      "learning_rate": 1.9470220675446844e-05,
+      "loss": 0.8872,
+      "step": 5559
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8118411842245203,
+      "learning_rate": 1.946997558312713e-05,
+      "loss": 0.9349,
+      "step": 5560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.440867192051577,
+      "learning_rate": 1.9469730435670227e-05,
+      "loss": 0.9461,
+      "step": 5561
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8660427102333117,
+      "learning_rate": 1.9469485233077554e-05,
+      "loss": 1.0098,
+      "step": 5562
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6166158810817404,
+      "learning_rate": 1.9469239975350538e-05,
+      "loss": 0.99,
+      "step": 5563
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.794632022429996,
+      "learning_rate": 1.946899466249061e-05,
+      "loss": 0.846,
+      "step": 5564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8173765260373199,
+      "learning_rate": 1.9468749294499196e-05,
+      "loss": 0.67,
+      "step": 5565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7676959593809638,
+      "learning_rate": 1.9468503871377728e-05,
+      "loss": 0.7176,
+      "step": 5566
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7702919985441266,
+      "learning_rate": 1.946825839312763e-05,
+      "loss": 0.8898,
+      "step": 5567
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.598849816536767,
+      "learning_rate": 1.9468012859750337e-05,
+      "loss": 1.0081,
+      "step": 5568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6017908846987712,
+      "learning_rate": 1.9467767271247275e-05,
+      "loss": 0.9374,
+      "step": 5569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6391998402118693,
+      "learning_rate": 1.9467521627619874e-05,
+      "loss": 0.8875,
+      "step": 5570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8790938440484277,
+      "learning_rate": 1.9467275928869565e-05,
+      "loss": 0.9741,
+      "step": 5571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.619630858287463,
+      "learning_rate": 1.9467030174997777e-05,
+      "loss": 1.0555,
+      "step": 5572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6025794220002647,
+      "learning_rate": 1.9466784366005943e-05,
+      "loss": 1.1279,
+      "step": 5573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0477072722228606,
+      "learning_rate": 1.9466538501895493e-05,
+      "loss": 0.8845,
+      "step": 5574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6794342144604713,
+      "learning_rate": 1.946629258266786e-05,
+      "loss": 1.0643,
+      "step": 5575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.285120221204473,
+      "learning_rate": 1.9466046608324472e-05,
+      "loss": 0.9894,
+      "step": 5576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8110390026498306,
+      "learning_rate": 1.9465800578866764e-05,
+      "loss": 0.9525,
+      "step": 5577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8580426922643691,
+      "learning_rate": 1.9465554494296173e-05,
+      "loss": 0.9253,
+      "step": 5578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2206155876286404,
+      "learning_rate": 1.9465308354614125e-05,
+      "loss": 0.8539,
+      "step": 5579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.177086229346537,
+      "learning_rate": 1.946506215982205e-05,
+      "loss": 0.6997,
+      "step": 5580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8473537065653152,
+      "learning_rate": 1.946481590992139e-05,
+      "loss": 0.9851,
+      "step": 5581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.994327072107944,
+      "learning_rate": 1.9464569604913575e-05,
+      "loss": 0.8211,
+      "step": 5582
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4407543734550787,
+      "learning_rate": 1.9464323244800042e-05,
+      "loss": 0.86,
+      "step": 5583
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.654607211112,
+      "learning_rate": 1.946407682958222e-05,
+      "loss": 0.8248,
+      "step": 5584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8532675412930968,
+      "learning_rate": 1.9463830359261547e-05,
+      "loss": 0.9122,
+      "step": 5585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9393101750606926,
+      "learning_rate": 1.946358383383946e-05,
+      "loss": 0.7753,
+      "step": 5586
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6354920469300134,
+      "learning_rate": 1.9463337253317385e-05,
+      "loss": 0.8788,
+      "step": 5587
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.794078119474391,
+      "learning_rate": 1.946309061769677e-05,
+      "loss": 0.7169,
+      "step": 5588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7687578553795709,
+      "learning_rate": 1.9462843926979044e-05,
+      "loss": 0.7775,
+      "step": 5589
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6367553130402661,
+      "learning_rate": 1.9462597181165647e-05,
+      "loss": 0.9984,
+      "step": 5590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6991757255723818,
+      "learning_rate": 1.946235038025801e-05,
+      "loss": 0.9255,
+      "step": 5591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.630212266340519,
+      "learning_rate": 1.9462103524257574e-05,
+      "loss": 0.9592,
+      "step": 5592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6294635197812979,
+      "learning_rate": 1.9461856613165775e-05,
+      "loss": 0.9656,
+      "step": 5593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6311316506660434,
+      "learning_rate": 1.946160964698405e-05,
+      "loss": 0.9171,
+      "step": 5594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8238146620138327,
+      "learning_rate": 1.9461362625713845e-05,
+      "loss": 0.9204,
+      "step": 5595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7707377164137268,
+      "learning_rate": 1.9461115549356588e-05,
+      "loss": 0.7367,
+      "step": 5596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7736078544799492,
+      "learning_rate": 1.9460868417913718e-05,
+      "loss": 0.9928,
+      "step": 5597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9763764989446524,
+      "learning_rate": 1.946062123138668e-05,
+      "loss": 0.9044,
+      "step": 5598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9080173090881365,
+      "learning_rate": 1.9460373989776905e-05,
+      "loss": 0.9908,
+      "step": 5599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4785057129978054,
+      "learning_rate": 1.9460126693085842e-05,
+      "loss": 0.9443,
+      "step": 5600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8946138985631393,
+      "learning_rate": 1.9459879341314927e-05,
+      "loss": 0.8865,
+      "step": 5601
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0728442922307817,
+      "learning_rate": 1.9459631934465596e-05,
+      "loss": 0.8918,
+      "step": 5602
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8389454629610706,
+      "learning_rate": 1.9459384472539297e-05,
+      "loss": 0.7736,
+      "step": 5603
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7417227652160945,
+      "learning_rate": 1.9459136955537465e-05,
+      "loss": 0.9631,
+      "step": 5604
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.94007952901365,
+      "learning_rate": 1.9458889383461543e-05,
+      "loss": 0.9717,
+      "step": 5605
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6427259504118135,
+      "learning_rate": 1.9458641756312973e-05,
+      "loss": 0.9683,
+      "step": 5606
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9370828450625788,
+      "learning_rate": 1.9458394074093198e-05,
+      "loss": 0.8726,
+      "step": 5607
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.571398232597539,
+      "learning_rate": 1.9458146336803653e-05,
+      "loss": 0.8585,
+      "step": 5608
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5820467656384316,
+      "learning_rate": 1.945789854444579e-05,
+      "loss": 0.9341,
+      "step": 5609
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.077112706379906,
+      "learning_rate": 1.9457650697021046e-05,
+      "loss": 0.8331,
+      "step": 5610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8514332776193387,
+      "learning_rate": 1.9457402794530864e-05,
+      "loss": 0.982,
+      "step": 5611
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7550686338407318,
+      "learning_rate": 1.945715483697669e-05,
+      "loss": 1.0676,
+      "step": 5612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9489642900910271,
+      "learning_rate": 1.9456906824359965e-05,
+      "loss": 0.9591,
+      "step": 5613
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.058054317175835,
+      "learning_rate": 1.9456658756682136e-05,
+      "loss": 0.9963,
+      "step": 5614
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1584490888076795,
+      "learning_rate": 1.9456410633944645e-05,
+      "loss": 0.9577,
+      "step": 5615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7721883462417571,
+      "learning_rate": 1.9456162456148938e-05,
+      "loss": 0.7187,
+      "step": 5616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9492839022825044,
+      "learning_rate": 1.9455914223296458e-05,
+      "loss": 0.9398,
+      "step": 5617
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.346342467306121,
+      "learning_rate": 1.9455665935388653e-05,
+      "loss": 0.7618,
+      "step": 5618
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7652770142730745,
+      "learning_rate": 1.945541759242697e-05,
+      "loss": 0.8462,
+      "step": 5619
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7610639551519955,
+      "learning_rate": 1.945516919441285e-05,
+      "loss": 0.779,
+      "step": 5620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6441609231550531,
+      "learning_rate": 1.945492074134774e-05,
+      "loss": 0.8012,
+      "step": 5621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.111692002069295,
+      "learning_rate": 1.945467223323309e-05,
+      "loss": 0.9012,
+      "step": 5622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.626301773443236,
+      "learning_rate": 1.945442367007034e-05,
+      "loss": 0.8932,
+      "step": 5623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.834256743256793,
+      "learning_rate": 1.945417505186095e-05,
+      "loss": 0.8814,
+      "step": 5624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.614601743016186,
+      "learning_rate": 1.9453926378606354e-05,
+      "loss": 0.8935,
+      "step": 5625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.508492330176203,
+      "learning_rate": 1.9453677650308007e-05,
+      "loss": 0.9789,
+      "step": 5626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7786379533051233,
+      "learning_rate": 1.945342886696736e-05,
+      "loss": 0.9372,
+      "step": 5627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1263598084923965,
+      "learning_rate": 1.945318002858585e-05,
+      "loss": 0.9461,
+      "step": 5628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.952475200608148,
+      "learning_rate": 1.9452931135164936e-05,
+      "loss": 0.8461,
+      "step": 5629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1911447564051567,
+      "learning_rate": 1.9452682186706067e-05,
+      "loss": 0.9953,
+      "step": 5630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8404640641334677,
+      "learning_rate": 1.945243318321069e-05,
+      "loss": 0.98,
+      "step": 5631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.763065950278773,
+      "learning_rate": 1.9452184124680247e-05,
+      "loss": 0.9802,
+      "step": 5632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0489464991322834,
+      "learning_rate": 1.94519350111162e-05,
+      "loss": 0.8182,
+      "step": 5633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3057906934579053,
+      "learning_rate": 1.9451685842519997e-05,
+      "loss": 0.819,
+      "step": 5634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7768396439166037,
+      "learning_rate": 1.9451436618893086e-05,
+      "loss": 0.8702,
+      "step": 5635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9510553080124287,
+      "learning_rate": 1.9451187340236918e-05,
+      "loss": 0.84,
+      "step": 5636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8530926538192378,
+      "learning_rate": 1.9450938006552943e-05,
+      "loss": 0.8272,
+      "step": 5637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1306928782720136,
+      "learning_rate": 1.945068861784262e-05,
+      "loss": 0.968,
+      "step": 5638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6319705004238534,
+      "learning_rate": 1.9450439174107388e-05,
+      "loss": 0.89,
+      "step": 5639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8883086726593596,
+      "learning_rate": 1.9450189675348712e-05,
+      "loss": 1.0238,
+      "step": 5640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.719821769365929,
+      "learning_rate": 1.944994012156804e-05,
+      "loss": 0.895,
+      "step": 5641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7764186851048218,
+      "learning_rate": 1.9449690512766825e-05,
+      "loss": 0.7836,
+      "step": 5642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6807102157584177,
+      "learning_rate": 1.944944084894652e-05,
+      "loss": 0.9731,
+      "step": 5643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1216678260596806,
+      "learning_rate": 1.944919113010858e-05,
+      "loss": 0.8061,
+      "step": 5644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.524137778927005,
+      "learning_rate": 1.9448941356254453e-05,
+      "loss": 0.9264,
+      "step": 5645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.291585328406027,
+      "learning_rate": 1.94486915273856e-05,
+      "loss": 0.9967,
+      "step": 5646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7940924600743445,
+      "learning_rate": 1.9448441643503473e-05,
+      "loss": 0.7707,
+      "step": 5647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8572871200392878,
+      "learning_rate": 1.944819170460953e-05,
+      "loss": 1.0956,
+      "step": 5648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7611278568688633,
+      "learning_rate": 1.9447941710705223e-05,
+      "loss": 0.9421,
+      "step": 5649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.874281896569941,
+      "learning_rate": 1.9447691661792007e-05,
+      "loss": 0.833,
+      "step": 5650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.487649163804734,
+      "learning_rate": 1.944744155787134e-05,
+      "loss": 1.1076,
+      "step": 5651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7976002158913345,
+      "learning_rate": 1.944719139894468e-05,
+      "loss": 0.8644,
+      "step": 5652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7058808417473112,
+      "learning_rate": 1.9446941185013474e-05,
+      "loss": 0.9312,
+      "step": 5653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7782242289673382,
+      "learning_rate": 1.944669091607919e-05,
+      "loss": 0.7261,
+      "step": 5654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6316487358249325,
+      "learning_rate": 1.944644059214328e-05,
+      "loss": 0.9945,
+      "step": 5655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9277218294043434,
+      "learning_rate": 1.94461902132072e-05,
+      "loss": 0.831,
+      "step": 5656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5218956936357921,
+      "learning_rate": 1.9445939779272415e-05,
+      "loss": 0.9995,
+      "step": 5657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7513764442038993,
+      "learning_rate": 1.9445689290340372e-05,
+      "loss": 0.7848,
+      "step": 5658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4595892253180889,
+      "learning_rate": 1.944543874641254e-05,
+      "loss": 0.9062,
+      "step": 5659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.977832978314693,
+      "learning_rate": 1.9445188147490373e-05,
+      "loss": 0.9399,
+      "step": 5660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.264632491080904,
+      "learning_rate": 1.9444937493575326e-05,
+      "loss": 1.006,
+      "step": 5661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.857058821033364,
+      "learning_rate": 1.9444686784668868e-05,
+      "loss": 0.903,
+      "step": 5662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7902126639654901,
+      "learning_rate": 1.944443602077245e-05,
+      "loss": 0.73,
+      "step": 5663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8021147703245752,
+      "learning_rate": 1.9444185201887536e-05,
+      "loss": 0.8647,
+      "step": 5664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7883965295192273,
+      "learning_rate": 1.9443934328015587e-05,
+      "loss": 0.7473,
+      "step": 5665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.759460564561498,
+      "learning_rate": 1.944368339915806e-05,
+      "loss": 0.8287,
+      "step": 5666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5779155519402759,
+      "learning_rate": 1.9443432415316424e-05,
+      "loss": 0.8634,
+      "step": 5667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.597350157785971,
+      "learning_rate": 1.944318137649213e-05,
+      "loss": 1.0445,
+      "step": 5668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1397411930619743,
+      "learning_rate": 1.9442930282686647e-05,
+      "loss": 0.9908,
+      "step": 5669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5163852488028597,
+      "learning_rate": 1.9442679133901433e-05,
+      "loss": 0.9799,
+      "step": 5670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6067447594144728,
+      "learning_rate": 1.9442427930137948e-05,
+      "loss": 0.913,
+      "step": 5671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6686371951825705,
+      "learning_rate": 1.9442176671397663e-05,
+      "loss": 0.9808,
+      "step": 5672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7065017612848097,
+      "learning_rate": 1.9441925357682033e-05,
+      "loss": 0.9159,
+      "step": 5673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.045070908663627,
+      "learning_rate": 1.9441673988992525e-05,
+      "loss": 0.8811,
+      "step": 5674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7486946490079842,
+      "learning_rate": 1.9441422565330606e-05,
+      "loss": 0.9502,
+      "step": 5675
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6029229582572917,
+      "learning_rate": 1.9441171086697732e-05,
+      "loss": 0.8546,
+      "step": 5676
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.627932717823064,
+      "learning_rate": 1.944091955309537e-05,
+      "loss": 0.9818,
+      "step": 5677
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7486741582606655,
+      "learning_rate": 1.9440667964524987e-05,
+      "loss": 0.8167,
+      "step": 5678
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7084460126449272,
+      "learning_rate": 1.9440416320988046e-05,
+      "loss": 0.8951,
+      "step": 5679
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5762629135540747,
+      "learning_rate": 1.9440164622486013e-05,
+      "loss": 0.9413,
+      "step": 5680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5913953265145058,
+      "learning_rate": 1.9439912869020353e-05,
+      "loss": 1.0678,
+      "step": 5681
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7391026327995562,
+      "learning_rate": 1.943966106059253e-05,
+      "loss": 0.9541,
+      "step": 5682
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8034830600668424,
+      "learning_rate": 1.9439409197204012e-05,
+      "loss": 0.7653,
+      "step": 5683
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9373341698674171,
+      "learning_rate": 1.9439157278856268e-05,
+      "loss": 0.9967,
+      "step": 5684
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9679276071568579,
+      "learning_rate": 1.9438905305550756e-05,
+      "loss": 0.8143,
+      "step": 5685
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.592258027863197,
+      "learning_rate": 1.9438653277288953e-05,
+      "loss": 0.9696,
+      "step": 5686
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.95841388164575,
+      "learning_rate": 1.9438401194072324e-05,
+      "loss": 0.8858,
+      "step": 5687
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8034273621980739,
+      "learning_rate": 1.9438149055902335e-05,
+      "loss": 0.9233,
+      "step": 5688
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8693001768658706,
+      "learning_rate": 1.943789686278045e-05,
+      "loss": 0.8727,
+      "step": 5689
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5182680228748628,
+      "learning_rate": 1.9437644614708143e-05,
+      "loss": 1.0014,
+      "step": 5690
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2429068474571743,
+      "learning_rate": 1.9437392311686884e-05,
+      "loss": 0.8478,
+      "step": 5691
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7315368107257325,
+      "learning_rate": 1.9437139953718137e-05,
+      "loss": 0.96,
+      "step": 5692
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2585236298435056,
+      "learning_rate": 1.9436887540803372e-05,
+      "loss": 0.8449,
+      "step": 5693
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2303914117369943,
+      "learning_rate": 1.943663507294406e-05,
+      "loss": 0.8996,
+      "step": 5694
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5800205009378832,
+      "learning_rate": 1.9436382550141675e-05,
+      "loss": 0.8711,
+      "step": 5695
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1387123116062776,
+      "learning_rate": 1.943612997239768e-05,
+      "loss": 0.9544,
+      "step": 5696
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6684803367301047,
+      "learning_rate": 1.943587733971355e-05,
+      "loss": 0.9319,
+      "step": 5697
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6455041396647647,
+      "learning_rate": 1.9435624652090754e-05,
+      "loss": 0.8827,
+      "step": 5698
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6490897286391282,
+      "learning_rate": 1.9435371909530766e-05,
+      "loss": 0.9328,
+      "step": 5699
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.344855377495478,
+      "learning_rate": 1.9435119112035054e-05,
+      "loss": 0.9577,
+      "step": 5700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7820838355701548,
+      "learning_rate": 1.943486625960509e-05,
+      "loss": 0.7115,
+      "step": 5701
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8057491880469692,
+      "learning_rate": 1.943461335224235e-05,
+      "loss": 0.8955,
+      "step": 5702
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8204058629504718,
+      "learning_rate": 1.9434360389948303e-05,
+      "loss": 0.785,
+      "step": 5703
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.744281355049709,
+      "learning_rate": 1.9434107372724424e-05,
+      "loss": 1.0047,
+      "step": 5704
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.546801324872568,
+      "learning_rate": 1.9433854300572184e-05,
+      "loss": 0.7936,
+      "step": 5705
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.045023393507789,
+      "learning_rate": 1.943360117349306e-05,
+      "loss": 0.8944,
+      "step": 5706
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7575464967620305,
+      "learning_rate": 1.943334799148852e-05,
+      "loss": 0.9787,
+      "step": 5707
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7545380658048566,
+      "learning_rate": 1.9433094754560042e-05,
+      "loss": 0.9411,
+      "step": 5708
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7123513141734756,
+      "learning_rate": 1.94328414627091e-05,
+      "loss": 0.9765,
+      "step": 5709
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6626576405607847,
+      "learning_rate": 1.943258811593717e-05,
+      "loss": 0.9732,
+      "step": 5710
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6784502803336645,
+      "learning_rate": 1.9432334714245723e-05,
+      "loss": 0.9821,
+      "step": 5711
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7713566085979705,
+      "learning_rate": 1.943208125763624e-05,
+      "loss": 0.7529,
+      "step": 5712
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9486116390336343,
+      "learning_rate": 1.943182774611019e-05,
+      "loss": 1.0135,
+      "step": 5713
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0683455473375267,
+      "learning_rate": 1.943157417966906e-05,
+      "loss": 0.7989,
+      "step": 5714
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6671251207050704,
+      "learning_rate": 1.9431320558314313e-05,
+      "loss": 0.8814,
+      "step": 5715
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8790646116128267,
+      "learning_rate": 1.9431066882047434e-05,
+      "loss": 0.9798,
+      "step": 5716
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6278711417425404,
+      "learning_rate": 1.9430813150869896e-05,
+      "loss": 1.0309,
+      "step": 5717
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8658735018174994,
+      "learning_rate": 1.943055936478318e-05,
+      "loss": 0.8654,
+      "step": 5718
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7730919708385164,
+      "learning_rate": 1.943030552378876e-05,
+      "loss": 0.9605,
+      "step": 5719
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9104051620527407,
+      "learning_rate": 1.9430051627888117e-05,
+      "loss": 0.7972,
+      "step": 5720
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9764256476214157,
+      "learning_rate": 1.942979767708273e-05,
+      "loss": 0.921,
+      "step": 5721
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6710683991643869,
+      "learning_rate": 1.942954367137407e-05,
+      "loss": 0.8503,
+      "step": 5722
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7105370194810705,
+      "learning_rate": 1.9429289610763624e-05,
+      "loss": 0.9535,
+      "step": 5723
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6100356399421571,
+      "learning_rate": 1.9429035495252872e-05,
+      "loss": 0.942,
+      "step": 5724
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.735219450473059,
+      "learning_rate": 1.9428781324843285e-05,
+      "loss": 0.6977,
+      "step": 5725
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8265027616917258,
+      "learning_rate": 1.942852709953635e-05,
+      "loss": 0.9305,
+      "step": 5726
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9652617958354872,
+      "learning_rate": 1.9428272819333545e-05,
+      "loss": 0.99,
+      "step": 5727
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.816159876398867,
+      "learning_rate": 1.9428018484236353e-05,
+      "loss": 0.9502,
+      "step": 5728
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.769503222354996,
+      "learning_rate": 1.9427764094246248e-05,
+      "loss": 1.0381,
+      "step": 5729
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7655071312568998,
+      "learning_rate": 1.942750964936472e-05,
+      "loss": 0.7643,
+      "step": 5730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7936086413717234,
+      "learning_rate": 1.9427255149593245e-05,
+      "loss": 0.7506,
+      "step": 5731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6122088641750054,
+      "learning_rate": 1.9427000594933305e-05,
+      "loss": 0.8431,
+      "step": 5732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.115561984721819,
+      "learning_rate": 1.9426745985386385e-05,
+      "loss": 0.9372,
+      "step": 5733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5393404620748414,
+      "learning_rate": 1.9426491320953963e-05,
+      "loss": 0.9973,
+      "step": 5734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.811578334139072,
+      "learning_rate": 1.9426236601637523e-05,
+      "loss": 0.7507,
+      "step": 5735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7400618622117658,
+      "learning_rate": 1.9425981827438552e-05,
+      "loss": 0.8309,
+      "step": 5736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9621704602253227,
+      "learning_rate": 1.942572699835853e-05,
+      "loss": 0.846,
+      "step": 5737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.60756622624242,
+      "learning_rate": 1.942547211439894e-05,
+      "loss": 0.9965,
+      "step": 5738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6652046334539956,
+      "learning_rate": 1.9425217175561268e-05,
+      "loss": 0.8869,
+      "step": 5739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2703675914350416,
+      "learning_rate": 1.9424962181847e-05,
+      "loss": 0.9083,
+      "step": 5740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6845949693271274,
+      "learning_rate": 1.9424707133257615e-05,
+      "loss": 0.8492,
+      "step": 5741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9343292505506275,
+      "learning_rate": 1.9424452029794603e-05,
+      "loss": 0.9282,
+      "step": 5742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6178912506985221,
+      "learning_rate": 1.9424196871459448e-05,
+      "loss": 0.7911,
+      "step": 5743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8445109135863259,
+      "learning_rate": 1.9423941658253634e-05,
+      "loss": 0.7381,
+      "step": 5744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7911606536249984,
+      "learning_rate": 1.942368639017865e-05,
+      "loss": 0.9342,
+      "step": 5745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5846396555862816,
+      "learning_rate": 1.9423431067235977e-05,
+      "loss": 0.7364,
+      "step": 5746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6820639519545628,
+      "learning_rate": 1.9423175689427105e-05,
+      "loss": 0.9792,
+      "step": 5747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1207377320395264,
+      "learning_rate": 1.9422920256753523e-05,
+      "loss": 0.8226,
+      "step": 5748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.570060752836287,
+      "learning_rate": 1.9422664769216716e-05,
+      "loss": 1.0906,
+      "step": 5749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4823050283090062,
+      "learning_rate": 1.9422409226818175e-05,
+      "loss": 1.0563,
+      "step": 5750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8073323163915219,
+      "learning_rate": 1.9422153629559376e-05,
+      "loss": 0.7633,
+      "step": 5751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.621584366113555,
+      "learning_rate": 1.942189797744182e-05,
+      "loss": 0.9371,
+      "step": 5752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.754385748059027,
+      "learning_rate": 1.942164227046699e-05,
+      "loss": 0.7498,
+      "step": 5753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7833215514789809,
+      "learning_rate": 1.9421386508636378e-05,
+      "loss": 0.7331,
+      "step": 5754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6929478270340657,
+      "learning_rate": 1.9421130691951472e-05,
+      "loss": 1.0284,
+      "step": 5755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.693875954024997,
+      "learning_rate": 1.942087482041376e-05,
+      "loss": 0.9443,
+      "step": 5756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5100441458865792,
+      "learning_rate": 1.9420618894024727e-05,
+      "loss": 0.8988,
+      "step": 5757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7524027077603634,
+      "learning_rate": 1.942036291278587e-05,
+      "loss": 0.9731,
+      "step": 5758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2454870682271886,
+      "learning_rate": 1.942010687669868e-05,
+      "loss": 0.8524,
+      "step": 5759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3553678114116834,
+      "learning_rate": 1.9419850785764645e-05,
+      "loss": 0.8444,
+      "step": 5760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9592661241385914,
+      "learning_rate": 1.9419594639985257e-05,
+      "loss": 0.9847,
+      "step": 5761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8597739100985293,
+      "learning_rate": 1.9419338439362007e-05,
+      "loss": 0.9481,
+      "step": 5762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6010861624110517,
+      "learning_rate": 1.9419082183896385e-05,
+      "loss": 0.894,
+      "step": 5763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7933785420440697,
+      "learning_rate": 1.9418825873589883e-05,
+      "loss": 0.7112,
+      "step": 5764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8630498144081242,
+      "learning_rate": 1.9418569508443998e-05,
+      "loss": 0.9105,
+      "step": 5765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4864838040029626,
+      "learning_rate": 1.941831308846022e-05,
+      "loss": 0.941,
+      "step": 5766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7099485246089603,
+      "learning_rate": 1.9418056613640038e-05,
+      "loss": 0.9448,
+      "step": 5767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9330984738108454,
+      "learning_rate": 1.941780008398495e-05,
+      "loss": 0.925,
+      "step": 5768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9834206503480492,
+      "learning_rate": 1.941754349949645e-05,
+      "loss": 0.9505,
+      "step": 5769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.785085450983868,
+      "learning_rate": 1.941728686017603e-05,
+      "loss": 0.9107,
+      "step": 5770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.776439796908757,
+      "learning_rate": 1.9417030166025182e-05,
+      "loss": 0.8718,
+      "step": 5771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8274554566309859,
+      "learning_rate": 1.9416773417045406e-05,
+      "loss": 0.775,
+      "step": 5772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.653165748318825,
+      "learning_rate": 1.9416516613238193e-05,
+      "loss": 1.0372,
+      "step": 5773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8021743565900934,
+      "learning_rate": 1.9416259754605034e-05,
+      "loss": 0.8988,
+      "step": 5774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5339179895969626,
+      "learning_rate": 1.9416002841147438e-05,
+      "loss": 0.9206,
+      "step": 5775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.715856478466814,
+      "learning_rate": 1.9415745872866887e-05,
+      "loss": 0.9207,
+      "step": 5776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8073296240596768,
+      "learning_rate": 1.9415488849764884e-05,
+      "loss": 0.867,
+      "step": 5777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.859732248958886,
+      "learning_rate": 1.9415231771842925e-05,
+      "loss": 1.0054,
+      "step": 5778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.029427710334952,
+      "learning_rate": 1.9414974639102505e-05,
+      "loss": 0.9267,
+      "step": 5779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6404467734248582,
+      "learning_rate": 1.941471745154512e-05,
+      "loss": 1.024,
+      "step": 5780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7108588168616778,
+      "learning_rate": 1.941446020917227e-05,
+      "loss": 0.9406,
+      "step": 5781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.456741840806907,
+      "learning_rate": 1.9414202911985453e-05,
+      "loss": 0.8602,
+      "step": 5782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.812061437127353,
+      "learning_rate": 1.9413945559986164e-05,
+      "loss": 0.8618,
+      "step": 5783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7005619949520618,
+      "learning_rate": 1.9413688153175907e-05,
+      "loss": 0.922,
+      "step": 5784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5999539443982844,
+      "learning_rate": 1.9413430691556175e-05,
+      "loss": 1.002,
+      "step": 5785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3683251168848276,
+      "learning_rate": 1.9413173175128472e-05,
+      "loss": 0.8305,
+      "step": 5786
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7025008211636279,
+      "learning_rate": 1.9412915603894294e-05,
+      "loss": 0.8919,
+      "step": 5787
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6861466135451724,
+      "learning_rate": 1.941265797785514e-05,
+      "loss": 0.9105,
+      "step": 5788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.737757904218016,
+      "learning_rate": 1.941240029701251e-05,
+      "loss": 0.8642,
+      "step": 5789
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5584500671091364,
+      "learning_rate": 1.9412142561367906e-05,
+      "loss": 0.94,
+      "step": 5790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8783007161383656,
+      "learning_rate": 1.941188477092283e-05,
+      "loss": 0.9333,
+      "step": 5791
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.904992293603824,
+      "learning_rate": 1.941162692567878e-05,
+      "loss": 0.794,
+      "step": 5792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8223953702218076,
+      "learning_rate": 1.941136902563726e-05,
+      "loss": 0.9244,
+      "step": 5793
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5327592189650372,
+      "learning_rate": 1.941111107079977e-05,
+      "loss": 0.9481,
+      "step": 5794
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5916672287952385,
+      "learning_rate": 1.9410853061167806e-05,
+      "loss": 0.9686,
+      "step": 5795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.658683138894266,
+      "learning_rate": 1.9410594996742884e-05,
+      "loss": 1.082,
+      "step": 5796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5939755930561317,
+      "learning_rate": 1.9410336877526494e-05,
+      "loss": 0.9631,
+      "step": 5797
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6275589859969968,
+      "learning_rate": 1.9410078703520147e-05,
+      "loss": 1.0382,
+      "step": 5798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.571030736183376,
+      "learning_rate": 1.9409820474725338e-05,
+      "loss": 0.9143,
+      "step": 5799
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7517327116977541,
+      "learning_rate": 1.9409562191143577e-05,
+      "loss": 0.9008,
+      "step": 5800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6749919099607367,
+      "learning_rate": 1.9409303852776365e-05,
+      "loss": 0.9029,
+      "step": 5801
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0134274822821006,
+      "learning_rate": 1.940904545962521e-05,
+      "loss": 0.9024,
+      "step": 5802
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9796047913652606,
+      "learning_rate": 1.940878701169161e-05,
+      "loss": 0.7718,
+      "step": 5803
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8097293397962755,
+      "learning_rate": 1.9408528508977076e-05,
+      "loss": 0.7181,
+      "step": 5804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0085507793069852,
+      "learning_rate": 1.940826995148311e-05,
+      "loss": 0.8793,
+      "step": 5805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5093157060922469,
+      "learning_rate": 1.9408011339211215e-05,
+      "loss": 1.0191,
+      "step": 5806
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.785399871307333,
+      "learning_rate": 1.94077526721629e-05,
+      "loss": 0.8489,
+      "step": 5807
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7771147642746192,
+      "learning_rate": 1.940749395033967e-05,
+      "loss": 0.7129,
+      "step": 5808
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6401195572248812,
+      "learning_rate": 1.9407235173743038e-05,
+      "loss": 0.9683,
+      "step": 5809
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5505627559031347,
+      "learning_rate": 1.94069763423745e-05,
+      "loss": 1.0787,
+      "step": 5810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7868945233150229,
+      "learning_rate": 1.9406717456235566e-05,
+      "loss": 0.9466,
+      "step": 5811
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7163677495519707,
+      "learning_rate": 1.9406458515327747e-05,
+      "loss": 1.0326,
+      "step": 5812
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7403590328469967,
+      "learning_rate": 1.940619951965255e-05,
+      "loss": 0.9314,
+      "step": 5813
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.600666980127437,
+      "learning_rate": 1.9405940469211478e-05,
+      "loss": 0.8652,
+      "step": 5814
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5010144048063985,
+      "learning_rate": 1.9405681364006047e-05,
+      "loss": 0.894,
+      "step": 5815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.801371070245817,
+      "learning_rate": 1.9405422204037758e-05,
+      "loss": 0.7643,
+      "step": 5816
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.703125993216787,
+      "learning_rate": 1.940516298930812e-05,
+      "loss": 0.9565,
+      "step": 5817
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4703973642787984,
+      "learning_rate": 1.940490371981865e-05,
+      "loss": 0.9265,
+      "step": 5818
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6212327432470333,
+      "learning_rate": 1.9404644395570852e-05,
+      "loss": 0.9761,
+      "step": 5819
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8406240215824008,
+      "learning_rate": 1.940438501656624e-05,
+      "loss": 0.7443,
+      "step": 5820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5913978262964183,
+      "learning_rate": 1.9404125582806316e-05,
+      "loss": 0.9907,
+      "step": 5821
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7656985801865082,
+      "learning_rate": 1.9403866094292598e-05,
+      "loss": 0.8612,
+      "step": 5822
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5876361958471676,
+      "learning_rate": 1.9403606551026596e-05,
+      "loss": 0.9129,
+      "step": 5823
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.505774423163642,
+      "learning_rate": 1.9403346953009817e-05,
+      "loss": 0.9009,
+      "step": 5824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6253017630317963,
+      "learning_rate": 1.9403087300243776e-05,
+      "loss": 0.8924,
+      "step": 5825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1929291416678627,
+      "learning_rate": 1.9402827592729984e-05,
+      "loss": 0.8692,
+      "step": 5826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.534801250968476,
+      "learning_rate": 1.9402567830469952e-05,
+      "loss": 0.9614,
+      "step": 5827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8419506051442542,
+      "learning_rate": 1.9402308013465192e-05,
+      "loss": 0.9387,
+      "step": 5828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6621458402103693,
+      "learning_rate": 1.9402048141717217e-05,
+      "loss": 0.9382,
+      "step": 5829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.618463651722485,
+      "learning_rate": 1.9401788215227547e-05,
+      "loss": 0.9976,
+      "step": 5830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8911528044260912,
+      "learning_rate": 1.9401528233997684e-05,
+      "loss": 0.8688,
+      "step": 5831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9781005742412388,
+      "learning_rate": 1.940126819802915e-05,
+      "loss": 0.9299,
+      "step": 5832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.710887247443324,
+      "learning_rate": 1.9401008107323457e-05,
+      "loss": 0.9987,
+      "step": 5833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1193470513207355,
+      "learning_rate": 1.9400747961882115e-05,
+      "loss": 0.7923,
+      "step": 5834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6447140424199145,
+      "learning_rate": 1.9400487761706646e-05,
+      "loss": 0.9807,
+      "step": 5835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4806367979994521,
+      "learning_rate": 1.9400227506798557e-05,
+      "loss": 0.9144,
+      "step": 5836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6460316280607725,
+      "learning_rate": 1.939996719715937e-05,
+      "loss": 0.8431,
+      "step": 5837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6964741899320819,
+      "learning_rate": 1.93997068327906e-05,
+      "loss": 1.0099,
+      "step": 5838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5911510900181778,
+      "learning_rate": 1.9399446413693756e-05,
+      "loss": 0.8811,
+      "step": 5839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9496713082156287,
+      "learning_rate": 1.939918593987036e-05,
+      "loss": 0.8641,
+      "step": 5840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5140191981010873,
+      "learning_rate": 1.939892541132193e-05,
+      "loss": 0.9571,
+      "step": 5841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.80611537753462,
+      "learning_rate": 1.9398664828049983e-05,
+      "loss": 0.8799,
+      "step": 5842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6256550960644736,
+      "learning_rate": 1.939840419005603e-05,
+      "loss": 0.9924,
+      "step": 5843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.950523130118516,
+      "learning_rate": 1.9398143497341593e-05,
+      "loss": 0.9974,
+      "step": 5844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7106022264049525,
+      "learning_rate": 1.939788274990819e-05,
+      "loss": 0.8794,
+      "step": 5845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9004389764936058,
+      "learning_rate": 1.9397621947757338e-05,
+      "loss": 0.9133,
+      "step": 5846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7835056760290917,
+      "learning_rate": 1.9397361090890554e-05,
+      "loss": 0.9692,
+      "step": 5847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.588883866413301,
+      "learning_rate": 1.939710017930936e-05,
+      "loss": 0.9293,
+      "step": 5848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7933144355623951,
+      "learning_rate": 1.939683921301527e-05,
+      "loss": 0.8438,
+      "step": 5849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7000147274201212,
+      "learning_rate": 1.9396578192009812e-05,
+      "loss": 0.9752,
+      "step": 5850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8328474944783144,
+      "learning_rate": 1.9396317116294502e-05,
+      "loss": 1.019,
+      "step": 5851
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7090774832464153,
+      "learning_rate": 1.9396055985870855e-05,
+      "loss": 0.8453,
+      "step": 5852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.91339100695914,
+      "learning_rate": 1.93957948007404e-05,
+      "loss": 0.9337,
+      "step": 5853
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7719642934637332,
+      "learning_rate": 1.939553356090465e-05,
+      "loss": 1.0992,
+      "step": 5854
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6571697514857997,
+      "learning_rate": 1.9395272266365127e-05,
+      "loss": 1.0131,
+      "step": 5855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6320010014509783,
+      "learning_rate": 1.9395010917123355e-05,
+      "loss": 0.8591,
+      "step": 5856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9278286998289813,
+      "learning_rate": 1.9394749513180857e-05,
+      "loss": 0.969,
+      "step": 5857
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8356023742883432,
+      "learning_rate": 1.9394488054539153e-05,
+      "loss": 0.9159,
+      "step": 5858
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5139773542614945,
+      "learning_rate": 1.9394226541199764e-05,
+      "loss": 0.9452,
+      "step": 5859
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.777988683558711,
+      "learning_rate": 1.9393964973164216e-05,
+      "loss": 0.8564,
+      "step": 5860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6817441171641827,
+      "learning_rate": 1.939370335043403e-05,
+      "loss": 0.9449,
+      "step": 5861
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.043899515955633,
+      "learning_rate": 1.9393441673010725e-05,
+      "loss": 1.0568,
+      "step": 5862
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4711757954080378,
+      "learning_rate": 1.939317994089583e-05,
+      "loss": 0.9136,
+      "step": 5863
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8109390529497547,
+      "learning_rate": 1.939291815409087e-05,
+      "loss": 0.8866,
+      "step": 5864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9109777742866432,
+      "learning_rate": 1.9392656312597365e-05,
+      "loss": 0.7759,
+      "step": 5865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2245431472236272,
+      "learning_rate": 1.939239441641684e-05,
+      "loss": 0.8869,
+      "step": 5866
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5119531475970067,
+      "learning_rate": 1.9392132465550824e-05,
+      "loss": 0.8703,
+      "step": 5867
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8323543113520988,
+      "learning_rate": 1.939187046000084e-05,
+      "loss": 0.8066,
+      "step": 5868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.039025561884893,
+      "learning_rate": 1.939160839976841e-05,
+      "loss": 0.9276,
+      "step": 5869
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5550233877438482,
+      "learning_rate": 1.939134628485506e-05,
+      "loss": 0.8866,
+      "step": 5870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7964210672597387,
+      "learning_rate": 1.9391084115262323e-05,
+      "loss": 0.7157,
+      "step": 5871
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.692839713393052,
+      "learning_rate": 1.9390821890991723e-05,
+      "loss": 0.9497,
+      "step": 5872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5900443964119468,
+      "learning_rate": 1.9390559612044783e-05,
+      "loss": 0.9321,
+      "step": 5873
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.58345438114236,
+      "learning_rate": 1.9390297278423028e-05,
+      "loss": 1.057,
+      "step": 5874
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7052992271349852,
+      "learning_rate": 1.9390034890127993e-05,
+      "loss": 0.953,
+      "step": 5875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9891527086367171,
+      "learning_rate": 1.93897724471612e-05,
+      "loss": 0.9111,
+      "step": 5876
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2642198402151283,
+      "learning_rate": 1.938950994952418e-05,
+      "loss": 0.9173,
+      "step": 5877
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3972362609692364,
+      "learning_rate": 1.938924739721846e-05,
+      "loss": 0.96,
+      "step": 5878
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.831234857087499,
+      "learning_rate": 1.9388984790245575e-05,
+      "loss": 1.0112,
+      "step": 5879
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7843904358373435,
+      "learning_rate": 1.938872212860704e-05,
+      "loss": 0.9981,
+      "step": 5880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.901010974531216,
+      "learning_rate": 1.9388459412304397e-05,
+      "loss": 0.9038,
+      "step": 5881
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7898815845769523,
+      "learning_rate": 1.938819664133917e-05,
+      "loss": 0.7469,
+      "step": 5882
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9467975501579045,
+      "learning_rate": 1.938793381571289e-05,
+      "loss": 0.7683,
+      "step": 5883
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7412912118236583,
+      "learning_rate": 1.938767093542708e-05,
+      "loss": 0.9427,
+      "step": 5884
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0883516916119937,
+      "learning_rate": 1.938740800048329e-05,
+      "loss": 0.968,
+      "step": 5885
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5276422385393609,
+      "learning_rate": 1.938714501088303e-05,
+      "loss": 0.8194,
+      "step": 5886
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7750663593652036,
+      "learning_rate": 1.9386881966627844e-05,
+      "loss": 0.9545,
+      "step": 5887
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1025189414676175,
+      "learning_rate": 1.9386618867719258e-05,
+      "loss": 1.1565,
+      "step": 5888
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.542155988652206,
+      "learning_rate": 1.9386355714158806e-05,
+      "loss": 0.9698,
+      "step": 5889
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.112095613957966,
+      "learning_rate": 1.9386092505948018e-05,
+      "loss": 0.8528,
+      "step": 5890
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9735624492749837,
+      "learning_rate": 1.9385829243088426e-05,
+      "loss": 0.8378,
+      "step": 5891
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7869005966570377,
+      "learning_rate": 1.938556592558157e-05,
+      "loss": 0.7959,
+      "step": 5892
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8621706425817246,
+      "learning_rate": 1.9385302553428974e-05,
+      "loss": 0.8959,
+      "step": 5893
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8090643370782358,
+      "learning_rate": 1.9385039126632176e-05,
+      "loss": 0.9737,
+      "step": 5894
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.331921275664172,
+      "learning_rate": 1.9384775645192707e-05,
+      "loss": 0.8316,
+      "step": 5895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7440993293386062,
+      "learning_rate": 1.9384512109112104e-05,
+      "loss": 0.8464,
+      "step": 5896
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7714017878160541,
+      "learning_rate": 1.9384248518391903e-05,
+      "loss": 0.7173,
+      "step": 5897
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9083233732769826,
+      "learning_rate": 1.9383984873033636e-05,
+      "loss": 0.8559,
+      "step": 5898
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.897320867626837,
+      "learning_rate": 1.9383721173038836e-05,
+      "loss": 0.9734,
+      "step": 5899
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7372774363375196,
+      "learning_rate": 1.938345741840904e-05,
+      "loss": 0.8985,
+      "step": 5900
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7475524260989526,
+      "learning_rate": 1.9383193609145786e-05,
+      "loss": 1.1041,
+      "step": 5901
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8164503764930904,
+      "learning_rate": 1.9382929745250605e-05,
+      "loss": 0.6812,
+      "step": 5902
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.384252479833271,
+      "learning_rate": 1.938266582672504e-05,
+      "loss": 1.0322,
+      "step": 5903
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7255440730779072,
+      "learning_rate": 1.9382401853570624e-05,
+      "loss": 0.8689,
+      "step": 5904
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1371693171097066,
+      "learning_rate": 1.938213782578889e-05,
+      "loss": 1.0203,
+      "step": 5905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.681058964274289,
+      "learning_rate": 1.9381873743381384e-05,
+      "loss": 1.0396,
+      "step": 5906
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.64563768189886,
+      "learning_rate": 1.9381609606349635e-05,
+      "loss": 1.0413,
+      "step": 5907
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.543429997299268,
+      "learning_rate": 1.9381345414695188e-05,
+      "loss": 0.9459,
+      "step": 5908
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.939957634963097,
+      "learning_rate": 1.9381081168419577e-05,
+      "loss": 0.7357,
+      "step": 5909
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7716176830338275,
+      "learning_rate": 1.938081686752434e-05,
+      "loss": 1.0498,
+      "step": 5910
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6310089865746662,
+      "learning_rate": 1.938055251201102e-05,
+      "loss": 0.909,
+      "step": 5911
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7866294365245999,
+      "learning_rate": 1.9380288101881152e-05,
+      "loss": 0.786,
+      "step": 5912
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7594920781579184,
+      "learning_rate": 1.9380023637136272e-05,
+      "loss": 0.758,
+      "step": 5913
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7595394613522861,
+      "learning_rate": 1.937975911777793e-05,
+      "loss": 0.7007,
+      "step": 5914
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8742025170158918,
+      "learning_rate": 1.937949454380766e-05,
+      "loss": 0.8866,
+      "step": 5915
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.763705869138619,
+      "learning_rate": 1.9379229915227006e-05,
+      "loss": 0.8999,
+      "step": 5916
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5699757345838998,
+      "learning_rate": 1.9378965232037503e-05,
+      "loss": 0.87,
+      "step": 5917
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1944688523098805,
+      "learning_rate": 1.9378700494240697e-05,
+      "loss": 0.9612,
+      "step": 5918
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9467757199675078,
+      "learning_rate": 1.9378435701838127e-05,
+      "loss": 0.9015,
+      "step": 5919
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.669455973200291,
+      "learning_rate": 1.937817085483133e-05,
+      "loss": 0.9035,
+      "step": 5920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.519972284615471,
+      "learning_rate": 1.937790595322186e-05,
+      "loss": 0.8475,
+      "step": 5921
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8038232601808663,
+      "learning_rate": 1.937764099701125e-05,
+      "loss": 0.7928,
+      "step": 5922
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8275896254258905,
+      "learning_rate": 1.9377375986201047e-05,
+      "loss": 0.894,
+      "step": 5923
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4970874896008097,
+      "learning_rate": 1.9377110920792788e-05,
+      "loss": 0.7786,
+      "step": 5924
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4892647621508555,
+      "learning_rate": 1.937684580078802e-05,
+      "loss": 0.8146,
+      "step": 5925
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7027019331872881,
+      "learning_rate": 1.9376580626188292e-05,
+      "loss": 0.8812,
+      "step": 5926
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.000713511482929,
+      "learning_rate": 1.937631539699514e-05,
+      "loss": 0.7853,
+      "step": 5927
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5145000774244515,
+      "learning_rate": 1.9376050113210113e-05,
+      "loss": 0.8075,
+      "step": 5928
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7857559688256741,
+      "learning_rate": 1.9375784774834754e-05,
+      "loss": 0.7831,
+      "step": 5929
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7347915173838915,
+      "learning_rate": 1.9375519381870608e-05,
+      "loss": 0.968,
+      "step": 5930
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5365757383590184,
+      "learning_rate": 1.9375253934319217e-05,
+      "loss": 0.8232,
+      "step": 5931
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0978440991406004,
+      "learning_rate": 1.9374988432182133e-05,
+      "loss": 0.8811,
+      "step": 5932
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9934994106427615,
+      "learning_rate": 1.9374722875460894e-05,
+      "loss": 0.8124,
+      "step": 5933
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8573790460775872,
+      "learning_rate": 1.9374457264157055e-05,
+      "loss": 0.9531,
+      "step": 5934
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8036589455582838,
+      "learning_rate": 1.9374191598272154e-05,
+      "loss": 0.7607,
+      "step": 5935
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3007595588036214,
+      "learning_rate": 1.9373925877807744e-05,
+      "loss": 0.8715,
+      "step": 5936
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8247367109176234,
+      "learning_rate": 1.937366010276537e-05,
+      "loss": 0.7389,
+      "step": 5937
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7874319585270586,
+      "learning_rate": 1.9373394273146577e-05,
+      "loss": 0.7479,
+      "step": 5938
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5421602407916521,
+      "learning_rate": 1.9373128388952914e-05,
+      "loss": 0.9489,
+      "step": 5939
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.816501843721117,
+      "learning_rate": 1.9372862450185934e-05,
+      "loss": 0.8989,
+      "step": 5940
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1988792487432294,
+      "learning_rate": 1.937259645684718e-05,
+      "loss": 0.9556,
+      "step": 5941
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5328905338959977,
+      "learning_rate": 1.9372330408938203e-05,
+      "loss": 0.893,
+      "step": 5942
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2335867803663096,
+      "learning_rate": 1.937206430646055e-05,
+      "loss": 0.8899,
+      "step": 5943
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7884288876470676,
+      "learning_rate": 1.937179814941577e-05,
+      "loss": 0.8054,
+      "step": 5944
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5930376011039518,
+      "learning_rate": 1.9371531937805417e-05,
+      "loss": 0.9133,
+      "step": 5945
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8647103174338132,
+      "learning_rate": 1.9371265671631038e-05,
+      "loss": 0.773,
+      "step": 5946
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5550011150854148,
+      "learning_rate": 1.9370999350894178e-05,
+      "loss": 0.9952,
+      "step": 5947
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.621507866813614,
+      "learning_rate": 1.93707329755964e-05,
+      "loss": 0.9557,
+      "step": 5948
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.851531882670472,
+      "learning_rate": 1.9370466545739246e-05,
+      "loss": 0.8811,
+      "step": 5949
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6337399450048364,
+      "learning_rate": 1.9370200061324267e-05,
+      "loss": 1.015,
+      "step": 5950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4416223393415546,
+      "learning_rate": 1.9369933522353017e-05,
+      "loss": 0.8999,
+      "step": 5951
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8142830076787433,
+      "learning_rate": 1.936966692882705e-05,
+      "loss": 1.1055,
+      "step": 5952
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.851636497232011,
+      "learning_rate": 1.9369400280747913e-05,
+      "loss": 0.9154,
+      "step": 5953
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9445671840522807,
+      "learning_rate": 1.9369133578117162e-05,
+      "loss": 0.8264,
+      "step": 5954
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8104300468664207,
+      "learning_rate": 1.936886682093635e-05,
+      "loss": 0.9411,
+      "step": 5955
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7640996568926121,
+      "learning_rate": 1.9368600009207027e-05,
+      "loss": 0.7225,
+      "step": 5956
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8381751441497904,
+      "learning_rate": 1.9368333142930754e-05,
+      "loss": 0.8983,
+      "step": 5957
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6406042527337328,
+      "learning_rate": 1.9368066222109073e-05,
+      "loss": 0.9041,
+      "step": 5958
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9995519803288149,
+      "learning_rate": 1.936779924674355e-05,
+      "loss": 0.7121,
+      "step": 5959
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0673321847872455,
+      "learning_rate": 1.936753221683573e-05,
+      "loss": 0.8404,
+      "step": 5960
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.529259067508205,
+      "learning_rate": 1.936726513238717e-05,
+      "loss": 0.8713,
+      "step": 5961
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7863285949134927,
+      "learning_rate": 1.9366997993399432e-05,
+      "loss": 0.7725,
+      "step": 5962
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7820156717764409,
+      "learning_rate": 1.9366730799874063e-05,
+      "loss": 0.7593,
+      "step": 5963
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9633165913250754,
+      "learning_rate": 1.9366463551812624e-05,
+      "loss": 1.0547,
+      "step": 5964
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6859634687220244,
+      "learning_rate": 1.936619624921667e-05,
+      "loss": 0.8908,
+      "step": 5965
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.690717296334189,
+      "learning_rate": 1.936592889208775e-05,
+      "loss": 0.8078,
+      "step": 5966
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8979813393246614,
+      "learning_rate": 1.936566148042743e-05,
+      "loss": 0.9324,
+      "step": 5967
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.619694354080087,
+      "learning_rate": 1.9365394014237268e-05,
+      "loss": 0.8566,
+      "step": 5968
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1382153777212185,
+      "learning_rate": 1.9365126493518813e-05,
+      "loss": 0.7998,
+      "step": 5969
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2268942862854093,
+      "learning_rate": 1.9364858918273626e-05,
+      "loss": 1.0167,
+      "step": 5970
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5604353596025133,
+      "learning_rate": 1.9364591288503267e-05,
+      "loss": 0.9156,
+      "step": 5971
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6421051349375986,
+      "learning_rate": 1.9364323604209292e-05,
+      "loss": 0.8399,
+      "step": 5972
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7074231204611667,
+      "learning_rate": 1.936405586539326e-05,
+      "loss": 1.0105,
+      "step": 5973
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8030028209843016,
+      "learning_rate": 1.936378807205673e-05,
+      "loss": 0.9277,
+      "step": 5974
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9986270647901525,
+      "learning_rate": 1.9363520224201263e-05,
+      "loss": 0.9455,
+      "step": 5975
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8030546367712994,
+      "learning_rate": 1.9363252321828417e-05,
+      "loss": 0.7588,
+      "step": 5976
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5909004716658548,
+      "learning_rate": 1.936298436493975e-05,
+      "loss": 1.0436,
+      "step": 5977
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.725050669874304,
+      "learning_rate": 1.9362716353536825e-05,
+      "loss": 0.969,
+      "step": 5978
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.250197348760882,
+      "learning_rate": 1.93624482876212e-05,
+      "loss": 0.8323,
+      "step": 5979
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6594632253190578,
+      "learning_rate": 1.9362180167194435e-05,
+      "loss": 0.9372,
+      "step": 5980
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5987671557244685,
+      "learning_rate": 1.9361911992258096e-05,
+      "loss": 0.9936,
+      "step": 5981
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6510269987978299,
+      "learning_rate": 1.936164376281374e-05,
+      "loss": 0.9974,
+      "step": 5982
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6905910485449362,
+      "learning_rate": 1.9361375478862933e-05,
+      "loss": 0.9406,
+      "step": 5983
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9603968244685406,
+      "learning_rate": 1.9361107140407233e-05,
+      "loss": 0.8777,
+      "step": 5984
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6314481664589473,
+      "learning_rate": 1.9360838747448202e-05,
+      "loss": 0.9001,
+      "step": 5985
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7999530066901042,
+      "learning_rate": 1.9360570299987405e-05,
+      "loss": 0.9551,
+      "step": 5986
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9550171617467011,
+      "learning_rate": 1.9360301798026404e-05,
+      "loss": 0.9186,
+      "step": 5987
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0416014070272683,
+      "learning_rate": 1.9360033241566764e-05,
+      "loss": 0.9665,
+      "step": 5988
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7826784592698371,
+      "learning_rate": 1.9359764630610044e-05,
+      "loss": 0.7467,
+      "step": 5989
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1905511268169464,
+      "learning_rate": 1.9359495965157814e-05,
+      "loss": 0.8993,
+      "step": 5990
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6447922549282772,
+      "learning_rate": 1.9359227245211633e-05,
+      "loss": 0.9797,
+      "step": 5991
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6999267111957417,
+      "learning_rate": 1.935895847077307e-05,
+      "loss": 0.8164,
+      "step": 5992
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8491375325812964,
+      "learning_rate": 1.9358689641843686e-05,
+      "loss": 0.8761,
+      "step": 5993
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1283854601777046,
+      "learning_rate": 1.935842075842505e-05,
+      "loss": 0.9528,
+      "step": 5994
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6167320704765975,
+      "learning_rate": 1.9358151820518725e-05,
+      "loss": 0.873,
+      "step": 5995
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.699043815904354,
+      "learning_rate": 1.9357882828126278e-05,
+      "loss": 0.9686,
+      "step": 5996
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9134768272402445,
+      "learning_rate": 1.935761378124927e-05,
+      "loss": 0.8999,
+      "step": 5997
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5694379164541739,
+      "learning_rate": 1.935734467988928e-05,
+      "loss": 0.779,
+      "step": 5998
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7348448153375176,
+      "learning_rate": 1.935707552404786e-05,
+      "loss": 0.8317,
+      "step": 5999
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.704983677851632,
+      "learning_rate": 1.9356806313726584e-05,
+      "loss": 0.9455,
+      "step": 6000
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1614751894731183,
+      "learning_rate": 1.9356537048927023e-05,
+      "loss": 0.8745,
+      "step": 6001
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.762298598021869,
+      "learning_rate": 1.9356267729650738e-05,
+      "loss": 0.8709,
+      "step": 6002
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8272212710989049,
+      "learning_rate": 1.93559983558993e-05,
+      "loss": 0.7552,
+      "step": 6003
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.959612719604652,
+      "learning_rate": 1.9355728927674284e-05,
+      "loss": 0.808,
+      "step": 6004
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6300759026122376,
+      "learning_rate": 1.9355459444977245e-05,
+      "loss": 0.9903,
+      "step": 6005
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7530486445092823,
+      "learning_rate": 1.935518990780976e-05,
+      "loss": 0.7599,
+      "step": 6006
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8463048558940296,
+      "learning_rate": 1.93549203161734e-05,
+      "loss": 0.8016,
+      "step": 6007
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2206232115310036,
+      "learning_rate": 1.9354650670069732e-05,
+      "loss": 0.9027,
+      "step": 6008
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8251098640417288,
+      "learning_rate": 1.9354380969500326e-05,
+      "loss": 0.789,
+      "step": 6009
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9806152120507732,
+      "learning_rate": 1.9354111214466752e-05,
+      "loss": 0.7801,
+      "step": 6010
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0088926729841234,
+      "learning_rate": 1.9353841404970582e-05,
+      "loss": 0.8589,
+      "step": 6011
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5150580726540193,
+      "learning_rate": 1.9353571541013382e-05,
+      "loss": 0.9058,
+      "step": 6012
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4736619151777988,
+      "learning_rate": 1.935330162259673e-05,
+      "loss": 0.8654,
+      "step": 6013
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.788849659194644,
+      "learning_rate": 1.9353031649722196e-05,
+      "loss": 0.9853,
+      "step": 6014
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8350944821604136,
+      "learning_rate": 1.9352761622391348e-05,
+      "loss": 0.8711,
+      "step": 6015
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.964513467597511,
+      "learning_rate": 1.935249154060576e-05,
+      "loss": 0.9917,
+      "step": 6016
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.827151026955217,
+      "learning_rate": 1.9352221404367012e-05,
+      "loss": 0.823,
+      "step": 6017
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7602506891834953,
+      "learning_rate": 1.9351951213676664e-05,
+      "loss": 1.1082,
+      "step": 6018
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9945096241590456,
+      "learning_rate": 1.9351680968536298e-05,
+      "loss": 0.8615,
+      "step": 6019
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6019024919464313,
+      "learning_rate": 1.9351410668947482e-05,
+      "loss": 0.9638,
+      "step": 6020
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6320577410936217,
+      "learning_rate": 1.9351140314911798e-05,
+      "loss": 0.9676,
+      "step": 6021
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6462676685476567,
+      "learning_rate": 1.9350869906430807e-05,
+      "loss": 0.8583,
+      "step": 6022
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4237807166467509,
+      "learning_rate": 1.9350599443506097e-05,
+      "loss": 0.9607,
+      "step": 6023
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7956773684813356,
+      "learning_rate": 1.935032892613923e-05,
+      "loss": 0.8818,
+      "step": 6024
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7793207155859205,
+      "learning_rate": 1.9350058354331796e-05,
+      "loss": 0.9202,
+      "step": 6025
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6805618425167979,
+      "learning_rate": 1.9349787728085357e-05,
+      "loss": 0.9254,
+      "step": 6026
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7692033599687746,
+      "learning_rate": 1.9349517047401493e-05,
+      "loss": 0.7493,
+      "step": 6027
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7209998387737726,
+      "learning_rate": 1.9349246312281783e-05,
+      "loss": 0.8887,
+      "step": 6028
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9297469939145393,
+      "learning_rate": 1.9348975522727802e-05,
+      "loss": 0.9584,
+      "step": 6029
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1130668970115227,
+      "learning_rate": 1.9348704678741125e-05,
+      "loss": 0.9571,
+      "step": 6030
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1142176290205317,
+      "learning_rate": 1.934843378032333e-05,
+      "loss": 1.0018,
+      "step": 6031
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9636703494939935,
+      "learning_rate": 1.934816282747599e-05,
+      "loss": 0.8329,
+      "step": 6032
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8230733412408113,
+      "learning_rate": 1.9347891820200692e-05,
+      "loss": 0.8859,
+      "step": 6033
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.399517176513889,
+      "learning_rate": 1.9347620758499003e-05,
+      "loss": 0.9179,
+      "step": 6034
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0027673430353894,
+      "learning_rate": 1.9347349642372512e-05,
+      "loss": 1.0274,
+      "step": 6035
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3510017452232717,
+      "learning_rate": 1.9347078471822786e-05,
+      "loss": 0.8342,
+      "step": 6036
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7057913986708917,
+      "learning_rate": 1.9346807246851417e-05,
+      "loss": 0.9067,
+      "step": 6037
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7822319828565032,
+      "learning_rate": 1.9346535967459973e-05,
+      "loss": 0.9155,
+      "step": 6038
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9418476723515148,
+      "learning_rate": 1.9346264633650036e-05,
+      "loss": 0.9551,
+      "step": 6039
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6496302869382986,
+      "learning_rate": 1.934599324542319e-05,
+      "loss": 1.024,
+      "step": 6040
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4886623694630132,
+      "learning_rate": 1.934572180278101e-05,
+      "loss": 0.9062,
+      "step": 6041
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.503011589988399,
+      "learning_rate": 1.934545030572508e-05,
+      "loss": 0.8893,
+      "step": 6042
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.771970609045252,
+      "learning_rate": 1.934517875425698e-05,
+      "loss": 0.9022,
+      "step": 6043
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5957170268291323,
+      "learning_rate": 1.934490714837829e-05,
+      "loss": 1.0382,
+      "step": 6044
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0256182169871457,
+      "learning_rate": 1.9344635488090593e-05,
+      "loss": 0.9636,
+      "step": 6045
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8038942686255407,
+      "learning_rate": 1.934436377339547e-05,
+      "loss": 0.7158,
+      "step": 6046
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.644442731906356,
+      "learning_rate": 1.9344092004294503e-05,
+      "loss": 0.9057,
+      "step": 6047
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1440006016050943,
+      "learning_rate": 1.9343820180789273e-05,
+      "loss": 0.9584,
+      "step": 6048
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9182018075124916,
+      "learning_rate": 1.9343548302881364e-05,
+      "loss": 1.0182,
+      "step": 6049
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.71686911063111,
+      "learning_rate": 1.9343276370572357e-05,
+      "loss": 0.9787,
+      "step": 6050
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7965475581953945,
+      "learning_rate": 1.9343004383863838e-05,
+      "loss": 0.7327,
+      "step": 6051
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.878260050257555,
+      "learning_rate": 1.9342732342757392e-05,
+      "loss": 0.8362,
+      "step": 6052
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8511551425579453,
+      "learning_rate": 1.9342460247254596e-05,
+      "loss": 0.8925,
+      "step": 6053
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4495168083231422,
+      "learning_rate": 1.9342188097357042e-05,
+      "loss": 0.9197,
+      "step": 6054
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1658549951545756,
+      "learning_rate": 1.934191589306631e-05,
+      "loss": 0.9805,
+      "step": 6055
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7348703078231806,
+      "learning_rate": 1.9341643634383986e-05,
+      "loss": 1.0479,
+      "step": 6056
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8546689692125284,
+      "learning_rate": 1.9341371321311654e-05,
+      "loss": 1.0232,
+      "step": 6057
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8693331158403919,
+      "learning_rate": 1.9341098953850905e-05,
+      "loss": 0.9129,
+      "step": 6058
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8408215520472478,
+      "learning_rate": 1.934082653200332e-05,
+      "loss": 0.8256,
+      "step": 6059
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.44865455451044,
+      "learning_rate": 1.934055405577048e-05,
+      "loss": 0.8639,
+      "step": 6060
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.476815586475801,
+      "learning_rate": 1.9340281525153983e-05,
+      "loss": 0.7961,
+      "step": 6061
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.06181663260874,
+      "learning_rate": 1.9340008940155407e-05,
+      "loss": 0.9136,
+      "step": 6062
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3431848987290027,
+      "learning_rate": 1.933973630077634e-05,
+      "loss": 0.9398,
+      "step": 6063
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7900820155411835,
+      "learning_rate": 1.9339463607018375e-05,
+      "loss": 0.7364,
+      "step": 6064
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7224089147320487,
+      "learning_rate": 1.9339190858883096e-05,
+      "loss": 0.9502,
+      "step": 6065
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.715446530173856,
+      "learning_rate": 1.933891805637209e-05,
+      "loss": 0.9343,
+      "step": 6066
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7158358272441063,
+      "learning_rate": 1.9338645199486947e-05,
+      "loss": 1.0061,
+      "step": 6067
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7663080842753778,
+      "learning_rate": 1.9338372288229253e-05,
+      "loss": 1.1064,
+      "step": 6068
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7199513108369782,
+      "learning_rate": 1.93380993226006e-05,
+      "loss": 0.9321,
+      "step": 6069
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8096331476356153,
+      "learning_rate": 1.933782630260258e-05,
+      "loss": 0.8192,
+      "step": 6070
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7701415681937236,
+      "learning_rate": 1.9337553228236772e-05,
+      "loss": 0.7308,
+      "step": 6071
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.177998980560812,
+      "learning_rate": 1.933728009950478e-05,
+      "loss": 0.8841,
+      "step": 6072
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9300783506942227,
+      "learning_rate": 1.9337006916408182e-05,
+      "loss": 1.0031,
+      "step": 6073
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.793198264551209,
+      "learning_rate": 1.9336733678948577e-05,
+      "loss": 0.888,
+      "step": 6074
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3279021144630714,
+      "learning_rate": 1.933646038712755e-05,
+      "loss": 0.9679,
+      "step": 6075
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.99044584265373,
+      "learning_rate": 1.9336187040946695e-05,
+      "loss": 0.8452,
+      "step": 6076
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9733147083587075,
+      "learning_rate": 1.9335913640407606e-05,
+      "loss": 1.0253,
+      "step": 6077
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6920075893127686,
+      "learning_rate": 1.9335640185511872e-05,
+      "loss": 0.9105,
+      "step": 6078
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9506598470553018,
+      "learning_rate": 1.933536667626108e-05,
+      "loss": 0.953,
+      "step": 6079
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5127736483868424,
+      "learning_rate": 1.933509311265683e-05,
+      "loss": 0.8899,
+      "step": 6080
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.707282406758299,
+      "learning_rate": 1.9334819494700714e-05,
+      "loss": 0.8597,
+      "step": 6081
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8045258328167473,
+      "learning_rate": 1.9334545822394327e-05,
+      "loss": 0.7006,
+      "step": 6082
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7786912239965027,
+      "learning_rate": 1.9334272095739252e-05,
+      "loss": 1.0035,
+      "step": 6083
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8179109315018989,
+      "learning_rate": 1.9333998314737093e-05,
+      "loss": 0.8049,
+      "step": 6084
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6393910704311394,
+      "learning_rate": 1.9333724479389443e-05,
+      "loss": 1.0429,
+      "step": 6085
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7330520378290861,
+      "learning_rate": 1.9333450589697887e-05,
+      "loss": 0.8531,
+      "step": 6086
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7047115830592876,
+      "learning_rate": 1.9333176645664035e-05,
+      "loss": 1.029,
+      "step": 6087
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7784830204996347,
+      "learning_rate": 1.9332902647289468e-05,
+      "loss": 0.8444,
+      "step": 6088
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8810003715274466,
+      "learning_rate": 1.9332628594575792e-05,
+      "loss": 0.8474,
+      "step": 6089
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6569795828928864,
+      "learning_rate": 1.933235448752459e-05,
+      "loss": 0.8419,
+      "step": 6090
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.672343768437566,
+      "learning_rate": 1.9332080326137476e-05,
+      "loss": 0.8494,
+      "step": 6091
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5431254359594213,
+      "learning_rate": 1.9331806110416027e-05,
+      "loss": 1.0079,
+      "step": 6092
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.516456955198217,
+      "learning_rate": 1.9331531840361852e-05,
+      "loss": 0.9438,
+      "step": 6093
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.151669716443165,
+      "learning_rate": 1.9331257515976545e-05,
+      "loss": 0.8884,
+      "step": 6094
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6924884479975237,
+      "learning_rate": 1.93309831372617e-05,
+      "loss": 0.9023,
+      "step": 6095
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9760098046822545,
+      "learning_rate": 1.933070870421892e-05,
+      "loss": 0.8757,
+      "step": 6096
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8354916371248517,
+      "learning_rate": 1.9330434216849796e-05,
+      "loss": 0.7604,
+      "step": 6097
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6626732856295101,
+      "learning_rate": 1.9330159675155933e-05,
+      "loss": 0.8133,
+      "step": 6098
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8363489599310281,
+      "learning_rate": 1.9329885079138927e-05,
+      "loss": 0.8194,
+      "step": 6099
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9752317189614685,
+      "learning_rate": 1.9329610428800376e-05,
+      "loss": 0.9827,
+      "step": 6100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8697019518641218,
+      "learning_rate": 1.9329335724141877e-05,
+      "loss": 0.7829,
+      "step": 6101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6346417863545835,
+      "learning_rate": 1.9329060965165032e-05,
+      "loss": 0.9386,
+      "step": 6102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.23691267383781,
+      "learning_rate": 1.9328786151871444e-05,
+      "loss": 0.9958,
+      "step": 6103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6752277228906758,
+      "learning_rate": 1.9328511284262705e-05,
+      "loss": 0.9246,
+      "step": 6104
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9016122206869368,
+      "learning_rate": 1.9328236362340425e-05,
+      "loss": 0.887,
+      "step": 6105
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.678485840065173,
+      "learning_rate": 1.9327961386106198e-05,
+      "loss": 0.9361,
+      "step": 6106
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.798441577294215,
+      "learning_rate": 1.9327686355561622e-05,
+      "loss": 0.9442,
+      "step": 6107
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5724845380459231,
+      "learning_rate": 1.932741127070831e-05,
+      "loss": 0.9339,
+      "step": 6108
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7165280103157585,
+      "learning_rate": 1.932713613154785e-05,
+      "loss": 0.9391,
+      "step": 6109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.672587366127263,
+      "learning_rate": 1.9326860938081854e-05,
+      "loss": 0.8678,
+      "step": 6110
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9729260010574114,
+      "learning_rate": 1.932658569031192e-05,
+      "loss": 0.9356,
+      "step": 6111
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.659351687626217,
+      "learning_rate": 1.9326310388239654e-05,
+      "loss": 0.9226,
+      "step": 6112
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.669019908713921,
+      "learning_rate": 1.9326035031866653e-05,
+      "loss": 0.9019,
+      "step": 6113
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7216423467122068,
+      "learning_rate": 1.9325759621194528e-05,
+      "loss": 0.8785,
+      "step": 6114
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6115807533907491,
+      "learning_rate": 1.9325484156224872e-05,
+      "loss": 0.883,
+      "step": 6115
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.857860635847524,
+      "learning_rate": 1.9325208636959297e-05,
+      "loss": 0.9577,
+      "step": 6116
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.119383661225691,
+      "learning_rate": 1.932493306339941e-05,
+      "loss": 0.9313,
+      "step": 6117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0622181211677506,
+      "learning_rate": 1.9324657435546802e-05,
+      "loss": 0.8899,
+      "step": 6118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4826793955854505,
+      "learning_rate": 1.932438175340309e-05,
+      "loss": 0.897,
+      "step": 6119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7481340672781707,
+      "learning_rate": 1.9324106016969875e-05,
+      "loss": 0.9259,
+      "step": 6120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8142934168533302,
+      "learning_rate": 1.9323830226248763e-05,
+      "loss": 0.8868,
+      "step": 6121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.593481521864774,
+      "learning_rate": 1.932355438124136e-05,
+      "loss": 0.9077,
+      "step": 6122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7527162842035318,
+      "learning_rate": 1.9323278481949274e-05,
+      "loss": 0.8515,
+      "step": 6123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9717714845607277,
+      "learning_rate": 1.9323002528374104e-05,
+      "loss": 0.9513,
+      "step": 6124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8450746523377057,
+      "learning_rate": 1.9322726520517465e-05,
+      "loss": 0.8885,
+      "step": 6125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.82862728518432,
+      "learning_rate": 1.932245045838096e-05,
+      "loss": 0.907,
+      "step": 6126
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6348292748097597,
+      "learning_rate": 1.9322174341966196e-05,
+      "loss": 1.0178,
+      "step": 6127
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9063538175478385,
+      "learning_rate": 1.9321898171274785e-05,
+      "loss": 0.8063,
+      "step": 6128
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7676443738272036,
+      "learning_rate": 1.9321621946308328e-05,
+      "loss": 0.9886,
+      "step": 6129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0824074640871246,
+      "learning_rate": 1.9321345667068435e-05,
+      "loss": 0.9615,
+      "step": 6130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6888494632410604,
+      "learning_rate": 1.9321069333556722e-05,
+      "loss": 0.993,
+      "step": 6131
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1509093343615353,
+      "learning_rate": 1.9320792945774788e-05,
+      "loss": 0.8932,
+      "step": 6132
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.755786124037853,
+      "learning_rate": 1.932051650372425e-05,
+      "loss": 0.9492,
+      "step": 6133
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8306354568888987,
+      "learning_rate": 1.932024000740671e-05,
+      "loss": 1.0071,
+      "step": 6134
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8074720950676059,
+      "learning_rate": 1.9319963456823788e-05,
+      "loss": 0.7394,
+      "step": 6135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9998571449308744,
+      "learning_rate": 1.931968685197708e-05,
+      "loss": 0.8829,
+      "step": 6136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7676800094201663,
+      "learning_rate": 1.931941019286821e-05,
+      "loss": 0.7976,
+      "step": 6137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6351383486052389,
+      "learning_rate": 1.931913347949878e-05,
+      "loss": 0.9421,
+      "step": 6138
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7151969053029334,
+      "learning_rate": 1.931885671187041e-05,
+      "loss": 0.9311,
+      "step": 6139
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.681860603221807,
+      "learning_rate": 1.9318579889984705e-05,
+      "loss": 0.836,
+      "step": 6140
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6108382406751443,
+      "learning_rate": 1.9318303013843272e-05,
+      "loss": 0.9715,
+      "step": 6141
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.74991367712624,
+      "learning_rate": 1.9318026083447733e-05,
+      "loss": 0.7628,
+      "step": 6142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5071867596339712,
+      "learning_rate": 1.9317749098799695e-05,
+      "loss": 0.8953,
+      "step": 6143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.525555960850665,
+      "learning_rate": 1.9317472059900773e-05,
+      "loss": 0.848,
+      "step": 6144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8015029053812857,
+      "learning_rate": 1.9317194966752576e-05,
+      "loss": 0.9924,
+      "step": 6145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0606728925125375,
+      "learning_rate": 1.9316917819356723e-05,
+      "loss": 0.8537,
+      "step": 6146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7689655616949764,
+      "learning_rate": 1.9316640617714823e-05,
+      "loss": 0.7184,
+      "step": 6147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5796641616422027,
+      "learning_rate": 1.9316363361828493e-05,
+      "loss": 0.9063,
+      "step": 6148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.144511501812707,
+      "learning_rate": 1.9316086051699345e-05,
+      "loss": 0.9186,
+      "step": 6149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6215184700412382,
+      "learning_rate": 1.9315808687328996e-05,
+      "loss": 0.8766,
+      "step": 6150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5115849201358316,
+      "learning_rate": 1.9315531268719058e-05,
+      "loss": 0.9447,
+      "step": 6151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2883186581823303,
+      "learning_rate": 1.931525379587115e-05,
+      "loss": 1.0474,
+      "step": 6152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9379193009290578,
+      "learning_rate": 1.9314976268786883e-05,
+      "loss": 0.9181,
+      "step": 6153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7930888569989415,
+      "learning_rate": 1.9314698687467876e-05,
+      "loss": 0.7383,
+      "step": 6154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7885550252671991,
+      "learning_rate": 1.9314421051915744e-05,
+      "loss": 0.7451,
+      "step": 6155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5779312423224332,
+      "learning_rate": 1.93141433621321e-05,
+      "loss": 0.9151,
+      "step": 6156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7234671272972377,
+      "learning_rate": 1.931386561811857e-05,
+      "loss": 0.8662,
+      "step": 6157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7286570161505166,
+      "learning_rate": 1.9313587819876765e-05,
+      "loss": 1.0258,
+      "step": 6158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9729899193431377,
+      "learning_rate": 1.93133099674083e-05,
+      "loss": 0.9325,
+      "step": 6159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7852426834462125,
+      "learning_rate": 1.9313032060714797e-05,
+      "loss": 0.9715,
+      "step": 6160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5058780035008992,
+      "learning_rate": 1.9312754099797876e-05,
+      "loss": 1.0393,
+      "step": 6161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8400570460976495,
+      "learning_rate": 1.931247608465915e-05,
+      "loss": 1.0212,
+      "step": 6162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5716459572154837,
+      "learning_rate": 1.931219801530024e-05,
+      "loss": 0.7586,
+      "step": 6163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.229003941843047,
+      "learning_rate": 1.9311919891722764e-05,
+      "loss": 0.9424,
+      "step": 6164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.026777534056569,
+      "learning_rate": 1.9311641713928342e-05,
+      "loss": 0.9401,
+      "step": 6165
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7185464591847486,
+      "learning_rate": 1.9311363481918594e-05,
+      "loss": 0.9523,
+      "step": 6166
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8543328892518728,
+      "learning_rate": 1.9311085195695142e-05,
+      "loss": 0.7318,
+      "step": 6167
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6847245114870455,
+      "learning_rate": 1.9310806855259604e-05,
+      "loss": 0.8976,
+      "step": 6168
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7997844592180838,
+      "learning_rate": 1.93105284606136e-05,
+      "loss": 0.7199,
+      "step": 6169
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4992165691144685,
+      "learning_rate": 1.9310250011758752e-05,
+      "loss": 0.9936,
+      "step": 6170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.017916068001455,
+      "learning_rate": 1.9309971508696675e-05,
+      "loss": 0.8341,
+      "step": 6171
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7659687562702704,
+      "learning_rate": 1.9309692951429003e-05,
+      "loss": 0.9848,
+      "step": 6172
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6041435243533466,
+      "learning_rate": 1.9309414339957355e-05,
+      "loss": 0.8825,
+      "step": 6173
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7624019854589037,
+      "learning_rate": 1.9309135674283343e-05,
+      "loss": 0.8714,
+      "step": 6174
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.212555471991551,
+      "learning_rate": 1.9308856954408594e-05,
+      "loss": 0.9899,
+      "step": 6175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9572243738741872,
+      "learning_rate": 1.9308578180334737e-05,
+      "loss": 0.9887,
+      "step": 6176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7283463362893151,
+      "learning_rate": 1.930829935206339e-05,
+      "loss": 0.6978,
+      "step": 6177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6296665501799532,
+      "learning_rate": 1.9308020469596177e-05,
+      "loss": 0.9131,
+      "step": 6178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7568986393077606,
+      "learning_rate": 1.930774153293472e-05,
+      "loss": 0.9006,
+      "step": 6179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.870073449286941,
+      "learning_rate": 1.9307462542080646e-05,
+      "loss": 0.887,
+      "step": 6180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4190671155315067,
+      "learning_rate": 1.930718349703558e-05,
+      "loss": 0.8695,
+      "step": 6181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9452544116834858,
+      "learning_rate": 1.9306904397801144e-05,
+      "loss": 0.9769,
+      "step": 6182
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9940612483058966,
+      "learning_rate": 1.9306625244378964e-05,
+      "loss": 0.8011,
+      "step": 6183
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6650828762141983,
+      "learning_rate": 1.9306346036770667e-05,
+      "loss": 0.9219,
+      "step": 6184
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.726591587731879,
+      "learning_rate": 1.9306066774977875e-05,
+      "loss": 0.9332,
+      "step": 6185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9591091281101907,
+      "learning_rate": 1.930578745900222e-05,
+      "loss": 0.935,
+      "step": 6186
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2595540259913864,
+      "learning_rate": 1.930550808884532e-05,
+      "loss": 0.7012,
+      "step": 6187
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.709406305084124,
+      "learning_rate": 1.9305228664508805e-05,
+      "loss": 0.8923,
+      "step": 6188
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8007502730800815,
+      "learning_rate": 1.9304949185994303e-05,
+      "loss": 0.8284,
+      "step": 6189
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7562570660388686,
+      "learning_rate": 1.9304669653303444e-05,
+      "loss": 0.8535,
+      "step": 6190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9153367628287952,
+      "learning_rate": 1.930439006643785e-05,
+      "loss": 0.9536,
+      "step": 6191
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9654122251977841,
+      "learning_rate": 1.930411042539915e-05,
+      "loss": 0.8049,
+      "step": 6192
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7843611469307926,
+      "learning_rate": 1.930383073018898e-05,
+      "loss": 0.8043,
+      "step": 6193
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8867707932536573,
+      "learning_rate": 1.9303550980808956e-05,
+      "loss": 0.8332,
+      "step": 6194
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8972400130604028,
+      "learning_rate": 1.9303271177260715e-05,
+      "loss": 0.8518,
+      "step": 6195
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6007823022485335,
+      "learning_rate": 1.9302991319545883e-05,
+      "loss": 1.0862,
+      "step": 6196
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6507883324538313,
+      "learning_rate": 1.930271140766609e-05,
+      "loss": 0.9591,
+      "step": 6197
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8175474222805559,
+      "learning_rate": 1.930243144162297e-05,
+      "loss": 0.8976,
+      "step": 6198
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6981557858845118,
+      "learning_rate": 1.930215142141814e-05,
+      "loss": 1.0564,
+      "step": 6199
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6511368477735444,
+      "learning_rate": 1.930187134705325e-05,
+      "loss": 0.9204,
+      "step": 6200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6410405021698722,
+      "learning_rate": 1.930159121852991e-05,
+      "loss": 0.9172,
+      "step": 6201
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7099852520054395,
+      "learning_rate": 1.930131103584977e-05,
+      "loss": 0.7924,
+      "step": 6202
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6508701309985196,
+      "learning_rate": 1.9301030799014445e-05,
+      "loss": 0.8729,
+      "step": 6203
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6713208985504837,
+      "learning_rate": 1.9300750508025577e-05,
+      "loss": 0.9393,
+      "step": 6204
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9914965142993128,
+      "learning_rate": 1.93004701628848e-05,
+      "loss": 0.8849,
+      "step": 6205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7839881956662393,
+      "learning_rate": 1.9300189763593736e-05,
+      "loss": 0.7275,
+      "step": 6206
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8197124982886146,
+      "learning_rate": 1.9299909310154025e-05,
+      "loss": 0.9514,
+      "step": 6207
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7889794395687584,
+      "learning_rate": 1.9299628802567293e-05,
+      "loss": 0.9756,
+      "step": 6208
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9062281550626063,
+      "learning_rate": 1.9299348240835182e-05,
+      "loss": 0.8552,
+      "step": 6209
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4317739429986647,
+      "learning_rate": 1.9299067624959322e-05,
+      "loss": 0.8132,
+      "step": 6210
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.062289660555459,
+      "learning_rate": 1.9298786954941344e-05,
+      "loss": 0.8824,
+      "step": 6211
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0695844842374034,
+      "learning_rate": 1.9298506230782886e-05,
+      "loss": 0.8923,
+      "step": 6212
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7892583911652988,
+      "learning_rate": 1.929822545248558e-05,
+      "loss": 0.943,
+      "step": 6213
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.032135582923966,
+      "learning_rate": 1.9297944620051063e-05,
+      "loss": 0.9301,
+      "step": 6214
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.508588134131592,
+      "learning_rate": 1.9297663733480966e-05,
+      "loss": 0.9359,
+      "step": 6215
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5564180931386398,
+      "learning_rate": 1.929738279277693e-05,
+      "loss": 0.8469,
+      "step": 6216
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.791110400247542,
+      "learning_rate": 1.9297101797940586e-05,
+      "loss": 0.8941,
+      "step": 6217
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8115627471882183,
+      "learning_rate": 1.9296820748973572e-05,
+      "loss": 0.9347,
+      "step": 6218
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7163476603431367,
+      "learning_rate": 1.9296539645877524e-05,
+      "loss": 0.9409,
+      "step": 6219
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5367926000576044,
+      "learning_rate": 1.9296258488654082e-05,
+      "loss": 0.9382,
+      "step": 6220
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.648972057915115,
+      "learning_rate": 1.9295977277304875e-05,
+      "loss": 0.9066,
+      "step": 6221
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.66717359134965,
+      "learning_rate": 1.9295696011831547e-05,
+      "loss": 0.9062,
+      "step": 6222
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.722177397901838,
+      "learning_rate": 1.9295414692235738e-05,
+      "loss": 1.0333,
+      "step": 6223
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7631871539275098,
+      "learning_rate": 1.9295133318519077e-05,
+      "loss": 0.7255,
+      "step": 6224
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5002642193118738,
+      "learning_rate": 1.9294851890683208e-05,
+      "loss": 0.964,
+      "step": 6225
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8126624776934711,
+      "learning_rate": 1.929457040872977e-05,
+      "loss": 0.7352,
+      "step": 6226
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7738441238247539,
+      "learning_rate": 1.92942888726604e-05,
+      "loss": 0.7593,
+      "step": 6227
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.61245074100175,
+      "learning_rate": 1.9294007282476737e-05,
+      "loss": 0.8403,
+      "step": 6228
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.561381492206762,
+      "learning_rate": 1.929372563818042e-05,
+      "loss": 0.978,
+      "step": 6229
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6247599724216693,
+      "learning_rate": 1.9293443939773092e-05,
+      "loss": 0.9803,
+      "step": 6230
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.616888679437041,
+      "learning_rate": 1.9293162187256392e-05,
+      "loss": 0.8953,
+      "step": 6231
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.542530276994592,
+      "learning_rate": 1.9292880380631956e-05,
+      "loss": 0.9812,
+      "step": 6232
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0067889804984973,
+      "learning_rate": 1.929259851990143e-05,
+      "loss": 0.8575,
+      "step": 6233
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5495295637760185,
+      "learning_rate": 1.9292316605066455e-05,
+      "loss": 0.9188,
+      "step": 6234
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.68730158002875,
+      "learning_rate": 1.929203463612867e-05,
+      "loss": 0.8186,
+      "step": 6235
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.989208084545542,
+      "learning_rate": 1.9291752613089717e-05,
+      "loss": 1.055,
+      "step": 6236
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8059139858863396,
+      "learning_rate": 1.929147053595124e-05,
+      "loss": 0.8923,
+      "step": 6237
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7255351835166326,
+      "learning_rate": 1.9291188404714876e-05,
+      "loss": 0.799,
+      "step": 6238
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.968167471140407,
+      "learning_rate": 1.9290906219382277e-05,
+      "loss": 0.8909,
+      "step": 6239
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.261007718966533,
+      "learning_rate": 1.9290623979955077e-05,
+      "loss": 0.8984,
+      "step": 6240
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.459805652289001,
+      "learning_rate": 1.9290341686434924e-05,
+      "loss": 0.8874,
+      "step": 6241
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6825955520337328,
+      "learning_rate": 1.929005933882346e-05,
+      "loss": 0.9349,
+      "step": 6242
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.684731483489434,
+      "learning_rate": 1.928977693712233e-05,
+      "loss": 1.0776,
+      "step": 6243
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.692554270625254,
+      "learning_rate": 1.9289494481333177e-05,
+      "loss": 0.9273,
+      "step": 6244
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1469350889179957,
+      "learning_rate": 1.9289211971457644e-05,
+      "loss": 0.9409,
+      "step": 6245
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9950461855513457,
+      "learning_rate": 1.928892940749738e-05,
+      "loss": 0.962,
+      "step": 6246
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7441541716714604,
+      "learning_rate": 1.9288646789454032e-05,
+      "loss": 0.9908,
+      "step": 6247
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.855522241433085,
+      "learning_rate": 1.9288364117329233e-05,
+      "loss": 0.9203,
+      "step": 6248
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4297129380508158,
+      "learning_rate": 1.9288081391124643e-05,
+      "loss": 0.9859,
+      "step": 6249
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4815745406957357,
+      "learning_rate": 1.9287798610841903e-05,
+      "loss": 0.9811,
+      "step": 6250
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.824456807310504,
+      "learning_rate": 1.928751577648266e-05,
+      "loss": 0.7288,
+      "step": 6251
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6435494700311895,
+      "learning_rate": 1.9287232888048554e-05,
+      "loss": 0.8112,
+      "step": 6252
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4719459545196936,
+      "learning_rate": 1.9286949945541243e-05,
+      "loss": 0.8656,
+      "step": 6253
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6750482887380622,
+      "learning_rate": 1.9286666948962368e-05,
+      "loss": 1.0146,
+      "step": 6254
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7190784836910282,
+      "learning_rate": 1.9286383898313578e-05,
+      "loss": 0.9382,
+      "step": 6255
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2215930608351044,
+      "learning_rate": 1.9286100793596523e-05,
+      "loss": 0.9498,
+      "step": 6256
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7626719406373546,
+      "learning_rate": 1.9285817634812844e-05,
+      "loss": 0.918,
+      "step": 6257
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8091230031314687,
+      "learning_rate": 1.92855344219642e-05,
+      "loss": 0.9606,
+      "step": 6258
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5916431977388348,
+      "learning_rate": 1.9285251155052233e-05,
+      "loss": 0.9052,
+      "step": 6259
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7815858126278783,
+      "learning_rate": 1.928496783407859e-05,
+      "loss": 1.0416,
+      "step": 6260
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7729290805143388,
+      "learning_rate": 1.928468445904493e-05,
+      "loss": 0.7048,
+      "step": 6261
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7929991785135477,
+      "learning_rate": 1.9284401029952898e-05,
+      "loss": 0.9391,
+      "step": 6262
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.801094347313833,
+      "learning_rate": 1.928411754680414e-05,
+      "loss": 1.0149,
+      "step": 6263
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0823229260340668,
+      "learning_rate": 1.928383400960031e-05,
+      "loss": 0.9676,
+      "step": 6264
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.756780006645932,
+      "learning_rate": 1.9283550418343063e-05,
+      "loss": 0.8643,
+      "step": 6265
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.109666540387926,
+      "learning_rate": 1.9283266773034047e-05,
+      "loss": 0.9633,
+      "step": 6266
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.776692402474345,
+      "learning_rate": 1.928298307367491e-05,
+      "loss": 0.7805,
+      "step": 6267
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7709793063270776,
+      "learning_rate": 1.928269932026731e-05,
+      "loss": 0.88,
+      "step": 6268
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8388468286817539,
+      "learning_rate": 1.928241551281289e-05,
+      "loss": 0.8141,
+      "step": 6269
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.620576204383095,
+      "learning_rate": 1.928213165131331e-05,
+      "loss": 0.8234,
+      "step": 6270
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.909873311027714,
+      "learning_rate": 1.9281847735770224e-05,
+      "loss": 0.8974,
+      "step": 6271
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7873035593411867,
+      "learning_rate": 1.928156376618528e-05,
+      "loss": 0.7574,
+      "step": 6272
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0648536181162216,
+      "learning_rate": 1.9281279742560134e-05,
+      "loss": 0.9088,
+      "step": 6273
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.718793977640799,
+      "learning_rate": 1.9280995664896435e-05,
+      "loss": 0.9844,
+      "step": 6274
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8787806009733528,
+      "learning_rate": 1.9280711533195843e-05,
+      "loss": 1.1067,
+      "step": 6275
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6053734189306212,
+      "learning_rate": 1.9280427347460013e-05,
+      "loss": 0.91,
+      "step": 6276
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4149472737844575,
+      "learning_rate": 1.9280143107690593e-05,
+      "loss": 0.948,
+      "step": 6277
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9366549224306087,
+      "learning_rate": 1.9279858813889246e-05,
+      "loss": 0.8542,
+      "step": 6278
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.08823185433039,
+      "learning_rate": 1.927957446605762e-05,
+      "loss": 0.9984,
+      "step": 6279
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9855265664403863,
+      "learning_rate": 1.9279290064197375e-05,
+      "loss": 0.8137,
+      "step": 6280
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9504215158270874,
+      "learning_rate": 1.9279005608310165e-05,
+      "loss": 0.8667,
+      "step": 6281
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8201254240552802,
+      "learning_rate": 1.9278721098397645e-05,
+      "loss": 1.0627,
+      "step": 6282
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6493778329415394,
+      "learning_rate": 1.9278436534461476e-05,
+      "loss": 0.7846,
+      "step": 6283
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7490494969158261,
+      "learning_rate": 1.927815191650331e-05,
+      "loss": 1.0517,
+      "step": 6284
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8083716431350805,
+      "learning_rate": 1.9277867244524805e-05,
+      "loss": 0.7994,
+      "step": 6285
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6604704631222256,
+      "learning_rate": 1.927758251852762e-05,
+      "loss": 0.8846,
+      "step": 6286
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8154162405910375,
+      "learning_rate": 1.9277297738513415e-05,
+      "loss": 0.8927,
+      "step": 6287
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8968500246050457,
+      "learning_rate": 1.9277012904483846e-05,
+      "loss": 0.944,
+      "step": 6288
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7115738550592372,
+      "learning_rate": 1.927672801644057e-05,
+      "loss": 0.9386,
+      "step": 6289
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.354330839095586,
+      "learning_rate": 1.9276443074385246e-05,
+      "loss": 1.1388,
+      "step": 6290
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.026112103492996,
+      "learning_rate": 1.927615807831953e-05,
+      "loss": 0.865,
+      "step": 6291
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7478632674180125,
+      "learning_rate": 1.927587302824509e-05,
+      "loss": 0.7886,
+      "step": 6292
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8841884544930687,
+      "learning_rate": 1.9275587924163578e-05,
+      "loss": 0.8948,
+      "step": 6293
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7434672182286023,
+      "learning_rate": 1.927530276607666e-05,
+      "loss": 0.8138,
+      "step": 6294
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8741622013868433,
+      "learning_rate": 1.9275017553985986e-05,
+      "loss": 1.022,
+      "step": 6295
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4972907617191504,
+      "learning_rate": 1.927473228789323e-05,
+      "loss": 0.8056,
+      "step": 6296
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8350062951586944,
+      "learning_rate": 1.927444696780004e-05,
+      "loss": 0.7418,
+      "step": 6297
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7984892546545246,
+      "learning_rate": 1.9274161593708086e-05,
+      "loss": 1.077,
+      "step": 6298
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7655461670121253,
+      "learning_rate": 1.927387616561903e-05,
+      "loss": 0.7443,
+      "step": 6299
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8025652007143426,
+      "learning_rate": 1.927359068353453e-05,
+      "loss": 0.7308,
+      "step": 6300
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7135560401445589,
+      "learning_rate": 1.9273305147456247e-05,
+      "loss": 0.8043,
+      "step": 6301
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.552673120272585,
+      "learning_rate": 1.9273019557385845e-05,
+      "loss": 1.0047,
+      "step": 6302
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8243347559834802,
+      "learning_rate": 1.927273391332499e-05,
+      "loss": 0.9595,
+      "step": 6303
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.884706858552018,
+      "learning_rate": 1.927244821527534e-05,
+      "loss": 0.8852,
+      "step": 6304
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.092190076905518,
+      "learning_rate": 1.9272162463238562e-05,
+      "loss": 0.8676,
+      "step": 6305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6405772774502254,
+      "learning_rate": 1.927187665721632e-05,
+      "loss": 0.9894,
+      "step": 6306
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7357659840597122,
+      "learning_rate": 1.9271590797210276e-05,
+      "loss": 0.7722,
+      "step": 6307
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8450492535599907,
+      "learning_rate": 1.9271304883222095e-05,
+      "loss": 1.0154,
+      "step": 6308
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.649007511506769,
+      "learning_rate": 1.927101891525344e-05,
+      "loss": 0.8467,
+      "step": 6309
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5311474269492937,
+      "learning_rate": 1.927073289330598e-05,
+      "loss": 1.1346,
+      "step": 6310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.159334890583792,
+      "learning_rate": 1.9270446817381377e-05,
+      "loss": 0.7812,
+      "step": 6311
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7657804735737662,
+      "learning_rate": 1.9270160687481296e-05,
+      "loss": 0.7408,
+      "step": 6312
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.179179550156821,
+      "learning_rate": 1.926987450360741e-05,
+      "loss": 1.0003,
+      "step": 6313
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7839604278089181,
+      "learning_rate": 1.9269588265761374e-05,
+      "loss": 0.7464,
+      "step": 6314
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0191198402546657,
+      "learning_rate": 1.9269301973944864e-05,
+      "loss": 0.9364,
+      "step": 6315
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1688603434215645,
+      "learning_rate": 1.926901562815954e-05,
+      "loss": 0.7332,
+      "step": 6316
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6749428154987664,
+      "learning_rate": 1.9268729228407076e-05,
+      "loss": 0.8999,
+      "step": 6317
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7297135086613925,
+      "learning_rate": 1.9268442774689132e-05,
+      "loss": 0.6874,
+      "step": 6318
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5990649437510316,
+      "learning_rate": 1.9268156267007383e-05,
+      "loss": 0.9849,
+      "step": 6319
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7843406929749452,
+      "learning_rate": 1.9267869705363493e-05,
+      "loss": 0.6958,
+      "step": 6320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.965851824035777,
+      "learning_rate": 1.926758308975913e-05,
+      "loss": 0.9817,
+      "step": 6321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1552470182206984,
+      "learning_rate": 1.9267296420195966e-05,
+      "loss": 0.8224,
+      "step": 6322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.806073809776275,
+      "learning_rate": 1.926700969667567e-05,
+      "loss": 0.7498,
+      "step": 6323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8292680421051857,
+      "learning_rate": 1.9266722919199907e-05,
+      "loss": 0.8213,
+      "step": 6324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.590610652811131,
+      "learning_rate": 1.926643608777035e-05,
+      "loss": 0.8536,
+      "step": 6325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7485311290120147,
+      "learning_rate": 1.926614920238867e-05,
+      "loss": 0.9752,
+      "step": 6326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8965551370384737,
+      "learning_rate": 1.926586226305653e-05,
+      "loss": 0.9687,
+      "step": 6327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.645020469266551,
+      "learning_rate": 1.9265575269775612e-05,
+      "loss": 0.9301,
+      "step": 6328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6385303629340109,
+      "learning_rate": 1.926528822254758e-05,
+      "loss": 0.8627,
+      "step": 6329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5096179444552953,
+      "learning_rate": 1.9265001121374108e-05,
+      "loss": 0.9671,
+      "step": 6330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7283594574465608,
+      "learning_rate": 1.9264713966256864e-05,
+      "loss": 0.928,
+      "step": 6331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.988885223295562,
+      "learning_rate": 1.9264426757197523e-05,
+      "loss": 0.8772,
+      "step": 6332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7846775520911489,
+      "learning_rate": 1.9264139494197755e-05,
+      "loss": 0.7871,
+      "step": 6333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7058670671456428,
+      "learning_rate": 1.9263852177259234e-05,
+      "loss": 0.8467,
+      "step": 6334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.594831606934272,
+      "learning_rate": 1.9263564806383637e-05,
+      "loss": 0.8548,
+      "step": 6335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1026324046100524,
+      "learning_rate": 1.9263277381572628e-05,
+      "loss": 0.7815,
+      "step": 6336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8899273148321902,
+      "learning_rate": 1.926298990282789e-05,
+      "loss": 0.917,
+      "step": 6337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6494056730278648,
+      "learning_rate": 1.9262702370151087e-05,
+      "loss": 0.7973,
+      "step": 6338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.847365990524411,
+      "learning_rate": 1.92624147835439e-05,
+      "loss": 0.7463,
+      "step": 6339
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8046291368683853,
+      "learning_rate": 1.9262127143008e-05,
+      "loss": 0.979,
+      "step": 6340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7745097681162137,
+      "learning_rate": 1.9261839448545066e-05,
+      "loss": 0.7829,
+      "step": 6341
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8474950574309149,
+      "learning_rate": 1.926155170015677e-05,
+      "loss": 0.7141,
+      "step": 6342
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9432628034913295,
+      "learning_rate": 1.926126389784479e-05,
+      "loss": 0.9966,
+      "step": 6343
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.829489872210292,
+      "learning_rate": 1.9260976041610797e-05,
+      "loss": 0.9622,
+      "step": 6344
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7475045491098282,
+      "learning_rate": 1.926068813145647e-05,
+      "loss": 0.7243,
+      "step": 6345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7637268414237588,
+      "learning_rate": 1.9260400167383485e-05,
+      "loss": 0.6913,
+      "step": 6346
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6597401793022355,
+      "learning_rate": 1.9260112149393517e-05,
+      "loss": 0.9245,
+      "step": 6347
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8093433330839542,
+      "learning_rate": 1.9259824077488248e-05,
+      "loss": 0.738,
+      "step": 6348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7821090774802124,
+      "learning_rate": 1.925953595166935e-05,
+      "loss": 0.7341,
+      "step": 6349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.977937949903086,
+      "learning_rate": 1.92592477719385e-05,
+      "loss": 0.8394,
+      "step": 6350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7562426070888364,
+      "learning_rate": 1.925895953829738e-05,
+      "loss": 0.9001,
+      "step": 6351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6901387421940586,
+      "learning_rate": 1.9258671250747665e-05,
+      "loss": 0.9064,
+      "step": 6352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.861902655502834,
+      "learning_rate": 1.9258382909291037e-05,
+      "loss": 0.8824,
+      "step": 6353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6454256335241813,
+      "learning_rate": 1.925809451392917e-05,
+      "loss": 0.9338,
+      "step": 6354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7181006079213885,
+      "learning_rate": 1.9257806064663746e-05,
+      "loss": 0.9062,
+      "step": 6355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6445602341843524,
+      "learning_rate": 1.9257517561496443e-05,
+      "loss": 0.9739,
+      "step": 6356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7801387899558161,
+      "learning_rate": 1.9257229004428944e-05,
+      "loss": 0.749,
+      "step": 6357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7330530258683021,
+      "learning_rate": 1.925694039346293e-05,
+      "loss": 0.9408,
+      "step": 6358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8489028226978637,
+      "learning_rate": 1.9256651728600073e-05,
+      "loss": 1.0163,
+      "step": 6359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9166277286292424,
+      "learning_rate": 1.925636300984206e-05,
+      "loss": 0.8957,
+      "step": 6360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.561610191943024,
+      "learning_rate": 1.9256074237190573e-05,
+      "loss": 0.9271,
+      "step": 6361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7361928985519406,
+      "learning_rate": 1.925578541064729e-05,
+      "loss": 0.9038,
+      "step": 6362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.655938093878207,
+      "learning_rate": 1.9255496530213895e-05,
+      "loss": 0.9054,
+      "step": 6363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5448408430730725,
+      "learning_rate": 1.925520759589207e-05,
+      "loss": 0.8129,
+      "step": 6364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8053244567454176,
+      "learning_rate": 1.9254918607683493e-05,
+      "loss": 0.88,
+      "step": 6365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.99328670189825,
+      "learning_rate": 1.925462956558985e-05,
+      "loss": 0.986,
+      "step": 6366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7212034126803122,
+      "learning_rate": 1.9254340469612826e-05,
+      "loss": 0.7652,
+      "step": 6367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6672399940553762,
+      "learning_rate": 1.9254051319754103e-05,
+      "loss": 0.9651,
+      "step": 6368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4611464124510534,
+      "learning_rate": 1.925376211601536e-05,
+      "loss": 1.0507,
+      "step": 6369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4855852427619778,
+      "learning_rate": 1.9253472858398284e-05,
+      "loss": 0.8675,
+      "step": 6370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0022390862374233,
+      "learning_rate": 1.925318354690456e-05,
+      "loss": 0.9585,
+      "step": 6371
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7950509070825656,
+      "learning_rate": 1.9252894181535874e-05,
+      "loss": 0.7146,
+      "step": 6372
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.616913058782382,
+      "learning_rate": 1.9252604762293908e-05,
+      "loss": 0.7571,
+      "step": 6373
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5774100456671412,
+      "learning_rate": 1.9252315289180348e-05,
+      "loss": 0.8366,
+      "step": 6374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.934813289160634,
+      "learning_rate": 1.9252025762196878e-05,
+      "loss": 0.918,
+      "step": 6375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.087950761456915,
+      "learning_rate": 1.9251736181345186e-05,
+      "loss": 0.9298,
+      "step": 6376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6580713196051404,
+      "learning_rate": 1.925144654662696e-05,
+      "loss": 0.882,
+      "step": 6377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6055970219569426,
+      "learning_rate": 1.9251156858043877e-05,
+      "loss": 0.987,
+      "step": 6378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.816219457640612,
+      "learning_rate": 1.925086711559763e-05,
+      "loss": 1.0877,
+      "step": 6379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6359961853746534,
+      "learning_rate": 1.9250577319289912e-05,
+      "loss": 0.9432,
+      "step": 6380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8409603608546574,
+      "learning_rate": 1.92502874691224e-05,
+      "loss": 0.8605,
+      "step": 6381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6921554707729527,
+      "learning_rate": 1.9249997565096783e-05,
+      "loss": 0.9776,
+      "step": 6382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3706255382794206,
+      "learning_rate": 1.9249707607214757e-05,
+      "loss": 0.931,
+      "step": 6383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7661003662258878,
+      "learning_rate": 1.9249417595478005e-05,
+      "loss": 0.8673,
+      "step": 6384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0126744497539923,
+      "learning_rate": 1.9249127529888212e-05,
+      "loss": 0.8964,
+      "step": 6385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6885352973053618,
+      "learning_rate": 1.924883741044707e-05,
+      "loss": 0.9608,
+      "step": 6386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1757907499243148,
+      "learning_rate": 1.924854723715627e-05,
+      "loss": 0.8272,
+      "step": 6387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6100724925777892,
+      "learning_rate": 1.92482570100175e-05,
+      "loss": 0.9484,
+      "step": 6388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2473890578905964,
+      "learning_rate": 1.924796672903245e-05,
+      "loss": 0.9076,
+      "step": 6389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7668097903432699,
+      "learning_rate": 1.924767639420281e-05,
+      "loss": 0.7431,
+      "step": 6390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7649115770712838,
+      "learning_rate": 1.924738600553027e-05,
+      "loss": 0.9692,
+      "step": 6391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5499827819363174,
+      "learning_rate": 1.924709556301652e-05,
+      "loss": 0.9507,
+      "step": 6392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.447985180702764,
+      "learning_rate": 1.9246805066663255e-05,
+      "loss": 0.8582,
+      "step": 6393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5233966190327453,
+      "learning_rate": 1.9246514516472157e-05,
+      "loss": 1.0283,
+      "step": 6394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1648026172800803,
+      "learning_rate": 1.9246223912444933e-05,
+      "loss": 0.9054,
+      "step": 6395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.263489400445162,
+      "learning_rate": 1.924593325458326e-05,
+      "loss": 0.8438,
+      "step": 6396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9575252773434177,
+      "learning_rate": 1.9245642542888836e-05,
+      "loss": 0.8911,
+      "step": 6397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5651496224616157,
+      "learning_rate": 1.9245351777363355e-05,
+      "loss": 1.0434,
+      "step": 6398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.829367453464796,
+      "learning_rate": 1.924506095800851e-05,
+      "loss": 0.8735,
+      "step": 6399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6504408722908617,
+      "learning_rate": 1.9244770084825993e-05,
+      "loss": 0.8464,
+      "step": 6400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8360358308428246,
+      "learning_rate": 1.92444791578175e-05,
+      "loss": 0.8551,
+      "step": 6401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.76850645320693,
+      "learning_rate": 1.924418817698472e-05,
+      "loss": 0.9617,
+      "step": 6402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6091447629597053,
+      "learning_rate": 1.9243897142329347e-05,
+      "loss": 0.8865,
+      "step": 6403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6247205852230482,
+      "learning_rate": 1.9243606053853084e-05,
+      "loss": 1.0187,
+      "step": 6404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.616095737028039,
+      "learning_rate": 1.9243314911557614e-05,
+      "loss": 0.8285,
+      "step": 6405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8805584531876283,
+      "learning_rate": 1.9243023715444646e-05,
+      "loss": 0.9013,
+      "step": 6406
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7769551670632597,
+      "learning_rate": 1.9242732465515862e-05,
+      "loss": 0.9172,
+      "step": 6407
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8679166479091998,
+      "learning_rate": 1.9242441161772965e-05,
+      "loss": 0.8764,
+      "step": 6408
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9457473694753014,
+      "learning_rate": 1.9242149804217648e-05,
+      "loss": 0.9433,
+      "step": 6409
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2327464840301015,
+      "learning_rate": 1.9241858392851612e-05,
+      "loss": 0.8294,
+      "step": 6410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5566625357974024,
+      "learning_rate": 1.9241566927676548e-05,
+      "loss": 0.8962,
+      "step": 6411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.189199175304309,
+      "learning_rate": 1.9241275408694156e-05,
+      "loss": 0.9126,
+      "step": 6412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6458995646187078,
+      "learning_rate": 1.924098383590613e-05,
+      "loss": 0.8821,
+      "step": 6413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.636096624932424,
+      "learning_rate": 1.9240692209314176e-05,
+      "loss": 0.9927,
+      "step": 6414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6723040554773774,
+      "learning_rate": 1.9240400528919983e-05,
+      "loss": 0.9168,
+      "step": 6415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8168444742561884,
+      "learning_rate": 1.9240108794725252e-05,
+      "loss": 1.0162,
+      "step": 6416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6412521121877657,
+      "learning_rate": 1.9239817006731684e-05,
+      "loss": 0.8117,
+      "step": 6417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9665018527796727,
+      "learning_rate": 1.9239525164940977e-05,
+      "loss": 0.8084,
+      "step": 6418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8148922704529729,
+      "learning_rate": 1.9239233269354828e-05,
+      "loss": 0.8233,
+      "step": 6419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.561783551939028,
+      "learning_rate": 1.9238941319974938e-05,
+      "loss": 0.9099,
+      "step": 6420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.767507540015841,
+      "learning_rate": 1.9238649316803004e-05,
+      "loss": 0.7552,
+      "step": 6421
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6282199394628052,
+      "learning_rate": 1.923835725984073e-05,
+      "loss": 1.026,
+      "step": 6422
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7755433642498017,
+      "learning_rate": 1.923806514908982e-05,
+      "loss": 0.8953,
+      "step": 6423
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6163923135846217,
+      "learning_rate": 1.9237772984551966e-05,
+      "loss": 0.8714,
+      "step": 6424
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8564469300552244,
+      "learning_rate": 1.9237480766228872e-05,
+      "loss": 0.8662,
+      "step": 6425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8847618735478675,
+      "learning_rate": 1.923718849412224e-05,
+      "loss": 0.8811,
+      "step": 6426
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1888380547496515,
+      "learning_rate": 1.9236896168233775e-05,
+      "loss": 1.0363,
+      "step": 6427
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6494529786863295,
+      "learning_rate": 1.9236603788565176e-05,
+      "loss": 0.9096,
+      "step": 6428
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0695992537862593,
+      "learning_rate": 1.9236311355118144e-05,
+      "loss": 0.8963,
+      "step": 6429
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7928644799032277,
+      "learning_rate": 1.9236018867894382e-05,
+      "loss": 0.7378,
+      "step": 6430
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6364793334386536,
+      "learning_rate": 1.9235726326895595e-05,
+      "loss": 0.98,
+      "step": 6431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6355156067592989,
+      "learning_rate": 1.9235433732123488e-05,
+      "loss": 0.8552,
+      "step": 6432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5607268370251013,
+      "learning_rate": 1.923514108357976e-05,
+      "loss": 0.8889,
+      "step": 6433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1928641929254566,
+      "learning_rate": 1.9234848381266117e-05,
+      "loss": 0.905,
+      "step": 6434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6188627971260237,
+      "learning_rate": 1.9234555625184265e-05,
+      "loss": 0.8895,
+      "step": 6435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6522820595910515,
+      "learning_rate": 1.9234262815335903e-05,
+      "loss": 0.9238,
+      "step": 6436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4750923662049773,
+      "learning_rate": 1.923396995172274e-05,
+      "loss": 0.8858,
+      "step": 6437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7160300543961804,
+      "learning_rate": 1.9233677034346484e-05,
+      "loss": 1.0015,
+      "step": 6438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8407782714138252,
+      "learning_rate": 1.923338406320883e-05,
+      "loss": 0.7528,
+      "step": 6439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5457062150285277,
+      "learning_rate": 1.92330910383115e-05,
+      "loss": 0.981,
+      "step": 6440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1394924586614974,
+      "learning_rate": 1.9232797959656184e-05,
+      "loss": 1.0855,
+      "step": 6441
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.663702665509608,
+      "learning_rate": 1.9232504827244594e-05,
+      "loss": 0.9257,
+      "step": 6442
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8059372411914878,
+      "learning_rate": 1.9232211641078443e-05,
+      "loss": 1.1357,
+      "step": 6443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0598858321840434,
+      "learning_rate": 1.9231918401159433e-05,
+      "loss": 0.7803,
+      "step": 6444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6349939005662888,
+      "learning_rate": 1.923162510748927e-05,
+      "loss": 0.9739,
+      "step": 6445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.350325429049454,
+      "learning_rate": 1.9231331760069662e-05,
+      "loss": 0.9903,
+      "step": 6446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6429595307249025,
+      "learning_rate": 1.9231038358902317e-05,
+      "loss": 0.8819,
+      "step": 6447
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.679077732844582,
+      "learning_rate": 1.9230744903988943e-05,
+      "loss": 0.7914,
+      "step": 6448
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.251506129289405,
+      "learning_rate": 1.9230451395331253e-05,
+      "loss": 0.8331,
+      "step": 6449
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.773938554038058,
+      "learning_rate": 1.9230157832930953e-05,
+      "loss": 0.7589,
+      "step": 6450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.677510743643162,
+      "learning_rate": 1.922986421678975e-05,
+      "loss": 0.9502,
+      "step": 6451
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7405263171244858,
+      "learning_rate": 1.9229570546909354e-05,
+      "loss": 0.9449,
+      "step": 6452
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.037064258327508,
+      "learning_rate": 1.922927682329148e-05,
+      "loss": 0.8947,
+      "step": 6453
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6062271164013326,
+      "learning_rate": 1.922898304593783e-05,
+      "loss": 0.8216,
+      "step": 6454
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6552918381251978,
+      "learning_rate": 1.922868921485012e-05,
+      "loss": 0.8937,
+      "step": 6455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5380969967141755,
+      "learning_rate": 1.9228395330030064e-05,
+      "loss": 0.9088,
+      "step": 6456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0709625805563805,
+      "learning_rate": 1.9228101391479366e-05,
+      "loss": 0.9836,
+      "step": 6457
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.761406899902661,
+      "learning_rate": 1.922780739919974e-05,
+      "loss": 0.7496,
+      "step": 6458
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9481153779121894,
+      "learning_rate": 1.9227513353192895e-05,
+      "loss": 0.8117,
+      "step": 6459
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4297811173409523,
+      "learning_rate": 1.9227219253460552e-05,
+      "loss": 0.9966,
+      "step": 6460
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9335452414458205,
+      "learning_rate": 1.9226925100004415e-05,
+      "loss": 0.8693,
+      "step": 6461
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7601202885147457,
+      "learning_rate": 1.9226630892826196e-05,
+      "loss": 0.7733,
+      "step": 6462
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5812109666155854,
+      "learning_rate": 1.9226336631927613e-05,
+      "loss": 1.0403,
+      "step": 6463
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6670921462586243,
+      "learning_rate": 1.9226042317310377e-05,
+      "loss": 0.9173,
+      "step": 6464
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8247906267093731,
+      "learning_rate": 1.9225747948976205e-05,
+      "loss": 0.7774,
+      "step": 6465
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7290493112804475,
+      "learning_rate": 1.92254535269268e-05,
+      "loss": 0.854,
+      "step": 6466
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0987013698130776,
+      "learning_rate": 1.9225159051163888e-05,
+      "loss": 1.0011,
+      "step": 6467
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.690138477267603,
+      "learning_rate": 1.9224864521689182e-05,
+      "loss": 0.9192,
+      "step": 6468
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6395838020532836,
+      "learning_rate": 1.922456993850439e-05,
+      "loss": 1.0024,
+      "step": 6469
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.592687989574276,
+      "learning_rate": 1.9224275301611234e-05,
+      "loss": 0.8553,
+      "step": 6470
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.491190064688413,
+      "learning_rate": 1.9223980611011424e-05,
+      "loss": 0.9348,
+      "step": 6471
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7393104082187858,
+      "learning_rate": 1.922368586670668e-05,
+      "loss": 0.9732,
+      "step": 6472
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7340349658653866,
+      "learning_rate": 1.9223391068698717e-05,
+      "loss": 1.0517,
+      "step": 6473
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5572116765848472,
+      "learning_rate": 1.9223096216989253e-05,
+      "loss": 1.0248,
+      "step": 6474
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7176144618112525,
+      "learning_rate": 1.922280131158e-05,
+      "loss": 1.0399,
+      "step": 6475
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5827154803952952,
+      "learning_rate": 1.9222506352472677e-05,
+      "loss": 0.8611,
+      "step": 6476
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6752589943138987,
+      "learning_rate": 1.9222211339669005e-05,
+      "loss": 0.8861,
+      "step": 6477
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9851955850610992,
+      "learning_rate": 1.9221916273170697e-05,
+      "loss": 0.9624,
+      "step": 6478
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8373563582041554,
+      "learning_rate": 1.9221621152979476e-05,
+      "loss": 0.9425,
+      "step": 6479
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9332534872964318,
+      "learning_rate": 1.9221325979097053e-05,
+      "loss": 0.8713,
+      "step": 6480
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7903888413661682,
+      "learning_rate": 1.9221030751525154e-05,
+      "loss": 0.7819,
+      "step": 6481
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6626944635582424,
+      "learning_rate": 1.9220735470265492e-05,
+      "loss": 0.8993,
+      "step": 6482
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7613456473888004,
+      "learning_rate": 1.9220440135319794e-05,
+      "loss": 0.7209,
+      "step": 6483
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.248581518452184,
+      "learning_rate": 1.9220144746689768e-05,
+      "loss": 0.9817,
+      "step": 6484
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.743835347813036,
+      "learning_rate": 1.9219849304377142e-05,
+      "loss": 0.9271,
+      "step": 6485
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4657666016951043,
+      "learning_rate": 1.9219553808383636e-05,
+      "loss": 0.8888,
+      "step": 6486
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8502530247985878,
+      "learning_rate": 1.921925825871097e-05,
+      "loss": 0.8026,
+      "step": 6487
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0400408258147325,
+      "learning_rate": 1.921896265536086e-05,
+      "loss": 1.0567,
+      "step": 6488
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.715291284026423,
+      "learning_rate": 1.9218666998335034e-05,
+      "loss": 0.9817,
+      "step": 6489
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.949944059901786,
+      "learning_rate": 1.921837128763521e-05,
+      "loss": 0.9413,
+      "step": 6490
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.192314077409652,
+      "learning_rate": 1.9218075523263104e-05,
+      "loss": 0.9132,
+      "step": 6491
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.573490762315899,
+      "learning_rate": 1.9217779705220452e-05,
+      "loss": 0.7861,
+      "step": 6492
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6274058527908526,
+      "learning_rate": 1.9217483833508962e-05,
+      "loss": 0.9016,
+      "step": 6493
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5935628155036579,
+      "learning_rate": 1.921718790813037e-05,
+      "loss": 0.9464,
+      "step": 6494
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6530251947249148,
+      "learning_rate": 1.9216891929086382e-05,
+      "loss": 0.8253,
+      "step": 6495
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7191480817700258,
+      "learning_rate": 1.9216595896378737e-05,
+      "loss": 0.9245,
+      "step": 6496
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7550111610355756,
+      "learning_rate": 1.9216299810009156e-05,
+      "loss": 0.8477,
+      "step": 6497
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6869342570524894,
+      "learning_rate": 1.9216003669979356e-05,
+      "loss": 0.8076,
+      "step": 6498
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8374401243113463,
+      "learning_rate": 1.9215707476291064e-05,
+      "loss": 0.9143,
+      "step": 6499
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.673817861563376,
+      "learning_rate": 1.9215411228946005e-05,
+      "loss": 0.9093,
+      "step": 6500
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.250887861684083,
+      "learning_rate": 1.9215114927945905e-05,
+      "loss": 1.0142,
+      "step": 6501
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9065555482200838,
+      "learning_rate": 1.9214818573292492e-05,
+      "loss": 0.9572,
+      "step": 6502
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6401149694227781,
+      "learning_rate": 1.9214522164987484e-05,
+      "loss": 0.9783,
+      "step": 6503
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9102853029063638,
+      "learning_rate": 1.921422570303261e-05,
+      "loss": 0.9554,
+      "step": 6504
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0529683116587125,
+      "learning_rate": 1.9213929187429602e-05,
+      "loss": 0.8475,
+      "step": 6505
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7072725072169765,
+      "learning_rate": 1.9213632618180175e-05,
+      "loss": 0.933,
+      "step": 6506
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7505796864518848,
+      "learning_rate": 1.9213335995286065e-05,
+      "loss": 0.8946,
+      "step": 6507
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8492962300139497,
+      "learning_rate": 1.9213039318748995e-05,
+      "loss": 0.8775,
+      "step": 6508
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.752999662978692,
+      "learning_rate": 1.9212742588570694e-05,
+      "loss": 0.9499,
+      "step": 6509
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8000290520149647,
+      "learning_rate": 1.921244580475289e-05,
+      "loss": 0.9062,
+      "step": 6510
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8026134735468389,
+      "learning_rate": 1.9212148967297307e-05,
+      "loss": 0.9775,
+      "step": 6511
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6096446588590296,
+      "learning_rate": 1.921185207620568e-05,
+      "loss": 0.7634,
+      "step": 6512
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.767384459618508,
+      "learning_rate": 1.9211555131479728e-05,
+      "loss": 0.718,
+      "step": 6513
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1729374845264733,
+      "learning_rate": 1.921125813312119e-05,
+      "loss": 0.8051,
+      "step": 6514
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8344589777818034,
+      "learning_rate": 1.921096108113179e-05,
+      "loss": 0.8088,
+      "step": 6515
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6112098682061329,
+      "learning_rate": 1.921066397551326e-05,
+      "loss": 0.8135,
+      "step": 6516
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5479087308538522,
+      "learning_rate": 1.9210366816267327e-05,
+      "loss": 0.8424,
+      "step": 6517
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.930936885339278,
+      "learning_rate": 1.9210069603395724e-05,
+      "loss": 0.9825,
+      "step": 6518
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3092162375870546,
+      "learning_rate": 1.920977233690018e-05,
+      "loss": 0.9623,
+      "step": 6519
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4914666128329164,
+      "learning_rate": 1.920947501678242e-05,
+      "loss": 0.9775,
+      "step": 6520
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5576544416764782,
+      "learning_rate": 1.9209177643044187e-05,
+      "loss": 1.0858,
+      "step": 6521
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.711927445577524,
+      "learning_rate": 1.9208880215687203e-05,
+      "loss": 0.8695,
+      "step": 6522
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7887102551899766,
+      "learning_rate": 1.9208582734713202e-05,
+      "loss": 0.7013,
+      "step": 6523
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6711866441069065,
+      "learning_rate": 1.920828520012392e-05,
+      "loss": 0.8827,
+      "step": 6524
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6370224031764353,
+      "learning_rate": 1.9207987611921085e-05,
+      "loss": 0.98,
+      "step": 6525
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.825934213840267,
+      "learning_rate": 1.9207689970106434e-05,
+      "loss": 0.8872,
+      "step": 6526
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.827866577196071,
+      "learning_rate": 1.9207392274681692e-05,
+      "loss": 0.9364,
+      "step": 6527
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5974186571636504,
+      "learning_rate": 1.92070945256486e-05,
+      "loss": 0.8459,
+      "step": 6528
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.58155107650669,
+      "learning_rate": 1.920679672300889e-05,
+      "loss": 1.019,
+      "step": 6529
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5582181996193762,
+      "learning_rate": 1.920649886676429e-05,
+      "loss": 0.9743,
+      "step": 6530
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6818782499915035,
+      "learning_rate": 1.9206200956916542e-05,
+      "loss": 0.8273,
+      "step": 6531
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7118677677658727,
+      "learning_rate": 1.9205902993467377e-05,
+      "loss": 0.9886,
+      "step": 6532
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9784347556722108,
+      "learning_rate": 1.920560497641853e-05,
+      "loss": 0.927,
+      "step": 6533
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7074439950178928,
+      "learning_rate": 1.9205306905771735e-05,
+      "loss": 0.8588,
+      "step": 6534
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7737500910667798,
+      "learning_rate": 1.920500878152873e-05,
+      "loss": 0.9751,
+      "step": 6535
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3640441961010836,
+      "learning_rate": 1.920471060369125e-05,
+      "loss": 0.9601,
+      "step": 6536
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8181284826456028,
+      "learning_rate": 1.920441237226103e-05,
+      "loss": 0.9016,
+      "step": 6537
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.464489239908827,
+      "learning_rate": 1.9204114087239806e-05,
+      "loss": 0.9538,
+      "step": 6538
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.719330656860265,
+      "learning_rate": 1.920381574862932e-05,
+      "loss": 0.874,
+      "step": 6539
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.902590114680806,
+      "learning_rate": 1.92035173564313e-05,
+      "loss": 1.0393,
+      "step": 6540
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.887097170538762,
+      "learning_rate": 1.9203218910647492e-05,
+      "loss": 0.8559,
+      "step": 6541
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8489856972752177,
+      "learning_rate": 1.920292041127963e-05,
+      "loss": 0.8284,
+      "step": 6542
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7587336293067923,
+      "learning_rate": 1.920262185832945e-05,
+      "loss": 0.8459,
+      "step": 6543
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8551103565930906,
+      "learning_rate": 1.920232325179869e-05,
+      "loss": 0.8557,
+      "step": 6544
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7631977516367567,
+      "learning_rate": 1.9202024591689094e-05,
+      "loss": 0.883,
+      "step": 6545
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6069046033903251,
+      "learning_rate": 1.9201725878002394e-05,
+      "loss": 0.9174,
+      "step": 6546
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4521707187591315,
+      "learning_rate": 1.9201427110740335e-05,
+      "loss": 1.0057,
+      "step": 6547
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.578340861567971,
+      "learning_rate": 1.9201128289904654e-05,
+      "loss": 0.9635,
+      "step": 6548
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6059050948299287,
+      "learning_rate": 1.9200829415497095e-05,
+      "loss": 0.9702,
+      "step": 6549
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7978892355107887,
+      "learning_rate": 1.920053048751939e-05,
+      "loss": 0.9241,
+      "step": 6550
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.736374042329456,
+      "learning_rate": 1.9200231505973282e-05,
+      "loss": 0.8302,
+      "step": 6551
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5591851754773194,
+      "learning_rate": 1.9199932470860516e-05,
+      "loss": 0.8199,
+      "step": 6552
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0343844615937967,
+      "learning_rate": 1.919963338218283e-05,
+      "loss": 0.8173,
+      "step": 6553
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8166465078039908,
+      "learning_rate": 1.9199334239941968e-05,
+      "loss": 0.8441,
+      "step": 6554
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9496998473619502,
+      "learning_rate": 1.9199035044139666e-05,
+      "loss": 0.894,
+      "step": 6555
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.048779476378459,
+      "learning_rate": 1.919873579477767e-05,
+      "loss": 0.9006,
+      "step": 6556
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7019540037722416,
+      "learning_rate": 1.9198436491857725e-05,
+      "loss": 1.0104,
+      "step": 6557
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6537966042507217,
+      "learning_rate": 1.9198137135381567e-05,
+      "loss": 0.8022,
+      "step": 6558
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1645409673424862,
+      "learning_rate": 1.9197837725350944e-05,
+      "loss": 0.9673,
+      "step": 6559
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6605431645955002,
+      "learning_rate": 1.91975382617676e-05,
+      "loss": 0.9475,
+      "step": 6560
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6706451006273468,
+      "learning_rate": 1.9197238744633275e-05,
+      "loss": 0.9025,
+      "step": 6561
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7265964836178727,
+      "learning_rate": 1.919693917394971e-05,
+      "loss": 0.9646,
+      "step": 6562
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9002066426977688,
+      "learning_rate": 1.9196639549718655e-05,
+      "loss": 1.0937,
+      "step": 6563
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.99452304500804,
+      "learning_rate": 1.919633987194186e-05,
+      "loss": 0.8787,
+      "step": 6564
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5977863558308436,
+      "learning_rate": 1.9196040140621056e-05,
+      "loss": 0.8966,
+      "step": 6565
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7623047192503893,
+      "learning_rate": 1.9195740355757998e-05,
+      "loss": 0.9276,
+      "step": 6566
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7600698049904733,
+      "learning_rate": 1.9195440517354424e-05,
+      "loss": 0.8837,
+      "step": 6567
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8082391254311627,
+      "learning_rate": 1.9195140625412087e-05,
+      "loss": 1.0062,
+      "step": 6568
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4922502121889991,
+      "learning_rate": 1.919484067993273e-05,
+      "loss": 0.9024,
+      "step": 6569
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9673521399160268,
+      "learning_rate": 1.91945406809181e-05,
+      "loss": 0.9504,
+      "step": 6570
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.851009611313168,
+      "learning_rate": 1.919424062836994e-05,
+      "loss": 0.7754,
+      "step": 6571
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6700462646682968,
+      "learning_rate": 1.9193940522290008e-05,
+      "loss": 0.9968,
+      "step": 6572
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8124431639926335,
+      "learning_rate": 1.9193640362680038e-05,
+      "loss": 1.003,
+      "step": 6573
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5903286098628475,
+      "learning_rate": 1.919334014954178e-05,
+      "loss": 0.9957,
+      "step": 6574
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5098273888762879,
+      "learning_rate": 1.919303988287699e-05,
+      "loss": 0.9903,
+      "step": 6575
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7934953049264817,
+      "learning_rate": 1.9192739562687412e-05,
+      "loss": 0.8662,
+      "step": 6576
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8291726704614515,
+      "learning_rate": 1.919243918897479e-05,
+      "loss": 0.8732,
+      "step": 6577
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6746548635160192,
+      "learning_rate": 1.9192138761740883e-05,
+      "loss": 0.9398,
+      "step": 6578
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8742385785522873,
+      "learning_rate": 1.9191838280987428e-05,
+      "loss": 0.9789,
+      "step": 6579
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8158089804298314,
+      "learning_rate": 1.919153774671618e-05,
+      "loss": 0.9425,
+      "step": 6580
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6766435378754045,
+      "learning_rate": 1.919123715892889e-05,
+      "loss": 0.8753,
+      "step": 6581
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7232561823840749,
+      "learning_rate": 1.919093651762731e-05,
+      "loss": 1.0315,
+      "step": 6582
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5328367402753391,
+      "learning_rate": 1.919063582281319e-05,
+      "loss": 0.9298,
+      "step": 6583
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8826251201361597,
+      "learning_rate": 1.9190335074488275e-05,
+      "loss": 0.871,
+      "step": 6584
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7755660611286286,
+      "learning_rate": 1.9190034272654316e-05,
+      "loss": 0.8982,
+      "step": 6585
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8032880258984199,
+      "learning_rate": 1.9189733417313073e-05,
+      "loss": 0.8842,
+      "step": 6586
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9358782006959616,
+      "learning_rate": 1.918943250846629e-05,
+      "loss": 0.9579,
+      "step": 6587
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.59497776100002,
+      "learning_rate": 1.9189131546115725e-05,
+      "loss": 0.9705,
+      "step": 6588
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5644684380912655,
+      "learning_rate": 1.9188830530263124e-05,
+      "loss": 0.8728,
+      "step": 6589
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7143659638887876,
+      "learning_rate": 1.9188529460910244e-05,
+      "loss": 0.8297,
+      "step": 6590
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8461012550115372,
+      "learning_rate": 1.9188228338058835e-05,
+      "loss": 0.8102,
+      "step": 6591
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6023961501234711,
+      "learning_rate": 1.918792716171065e-05,
+      "loss": 0.94,
+      "step": 6592
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6139706529949696,
+      "learning_rate": 1.9187625931867448e-05,
+      "loss": 0.9033,
+      "step": 6593
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.87468480922999,
+      "learning_rate": 1.9187324648530977e-05,
+      "loss": 0.7986,
+      "step": 6594
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7544001787187937,
+      "learning_rate": 1.9187023311702996e-05,
+      "loss": 0.7695,
+      "step": 6595
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7452029762544432,
+      "learning_rate": 1.9186721921385257e-05,
+      "loss": 0.8179,
+      "step": 6596
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7244885687254907,
+      "learning_rate": 1.9186420477579508e-05,
+      "loss": 0.9259,
+      "step": 6597
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.606082972973968,
+      "learning_rate": 1.918611898028752e-05,
+      "loss": 0.9095,
+      "step": 6598
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0698846168765677,
+      "learning_rate": 1.9185817429511033e-05,
+      "loss": 0.8625,
+      "step": 6599
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.101047581330499,
+      "learning_rate": 1.9185515825251806e-05,
+      "loss": 0.8909,
+      "step": 6600
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6511300251954444,
+      "learning_rate": 1.91852141675116e-05,
+      "loss": 0.8996,
+      "step": 6601
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.053592419398753,
+      "learning_rate": 1.9184912456292172e-05,
+      "loss": 0.8154,
+      "step": 6602
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9373211724202923,
+      "learning_rate": 1.9184610691595273e-05,
+      "loss": 0.8912,
+      "step": 6603
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7841666904621791,
+      "learning_rate": 1.9184308873422666e-05,
+      "loss": 0.695,
+      "step": 6604
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.849810891983833,
+      "learning_rate": 1.9184007001776103e-05,
+      "loss": 0.8954,
+      "step": 6605
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8047699845248968,
+      "learning_rate": 1.9183705076657343e-05,
+      "loss": 0.9078,
+      "step": 6606
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7704984567158837,
+      "learning_rate": 1.9183403098068145e-05,
+      "loss": 1.0444,
+      "step": 6607
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.872304322997149,
+      "learning_rate": 1.918310106601027e-05,
+      "loss": 0.8701,
+      "step": 6608
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8425258708186146,
+      "learning_rate": 1.9182798980485468e-05,
+      "loss": 0.7518,
+      "step": 6609
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.606755277657322,
+      "learning_rate": 1.9182496841495504e-05,
+      "loss": 0.8866,
+      "step": 6610
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.354106361103698,
+      "learning_rate": 1.9182194649042138e-05,
+      "loss": 0.9534,
+      "step": 6611
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6045569670633242,
+      "learning_rate": 1.918189240312713e-05,
+      "loss": 0.9392,
+      "step": 6612
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3905462028827644,
+      "learning_rate": 1.918159010375223e-05,
+      "loss": 0.8845,
+      "step": 6613
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5686432848798335,
+      "learning_rate": 1.9181287750919214e-05,
+      "loss": 0.9203,
+      "step": 6614
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.54091580265732,
+      "learning_rate": 1.918098534462983e-05,
+      "loss": 0.853,
+      "step": 6615
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0066664650286254,
+      "learning_rate": 1.918068288488584e-05,
+      "loss": 0.7618,
+      "step": 6616
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7454382672882083,
+      "learning_rate": 1.918038037168901e-05,
+      "loss": 0.7257,
+      "step": 6617
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7670313102488315,
+      "learning_rate": 1.9180077805041103e-05,
+      "loss": 0.76,
+      "step": 6618
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.818004732409564,
+      "learning_rate": 1.917977518494387e-05,
+      "loss": 0.9652,
+      "step": 6619
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.599950937337518,
+      "learning_rate": 1.9179472511399084e-05,
+      "loss": 0.8822,
+      "step": 6620
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6443951868413806,
+      "learning_rate": 1.91791697844085e-05,
+      "loss": 0.9089,
+      "step": 6621
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8293382015385915,
+      "learning_rate": 1.917886700397389e-05,
+      "loss": 0.6643,
+      "step": 6622
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.856657894717166,
+      "learning_rate": 1.9178564170097004e-05,
+      "loss": 1.0913,
+      "step": 6623
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.806504782969547,
+      "learning_rate": 1.9178261282779614e-05,
+      "loss": 0.9481,
+      "step": 6624
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9403771143062165,
+      "learning_rate": 1.9177958342023478e-05,
+      "loss": 0.7698,
+      "step": 6625
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7799929304924638,
+      "learning_rate": 1.9177655347830368e-05,
+      "loss": 0.938,
+      "step": 6626
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7927674887829437,
+      "learning_rate": 1.9177352300202038e-05,
+      "loss": 0.8302,
+      "step": 6627
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7115090200456486,
+      "learning_rate": 1.917704919914026e-05,
+      "loss": 1.0303,
+      "step": 6628
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.712283750432488,
+      "learning_rate": 1.9176746044646796e-05,
+      "loss": 1.0364,
+      "step": 6629
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.601131947294626,
+      "learning_rate": 1.917644283672341e-05,
+      "loss": 0.8655,
+      "step": 6630
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.768634644675835,
+      "learning_rate": 1.9176139575371867e-05,
+      "loss": 0.7585,
+      "step": 6631
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7998912196242225,
+      "learning_rate": 1.9175836260593937e-05,
+      "loss": 0.8311,
+      "step": 6632
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7748036670856894,
+      "learning_rate": 1.9175532892391383e-05,
+      "loss": 0.7734,
+      "step": 6633
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7946141378310296,
+      "learning_rate": 1.9175229470765972e-05,
+      "loss": 0.7731,
+      "step": 6634
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2026145890237725,
+      "learning_rate": 1.917492599571947e-05,
+      "loss": 0.9989,
+      "step": 6635
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8170847468176028,
+      "learning_rate": 1.9174622467253644e-05,
+      "loss": 0.9416,
+      "step": 6636
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6821717086918924,
+      "learning_rate": 1.917431888537026e-05,
+      "loss": 1.0064,
+      "step": 6637
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8569113782936304,
+      "learning_rate": 1.917401525007109e-05,
+      "loss": 0.8502,
+      "step": 6638
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.684506787176948,
+      "learning_rate": 1.9173711561357895e-05,
+      "loss": 0.916,
+      "step": 6639
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5426735139634118,
+      "learning_rate": 1.917340781923245e-05,
+      "loss": 0.8432,
+      "step": 6640
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4731937274740852,
+      "learning_rate": 1.917310402369652e-05,
+      "loss": 0.9176,
+      "step": 6641
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7620073887096576,
+      "learning_rate": 1.9172800174751873e-05,
+      "loss": 0.7683,
+      "step": 6642
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6638651540043348,
+      "learning_rate": 1.9172496272400282e-05,
+      "loss": 0.8808,
+      "step": 6643
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.928059226278064,
+      "learning_rate": 1.917219231664351e-05,
+      "loss": 0.9553,
+      "step": 6644
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.792375880524863,
+      "learning_rate": 1.917188830748333e-05,
+      "loss": 0.9499,
+      "step": 6645
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7440719653719685,
+      "learning_rate": 1.9171584244921516e-05,
+      "loss": 0.9022,
+      "step": 6646
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7458872332733715,
+      "learning_rate": 1.9171280128959834e-05,
+      "loss": 0.846,
+      "step": 6647
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.017791466267315,
+      "learning_rate": 1.9170975959600053e-05,
+      "loss": 0.8791,
+      "step": 6648
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0345081387466943,
+      "learning_rate": 1.917067173684395e-05,
+      "loss": 0.8515,
+      "step": 6649
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8332481543628218,
+      "learning_rate": 1.917036746069329e-05,
+      "loss": 0.9089,
+      "step": 6650
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6855102836705413,
+      "learning_rate": 1.917006313114985e-05,
+      "loss": 0.9042,
+      "step": 6651
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.660672190536663,
+      "learning_rate": 1.9169758748215396e-05,
+      "loss": 0.8307,
+      "step": 6652
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5825104767617737,
+      "learning_rate": 1.9169454311891705e-05,
+      "loss": 0.8692,
+      "step": 6653
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6451152773012934,
+      "learning_rate": 1.916914982218055e-05,
+      "loss": 0.8173,
+      "step": 6654
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5201729433577995,
+      "learning_rate": 1.9168845279083697e-05,
+      "loss": 0.8401,
+      "step": 6655
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.914505319498248,
+      "learning_rate": 1.9168540682602928e-05,
+      "loss": 0.949,
+      "step": 6656
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5829845801598366,
+      "learning_rate": 1.916823603274001e-05,
+      "loss": 0.9904,
+      "step": 6657
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6515073560079319,
+      "learning_rate": 1.916793132949672e-05,
+      "loss": 0.9143,
+      "step": 6658
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.871958648153959,
+      "learning_rate": 1.916762657287483e-05,
+      "loss": 0.9429,
+      "step": 6659
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6232037956849683,
+      "learning_rate": 1.916732176287612e-05,
+      "loss": 0.9817,
+      "step": 6660
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6980676149457123,
+      "learning_rate": 1.9167016899502357e-05,
+      "loss": 0.8961,
+      "step": 6661
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1405144142395356,
+      "learning_rate": 1.916671198275532e-05,
+      "loss": 0.914,
+      "step": 6662
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4630534168380194,
+      "learning_rate": 1.9166407012636785e-05,
+      "loss": 0.8483,
+      "step": 6663
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7475698676016125,
+      "learning_rate": 1.9166101989148527e-05,
+      "loss": 0.9135,
+      "step": 6664
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8880901664679326,
+      "learning_rate": 1.9165796912292317e-05,
+      "loss": 0.9887,
+      "step": 6665
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.676677218817437,
+      "learning_rate": 1.9165491782069942e-05,
+      "loss": 0.8791,
+      "step": 6666
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.812008265885456,
+      "learning_rate": 1.9165186598483166e-05,
+      "loss": 1.006,
+      "step": 6667
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8983170635386106,
+      "learning_rate": 1.9164881361533775e-05,
+      "loss": 0.8029,
+      "step": 6668
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1119650492559994,
+      "learning_rate": 1.9164576071223542e-05,
+      "loss": 1.1212,
+      "step": 6669
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.610667261984587,
+      "learning_rate": 1.9164270727554247e-05,
+      "loss": 0.8436,
+      "step": 6670
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7390432425899125,
+      "learning_rate": 1.9163965330527664e-05,
+      "loss": 1.0517,
+      "step": 6671
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.712595605570291,
+      "learning_rate": 1.9163659880145575e-05,
+      "loss": 0.9942,
+      "step": 6672
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9352209203042918,
+      "learning_rate": 1.9163354376409757e-05,
+      "loss": 1.0375,
+      "step": 6673
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.700485069056028,
+      "learning_rate": 1.9163048819321987e-05,
+      "loss": 0.9074,
+      "step": 6674
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5242868787142523,
+      "learning_rate": 1.916274320888405e-05,
+      "loss": 0.8052,
+      "step": 6675
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6853509864892902,
+      "learning_rate": 1.916243754509772e-05,
+      "loss": 0.9419,
+      "step": 6676
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6017196495853925,
+      "learning_rate": 1.9162131827964774e-05,
+      "loss": 0.9713,
+      "step": 6677
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.10149625290785,
+      "learning_rate": 1.9161826057487e-05,
+      "loss": 0.8725,
+      "step": 6678
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9639572011215323,
+      "learning_rate": 1.9161520233666172e-05,
+      "loss": 0.8695,
+      "step": 6679
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6402718350864764,
+      "learning_rate": 1.9161214356504072e-05,
+      "loss": 0.963,
+      "step": 6680
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9934945710049665,
+      "learning_rate": 1.916090842600248e-05,
+      "loss": 0.9885,
+      "step": 6681
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4198303188219434,
+      "learning_rate": 1.9160602442163182e-05,
+      "loss": 0.9093,
+      "step": 6682
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2006685478436916,
+      "learning_rate": 1.916029640498796e-05,
+      "loss": 0.837,
+      "step": 6683
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9029119542457582,
+      "learning_rate": 1.915999031447859e-05,
+      "loss": 0.9425,
+      "step": 6684
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7863492269093001,
+      "learning_rate": 1.915968417063685e-05,
+      "loss": 0.7384,
+      "step": 6685
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.642478441357304,
+      "learning_rate": 1.9159377973464536e-05,
+      "loss": 1.0674,
+      "step": 6686
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6964269596354424,
+      "learning_rate": 1.9159071722963422e-05,
+      "loss": 0.9232,
+      "step": 6687
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8661086852542943,
+      "learning_rate": 1.915876541913529e-05,
+      "loss": 0.8086,
+      "step": 6688
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6126996388724695,
+      "learning_rate": 1.9158459061981928e-05,
+      "loss": 0.8934,
+      "step": 6689
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7120113859921038,
+      "learning_rate": 1.9158152651505116e-05,
+      "loss": 1.0236,
+      "step": 6690
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8841054886016746,
+      "learning_rate": 1.9157846187706642e-05,
+      "loss": 0.8576,
+      "step": 6691
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7578073341903016,
+      "learning_rate": 1.9157539670588287e-05,
+      "loss": 0.7219,
+      "step": 6692
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7441119039183555,
+      "learning_rate": 1.915723310015184e-05,
+      "loss": 0.8266,
+      "step": 6693
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5896677284720009,
+      "learning_rate": 1.915692647639908e-05,
+      "loss": 0.9392,
+      "step": 6694
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5689130729594818,
+      "learning_rate": 1.9156619799331796e-05,
+      "loss": 0.9515,
+      "step": 6695
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6878346981284504,
+      "learning_rate": 1.9156313068951772e-05,
+      "loss": 1.0145,
+      "step": 6696
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7384821778672161,
+      "learning_rate": 1.9156006285260792e-05,
+      "loss": 0.891,
+      "step": 6697
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.758767658378224,
+      "learning_rate": 1.9155699448260648e-05,
+      "loss": 0.8933,
+      "step": 6698
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7798999378327718,
+      "learning_rate": 1.9155392557953126e-05,
+      "loss": 0.8131,
+      "step": 6699
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7592146239065356,
+      "learning_rate": 1.9155085614340004e-05,
+      "loss": 0.7207,
+      "step": 6700
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5634497482401293,
+      "learning_rate": 1.915477861742308e-05,
+      "loss": 0.7899,
+      "step": 6701
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8098157982291843,
+      "learning_rate": 1.9154471567204132e-05,
+      "loss": 0.7348,
+      "step": 6702
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6942645358320205,
+      "learning_rate": 1.9154164463684953e-05,
+      "loss": 0.9372,
+      "step": 6703
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.594472502667506,
+      "learning_rate": 1.915385730686733e-05,
+      "loss": 0.9761,
+      "step": 6704
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6967387791580368,
+      "learning_rate": 1.9153550096753056e-05,
+      "loss": 0.9871,
+      "step": 6705
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8625044606772632,
+      "learning_rate": 1.9153242833343912e-05,
+      "loss": 0.9731,
+      "step": 6706
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0102545172952406,
+      "learning_rate": 1.915293551664169e-05,
+      "loss": 1.018,
+      "step": 6707
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1725503164722832,
+      "learning_rate": 1.9152628146648183e-05,
+      "loss": 0.8406,
+      "step": 6708
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.407919014593484,
+      "learning_rate": 1.9152320723365176e-05,
+      "loss": 0.9093,
+      "step": 6709
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6690296547253327,
+      "learning_rate": 1.9152013246794455e-05,
+      "loss": 0.9435,
+      "step": 6710
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.76134728864101,
+      "learning_rate": 1.915170571693782e-05,
+      "loss": 1.055,
+      "step": 6711
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6689954619227905,
+      "learning_rate": 1.915139813379706e-05,
+      "loss": 0.8721,
+      "step": 6712
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7998027093394446,
+      "learning_rate": 1.9151090497373955e-05,
+      "loss": 0.86,
+      "step": 6713
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6455367587724359,
+      "learning_rate": 1.915078280767031e-05,
+      "loss": 0.8308,
+      "step": 6714
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.98930130108918,
+      "learning_rate": 1.9150475064687907e-05,
+      "loss": 0.8761,
+      "step": 6715
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7751908809983803,
+      "learning_rate": 1.9150167268428544e-05,
+      "loss": 0.7534,
+      "step": 6716
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8704316478306561,
+      "learning_rate": 1.9149859418894007e-05,
+      "loss": 0.7835,
+      "step": 6717
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6371718063392058,
+      "learning_rate": 1.9149551516086095e-05,
+      "loss": 0.8418,
+      "step": 6718
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6270487297942808,
+      "learning_rate": 1.9149243560006598e-05,
+      "loss": 1.069,
+      "step": 6719
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9492636612688943,
+      "learning_rate": 1.9148935550657306e-05,
+      "loss": 0.8183,
+      "step": 6720
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4751377118228837,
+      "learning_rate": 1.9148627488040014e-05,
+      "loss": 1.0895,
+      "step": 6721
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7339517882772595,
+      "learning_rate": 1.914831937215652e-05,
+      "loss": 0.8924,
+      "step": 6722
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8681902700080009,
+      "learning_rate": 1.9148011203008607e-05,
+      "loss": 0.961,
+      "step": 6723
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.740638514126305,
+      "learning_rate": 1.9147702980598085e-05,
+      "loss": 0.9369,
+      "step": 6724
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.446528032913304,
+      "learning_rate": 1.9147394704926734e-05,
+      "loss": 0.8004,
+      "step": 6725
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6245354398101144,
+      "learning_rate": 1.914708637599636e-05,
+      "loss": 0.9027,
+      "step": 6726
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.944433422314569,
+      "learning_rate": 1.914677799380875e-05,
+      "loss": 0.8898,
+      "step": 6727
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5571999107902645,
+      "learning_rate": 1.9146469558365703e-05,
+      "loss": 0.8692,
+      "step": 6728
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6812791162664806,
+      "learning_rate": 1.9146161069669015e-05,
+      "loss": 0.8359,
+      "step": 6729
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7630031006013615,
+      "learning_rate": 1.914585252772048e-05,
+      "loss": 0.7024,
+      "step": 6730
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7656453899908978,
+      "learning_rate": 1.9145543932521898e-05,
+      "loss": 0.7052,
+      "step": 6731
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.472666503880196,
+      "learning_rate": 1.9145235284075064e-05,
+      "loss": 0.942,
+      "step": 6732
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7298220470035257,
+      "learning_rate": 1.9144926582381777e-05,
+      "loss": 0.9047,
+      "step": 6733
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5458286790815903,
+      "learning_rate": 1.9144617827443826e-05,
+      "loss": 0.8329,
+      "step": 6734
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.561002038051623,
+      "learning_rate": 1.914430901926302e-05,
+      "loss": 0.8633,
+      "step": 6735
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.658436475754063,
+      "learning_rate": 1.9144000157841146e-05,
+      "loss": 0.8857,
+      "step": 6736
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6697048810326933,
+      "learning_rate": 1.9143691243180015e-05,
+      "loss": 0.8899,
+      "step": 6737
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5904670211290088,
+      "learning_rate": 1.9143382275281416e-05,
+      "loss": 1.025,
+      "step": 6738
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6831546955529246,
+      "learning_rate": 1.914307325414715e-05,
+      "loss": 0.7532,
+      "step": 6739
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9262003900263167,
+      "learning_rate": 1.914276417977902e-05,
+      "loss": 0.8458,
+      "step": 6740
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.690688513061978,
+      "learning_rate": 1.9142455052178817e-05,
+      "loss": 0.9388,
+      "step": 6741
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5467657162351032,
+      "learning_rate": 1.914214587134835e-05,
+      "loss": 0.9556,
+      "step": 6742
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9722913540312033,
+      "learning_rate": 1.9141836637289414e-05,
+      "loss": 0.9804,
+      "step": 6743
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7638480271873743,
+      "learning_rate": 1.9141527350003812e-05,
+      "loss": 0.6925,
+      "step": 6744
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8257325749430691,
+      "learning_rate": 1.9141218009493344e-05,
+      "loss": 0.7331,
+      "step": 6745
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9980633969546753,
+      "learning_rate": 1.9140908615759807e-05,
+      "loss": 0.9233,
+      "step": 6746
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6669822121771736,
+      "learning_rate": 1.914059916880501e-05,
+      "loss": 0.9863,
+      "step": 6747
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7858959135004536,
+      "learning_rate": 1.9140289668630753e-05,
+      "loss": 0.7496,
+      "step": 6748
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6877138792143684,
+      "learning_rate": 1.9139980115238827e-05,
+      "loss": 0.8513,
+      "step": 6749
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7432063196106542,
+      "learning_rate": 1.913967050863105e-05,
+      "loss": 1.0033,
+      "step": 6750
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5608381805255624,
+      "learning_rate": 1.9139360848809215e-05,
+      "loss": 0.9833,
+      "step": 6751
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8673718601310867,
+      "learning_rate": 1.913905113577513e-05,
+      "loss": 0.9611,
+      "step": 6752
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8690375029590218,
+      "learning_rate": 1.9138741369530597e-05,
+      "loss": 0.9587,
+      "step": 6753
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5474566099424611,
+      "learning_rate": 1.9138431550077414e-05,
+      "loss": 0.9139,
+      "step": 6754
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.768002065774726,
+      "learning_rate": 1.913812167741739e-05,
+      "loss": 0.9101,
+      "step": 6755
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6339674063052427,
+      "learning_rate": 1.9137811751552332e-05,
+      "loss": 1.066,
+      "step": 6756
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5111939389445617,
+      "learning_rate": 1.913750177248404e-05,
+      "loss": 0.853,
+      "step": 6757
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4641832521588598,
+      "learning_rate": 1.913719174021432e-05,
+      "loss": 0.9305,
+      "step": 6758
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.798036419521776,
+      "learning_rate": 1.9136881654744973e-05,
+      "loss": 0.8371,
+      "step": 6759
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6324069651052928,
+      "learning_rate": 1.913657151607781e-05,
+      "loss": 0.9684,
+      "step": 6760
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.642923474648818,
+      "learning_rate": 1.913626132421464e-05,
+      "loss": 0.9519,
+      "step": 6761
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6945039505065396,
+      "learning_rate": 1.913595107915726e-05,
+      "loss": 0.9633,
+      "step": 6762
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4950775180656692,
+      "learning_rate": 1.913564078090748e-05,
+      "loss": 0.9401,
+      "step": 6763
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.589938908457207,
+      "learning_rate": 1.9135330429467105e-05,
+      "loss": 0.9571,
+      "step": 6764
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7991621969074523,
+      "learning_rate": 1.913502002483795e-05,
+      "loss": 0.7607,
+      "step": 6765
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6488110216183005,
+      "learning_rate": 1.9134709567021808e-05,
+      "loss": 0.9989,
+      "step": 6766
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.686951033902281,
+      "learning_rate": 1.9134399056020504e-05,
+      "loss": 0.8562,
+      "step": 6767
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7408933791216492,
+      "learning_rate": 1.913408849183583e-05,
+      "loss": 0.9619,
+      "step": 6768
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6869117207742506,
+      "learning_rate": 1.9133777874469605e-05,
+      "loss": 0.9717,
+      "step": 6769
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4744431401572302,
+      "learning_rate": 1.913346720392363e-05,
+      "loss": 0.9819,
+      "step": 6770
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.626603867082836,
+      "learning_rate": 1.913315648019972e-05,
+      "loss": 0.8965,
+      "step": 6771
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8712861723325585,
+      "learning_rate": 1.9132845703299683e-05,
+      "loss": 1.0111,
+      "step": 6772
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4524208573630766,
+      "learning_rate": 1.9132534873225323e-05,
+      "loss": 0.9162,
+      "step": 6773
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6670868324149013,
+      "learning_rate": 1.9132223989978457e-05,
+      "loss": 0.8508,
+      "step": 6774
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5081941870509397,
+      "learning_rate": 1.913191305356089e-05,
+      "loss": 1.0066,
+      "step": 6775
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7638303034823142,
+      "learning_rate": 1.9131602063974434e-05,
+      "loss": 0.8621,
+      "step": 6776
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.066993018573765,
+      "learning_rate": 1.9131291021220902e-05,
+      "loss": 0.8874,
+      "step": 6777
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1401197046908673,
+      "learning_rate": 1.91309799253021e-05,
+      "loss": 0.9947,
+      "step": 6778
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0190769369134776,
+      "learning_rate": 1.9130668776219843e-05,
+      "loss": 0.8244,
+      "step": 6779
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7286116225517134,
+      "learning_rate": 1.913035757397594e-05,
+      "loss": 0.767,
+      "step": 6780
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.216113666459922,
+      "learning_rate": 1.913004631857221e-05,
+      "loss": 0.849,
+      "step": 6781
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7115561621497994,
+      "learning_rate": 1.9129735010010452e-05,
+      "loss": 1.0186,
+      "step": 6782
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5446579638578974,
+      "learning_rate": 1.9129423648292492e-05,
+      "loss": 1.049,
+      "step": 6783
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5285503096322963,
+      "learning_rate": 1.9129112233420138e-05,
+      "loss": 0.8778,
+      "step": 6784
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7784924675535075,
+      "learning_rate": 1.9128800765395196e-05,
+      "loss": 0.7349,
+      "step": 6785
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8579948488073144,
+      "learning_rate": 1.9128489244219492e-05,
+      "loss": 1.0701,
+      "step": 6786
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8914784934603481,
+      "learning_rate": 1.9128177669894832e-05,
+      "loss": 0.8921,
+      "step": 6787
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.556407490883191,
+      "learning_rate": 1.9127866042423033e-05,
+      "loss": 0.8125,
+      "step": 6788
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 19.531606996578702,
+      "learning_rate": 1.9127554361805902e-05,
+      "loss": 0.9981,
+      "step": 6789
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9951831747897588,
+      "learning_rate": 1.9127242628045268e-05,
+      "loss": 0.9185,
+      "step": 6790
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.823096101006913,
+      "learning_rate": 1.912693084114293e-05,
+      "loss": 0.907,
+      "step": 6791
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6603529927135716,
+      "learning_rate": 1.9126619001100715e-05,
+      "loss": 0.9064,
+      "step": 6792
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7367253677981949,
+      "learning_rate": 1.9126307107920433e-05,
+      "loss": 0.8958,
+      "step": 6793
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.000848818774381,
+      "learning_rate": 1.9125995161603904e-05,
+      "loss": 0.9662,
+      "step": 6794
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4896674446442837,
+      "learning_rate": 1.912568316215294e-05,
+      "loss": 0.8329,
+      "step": 6795
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6031138527010136,
+      "learning_rate": 1.9125371109569362e-05,
+      "loss": 0.9725,
+      "step": 6796
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6818535321723456,
+      "learning_rate": 1.912505900385498e-05,
+      "loss": 1.0454,
+      "step": 6797
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5083514113911098,
+      "learning_rate": 1.912474684501162e-05,
+      "loss": 0.9966,
+      "step": 6798
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.718037792611767,
+      "learning_rate": 1.9124434633041097e-05,
+      "loss": 1.021,
+      "step": 6799
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8287156456140934,
+      "learning_rate": 1.9124122367945223e-05,
+      "loss": 0.8746,
+      "step": 6800
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6166151564454454,
+      "learning_rate": 1.912381004972582e-05,
+      "loss": 0.9472,
+      "step": 6801
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7764112404550751,
+      "learning_rate": 1.9123497678384706e-05,
+      "loss": 0.9113,
+      "step": 6802
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.832559798152361,
+      "learning_rate": 1.91231852539237e-05,
+      "loss": 0.8415,
+      "step": 6803
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8026199701021839,
+      "learning_rate": 1.9122872776344622e-05,
+      "loss": 0.855,
+      "step": 6804
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.789848668473047,
+      "learning_rate": 1.912256024564929e-05,
+      "loss": 1.0213,
+      "step": 6805
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.694208122611021,
+      "learning_rate": 1.9122247661839524e-05,
+      "loss": 0.9519,
+      "step": 6806
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6380288900069424,
+      "learning_rate": 1.9121935024917144e-05,
+      "loss": 0.9147,
+      "step": 6807
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7894047841073352,
+      "learning_rate": 1.9121622334883973e-05,
+      "loss": 0.7474,
+      "step": 6808
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5845036212308274,
+      "learning_rate": 1.9121309591741822e-05,
+      "loss": 0.9348,
+      "step": 6809
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.442410527918675,
+      "learning_rate": 1.9120996795492525e-05,
+      "loss": 0.9834,
+      "step": 6810
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6611364601530834,
+      "learning_rate": 1.9120683946137894e-05,
+      "loss": 0.9188,
+      "step": 6811
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2646800719880424,
+      "learning_rate": 1.9120371043679756e-05,
+      "loss": 0.9637,
+      "step": 6812
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9472371780868993,
+      "learning_rate": 1.912005808811993e-05,
+      "loss": 0.8878,
+      "step": 6813
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.081996066341777,
+      "learning_rate": 1.9119745079460238e-05,
+      "loss": 0.9104,
+      "step": 6814
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9124295768680735,
+      "learning_rate": 1.91194320177025e-05,
+      "loss": 0.9429,
+      "step": 6815
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6493714806779673,
+      "learning_rate": 1.9119118902848546e-05,
+      "loss": 0.9715,
+      "step": 6816
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8100761583370124,
+      "learning_rate": 1.9118805734900192e-05,
+      "loss": 0.7282,
+      "step": 6817
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.034579718819595,
+      "learning_rate": 1.9118492513859264e-05,
+      "loss": 1.0057,
+      "step": 6818
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7609983272849719,
+      "learning_rate": 1.911817923972759e-05,
+      "loss": 1.0022,
+      "step": 6819
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5625197437089502,
+      "learning_rate": 1.9117865912506984e-05,
+      "loss": 0.9822,
+      "step": 6820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.612972983660714,
+      "learning_rate": 1.9117552532199278e-05,
+      "loss": 0.7862,
+      "step": 6821
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4869306532965052,
+      "learning_rate": 1.9117239098806296e-05,
+      "loss": 0.884,
+      "step": 6822
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7118037825857562,
+      "learning_rate": 1.911692561232986e-05,
+      "loss": 1.0055,
+      "step": 6823
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8469888480595026,
+      "learning_rate": 1.9116612072771798e-05,
+      "loss": 0.7661,
+      "step": 6824
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2359588048696417,
+      "learning_rate": 1.911629848013393e-05,
+      "loss": 0.8772,
+      "step": 6825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.61973436871601,
+      "learning_rate": 1.911598483441809e-05,
+      "loss": 0.9079,
+      "step": 6826
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.685461704855619,
+      "learning_rate": 1.91156711356261e-05,
+      "loss": 0.8979,
+      "step": 6827
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5932890419936312,
+      "learning_rate": 1.911535738375979e-05,
+      "loss": 0.898,
+      "step": 6828
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6499914795857513,
+      "learning_rate": 1.9115043578820978e-05,
+      "loss": 0.8886,
+      "step": 6829
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5609835487371884,
+      "learning_rate": 1.9114729720811497e-05,
+      "loss": 0.8927,
+      "step": 6830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6043062259038383,
+      "learning_rate": 1.9114415809733176e-05,
+      "loss": 0.9046,
+      "step": 6831
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.386063259709077,
+      "learning_rate": 1.911410184558784e-05,
+      "loss": 0.9499,
+      "step": 6832
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7946267718133744,
+      "learning_rate": 1.911378782837732e-05,
+      "loss": 0.7758,
+      "step": 6833
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9172587072197427,
+      "learning_rate": 1.9113473758103435e-05,
+      "loss": 0.8711,
+      "step": 6834
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8048958217211241,
+      "learning_rate": 1.9113159634768025e-05,
+      "loss": 0.733,
+      "step": 6835
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6819991851217087,
+      "learning_rate": 1.9112845458372914e-05,
+      "loss": 1.079,
+      "step": 6836
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6663392927418672,
+      "learning_rate": 1.9112531228919933e-05,
+      "loss": 0.9052,
+      "step": 6837
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6427969941596168,
+      "learning_rate": 1.911221694641091e-05,
+      "loss": 1.1119,
+      "step": 6838
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7968071180490675,
+      "learning_rate": 1.9111902610847674e-05,
+      "loss": 0.9056,
+      "step": 6839
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6127591885691017,
+      "learning_rate": 1.9111588222232056e-05,
+      "loss": 1.0381,
+      "step": 6840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6392591280619724,
+      "learning_rate": 1.9111273780565886e-05,
+      "loss": 0.8678,
+      "step": 6841
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.432154355886258,
+      "learning_rate": 1.9110959285850995e-05,
+      "loss": 0.9163,
+      "step": 6842
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7271139196819916,
+      "learning_rate": 1.9110644738089216e-05,
+      "loss": 0.8986,
+      "step": 6843
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4878345344475865,
+      "learning_rate": 1.911033013728238e-05,
+      "loss": 0.9176,
+      "step": 6844
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9703669458006268,
+      "learning_rate": 1.9110015483432318e-05,
+      "loss": 0.9261,
+      "step": 6845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0790241066690984,
+      "learning_rate": 1.9109700776540857e-05,
+      "loss": 0.9774,
+      "step": 6846
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0687499692742333,
+      "learning_rate": 1.910938601660984e-05,
+      "loss": 0.9827,
+      "step": 6847
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8276404978047813,
+      "learning_rate": 1.910907120364109e-05,
+      "loss": 0.9004,
+      "step": 6848
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6041366711994374,
+      "learning_rate": 1.9108756337636446e-05,
+      "loss": 0.889,
+      "step": 6849
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7099122841676246,
+      "learning_rate": 1.9108441418597737e-05,
+      "loss": 1.0112,
+      "step": 6850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7898630336705045,
+      "learning_rate": 1.9108126446526797e-05,
+      "loss": 0.7132,
+      "step": 6851
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7958849963014178,
+      "learning_rate": 1.9107811421425465e-05,
+      "loss": 0.7383,
+      "step": 6852
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4125441435337631,
+      "learning_rate": 1.9107496343295567e-05,
+      "loss": 0.9381,
+      "step": 6853
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4881356916051736,
+      "learning_rate": 1.9107181212138945e-05,
+      "loss": 0.7395,
+      "step": 6854
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1183848272454777,
+      "learning_rate": 1.910686602795743e-05,
+      "loss": 0.9495,
+      "step": 6855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7121934846655904,
+      "learning_rate": 1.9106550790752857e-05,
+      "loss": 0.8964,
+      "step": 6856
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8309503508723208,
+      "learning_rate": 1.9106235500527065e-05,
+      "loss": 0.9487,
+      "step": 6857
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.702192817042143,
+      "learning_rate": 1.9105920157281883e-05,
+      "loss": 0.9006,
+      "step": 6858
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.79910686135208,
+      "learning_rate": 1.9105604761019153e-05,
+      "loss": 0.8478,
+      "step": 6859
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7289236503269576,
+      "learning_rate": 1.910528931174071e-05,
+      "loss": 1.0025,
+      "step": 6860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6990582405046126,
+      "learning_rate": 1.9104973809448388e-05,
+      "loss": 1.01,
+      "step": 6861
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8015634617520917,
+      "learning_rate": 1.9104658254144025e-05,
+      "loss": 0.9476,
+      "step": 6862
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9990578550570477,
+      "learning_rate": 1.9104342645829464e-05,
+      "loss": 0.8492,
+      "step": 6863
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6280623528086513,
+      "learning_rate": 1.9104026984506533e-05,
+      "loss": 0.9697,
+      "step": 6864
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.949380798826889,
+      "learning_rate": 1.9103711270177077e-05,
+      "loss": 0.8536,
+      "step": 6865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7542120150069636,
+      "learning_rate": 1.9103395502842932e-05,
+      "loss": 0.7547,
+      "step": 6866
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6857484303177508,
+      "learning_rate": 1.9103079682505938e-05,
+      "loss": 0.9264,
+      "step": 6867
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7291432821806167,
+      "learning_rate": 1.9102763809167927e-05,
+      "loss": 0.9348,
+      "step": 6868
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7036188744764498,
+      "learning_rate": 1.9102447882830745e-05,
+      "loss": 0.8862,
+      "step": 6869
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.643498103726902,
+      "learning_rate": 1.910213190349623e-05,
+      "loss": 1.0668,
+      "step": 6870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8051146331038967,
+      "learning_rate": 1.9101815871166222e-05,
+      "loss": 0.8439,
+      "step": 6871
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9166721568087552,
+      "learning_rate": 1.9101499785842563e-05,
+      "loss": 0.93,
+      "step": 6872
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6569241838202895,
+      "learning_rate": 1.9101183647527086e-05,
+      "loss": 0.9554,
+      "step": 6873
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8663879403039154,
+      "learning_rate": 1.910086745622164e-05,
+      "loss": 0.7715,
+      "step": 6874
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9351096359347444,
+      "learning_rate": 1.910055121192806e-05,
+      "loss": 0.983,
+      "step": 6875
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8623745792138195,
+      "learning_rate": 1.910023491464819e-05,
+      "loss": 1.0907,
+      "step": 6876
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7197171832215554,
+      "learning_rate": 1.9099918564383873e-05,
+      "loss": 1.0077,
+      "step": 6877
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.711538492489643,
+      "learning_rate": 1.9099602161136947e-05,
+      "loss": 0.987,
+      "step": 6878
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.544157928118615,
+      "learning_rate": 1.9099285704909256e-05,
+      "loss": 0.9309,
+      "step": 6879
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7155789674298043,
+      "learning_rate": 1.9098969195702646e-05,
+      "loss": 0.8929,
+      "step": 6880
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.789867892571983,
+      "learning_rate": 1.9098652633518956e-05,
+      "loss": 0.7536,
+      "step": 6881
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.487829289894396,
+      "learning_rate": 1.9098336018360027e-05,
+      "loss": 0.9032,
+      "step": 6882
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0788934427322543,
+      "learning_rate": 1.9098019350227705e-05,
+      "loss": 0.9924,
+      "step": 6883
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7222314979163929,
+      "learning_rate": 1.9097702629123836e-05,
+      "loss": 0.9095,
+      "step": 6884
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8654070535525709,
+      "learning_rate": 1.9097385855050263e-05,
+      "loss": 0.9489,
+      "step": 6885
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.330942107463734,
+      "learning_rate": 1.9097069028008827e-05,
+      "loss": 0.9569,
+      "step": 6886
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6078847939568655,
+      "learning_rate": 1.9096752148001378e-05,
+      "loss": 0.8887,
+      "step": 6887
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.650806457039182,
+      "learning_rate": 1.9096435215029757e-05,
+      "loss": 0.8367,
+      "step": 6888
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1461457211324517,
+      "learning_rate": 1.9096118229095806e-05,
+      "loss": 0.8478,
+      "step": 6889
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.571073815606921,
+      "learning_rate": 1.909580119020138e-05,
+      "loss": 0.9969,
+      "step": 6890
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.838231214193823,
+      "learning_rate": 1.9095484098348314e-05,
+      "loss": 0.7149,
+      "step": 6891
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6471651365858868,
+      "learning_rate": 1.9095166953538467e-05,
+      "loss": 0.8778,
+      "step": 6892
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7310300768402649,
+      "learning_rate": 1.9094849755773675e-05,
+      "loss": 0.9479,
+      "step": 6893
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1518142287842834,
+      "learning_rate": 1.909453250505579e-05,
+      "loss": 0.889,
+      "step": 6894
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9678008708567925,
+      "learning_rate": 1.9094215201386656e-05,
+      "loss": 1.0479,
+      "step": 6895
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7029150236797617,
+      "learning_rate": 1.9093897844768122e-05,
+      "loss": 0.9259,
+      "step": 6896
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6025029149394543,
+      "learning_rate": 1.9093580435202037e-05,
+      "loss": 0.8277,
+      "step": 6897
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.515621883157777,
+      "learning_rate": 1.9093262972690246e-05,
+      "loss": 0.8991,
+      "step": 6898
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.233080954441332,
+      "learning_rate": 1.90929454572346e-05,
+      "loss": 0.8782,
+      "step": 6899
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7542819407463575,
+      "learning_rate": 1.9092627888836945e-05,
+      "loss": 0.7715,
+      "step": 6900
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6802769575125935,
+      "learning_rate": 1.9092310267499135e-05,
+      "loss": 0.8909,
+      "step": 6901
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7700687541178175,
+      "learning_rate": 1.9091992593223015e-05,
+      "loss": 0.7591,
+      "step": 6902
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7209527000988676,
+      "learning_rate": 1.9091674866010434e-05,
+      "loss": 0.936,
+      "step": 6903
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6147282785137114,
+      "learning_rate": 1.9091357085863246e-05,
+      "loss": 0.94,
+      "step": 6904
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7546066293497768,
+      "learning_rate": 1.9091039252783296e-05,
+      "loss": 0.7524,
+      "step": 6905
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4805674338972514,
+      "learning_rate": 1.909072136677244e-05,
+      "loss": 0.8986,
+      "step": 6906
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7179381299794816,
+      "learning_rate": 1.909040342783252e-05,
+      "loss": 0.8681,
+      "step": 6907
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5739260236929926,
+      "learning_rate": 1.90900854359654e-05,
+      "loss": 0.9454,
+      "step": 6908
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5486311376875077,
+      "learning_rate": 1.9089767391172925e-05,
+      "loss": 0.8735,
+      "step": 6909
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7410066472435457,
+      "learning_rate": 1.9089449293456944e-05,
+      "loss": 0.8388,
+      "step": 6910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5474594275894067,
+      "learning_rate": 1.908913114281931e-05,
+      "loss": 0.9232,
+      "step": 6911
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9402834078084523,
+      "learning_rate": 1.908881293926188e-05,
+      "loss": 1.0604,
+      "step": 6912
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.769645403772883,
+      "learning_rate": 1.9088494682786502e-05,
+      "loss": 0.7438,
+      "step": 6913
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5379564915787582,
+      "learning_rate": 1.908817637339503e-05,
+      "loss": 0.9576,
+      "step": 6914
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1066905576998347,
+      "learning_rate": 1.908785801108932e-05,
+      "loss": 0.9084,
+      "step": 6915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.236702436195033,
+      "learning_rate": 1.9087539595871222e-05,
+      "loss": 0.8818,
+      "step": 6916
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6886830065622807,
+      "learning_rate": 1.908722112774259e-05,
+      "loss": 1.0315,
+      "step": 6917
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9204956385167136,
+      "learning_rate": 1.9086902606705286e-05,
+      "loss": 0.8753,
+      "step": 6918
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8359367155329571,
+      "learning_rate": 1.9086584032761148e-05,
+      "loss": 0.825,
+      "step": 6919
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.676227352246852,
+      "learning_rate": 1.9086265405912047e-05,
+      "loss": 0.9373,
+      "step": 6920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7528110376331698,
+      "learning_rate": 1.9085946726159833e-05,
+      "loss": 0.836,
+      "step": 6921
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7889829335618939,
+      "learning_rate": 1.908562799350636e-05,
+      "loss": 0.7965,
+      "step": 6922
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6668450292531218,
+      "learning_rate": 1.9085309207953483e-05,
+      "loss": 0.9511,
+      "step": 6923
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6372274272706595,
+      "learning_rate": 1.908499036950306e-05,
+      "loss": 0.8517,
+      "step": 6924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1357431661796986,
+      "learning_rate": 1.9084671478156947e-05,
+      "loss": 0.8062,
+      "step": 6925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7420251580836645,
+      "learning_rate": 1.9084352533917e-05,
+      "loss": 0.8981,
+      "step": 6926
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7696393732393751,
+      "learning_rate": 1.908403353678508e-05,
+      "loss": 0.9111,
+      "step": 6927
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.219876069160786,
+      "learning_rate": 1.9083714486763035e-05,
+      "loss": 0.8541,
+      "step": 6928
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8715274848964836,
+      "learning_rate": 1.9083395383852733e-05,
+      "loss": 0.9809,
+      "step": 6929
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6132753818904015,
+      "learning_rate": 1.9083076228056023e-05,
+      "loss": 0.9729,
+      "step": 6930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5342452458968552,
+      "learning_rate": 1.908275701937477e-05,
+      "loss": 0.9211,
+      "step": 6931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8881163094193496,
+      "learning_rate": 1.9082437757810827e-05,
+      "loss": 0.9755,
+      "step": 6932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7101203071916817,
+      "learning_rate": 1.908211844336606e-05,
+      "loss": 0.8755,
+      "step": 6933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8883881276865215,
+      "learning_rate": 1.9081799076042323e-05,
+      "loss": 0.9548,
+      "step": 6934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6607960800013715,
+      "learning_rate": 1.9081479655841476e-05,
+      "loss": 0.809,
+      "step": 6935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8627378007017745,
+      "learning_rate": 1.908116018276538e-05,
+      "loss": 1.0578,
+      "step": 6936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5647915389187512,
+      "learning_rate": 1.9080840656815895e-05,
+      "loss": 0.9221,
+      "step": 6937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6929592236883522,
+      "learning_rate": 1.908052107799488e-05,
+      "loss": 1.0012,
+      "step": 6938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6003840249038441,
+      "learning_rate": 1.9080201446304198e-05,
+      "loss": 0.9109,
+      "step": 6939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.742188554841522,
+      "learning_rate": 1.9079881761745706e-05,
+      "loss": 0.8932,
+      "step": 6940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5391036067625072,
+      "learning_rate": 1.907956202432127e-05,
+      "loss": 0.8457,
+      "step": 6941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6150031167119885,
+      "learning_rate": 1.9079242234032747e-05,
+      "loss": 1.0135,
+      "step": 6942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8348425604114538,
+      "learning_rate": 1.9078922390882e-05,
+      "loss": 0.9131,
+      "step": 6943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.686011681488566,
+      "learning_rate": 1.9078602494870894e-05,
+      "loss": 0.8632,
+      "step": 6944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6531759210898016,
+      "learning_rate": 1.907828254600129e-05,
+      "loss": 0.9727,
+      "step": 6945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1960195999039898,
+      "learning_rate": 1.9077962544275053e-05,
+      "loss": 0.9367,
+      "step": 6946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.106278219048578,
+      "learning_rate": 1.9077642489694043e-05,
+      "loss": 0.981,
+      "step": 6947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.641050063143954,
+      "learning_rate": 1.9077322382260123e-05,
+      "loss": 0.875,
+      "step": 6948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6657319883298838,
+      "learning_rate": 1.907700222197516e-05,
+      "loss": 0.9131,
+      "step": 6949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0268579649783502,
+      "learning_rate": 1.9076682008841015e-05,
+      "loss": 0.8554,
+      "step": 6950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6889526869951739,
+      "learning_rate": 1.9076361742859556e-05,
+      "loss": 0.9576,
+      "step": 6951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8200859585089032,
+      "learning_rate": 1.9076041424032642e-05,
+      "loss": 0.8325,
+      "step": 6952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7984311118271903,
+      "learning_rate": 1.9075721052362145e-05,
+      "loss": 0.908,
+      "step": 6953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.880037104085133,
+      "learning_rate": 1.9075400627849925e-05,
+      "loss": 0.7811,
+      "step": 6954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6309142883165615,
+      "learning_rate": 1.9075080150497848e-05,
+      "loss": 0.9096,
+      "step": 6955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.698816306350099,
+      "learning_rate": 1.907475962030778e-05,
+      "loss": 0.9883,
+      "step": 6956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6116050103559048,
+      "learning_rate": 1.9074439037281594e-05,
+      "loss": 0.8324,
+      "step": 6957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8583059272760772,
+      "learning_rate": 1.907411840142115e-05,
+      "loss": 0.7945,
+      "step": 6958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.415715219802404,
+      "learning_rate": 1.907379771272831e-05,
+      "loss": 0.9098,
+      "step": 6959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.095424514592777,
+      "learning_rate": 1.9073476971204952e-05,
+      "loss": 0.8587,
+      "step": 6960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.10269781652322,
+      "learning_rate": 1.9073156176852936e-05,
+      "loss": 0.9157,
+      "step": 6961
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7386896592751957,
+      "learning_rate": 1.9072835329674134e-05,
+      "loss": 0.8924,
+      "step": 6962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6080775233369582,
+      "learning_rate": 1.907251442967041e-05,
+      "loss": 1.0117,
+      "step": 6963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.636745402252411,
+      "learning_rate": 1.9072193476843636e-05,
+      "loss": 0.8144,
+      "step": 6964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.71760909122891,
+      "learning_rate": 1.907187247119568e-05,
+      "loss": 0.9001,
+      "step": 6965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4605159709066204,
+      "learning_rate": 1.907155141272841e-05,
+      "loss": 0.8703,
+      "step": 6966
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.822763169615348,
+      "learning_rate": 1.9071230301443698e-05,
+      "loss": 0.8712,
+      "step": 6967
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3178867328491357,
+      "learning_rate": 1.907090913734341e-05,
+      "loss": 0.8538,
+      "step": 6968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.207872384933053,
+      "learning_rate": 1.9070587920429414e-05,
+      "loss": 0.9414,
+      "step": 6969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7761387103822222,
+      "learning_rate": 1.9070266650703585e-05,
+      "loss": 0.9557,
+      "step": 6970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8027890738307734,
+      "learning_rate": 1.9069945328167792e-05,
+      "loss": 0.7469,
+      "step": 6971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5834261917467691,
+      "learning_rate": 1.906962395282391e-05,
+      "loss": 0.8328,
+      "step": 6972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7813969142441293,
+      "learning_rate": 1.90693025246738e-05,
+      "loss": 0.866,
+      "step": 6973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5241782287329741,
+      "learning_rate": 1.9068981043719343e-05,
+      "loss": 0.9174,
+      "step": 6974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8023059411002829,
+      "learning_rate": 1.9068659509962407e-05,
+      "loss": 1.0014,
+      "step": 6975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7453081248034792,
+      "learning_rate": 1.9068337923404863e-05,
+      "loss": 0.9376,
+      "step": 6976
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.840924575873335,
+      "learning_rate": 1.9068016284048584e-05,
+      "loss": 0.9065,
+      "step": 6977
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7409615771853717,
+      "learning_rate": 1.9067694591895443e-05,
+      "loss": 0.8857,
+      "step": 6978
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7748101280935574,
+      "learning_rate": 1.9067372846947315e-05,
+      "loss": 0.7383,
+      "step": 6979
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1513109643585446,
+      "learning_rate": 1.906705104920607e-05,
+      "loss": 0.8154,
+      "step": 6980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4645083541359483,
+      "learning_rate": 1.9066729198673588e-05,
+      "loss": 0.8598,
+      "step": 6981
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.587011725244856,
+      "learning_rate": 1.9066407295351732e-05,
+      "loss": 0.9981,
+      "step": 6982
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5858296716971247,
+      "learning_rate": 1.9066085339242387e-05,
+      "loss": 1.0393,
+      "step": 6983
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6159551787632338,
+      "learning_rate": 1.906576333034742e-05,
+      "loss": 1.0934,
+      "step": 6984
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4676255261973996,
+      "learning_rate": 1.906544126866871e-05,
+      "loss": 0.9204,
+      "step": 6985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4742845328070886,
+      "learning_rate": 1.906511915420813e-05,
+      "loss": 0.8202,
+      "step": 6986
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8515123943561467,
+      "learning_rate": 1.9064796986967557e-05,
+      "loss": 0.9592,
+      "step": 6987
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5986237552991545,
+      "learning_rate": 1.9064474766948864e-05,
+      "loss": 0.8758,
+      "step": 6988
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3997616964845885,
+      "learning_rate": 1.906415249415393e-05,
+      "loss": 0.909,
+      "step": 6989
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4910865455990856,
+      "learning_rate": 1.9063830168584633e-05,
+      "loss": 0.905,
+      "step": 6990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.846005708216825,
+      "learning_rate": 1.9063507790242845e-05,
+      "loss": 0.9499,
+      "step": 6991
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8450143685908709,
+      "learning_rate": 1.9063185359130446e-05,
+      "loss": 0.8983,
+      "step": 6992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7661491199362566,
+      "learning_rate": 1.906286287524931e-05,
+      "loss": 0.7076,
+      "step": 6993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8256535954853517,
+      "learning_rate": 1.9062540338601324e-05,
+      "loss": 0.9755,
+      "step": 6994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5795100042678036,
+      "learning_rate": 1.906221774918835e-05,
+      "loss": 0.924,
+      "step": 6995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.576755660451162,
+      "learning_rate": 1.906189510701228e-05,
+      "loss": 0.9614,
+      "step": 6996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5852068136629598,
+      "learning_rate": 1.906157241207499e-05,
+      "loss": 0.9444,
+      "step": 6997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8748149255006349,
+      "learning_rate": 1.9061249664378353e-05,
+      "loss": 0.7272,
+      "step": 6998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6741329297282246,
+      "learning_rate": 1.906092686392425e-05,
+      "loss": 0.9639,
+      "step": 6999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3115879360654787,
+      "learning_rate": 1.9060604010714566e-05,
+      "loss": 0.8224,
+      "step": 7000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.572391909416782,
+      "learning_rate": 1.9060281104751178e-05,
+      "loss": 0.8909,
+      "step": 7001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5173906477792338,
+      "learning_rate": 1.9059958146035963e-05,
+      "loss": 0.8189,
+      "step": 7002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6863469198579348,
+      "learning_rate": 1.90596351345708e-05,
+      "loss": 0.9641,
+      "step": 7003
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.512925213174594,
+      "learning_rate": 1.9059312070357577e-05,
+      "loss": 0.9204,
+      "step": 7004
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5603742752452685,
+      "learning_rate": 1.9058988953398166e-05,
+      "loss": 0.9379,
+      "step": 7005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.754036485225264,
+      "learning_rate": 1.9058665783694458e-05,
+      "loss": 0.9249,
+      "step": 7006
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.50176017731328,
+      "learning_rate": 1.9058342561248332e-05,
+      "loss": 0.9424,
+      "step": 7007
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7972299655580594,
+      "learning_rate": 1.9058019286061662e-05,
+      "loss": 0.7071,
+      "step": 7008
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7696261644960543,
+      "learning_rate": 1.905769595813634e-05,
+      "loss": 0.874,
+      "step": 7009
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.725426822766665,
+      "learning_rate": 1.9057372577474244e-05,
+      "loss": 0.9777,
+      "step": 7010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.789661217560472,
+      "learning_rate": 1.9057049144077256e-05,
+      "loss": 0.7508,
+      "step": 7011
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6276610573442152,
+      "learning_rate": 1.905672565794726e-05,
+      "loss": 0.9789,
+      "step": 7012
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5786187635219482,
+      "learning_rate": 1.9056402119086142e-05,
+      "loss": 0.8661,
+      "step": 7013
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7014186164005702,
+      "learning_rate": 1.9056078527495784e-05,
+      "loss": 0.8243,
+      "step": 7014
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8089728323255037,
+      "learning_rate": 1.9055754883178068e-05,
+      "loss": 0.9398,
+      "step": 7015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1328923256641716,
+      "learning_rate": 1.905543118613488e-05,
+      "loss": 0.8858,
+      "step": 7016
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1876445711846664,
+      "learning_rate": 1.9055107436368104e-05,
+      "loss": 0.8648,
+      "step": 7017
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8041033484954796,
+      "learning_rate": 1.905478363387963e-05,
+      "loss": 0.9382,
+      "step": 7018
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7729003648394943,
+      "learning_rate": 1.9054459778671334e-05,
+      "loss": 0.8092,
+      "step": 7019
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8080776259875613,
+      "learning_rate": 1.9054135870745106e-05,
+      "loss": 0.8242,
+      "step": 7020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7897976765260312,
+      "learning_rate": 1.9053811910102838e-05,
+      "loss": 0.7639,
+      "step": 7021
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5739577729836631,
+      "learning_rate": 1.9053487896746406e-05,
+      "loss": 0.9309,
+      "step": 7022
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8029388735888753,
+      "learning_rate": 1.9053163830677705e-05,
+      "loss": 0.7452,
+      "step": 7023
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0195105018854997,
+      "learning_rate": 1.9052839711898614e-05,
+      "loss": 0.9335,
+      "step": 7024
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8682978624212322,
+      "learning_rate": 1.9052515540411027e-05,
+      "loss": 0.8057,
+      "step": 7025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7043727281057854,
+      "learning_rate": 1.905219131621683e-05,
+      "loss": 0.8551,
+      "step": 7026
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5735677747792352,
+      "learning_rate": 1.9051867039317903e-05,
+      "loss": 0.894,
+      "step": 7027
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.640709293242207,
+      "learning_rate": 1.9051542709716147e-05,
+      "loss": 0.9615,
+      "step": 7028
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6231275569726675,
+      "learning_rate": 1.9051218327413438e-05,
+      "loss": 0.9408,
+      "step": 7029
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.782128252361875,
+      "learning_rate": 1.9050893892411675e-05,
+      "loss": 0.9077,
+      "step": 7030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.168567760284402,
+      "learning_rate": 1.905056940471274e-05,
+      "loss": 0.9577,
+      "step": 7031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8242735605806338,
+      "learning_rate": 1.9050244864318524e-05,
+      "loss": 0.6788,
+      "step": 7032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7663861367433779,
+      "learning_rate": 1.904992027123092e-05,
+      "loss": 0.7357,
+      "step": 7033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9741301984822348,
+      "learning_rate": 1.9049595625451813e-05,
+      "loss": 1.0526,
+      "step": 7034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1567255566126393,
+      "learning_rate": 1.9049270926983095e-05,
+      "loss": 0.8746,
+      "step": 7035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.30776705933299,
+      "learning_rate": 1.9048946175826658e-05,
+      "loss": 0.9577,
+      "step": 7036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7610919598566903,
+      "learning_rate": 1.904862137198439e-05,
+      "loss": 0.7244,
+      "step": 7037
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8811833691775066,
+      "learning_rate": 1.904829651545818e-05,
+      "loss": 0.9996,
+      "step": 7038
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9030432176458902,
+      "learning_rate": 1.904797160624993e-05,
+      "loss": 0.797,
+      "step": 7039
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.644215444958605,
+      "learning_rate": 1.904764664436152e-05,
+      "loss": 0.9944,
+      "step": 7040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9430301223804152,
+      "learning_rate": 1.904732162979485e-05,
+      "loss": 0.923,
+      "step": 7041
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.11832946913989,
+      "learning_rate": 1.904699656255181e-05,
+      "loss": 0.8874,
+      "step": 7042
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6602550634786308,
+      "learning_rate": 1.9046671442634288e-05,
+      "loss": 1.0163,
+      "step": 7043
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9653304118688795,
+      "learning_rate": 1.9046346270044183e-05,
+      "loss": 0.8073,
+      "step": 7044
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6118854293522757,
+      "learning_rate": 1.904602104478339e-05,
+      "loss": 1.0932,
+      "step": 7045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.19912284561857,
+      "learning_rate": 1.904569576685379e-05,
+      "loss": 0.895,
+      "step": 7046
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0494745024306544,
+      "learning_rate": 1.904537043625729e-05,
+      "loss": 0.8874,
+      "step": 7047
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4470680802170441,
+      "learning_rate": 1.9045045052995776e-05,
+      "loss": 0.8572,
+      "step": 7048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.771835005103977,
+      "learning_rate": 1.904471961707115e-05,
+      "loss": 0.836,
+      "step": 7049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5905942916935873,
+      "learning_rate": 1.9044394128485303e-05,
+      "loss": 0.8941,
+      "step": 7050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8551598616310467,
+      "learning_rate": 1.9044068587240128e-05,
+      "loss": 0.9662,
+      "step": 7051
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.634604966187208,
+      "learning_rate": 1.9043742993337523e-05,
+      "loss": 0.9424,
+      "step": 7052
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7316217673793397,
+      "learning_rate": 1.9043417346779386e-05,
+      "loss": 0.8012,
+      "step": 7053
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4896596810858793,
+      "learning_rate": 1.9043091647567608e-05,
+      "loss": 0.758,
+      "step": 7054
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.046909854955356,
+      "learning_rate": 1.9042765895704085e-05,
+      "loss": 0.9062,
+      "step": 7055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6434527159473888,
+      "learning_rate": 1.904244009119072e-05,
+      "loss": 1.0131,
+      "step": 7056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9167558508388454,
+      "learning_rate": 1.90421142340294e-05,
+      "loss": 0.945,
+      "step": 7057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8399038378918983,
+      "learning_rate": 1.9041788324222035e-05,
+      "loss": 0.7663,
+      "step": 7058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.055094892529932,
+      "learning_rate": 1.9041462361770513e-05,
+      "loss": 0.8352,
+      "step": 7059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2395091426565124,
+      "learning_rate": 1.9041136346676733e-05,
+      "loss": 0.8647,
+      "step": 7060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6892349142725878,
+      "learning_rate": 1.9040810278942596e-05,
+      "loss": 0.9536,
+      "step": 7061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8779080146704625,
+      "learning_rate": 1.904048415857e-05,
+      "loss": 1.0016,
+      "step": 7062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4467605705339093,
+      "learning_rate": 1.9040157985560842e-05,
+      "loss": 0.7648,
+      "step": 7063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0524014552573444,
+      "learning_rate": 1.9039831759917023e-05,
+      "loss": 0.9106,
+      "step": 7064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.621576095722102,
+      "learning_rate": 1.9039505481640447e-05,
+      "loss": 0.9497,
+      "step": 7065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7514100411658653,
+      "learning_rate": 1.9039179150733e-05,
+      "loss": 0.8816,
+      "step": 7066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7525319741571103,
+      "learning_rate": 1.9038852767196592e-05,
+      "loss": 0.7349,
+      "step": 7067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0163397401109937,
+      "learning_rate": 1.9038526331033122e-05,
+      "loss": 0.9357,
+      "step": 7068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7825674388303554,
+      "learning_rate": 1.903819984224449e-05,
+      "loss": 0.948,
+      "step": 7069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6053665525353433,
+      "learning_rate": 1.90378733008326e-05,
+      "loss": 0.9001,
+      "step": 7070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6525893374965193,
+      "learning_rate": 1.903754670679935e-05,
+      "loss": 0.816,
+      "step": 7071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8658566781054082,
+      "learning_rate": 1.9037220060146637e-05,
+      "loss": 0.9903,
+      "step": 7072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.438310043612567,
+      "learning_rate": 1.9036893360876373e-05,
+      "loss": 1.0126,
+      "step": 7073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8874517438050336,
+      "learning_rate": 1.9036566608990452e-05,
+      "loss": 0.9587,
+      "step": 7074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.995592287068795,
+      "learning_rate": 1.9036239804490778e-05,
+      "loss": 0.8915,
+      "step": 7075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.42582622077715,
+      "learning_rate": 1.9035912947379258e-05,
+      "loss": 0.8895,
+      "step": 7076
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5752480177921633,
+      "learning_rate": 1.903558603765779e-05,
+      "loss": 0.9134,
+      "step": 7077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.414217878292637,
+      "learning_rate": 1.9035259075328278e-05,
+      "loss": 0.8747,
+      "step": 7078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.882395609551371,
+      "learning_rate": 1.9034932060392633e-05,
+      "loss": 0.9235,
+      "step": 7079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.889673165749745,
+      "learning_rate": 1.903460499285275e-05,
+      "loss": 0.8143,
+      "step": 7080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.874820652672424,
+      "learning_rate": 1.9034277872710533e-05,
+      "loss": 1.0451,
+      "step": 7081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7872795590331712,
+      "learning_rate": 1.9033950699967896e-05,
+      "loss": 0.7313,
+      "step": 7082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.987247693336479,
+      "learning_rate": 1.9033623474626736e-05,
+      "loss": 0.817,
+      "step": 7083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9980946917442872,
+      "learning_rate": 1.9033296196688957e-05,
+      "loss": 0.8586,
+      "step": 7084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.023974572578324,
+      "learning_rate": 1.903296886615647e-05,
+      "loss": 0.9206,
+      "step": 7085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8610248227713821,
+      "learning_rate": 1.9032641483031177e-05,
+      "loss": 0.9152,
+      "step": 7086
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2082414035060114,
+      "learning_rate": 1.9032314047314985e-05,
+      "loss": 0.8643,
+      "step": 7087
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9117963493148742,
+      "learning_rate": 1.9031986559009804e-05,
+      "loss": 0.8942,
+      "step": 7088
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.064230515504794,
+      "learning_rate": 1.9031659018117537e-05,
+      "loss": 0.9613,
+      "step": 7089
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2637024167616504,
+      "learning_rate": 1.903133142464009e-05,
+      "loss": 0.8503,
+      "step": 7090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9129568320692636,
+      "learning_rate": 1.9031003778579376e-05,
+      "loss": 1.0228,
+      "step": 7091
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6865783708966096,
+      "learning_rate": 1.9030676079937296e-05,
+      "loss": 1.0822,
+      "step": 7092
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6940614445332989,
+      "learning_rate": 1.903034832871576e-05,
+      "loss": 0.9605,
+      "step": 7093
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.769678496155524,
+      "learning_rate": 1.903002052491668e-05,
+      "loss": 1.0124,
+      "step": 7094
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9527856437790518,
+      "learning_rate": 1.902969266854196e-05,
+      "loss": 0.97,
+      "step": 7095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9675472012034292,
+      "learning_rate": 1.902936475959351e-05,
+      "loss": 0.8493,
+      "step": 7096
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8004750861622296,
+      "learning_rate": 1.902903679807324e-05,
+      "loss": 0.9173,
+      "step": 7097
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7633473114919707,
+      "learning_rate": 1.902870878398306e-05,
+      "loss": 0.9109,
+      "step": 7098
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.469310955505181,
+      "learning_rate": 1.9028380717324875e-05,
+      "loss": 0.9421,
+      "step": 7099
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.821533061499054,
+      "learning_rate": 1.9028052598100607e-05,
+      "loss": 0.9527,
+      "step": 7100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7739191968358843,
+      "learning_rate": 1.902772442631215e-05,
+      "loss": 0.9188,
+      "step": 7101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.654418367833856,
+      "learning_rate": 1.902739620196143e-05,
+      "loss": 0.9103,
+      "step": 7102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.45949409354631,
+      "learning_rate": 1.902706792505035e-05,
+      "loss": 0.9013,
+      "step": 7103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.283866474133186,
+      "learning_rate": 1.902673959558082e-05,
+      "loss": 0.9939,
+      "step": 7104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7125625661504458,
+      "learning_rate": 1.9026411213554758e-05,
+      "loss": 0.8666,
+      "step": 7105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9793335494240392,
+      "learning_rate": 1.9026082778974065e-05,
+      "loss": 0.858,
+      "step": 7106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7317390027077237,
+      "learning_rate": 1.902575429184067e-05,
+      "loss": 0.991,
+      "step": 7107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6969114687450069,
+      "learning_rate": 1.902542575215647e-05,
+      "loss": 0.7861,
+      "step": 7108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.732715972180481,
+      "learning_rate": 1.902509715992338e-05,
+      "loss": 0.8554,
+      "step": 7109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6669323395562847,
+      "learning_rate": 1.9024768515143325e-05,
+      "loss": 1.0073,
+      "step": 7110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7084947769933048,
+      "learning_rate": 1.902443981781821e-05,
+      "loss": 0.9374,
+      "step": 7111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.020290404456947,
+      "learning_rate": 1.9024111067949945e-05,
+      "loss": 0.6675,
+      "step": 7112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.28701842530952,
+      "learning_rate": 1.902378226554045e-05,
+      "loss": 0.8748,
+      "step": 7113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7384809772231349,
+      "learning_rate": 1.902345341059164e-05,
+      "loss": 0.9876,
+      "step": 7114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6447812280300764,
+      "learning_rate": 1.9023124503105424e-05,
+      "loss": 0.8642,
+      "step": 7115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7639593881666722,
+      "learning_rate": 1.9022795543083722e-05,
+      "loss": 1.023,
+      "step": 7116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7371547876811735,
+      "learning_rate": 1.9022466530528448e-05,
+      "loss": 0.9041,
+      "step": 7117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1605761403369943,
+      "learning_rate": 1.9022137465441518e-05,
+      "loss": 0.8318,
+      "step": 7118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7668564413505246,
+      "learning_rate": 1.9021808347824847e-05,
+      "loss": 0.9446,
+      "step": 7119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.013585866291238,
+      "learning_rate": 1.902147917768035e-05,
+      "loss": 0.7957,
+      "step": 7120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1686469184354173,
+      "learning_rate": 1.9021149955009943e-05,
+      "loss": 0.881,
+      "step": 7121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4657326559972725,
+      "learning_rate": 1.902082067981555e-05,
+      "loss": 0.9029,
+      "step": 7122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9336426998604956,
+      "learning_rate": 1.902049135209908e-05,
+      "loss": 0.9521,
+      "step": 7123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7364704376274425,
+      "learning_rate": 1.9020161971862454e-05,
+      "loss": 0.8771,
+      "step": 7124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7041297880372401,
+      "learning_rate": 1.9019832539107592e-05,
+      "loss": 0.91,
+      "step": 7125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7822716509164838,
+      "learning_rate": 1.9019503053836406e-05,
+      "loss": 0.9098,
+      "step": 7126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.27394565885779,
+      "learning_rate": 1.901917351605082e-05,
+      "loss": 0.9459,
+      "step": 7127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9022126720552774,
+      "learning_rate": 1.901884392575275e-05,
+      "loss": 0.8884,
+      "step": 7128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5017536689373634,
+      "learning_rate": 1.9018514282944115e-05,
+      "loss": 0.8037,
+      "step": 7129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3351905503995134,
+      "learning_rate": 1.901818458762683e-05,
+      "loss": 0.9479,
+      "step": 7130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6081388688892955,
+      "learning_rate": 1.9017854839802824e-05,
+      "loss": 0.8645,
+      "step": 7131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.010000144993853,
+      "learning_rate": 1.9017525039474013e-05,
+      "loss": 0.7824,
+      "step": 7132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9406862470335247,
+      "learning_rate": 1.9017195186642315e-05,
+      "loss": 0.7351,
+      "step": 7133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.014915002779902,
+      "learning_rate": 1.9016865281309652e-05,
+      "loss": 0.8555,
+      "step": 7134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.299559502733518,
+      "learning_rate": 1.9016535323477945e-05,
+      "loss": 0.8475,
+      "step": 7135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9276876633851174,
+      "learning_rate": 1.9016205313149114e-05,
+      "loss": 0.8916,
+      "step": 7136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.732395822668889,
+      "learning_rate": 1.901587525032508e-05,
+      "loss": 0.946,
+      "step": 7137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9404844526566891,
+      "learning_rate": 1.9015545135007766e-05,
+      "loss": 0.8961,
+      "step": 7138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5877578556200067,
+      "learning_rate": 1.9015214967199095e-05,
+      "loss": 0.9778,
+      "step": 7139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.197843154136712,
+      "learning_rate": 1.901488474690099e-05,
+      "loss": 0.8808,
+      "step": 7140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7549251589665846,
+      "learning_rate": 1.901455447411537e-05,
+      "loss": 0.7228,
+      "step": 7141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6372735171658384,
+      "learning_rate": 1.901422414884416e-05,
+      "loss": 0.9785,
+      "step": 7142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7656820812884568,
+      "learning_rate": 1.9013893771089283e-05,
+      "loss": 0.9234,
+      "step": 7143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.931642872301724,
+      "learning_rate": 1.9013563340852664e-05,
+      "loss": 0.9357,
+      "step": 7144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.174798732487337,
+      "learning_rate": 1.9013232858136225e-05,
+      "loss": 0.9684,
+      "step": 7145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7951383699461878,
+      "learning_rate": 1.901290232294189e-05,
+      "loss": 0.9469,
+      "step": 7146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.72051821853079,
+      "learning_rate": 1.9012571735271584e-05,
+      "loss": 0.7493,
+      "step": 7147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6978426808903355,
+      "learning_rate": 1.9012241095127234e-05,
+      "loss": 0.9823,
+      "step": 7148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1495080117603385,
+      "learning_rate": 1.9011910402510762e-05,
+      "loss": 1.0281,
+      "step": 7149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.858303152068104,
+      "learning_rate": 1.9011579657424092e-05,
+      "loss": 1.0002,
+      "step": 7150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8431122816688821,
+      "learning_rate": 1.9011248859869155e-05,
+      "loss": 0.7222,
+      "step": 7151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7723004897391221,
+      "learning_rate": 1.9010918009847877e-05,
+      "loss": 1.0177,
+      "step": 7152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8196526596037734,
+      "learning_rate": 1.9010587107362177e-05,
+      "loss": 0.8399,
+      "step": 7153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5579053633320001,
+      "learning_rate": 1.9010256152413987e-05,
+      "loss": 0.9368,
+      "step": 7154
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.052473017600051,
+      "learning_rate": 1.900992514500523e-05,
+      "loss": 0.7785,
+      "step": 7155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7172914662604579,
+      "learning_rate": 1.900959408513784e-05,
+      "loss": 0.7045,
+      "step": 7156
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6498383235032956,
+      "learning_rate": 1.9009262972813746e-05,
+      "loss": 0.9747,
+      "step": 7157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.04995104713102,
+      "learning_rate": 1.9008931808034864e-05,
+      "loss": 0.9087,
+      "step": 7158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6555309977913641,
+      "learning_rate": 1.900860059080313e-05,
+      "loss": 0.9577,
+      "step": 7159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.553605185146003,
+      "learning_rate": 1.900826932112047e-05,
+      "loss": 0.7677,
+      "step": 7160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.714478082188255,
+      "learning_rate": 1.9007937998988818e-05,
+      "loss": 0.9461,
+      "step": 7161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1231425622335944,
+      "learning_rate": 1.9007606624410094e-05,
+      "loss": 0.9045,
+      "step": 7162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8365363407344772,
+      "learning_rate": 1.9007275197386238e-05,
+      "loss": 0.9134,
+      "step": 7163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4803028725203646,
+      "learning_rate": 1.900694371791917e-05,
+      "loss": 1.1388,
+      "step": 7164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6358447892478238,
+      "learning_rate": 1.9006612186010825e-05,
+      "loss": 1.057,
+      "step": 7165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7063959077098994,
+      "learning_rate": 1.9006280601663132e-05,
+      "loss": 0.8472,
+      "step": 7166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.756113416713315,
+      "learning_rate": 1.9005948964878024e-05,
+      "loss": 0.9767,
+      "step": 7167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7486110267727526,
+      "learning_rate": 1.900561727565743e-05,
+      "loss": 0.8843,
+      "step": 7168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5455697723343365,
+      "learning_rate": 1.900528553400328e-05,
+      "loss": 0.9092,
+      "step": 7169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4793543820278376,
+      "learning_rate": 1.9004953739917503e-05,
+      "loss": 0.8283,
+      "step": 7170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8143452628598388,
+      "learning_rate": 1.900462189340204e-05,
+      "loss": 0.7253,
+      "step": 7171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9160990173135939,
+      "learning_rate": 1.9004289994458815e-05,
+      "loss": 0.965,
+      "step": 7172
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1141287351833893,
+      "learning_rate": 1.9003958043089763e-05,
+      "loss": 0.9705,
+      "step": 7173
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7126029833416698,
+      "learning_rate": 1.900362603929682e-05,
+      "loss": 0.8125,
+      "step": 7174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.070069036614125,
+      "learning_rate": 1.900329398308191e-05,
+      "loss": 0.9256,
+      "step": 7175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.094362345164431,
+      "learning_rate": 1.9002961874446973e-05,
+      "loss": 0.996,
+      "step": 7176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.731416521906107,
+      "learning_rate": 1.9002629713393946e-05,
+      "loss": 0.8649,
+      "step": 7177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3822417296855622,
+      "learning_rate": 1.9002297499924752e-05,
+      "loss": 0.9445,
+      "step": 7178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6141766209991295,
+      "learning_rate": 1.900196523404134e-05,
+      "loss": 0.9042,
+      "step": 7179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7285671122070307,
+      "learning_rate": 1.900163291574563e-05,
+      "loss": 0.9005,
+      "step": 7180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6412261527999938,
+      "learning_rate": 1.9001300545039566e-05,
+      "loss": 0.9159,
+      "step": 7181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5564111404476868,
+      "learning_rate": 1.900096812192508e-05,
+      "loss": 0.9303,
+      "step": 7182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.658288020604283,
+      "learning_rate": 1.9000635646404108e-05,
+      "loss": 0.9539,
+      "step": 7183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7166468870571159,
+      "learning_rate": 1.9000303118478584e-05,
+      "loss": 0.8712,
+      "step": 7184
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6567314904172432,
+      "learning_rate": 1.899997053815045e-05,
+      "loss": 0.8836,
+      "step": 7185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5414252880409973,
+      "learning_rate": 1.8999637905421633e-05,
+      "loss": 0.9359,
+      "step": 7186
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7344080624752924,
+      "learning_rate": 1.899930522029408e-05,
+      "loss": 0.9694,
+      "step": 7187
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8011891743842493,
+      "learning_rate": 1.899897248276972e-05,
+      "loss": 0.7283,
+      "step": 7188
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.075191919253885,
+      "learning_rate": 1.8998639692850492e-05,
+      "loss": 1.0114,
+      "step": 7189
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8560788310463785,
+      "learning_rate": 1.899830685053834e-05,
+      "loss": 0.965,
+      "step": 7190
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.882895377167001,
+      "learning_rate": 1.8997973955835193e-05,
+      "loss": 1.0409,
+      "step": 7191
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6029719062415198,
+      "learning_rate": 1.899764100874299e-05,
+      "loss": 0.9345,
+      "step": 7192
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.665576798167966,
+      "learning_rate": 1.8997308009263675e-05,
+      "loss": 0.9398,
+      "step": 7193
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6352452349525421,
+      "learning_rate": 1.8996974957399186e-05,
+      "loss": 0.8627,
+      "step": 7194
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7776250890189658,
+      "learning_rate": 1.8996641853151462e-05,
+      "loss": 0.9273,
+      "step": 7195
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8613661933718129,
+      "learning_rate": 1.8996308696522435e-05,
+      "loss": 0.9692,
+      "step": 7196
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7647288216366053,
+      "learning_rate": 1.8995975487514052e-05,
+      "loss": 1.0812,
+      "step": 7197
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8029946884354537,
+      "learning_rate": 1.8995642226128256e-05,
+      "loss": 0.8167,
+      "step": 7198
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5492447715327249,
+      "learning_rate": 1.899530891236698e-05,
+      "loss": 0.9391,
+      "step": 7199
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.54083033688551,
+      "learning_rate": 1.8994975546232166e-05,
+      "loss": 0.7976,
+      "step": 7200
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.61890442932836,
+      "learning_rate": 1.899464212772576e-05,
+      "loss": 0.8125,
+      "step": 7201
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8258408806080278,
+      "learning_rate": 1.8994308656849697e-05,
+      "loss": 0.7582,
+      "step": 7202
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0065372526694785,
+      "learning_rate": 1.8993975133605925e-05,
+      "loss": 0.7984,
+      "step": 7203
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8345852161446259,
+      "learning_rate": 1.8993641557996377e-05,
+      "loss": 1.0624,
+      "step": 7204
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9156457327229592,
+      "learning_rate": 1.8993307930023005e-05,
+      "loss": 0.9757,
+      "step": 7205
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9581279636170563,
+      "learning_rate": 1.8992974249687747e-05,
+      "loss": 1.0083,
+      "step": 7206
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7085562577792366,
+      "learning_rate": 1.8992640516992544e-05,
+      "loss": 0.8854,
+      "step": 7207
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6673873624388227,
+      "learning_rate": 1.899230673193934e-05,
+      "loss": 0.9658,
+      "step": 7208
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.681113887390508,
+      "learning_rate": 1.899197289453008e-05,
+      "loss": 0.8162,
+      "step": 7209
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8181211356938011,
+      "learning_rate": 1.8991639004766708e-05,
+      "loss": 0.7613,
+      "step": 7210
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6983856640262323,
+      "learning_rate": 1.8991305062651167e-05,
+      "loss": 0.826,
+      "step": 7211
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7358272521709467,
+      "learning_rate": 1.8990971068185402e-05,
+      "loss": 0.7398,
+      "step": 7212
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8028340625135464,
+      "learning_rate": 1.8990637021371355e-05,
+      "loss": 0.9121,
+      "step": 7213
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8322771818290764,
+      "learning_rate": 1.8990302922210976e-05,
+      "loss": 0.9862,
+      "step": 7214
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4507534966379745,
+      "learning_rate": 1.89899687707062e-05,
+      "loss": 0.8307,
+      "step": 7215
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9848109586647587,
+      "learning_rate": 1.8989634566858987e-05,
+      "loss": 0.9205,
+      "step": 7216
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.291812606374404,
+      "learning_rate": 1.8989300310671273e-05,
+      "loss": 1.0045,
+      "step": 7217
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9955385231085714,
+      "learning_rate": 1.8988966002145007e-05,
+      "loss": 0.9244,
+      "step": 7218
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6163263996916284,
+      "learning_rate": 1.8988631641282134e-05,
+      "loss": 0.9852,
+      "step": 7219
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9225289890397455,
+      "learning_rate": 1.89882972280846e-05,
+      "loss": 0.8232,
+      "step": 7220
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7875246194658074,
+      "learning_rate": 1.8987962762554358e-05,
+      "loss": 1.1074,
+      "step": 7221
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8280919203009462,
+      "learning_rate": 1.898762824469335e-05,
+      "loss": 0.9943,
+      "step": 7222
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6517326553240999,
+      "learning_rate": 1.8987293674503525e-05,
+      "loss": 0.9858,
+      "step": 7223
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7917370292206939,
+      "learning_rate": 1.898695905198683e-05,
+      "loss": 0.7364,
+      "step": 7224
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5010799907563797,
+      "learning_rate": 1.8986624377145216e-05,
+      "loss": 0.777,
+      "step": 7225
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8612986659686361,
+      "learning_rate": 1.8986289649980627e-05,
+      "loss": 0.913,
+      "step": 7226
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8405774265211157,
+      "learning_rate": 1.8985954870495016e-05,
+      "loss": 0.8858,
+      "step": 7227
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2576133953883644,
+      "learning_rate": 1.8985620038690332e-05,
+      "loss": 0.8932,
+      "step": 7228
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6357199898679733,
+      "learning_rate": 1.8985285154568522e-05,
+      "loss": 0.8895,
+      "step": 7229
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.849020093431488,
+      "learning_rate": 1.8984950218131538e-05,
+      "loss": 0.9259,
+      "step": 7230
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3216458562668687,
+      "learning_rate": 1.8984615229381325e-05,
+      "loss": 0.8759,
+      "step": 7231
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4673898029617856,
+      "learning_rate": 1.8984280188319845e-05,
+      "loss": 0.8422,
+      "step": 7232
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.336256152131911,
+      "learning_rate": 1.8983945094949035e-05,
+      "loss": 0.9326,
+      "step": 7233
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8021034982312543,
+      "learning_rate": 1.8983609949270854e-05,
+      "loss": 0.9636,
+      "step": 7234
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5974985729961904,
+      "learning_rate": 1.8983274751287254e-05,
+      "loss": 0.9271,
+      "step": 7235
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.78832711588969,
+      "learning_rate": 1.8982939501000182e-05,
+      "loss": 0.823,
+      "step": 7236
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6896971412311954,
+      "learning_rate": 1.8982604198411594e-05,
+      "loss": 0.9397,
+      "step": 7237
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.768421844146923,
+      "learning_rate": 1.898226884352344e-05,
+      "loss": 0.916,
+      "step": 7238
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8060070045367107,
+      "learning_rate": 1.8981933436337675e-05,
+      "loss": 0.7441,
+      "step": 7239
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9663988561116137,
+      "learning_rate": 1.8981597976856247e-05,
+      "loss": 0.8777,
+      "step": 7240
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7444188587010743,
+      "learning_rate": 1.8981262465081112e-05,
+      "loss": 0.7635,
+      "step": 7241
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6071910043079074,
+      "learning_rate": 1.8980926901014226e-05,
+      "loss": 0.8634,
+      "step": 7242
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5926247918189473,
+      "learning_rate": 1.8980591284657536e-05,
+      "loss": 0.9757,
+      "step": 7243
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.205908333031114,
+      "learning_rate": 1.8980255616013004e-05,
+      "loss": 0.9923,
+      "step": 7244
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0170605798935726,
+      "learning_rate": 1.8979919895082578e-05,
+      "loss": 0.8979,
+      "step": 7245
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7039720364965065,
+      "learning_rate": 1.8979584121868217e-05,
+      "loss": 0.8964,
+      "step": 7246
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5126419891064178,
+      "learning_rate": 1.8979248296371876e-05,
+      "loss": 0.9228,
+      "step": 7247
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.66333890224229,
+      "learning_rate": 1.897891241859551e-05,
+      "loss": 0.8194,
+      "step": 7248
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7623757663652864,
+      "learning_rate": 1.8978576488541068e-05,
+      "loss": 0.9349,
+      "step": 7249
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3286242269067565,
+      "learning_rate": 1.897824050621051e-05,
+      "loss": 0.9513,
+      "step": 7250
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9238730741528893,
+      "learning_rate": 1.8977904471605798e-05,
+      "loss": 0.888,
+      "step": 7251
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7751262844910634,
+      "learning_rate": 1.8977568384728885e-05,
+      "loss": 0.9933,
+      "step": 7252
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5691045010059688,
+      "learning_rate": 1.8977232245581722e-05,
+      "loss": 0.8734,
+      "step": 7253
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7310349605259527,
+      "learning_rate": 1.8976896054166273e-05,
+      "loss": 0.9943,
+      "step": 7254
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6767883862797337,
+      "learning_rate": 1.8976559810484494e-05,
+      "loss": 1.0156,
+      "step": 7255
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1538885055971346,
+      "learning_rate": 1.8976223514538342e-05,
+      "loss": 0.7617,
+      "step": 7256
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9146814370745666,
+      "learning_rate": 1.8975887166329773e-05,
+      "loss": 0.8511,
+      "step": 7257
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9729891634803494,
+      "learning_rate": 1.8975550765860745e-05,
+      "loss": 0.8217,
+      "step": 7258
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.770226810321702,
+      "learning_rate": 1.8975214313133223e-05,
+      "loss": 1.0247,
+      "step": 7259
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.650806385831335,
+      "learning_rate": 1.8974877808149162e-05,
+      "loss": 0.8744,
+      "step": 7260
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.92507721282735,
+      "learning_rate": 1.897454125091052e-05,
+      "loss": 0.9909,
+      "step": 7261
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.722748711183635,
+      "learning_rate": 1.8974204641419256e-05,
+      "loss": 1.0574,
+      "step": 7262
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5781142810862347,
+      "learning_rate": 1.8973867979677332e-05,
+      "loss": 0.8171,
+      "step": 7263
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6027864090254798,
+      "learning_rate": 1.8973531265686706e-05,
+      "loss": 0.8112,
+      "step": 7264
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6001063649068903,
+      "learning_rate": 1.8973194499449342e-05,
+      "loss": 0.8138,
+      "step": 7265
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8077820538045893,
+      "learning_rate": 1.89728576809672e-05,
+      "loss": 0.7799,
+      "step": 7266
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8243930511170279,
+      "learning_rate": 1.897252081024224e-05,
+      "loss": 0.9075,
+      "step": 7267
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7046452488243353,
+      "learning_rate": 1.8972183887276422e-05,
+      "loss": 0.8545,
+      "step": 7268
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9071550707906284,
+      "learning_rate": 1.8971846912071707e-05,
+      "loss": 1.0595,
+      "step": 7269
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8075529548189089,
+      "learning_rate": 1.897150988463006e-05,
+      "loss": 0.8648,
+      "step": 7270
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8079827774647437,
+      "learning_rate": 1.8971172804953444e-05,
+      "loss": 0.785,
+      "step": 7271
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.166167375407458,
+      "learning_rate": 1.897083567304382e-05,
+      "loss": 0.9277,
+      "step": 7272
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7595038596721237,
+      "learning_rate": 1.8970498488903147e-05,
+      "loss": 0.8691,
+      "step": 7273
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.078202531984317,
+      "learning_rate": 1.8970161252533393e-05,
+      "loss": 0.9609,
+      "step": 7274
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.535451065470087,
+      "learning_rate": 1.8969823963936524e-05,
+      "loss": 0.8786,
+      "step": 7275
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7344268372439449,
+      "learning_rate": 1.8969486623114496e-05,
+      "loss": 0.8293,
+      "step": 7276
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6947152606701676,
+      "learning_rate": 1.8969149230069278e-05,
+      "loss": 0.8113,
+      "step": 7277
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6708730394577915,
+      "learning_rate": 1.8968811784802833e-05,
+      "loss": 0.9524,
+      "step": 7278
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.816915269994544,
+      "learning_rate": 1.896847428731713e-05,
+      "loss": 0.8515,
+      "step": 7279
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8355081324212592,
+      "learning_rate": 1.8968136737614123e-05,
+      "loss": 0.8027,
+      "step": 7280
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5548984534640182,
+      "learning_rate": 1.8967799135695788e-05,
+      "loss": 0.8552,
+      "step": 7281
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.48061929702214,
+      "learning_rate": 1.896746148156409e-05,
+      "loss": 1.0,
+      "step": 7282
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9053869449324825,
+      "learning_rate": 1.896712377522099e-05,
+      "loss": 1.0486,
+      "step": 7283
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4646956995889493,
+      "learning_rate": 1.8966786016668458e-05,
+      "loss": 1.0143,
+      "step": 7284
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.885781764599685,
+      "learning_rate": 1.8966448205908458e-05,
+      "loss": 0.8756,
+      "step": 7285
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8243071304750866,
+      "learning_rate": 1.8966110342942954e-05,
+      "loss": 0.9252,
+      "step": 7286
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.754176551700751,
+      "learning_rate": 1.8965772427773923e-05,
+      "loss": 0.8264,
+      "step": 7287
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8385444470168621,
+      "learning_rate": 1.8965434460403325e-05,
+      "loss": 0.9383,
+      "step": 7288
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1497285810660554,
+      "learning_rate": 1.896509644083313e-05,
+      "loss": 1.0047,
+      "step": 7289
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7041856081343498,
+      "learning_rate": 1.8964758369065303e-05,
+      "loss": 1.0598,
+      "step": 7290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6754449877989255,
+      "learning_rate": 1.8964420245101816e-05,
+      "loss": 0.8436,
+      "step": 7291
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.895660876728755,
+      "learning_rate": 1.8964082068944634e-05,
+      "loss": 1.0134,
+      "step": 7292
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8261414097760267,
+      "learning_rate": 1.896374384059573e-05,
+      "loss": 0.9272,
+      "step": 7293
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8087071668773701,
+      "learning_rate": 1.896340556005707e-05,
+      "loss": 0.775,
+      "step": 7294
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6051848267024023,
+      "learning_rate": 1.8963067227330624e-05,
+      "loss": 0.9783,
+      "step": 7295
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0356027290783265,
+      "learning_rate": 1.8962728842418366e-05,
+      "loss": 0.874,
+      "step": 7296
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8788340802611398,
+      "learning_rate": 1.8962390405322263e-05,
+      "loss": 0.8825,
+      "step": 7297
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8555680518180426,
+      "learning_rate": 1.8962051916044288e-05,
+      "loss": 1.0032,
+      "step": 7298
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8828380232593145,
+      "learning_rate": 1.8961713374586403e-05,
+      "loss": 0.9621,
+      "step": 7299
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.099606637093511,
+      "learning_rate": 1.8961374780950588e-05,
+      "loss": 0.8645,
+      "step": 7300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4829815541442397,
+      "learning_rate": 1.8961036135138813e-05,
+      "loss": 0.8691,
+      "step": 7301
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8761172763112328,
+      "learning_rate": 1.8960697437153048e-05,
+      "loss": 0.9524,
+      "step": 7302
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.017183653046305,
+      "learning_rate": 1.8960358686995265e-05,
+      "loss": 0.9355,
+      "step": 7303
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7277155920270026,
+      "learning_rate": 1.896001988466744e-05,
+      "loss": 1.0054,
+      "step": 7304
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.718920178177351,
+      "learning_rate": 1.895968103017154e-05,
+      "loss": 0.8806,
+      "step": 7305
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4877755028431636,
+      "learning_rate": 1.8959342123509538e-05,
+      "loss": 0.877,
+      "step": 7306
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.803049630981803,
+      "learning_rate": 1.8959003164683416e-05,
+      "loss": 0.8417,
+      "step": 7307
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.628156962387108,
+      "learning_rate": 1.8958664153695138e-05,
+      "loss": 1.0239,
+      "step": 7308
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7119028061275272,
+      "learning_rate": 1.895832509054668e-05,
+      "loss": 0.8676,
+      "step": 7309
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.301137472178733,
+      "learning_rate": 1.8957985975240016e-05,
+      "loss": 0.8346,
+      "step": 7310
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8757227702887382,
+      "learning_rate": 1.8957646807777123e-05,
+      "loss": 0.9839,
+      "step": 7311
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.698330631720405,
+      "learning_rate": 1.8957307588159978e-05,
+      "loss": 1.0187,
+      "step": 7312
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7762477409041144,
+      "learning_rate": 1.8956968316390547e-05,
+      "loss": 1.0783,
+      "step": 7313
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7935669522992648,
+      "learning_rate": 1.8956628992470812e-05,
+      "loss": 0.9933,
+      "step": 7314
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8336305421856224,
+      "learning_rate": 1.895628961640275e-05,
+      "loss": 0.7294,
+      "step": 7315
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7053602703176767,
+      "learning_rate": 1.895595018818833e-05,
+      "loss": 0.9816,
+      "step": 7316
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.166836451705553,
+      "learning_rate": 1.8955610707829536e-05,
+      "loss": 0.7722,
+      "step": 7317
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.762401227148261,
+      "learning_rate": 1.895527117532834e-05,
+      "loss": 0.8517,
+      "step": 7318
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5340631073502653,
+      "learning_rate": 1.8954931590686722e-05,
+      "loss": 0.7978,
+      "step": 7319
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6938564741234192,
+      "learning_rate": 1.8954591953906657e-05,
+      "loss": 0.809,
+      "step": 7320
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.575579430663909,
+      "learning_rate": 1.8954252264990122e-05,
+      "loss": 0.9509,
+      "step": 7321
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6917621771159048,
+      "learning_rate": 1.895391252393909e-05,
+      "loss": 0.9626,
+      "step": 7322
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2564936468163346,
+      "learning_rate": 1.8953572730755552e-05,
+      "loss": 0.8589,
+      "step": 7323
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9614239656296566,
+      "learning_rate": 1.8953232885441474e-05,
+      "loss": 0.9211,
+      "step": 7324
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.701288563205997,
+      "learning_rate": 1.8952892987998844e-05,
+      "loss": 1.0817,
+      "step": 7325
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6995871086726242,
+      "learning_rate": 1.8952553038429634e-05,
+      "loss": 0.9286,
+      "step": 7326
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7582113281935872,
+      "learning_rate": 1.8952213036735826e-05,
+      "loss": 1.0314,
+      "step": 7327
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0049895080139497,
+      "learning_rate": 1.8951872982919398e-05,
+      "loss": 0.8469,
+      "step": 7328
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.627718261494695,
+      "learning_rate": 1.8951532876982335e-05,
+      "loss": 0.8938,
+      "step": 7329
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6068743571782382,
+      "learning_rate": 1.895119271892661e-05,
+      "loss": 0.9588,
+      "step": 7330
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6137075035709405,
+      "learning_rate": 1.895085250875421e-05,
+      "loss": 0.9789,
+      "step": 7331
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9718789042042746,
+      "learning_rate": 1.8950512246467113e-05,
+      "loss": 0.9329,
+      "step": 7332
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6416889168992064,
+      "learning_rate": 1.8950171932067297e-05,
+      "loss": 0.9311,
+      "step": 7333
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6948628558056376,
+      "learning_rate": 1.8949831565556747e-05,
+      "loss": 0.9688,
+      "step": 7334
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.729292151023185,
+      "learning_rate": 1.8949491146937446e-05,
+      "loss": 0.881,
+      "step": 7335
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7492208885389413,
+      "learning_rate": 1.8949150676211374e-05,
+      "loss": 1.0265,
+      "step": 7336
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6306681024719747,
+      "learning_rate": 1.8948810153380514e-05,
+      "loss": 0.8941,
+      "step": 7337
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6656500716398568,
+      "learning_rate": 1.8948469578446845e-05,
+      "loss": 0.8783,
+      "step": 7338
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8409332275446,
+      "learning_rate": 1.8948128951412356e-05,
+      "loss": 0.781,
+      "step": 7339
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7501722121793513,
+      "learning_rate": 1.8947788272279028e-05,
+      "loss": 0.9462,
+      "step": 7340
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7600413819477232,
+      "learning_rate": 1.8947447541048844e-05,
+      "loss": 0.7279,
+      "step": 7341
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6022319205884819,
+      "learning_rate": 1.8947106757723782e-05,
+      "loss": 0.951,
+      "step": 7342
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.757504444149762,
+      "learning_rate": 1.8946765922305835e-05,
+      "loss": 0.808,
+      "step": 7343
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.893251355163327,
+      "learning_rate": 1.8946425034796988e-05,
+      "loss": 0.9605,
+      "step": 7344
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8596535231334153,
+      "learning_rate": 1.8946084095199216e-05,
+      "loss": 0.9589,
+      "step": 7345
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.67237131482304,
+      "learning_rate": 1.8945743103514516e-05,
+      "loss": 0.8526,
+      "step": 7346
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.64544613636652,
+      "learning_rate": 1.8945402059744865e-05,
+      "loss": 0.9249,
+      "step": 7347
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7609292963517321,
+      "learning_rate": 1.894506096389225e-05,
+      "loss": 0.7089,
+      "step": 7348
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0194555498897486,
+      "learning_rate": 1.894471981595866e-05,
+      "loss": 0.8641,
+      "step": 7349
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5392746756694384,
+      "learning_rate": 1.8944378615946077e-05,
+      "loss": 1.053,
+      "step": 7350
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.610445895310708,
+      "learning_rate": 1.8944037363856492e-05,
+      "loss": 0.8665,
+      "step": 7351
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.031611143186199,
+      "learning_rate": 1.894369605969189e-05,
+      "loss": 0.9047,
+      "step": 7352
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.690013468227407,
+      "learning_rate": 1.894335470345426e-05,
+      "loss": 0.8509,
+      "step": 7353
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7474809906452592,
+      "learning_rate": 1.894301329514558e-05,
+      "loss": 0.7164,
+      "step": 7354
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.78769548933036,
+      "learning_rate": 1.894267183476785e-05,
+      "loss": 0.9931,
+      "step": 7355
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8545554169013029,
+      "learning_rate": 1.8942330322323055e-05,
+      "loss": 0.6738,
+      "step": 7356
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8608057210213997,
+      "learning_rate": 1.894198875781318e-05,
+      "loss": 0.9749,
+      "step": 7357
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8086450216585737,
+      "learning_rate": 1.8941647141240213e-05,
+      "loss": 1.0112,
+      "step": 7358
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8765566201265798,
+      "learning_rate": 1.894130547260615e-05,
+      "loss": 0.8833,
+      "step": 7359
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7821583090410982,
+      "learning_rate": 1.8940963751912976e-05,
+      "loss": 0.8373,
+      "step": 7360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.636680789509576,
+      "learning_rate": 1.8940621979162678e-05,
+      "loss": 0.9141,
+      "step": 7361
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7888314461870392,
+      "learning_rate": 1.8940280154357248e-05,
+      "loss": 0.858,
+      "step": 7362
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8994386880223402,
+      "learning_rate": 1.8939938277498675e-05,
+      "loss": 0.8809,
+      "step": 7363
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9577674470849153,
+      "learning_rate": 1.8939596348588955e-05,
+      "loss": 0.8835,
+      "step": 7364
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7781011386201615,
+      "learning_rate": 1.893925436763007e-05,
+      "loss": 0.7235,
+      "step": 7365
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.526008756199567,
+      "learning_rate": 1.893891233462402e-05,
+      "loss": 0.963,
+      "step": 7366
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7921304925772178,
+      "learning_rate": 1.8938570249572793e-05,
+      "loss": 0.7457,
+      "step": 7367
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.273440229893652,
+      "learning_rate": 1.8938228112478377e-05,
+      "loss": 1.0972,
+      "step": 7368
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.615156600471727,
+      "learning_rate": 1.8937885923342768e-05,
+      "loss": 0.843,
+      "step": 7369
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7717841215497034,
+      "learning_rate": 1.893754368216796e-05,
+      "loss": 0.6897,
+      "step": 7370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6776723849895159,
+      "learning_rate": 1.893720138895594e-05,
+      "loss": 0.8635,
+      "step": 7371
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8117170814842771,
+      "learning_rate": 1.8936859043708705e-05,
+      "loss": 0.8707,
+      "step": 7372
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8607135341554453,
+      "learning_rate": 1.893651664642825e-05,
+      "loss": 0.8518,
+      "step": 7373
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.586083082727834,
+      "learning_rate": 1.893617419711656e-05,
+      "loss": 0.962,
+      "step": 7374
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9236269101490147,
+      "learning_rate": 1.893583169577564e-05,
+      "loss": 0.8618,
+      "step": 7375
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6454216071243843,
+      "learning_rate": 1.8935489142407474e-05,
+      "loss": 0.957,
+      "step": 7376
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8290745172414347,
+      "learning_rate": 1.8935146537014064e-05,
+      "loss": 1.0019,
+      "step": 7377
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1587665002169816,
+      "learning_rate": 1.8934803879597402e-05,
+      "loss": 0.8852,
+      "step": 7378
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7453277547376362,
+      "learning_rate": 1.8934461170159483e-05,
+      "loss": 1.0022,
+      "step": 7379
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7018815695089622,
+      "learning_rate": 1.8934118408702306e-05,
+      "loss": 0.9269,
+      "step": 7380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.937723853657035,
+      "learning_rate": 1.8933775595227856e-05,
+      "loss": 0.88,
+      "step": 7381
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.700729721068451,
+      "learning_rate": 1.893343272973814e-05,
+      "loss": 0.9618,
+      "step": 7382
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.337042781449766,
+      "learning_rate": 1.8933089812235153e-05,
+      "loss": 0.8635,
+      "step": 7383
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.133722117850204,
+      "learning_rate": 1.8932746842720884e-05,
+      "loss": 0.9818,
+      "step": 7384
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9547302363394652,
+      "learning_rate": 1.8932403821197338e-05,
+      "loss": 0.8483,
+      "step": 7385
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7662939483208937,
+      "learning_rate": 1.8932060747666504e-05,
+      "loss": 1.0126,
+      "step": 7386
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7058253549351006,
+      "learning_rate": 1.8931717622130392e-05,
+      "loss": 0.9588,
+      "step": 7387
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9022148714731355,
+      "learning_rate": 1.8931374444590987e-05,
+      "loss": 0.8625,
+      "step": 7388
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7402840394434718,
+      "learning_rate": 1.8931031215050295e-05,
+      "loss": 0.9277,
+      "step": 7389
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.001296977059854,
+      "learning_rate": 1.893068793351031e-05,
+      "loss": 0.8674,
+      "step": 7390
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8878227024453584,
+      "learning_rate": 1.8930344599973035e-05,
+      "loss": 0.9823,
+      "step": 7391
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6806190166124935,
+      "learning_rate": 1.8930001214440464e-05,
+      "loss": 0.8949,
+      "step": 7392
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.649801388914835,
+      "learning_rate": 1.89296577769146e-05,
+      "loss": 0.8754,
+      "step": 7393
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6389419260644826,
+      "learning_rate": 1.892931428739744e-05,
+      "loss": 0.8074,
+      "step": 7394
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.679868798933571,
+      "learning_rate": 1.8928970745890985e-05,
+      "loss": 0.8264,
+      "step": 7395
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6262552300173816,
+      "learning_rate": 1.8928627152397237e-05,
+      "loss": 0.9471,
+      "step": 7396
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6460659757332055,
+      "learning_rate": 1.8928283506918194e-05,
+      "loss": 1.1615,
+      "step": 7397
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6772013744350673,
+      "learning_rate": 1.892793980945586e-05,
+      "loss": 1.0196,
+      "step": 7398
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6509504662365504,
+      "learning_rate": 1.892759606001223e-05,
+      "loss": 0.8365,
+      "step": 7399
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7519713791653482,
+      "learning_rate": 1.8927252258589312e-05,
+      "loss": 0.7662,
+      "step": 7400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7334530169161722,
+      "learning_rate": 1.8926908405189105e-05,
+      "loss": 0.9055,
+      "step": 7401
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7672279921829426,
+      "learning_rate": 1.8926564499813608e-05,
+      "loss": 0.734,
+      "step": 7402
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6479921688911305,
+      "learning_rate": 1.8926220542464833e-05,
+      "loss": 0.8873,
+      "step": 7403
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8000816159736867,
+      "learning_rate": 1.8925876533144772e-05,
+      "loss": 0.7164,
+      "step": 7404
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.617110556509652,
+      "learning_rate": 1.892553247185543e-05,
+      "loss": 0.9338,
+      "step": 7405
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.706827203109585,
+      "learning_rate": 1.8925188358598815e-05,
+      "loss": 0.9002,
+      "step": 7406
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.879316452867438,
+      "learning_rate": 1.8924844193376926e-05,
+      "loss": 0.8181,
+      "step": 7407
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8964599532821025,
+      "learning_rate": 1.892449997619177e-05,
+      "loss": 1.0205,
+      "step": 7408
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6814120145717213,
+      "learning_rate": 1.8924155707045348e-05,
+      "loss": 0.9512,
+      "step": 7409
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6392862532731887,
+      "learning_rate": 1.8923811385939666e-05,
+      "loss": 0.9801,
+      "step": 7410
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8526486145302137,
+      "learning_rate": 1.892346701287673e-05,
+      "loss": 0.9809,
+      "step": 7411
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1751339585641967,
+      "learning_rate": 1.8923122587858543e-05,
+      "loss": 1.0131,
+      "step": 7412
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9883255069515526,
+      "learning_rate": 1.8922778110887114e-05,
+      "loss": 0.9441,
+      "step": 7413
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.534602485458821,
+      "learning_rate": 1.8922433581964443e-05,
+      "loss": 0.9759,
+      "step": 7414
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2321496943955568,
+      "learning_rate": 1.892208900109254e-05,
+      "loss": 1.0154,
+      "step": 7415
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6620003992707284,
+      "learning_rate": 1.892174436827341e-05,
+      "loss": 0.9535,
+      "step": 7416
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8042178944742296,
+      "learning_rate": 1.892139968350906e-05,
+      "loss": 0.7122,
+      "step": 7417
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4880674806582532,
+      "learning_rate": 1.8921054946801497e-05,
+      "loss": 1.0625,
+      "step": 7418
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9840384808502936,
+      "learning_rate": 1.8920710158152727e-05,
+      "loss": 0.7672,
+      "step": 7419
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7279144505156367,
+      "learning_rate": 1.8920365317564758e-05,
+      "loss": 0.8287,
+      "step": 7420
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7627260234992628,
+      "learning_rate": 1.8920020425039598e-05,
+      "loss": 0.7134,
+      "step": 7421
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7533410337487234,
+      "learning_rate": 1.8919675480579256e-05,
+      "loss": 0.9747,
+      "step": 7422
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6766536826078964,
+      "learning_rate": 1.891933048418574e-05,
+      "loss": 1.0608,
+      "step": 7423
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7787145841520661,
+      "learning_rate": 1.8918985435861056e-05,
+      "loss": 0.7645,
+      "step": 7424
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7727361115028724,
+      "learning_rate": 1.8918640335607215e-05,
+      "loss": 0.6972,
+      "step": 7425
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7117320077219251,
+      "learning_rate": 1.8918295183426226e-05,
+      "loss": 0.9194,
+      "step": 7426
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7445877825307354,
+      "learning_rate": 1.89179499793201e-05,
+      "loss": 0.9319,
+      "step": 7427
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.681547388047658,
+      "learning_rate": 1.891760472329085e-05,
+      "loss": 0.812,
+      "step": 7428
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.010988676562553,
+      "learning_rate": 1.891725941534048e-05,
+      "loss": 0.8744,
+      "step": 7429
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7387923532517158,
+      "learning_rate": 1.8916914055471e-05,
+      "loss": 0.7915,
+      "step": 7430
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.541186544229585,
+      "learning_rate": 1.891656864368442e-05,
+      "loss": 0.9141,
+      "step": 7431
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9296257456136072,
+      "learning_rate": 1.891622317998276e-05,
+      "loss": 0.9512,
+      "step": 7432
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6804262646354917,
+      "learning_rate": 1.8915877664368026e-05,
+      "loss": 0.925,
+      "step": 7433
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3094332548236176,
+      "learning_rate": 1.8915532096842225e-05,
+      "loss": 0.8556,
+      "step": 7434
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8490813597210696,
+      "learning_rate": 1.8915186477407376e-05,
+      "loss": 1.0565,
+      "step": 7435
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8476956147092283,
+      "learning_rate": 1.891484080606549e-05,
+      "loss": 0.7545,
+      "step": 7436
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6526004907822545,
+      "learning_rate": 1.8914495082818578e-05,
+      "loss": 0.8699,
+      "step": 7437
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9067883136787518,
+      "learning_rate": 1.891414930766865e-05,
+      "loss": 0.9376,
+      "step": 7438
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.348878668927802,
+      "learning_rate": 1.891380348061773e-05,
+      "loss": 0.9407,
+      "step": 7439
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9376352921419733,
+      "learning_rate": 1.891345760166782e-05,
+      "loss": 0.9692,
+      "step": 7440
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8682377855257088,
+      "learning_rate": 1.8913111670820936e-05,
+      "loss": 1.0124,
+      "step": 7441
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3872522802980762,
+      "learning_rate": 1.8912765688079097e-05,
+      "loss": 0.8166,
+      "step": 7442
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8388970953038224,
+      "learning_rate": 1.8912419653444313e-05,
+      "loss": 0.8075,
+      "step": 7443
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.640477862032805,
+      "learning_rate": 1.8912073566918597e-05,
+      "loss": 0.8133,
+      "step": 7444
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8060368968691334,
+      "learning_rate": 1.891172742850397e-05,
+      "loss": 0.7685,
+      "step": 7445
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9350050745663674,
+      "learning_rate": 1.8911381238202447e-05,
+      "loss": 0.8781,
+      "step": 7446
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8155126517136624,
+      "learning_rate": 1.891103499601604e-05,
+      "loss": 0.7786,
+      "step": 7447
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5343339971324843,
+      "learning_rate": 1.8910688701946763e-05,
+      "loss": 0.9489,
+      "step": 7448
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.321596487414115,
+      "learning_rate": 1.8910342355996637e-05,
+      "loss": 0.8551,
+      "step": 7449
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.122340785571428,
+      "learning_rate": 1.8909995958167678e-05,
+      "loss": 0.9209,
+      "step": 7450
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7116693587636178,
+      "learning_rate": 1.89096495084619e-05,
+      "loss": 0.9485,
+      "step": 7451
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1786913380100357,
+      "learning_rate": 1.8909303006881323e-05,
+      "loss": 0.9615,
+      "step": 7452
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7844656902862894,
+      "learning_rate": 1.8908956453427963e-05,
+      "loss": 0.9891,
+      "step": 7453
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7127706650755,
+      "learning_rate": 1.890860984810384e-05,
+      "loss": 0.9102,
+      "step": 7454
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8270905455955226,
+      "learning_rate": 1.8908263190910966e-05,
+      "loss": 0.8346,
+      "step": 7455
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.411203239766871,
+      "learning_rate": 1.8907916481851364e-05,
+      "loss": 0.9104,
+      "step": 7456
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.707649160176564,
+      "learning_rate": 1.8907569720927055e-05,
+      "loss": 1.0091,
+      "step": 7457
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8197245323769495,
+      "learning_rate": 1.8907222908140053e-05,
+      "loss": 0.8946,
+      "step": 7458
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.262597233722515,
+      "learning_rate": 1.8906876043492382e-05,
+      "loss": 0.9705,
+      "step": 7459
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4358650300780131,
+      "learning_rate": 1.8906529126986057e-05,
+      "loss": 0.9543,
+      "step": 7460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6166127222476916,
+      "learning_rate": 1.8906182158623097e-05,
+      "loss": 0.9421,
+      "step": 7461
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6311813249430296,
+      "learning_rate": 1.8905835138405526e-05,
+      "loss": 0.901,
+      "step": 7462
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0628481933189065,
+      "learning_rate": 1.8905488066335366e-05,
+      "loss": 0.8769,
+      "step": 7463
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8611481889717698,
+      "learning_rate": 1.8905140942414633e-05,
+      "loss": 0.9793,
+      "step": 7464
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5461248730441326,
+      "learning_rate": 1.890479376664535e-05,
+      "loss": 0.9728,
+      "step": 7465
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.811603212029055,
+      "learning_rate": 1.890444653902954e-05,
+      "loss": 0.7084,
+      "step": 7466
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.365832115186372,
+      "learning_rate": 1.890409925956922e-05,
+      "loss": 0.8613,
+      "step": 7467
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8005842482809185,
+      "learning_rate": 1.8903751928266415e-05,
+      "loss": 0.7365,
+      "step": 7468
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9163846852036983,
+      "learning_rate": 1.890340454512315e-05,
+      "loss": 0.9821,
+      "step": 7469
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7257697599191972,
+      "learning_rate": 1.8903057110141443e-05,
+      "loss": 0.9879,
+      "step": 7470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9753190074221794,
+      "learning_rate": 1.8902709623323322e-05,
+      "loss": 0.9086,
+      "step": 7471
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6007182298696219,
+      "learning_rate": 1.8902362084670806e-05,
+      "loss": 0.9446,
+      "step": 7472
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.84120879249596,
+      "learning_rate": 1.8902014494185918e-05,
+      "loss": 0.9749,
+      "step": 7473
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8193159989709021,
+      "learning_rate": 1.8901666851870684e-05,
+      "loss": 0.7602,
+      "step": 7474
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.589721630879703,
+      "learning_rate": 1.8901319157727127e-05,
+      "loss": 0.9448,
+      "step": 7475
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.398772178499144,
+      "learning_rate": 1.8900971411757273e-05,
+      "loss": 1.0227,
+      "step": 7476
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0982149588441286,
+      "learning_rate": 1.8900623613963144e-05,
+      "loss": 0.9546,
+      "step": 7477
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5973983827935758,
+      "learning_rate": 1.890027576434677e-05,
+      "loss": 0.847,
+      "step": 7478
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7496491890129738,
+      "learning_rate": 1.889992786291017e-05,
+      "loss": 0.7967,
+      "step": 7479
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8657790309719642,
+      "learning_rate": 1.889957990965537e-05,
+      "loss": 1.0087,
+      "step": 7480
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6473245566397732,
+      "learning_rate": 1.88992319045844e-05,
+      "loss": 0.7724,
+      "step": 7481
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6156082036048107,
+      "learning_rate": 1.8898883847699284e-05,
+      "loss": 1.0441,
+      "step": 7482
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9124606583835015,
+      "learning_rate": 1.889853573900205e-05,
+      "loss": 0.8686,
+      "step": 7483
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7500696127720707,
+      "learning_rate": 1.8898187578494723e-05,
+      "loss": 0.9285,
+      "step": 7484
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7432212525524304,
+      "learning_rate": 1.889783936617933e-05,
+      "loss": 0.7755,
+      "step": 7485
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5194055677959288,
+      "learning_rate": 1.88974911020579e-05,
+      "loss": 1.001,
+      "step": 7486
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8287410335740237,
+      "learning_rate": 1.889714278613246e-05,
+      "loss": 0.8607,
+      "step": 7487
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6446043830075512,
+      "learning_rate": 1.889679441840504e-05,
+      "loss": 0.9871,
+      "step": 7488
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7774046376538096,
+      "learning_rate": 1.8896445998877664e-05,
+      "loss": 0.7925,
+      "step": 7489
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.760780809332476,
+      "learning_rate": 1.8896097527552362e-05,
+      "loss": 1.0509,
+      "step": 7490
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7224339326980511,
+      "learning_rate": 1.8895749004431168e-05,
+      "loss": 0.9909,
+      "step": 7491
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7417276132054782,
+      "learning_rate": 1.88954004295161e-05,
+      "loss": 0.9936,
+      "step": 7492
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4781124872513873,
+      "learning_rate": 1.8895051802809202e-05,
+      "loss": 0.9094,
+      "step": 7493
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5657807757269713,
+      "learning_rate": 1.8894703124312492e-05,
+      "loss": 0.9586,
+      "step": 7494
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6469800314291403,
+      "learning_rate": 1.8894354394028005e-05,
+      "loss": 0.7762,
+      "step": 7495
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.319861602793114,
+      "learning_rate": 1.889400561195777e-05,
+      "loss": 1.0119,
+      "step": 7496
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2300743589825913,
+      "learning_rate": 1.889365677810382e-05,
+      "loss": 0.9572,
+      "step": 7497
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.67139442206231,
+      "learning_rate": 1.8893307892468184e-05,
+      "loss": 0.9899,
+      "step": 7498
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4725281037070321,
+      "learning_rate": 1.8892958955052893e-05,
+      "loss": 0.9839,
+      "step": 7499
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7167337460786027,
+      "learning_rate": 1.889260996585998e-05,
+      "loss": 0.9192,
+      "step": 7500
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6369789883691825,
+      "learning_rate": 1.889226092489148e-05,
+      "loss": 0.9522,
+      "step": 7501
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1608559062038335,
+      "learning_rate": 1.8891911832149418e-05,
+      "loss": 0.9314,
+      "step": 7502
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7114727553574993,
+      "learning_rate": 1.889156268763583e-05,
+      "loss": 0.9046,
+      "step": 7503
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4980880542883879,
+      "learning_rate": 1.889121349135275e-05,
+      "loss": 0.9172,
+      "step": 7504
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.566451715512264,
+      "learning_rate": 1.889086424330221e-05,
+      "loss": 0.9832,
+      "step": 7505
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.728712043539893,
+      "learning_rate": 1.8890514943486242e-05,
+      "loss": 0.9173,
+      "step": 7506
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7750118694723929,
+      "learning_rate": 1.8890165591906883e-05,
+      "loss": 0.9101,
+      "step": 7507
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8883784963665342,
+      "learning_rate": 1.8889816188566163e-05,
+      "loss": 0.9058,
+      "step": 7508
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5650898187765172,
+      "learning_rate": 1.8889466733466124e-05,
+      "loss": 0.9641,
+      "step": 7509
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0241696819754567,
+      "learning_rate": 1.888911722660879e-05,
+      "loss": 0.8717,
+      "step": 7510
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1413204270978077,
+      "learning_rate": 1.88887676679962e-05,
+      "loss": 0.8476,
+      "step": 7511
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.818039255026044,
+      "learning_rate": 1.8888418057630397e-05,
+      "loss": 0.9096,
+      "step": 7512
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9972563050879333,
+      "learning_rate": 1.8888068395513407e-05,
+      "loss": 0.9558,
+      "step": 7513
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.669623351062067,
+      "learning_rate": 1.8887718681647265e-05,
+      "loss": 0.854,
+      "step": 7514
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7900380484403307,
+      "learning_rate": 1.8887368916034017e-05,
+      "loss": 0.9251,
+      "step": 7515
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7974678910879781,
+      "learning_rate": 1.8887019098675692e-05,
+      "loss": 0.7395,
+      "step": 7516
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1201141743197085,
+      "learning_rate": 1.8886669229574326e-05,
+      "loss": 0.9242,
+      "step": 7517
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1147611258751353,
+      "learning_rate": 1.888631930873196e-05,
+      "loss": 0.9137,
+      "step": 7518
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9116763150495868,
+      "learning_rate": 1.888596933615063e-05,
+      "loss": 0.9061,
+      "step": 7519
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.525002845835919,
+      "learning_rate": 1.888561931183237e-05,
+      "loss": 0.9728,
+      "step": 7520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1210691449529007,
+      "learning_rate": 1.8885269235779226e-05,
+      "loss": 0.9719,
+      "step": 7521
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8348518987762243,
+      "learning_rate": 1.888491910799323e-05,
+      "loss": 1.0013,
+      "step": 7522
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6096101862159156,
+      "learning_rate": 1.888456892847642e-05,
+      "loss": 1.0024,
+      "step": 7523
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7087296958086844,
+      "learning_rate": 1.8884218697230838e-05,
+      "loss": 0.7411,
+      "step": 7524
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8246900677562352,
+      "learning_rate": 1.8883868414258525e-05,
+      "loss": 0.9645,
+      "step": 7525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.329920921764304,
+      "learning_rate": 1.8883518079561512e-05,
+      "loss": 1.0543,
+      "step": 7526
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.732548874957755,
+      "learning_rate": 1.8883167693141845e-05,
+      "loss": 1.0415,
+      "step": 7527
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8200230690105063,
+      "learning_rate": 1.8882817255001567e-05,
+      "loss": 0.73,
+      "step": 7528
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3980635521274345,
+      "learning_rate": 1.8882466765142715e-05,
+      "loss": 0.8934,
+      "step": 7529
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.847671798618957,
+      "learning_rate": 1.8882116223567327e-05,
+      "loss": 0.8756,
+      "step": 7530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.010973426022399,
+      "learning_rate": 1.8881765630277447e-05,
+      "loss": 0.9031,
+      "step": 7531
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9949314113620658,
+      "learning_rate": 1.8881414985275114e-05,
+      "loss": 0.8793,
+      "step": 7532
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.654934065625701,
+      "learning_rate": 1.8881064288562373e-05,
+      "loss": 1.0307,
+      "step": 7533
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6271516820013034,
+      "learning_rate": 1.888071354014126e-05,
+      "loss": 0.7947,
+      "step": 7534
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7255194166415457,
+      "learning_rate": 1.8880362740013827e-05,
+      "loss": 0.9065,
+      "step": 7535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7276443817046895,
+      "learning_rate": 1.8880011888182108e-05,
+      "loss": 0.7705,
+      "step": 7536
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7659052519354106,
+      "learning_rate": 1.887966098464815e-05,
+      "loss": 0.8963,
+      "step": 7537
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6245082021796449,
+      "learning_rate": 1.8879310029413993e-05,
+      "loss": 0.8197,
+      "step": 7538
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5564930251417266,
+      "learning_rate": 1.8878959022481682e-05,
+      "loss": 0.9611,
+      "step": 7539
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.693096066684766,
+      "learning_rate": 1.887860796385326e-05,
+      "loss": 0.8172,
+      "step": 7540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.375132317104774,
+      "learning_rate": 1.8878256853530772e-05,
+      "loss": 0.9649,
+      "step": 7541
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.050562833865823,
+      "learning_rate": 1.8877905691516263e-05,
+      "loss": 0.9742,
+      "step": 7542
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7556285867095756,
+      "learning_rate": 1.8877554477811775e-05,
+      "loss": 1.0182,
+      "step": 7543
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8249219595602826,
+      "learning_rate": 1.8877203212419354e-05,
+      "loss": 0.7363,
+      "step": 7544
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0573706972892585,
+      "learning_rate": 1.8876851895341044e-05,
+      "loss": 0.9087,
+      "step": 7545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6868046849512222,
+      "learning_rate": 1.8876500526578897e-05,
+      "loss": 0.8614,
+      "step": 7546
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8790395400160211,
+      "learning_rate": 1.8876149106134947e-05,
+      "loss": 0.8686,
+      "step": 7547
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6228733996620945,
+      "learning_rate": 1.8875797634011254e-05,
+      "loss": 0.8603,
+      "step": 7548
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.063909143677601,
+      "learning_rate": 1.887544611020985e-05,
+      "loss": 0.939,
+      "step": 7549
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9376344852221614,
+      "learning_rate": 1.8875094534732793e-05,
+      "loss": 0.9465,
+      "step": 7550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5921095790688704,
+      "learning_rate": 1.8874742907582124e-05,
+      "loss": 0.9679,
+      "step": 7551
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0667881282922056,
+      "learning_rate": 1.8874391228759894e-05,
+      "loss": 0.8912,
+      "step": 7552
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7823105372535639,
+      "learning_rate": 1.8874039498268147e-05,
+      "loss": 0.7469,
+      "step": 7553
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7763439721419705,
+      "learning_rate": 1.8873687716108935e-05,
+      "loss": 0.844,
+      "step": 7554
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.844543916228169,
+      "learning_rate": 1.88733358822843e-05,
+      "loss": 0.9734,
+      "step": 7555
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2171670296133072,
+      "learning_rate": 1.8872983996796298e-05,
+      "loss": 1.0998,
+      "step": 7556
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7670027870040683,
+      "learning_rate": 1.8872632059646973e-05,
+      "loss": 0.8103,
+      "step": 7557
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8809812996062345,
+      "learning_rate": 1.8872280070838373e-05,
+      "loss": 0.8715,
+      "step": 7558
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6351617417016036,
+      "learning_rate": 1.8871928030372553e-05,
+      "loss": 0.8156,
+      "step": 7559
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5198749522072128,
+      "learning_rate": 1.8871575938251557e-05,
+      "loss": 0.7597,
+      "step": 7560
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6417387257436098,
+      "learning_rate": 1.887122379447744e-05,
+      "loss": 0.9246,
+      "step": 7561
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6960179064702072,
+      "learning_rate": 1.8870871599052244e-05,
+      "loss": 0.8505,
+      "step": 7562
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7529217223523519,
+      "learning_rate": 1.8870519351978026e-05,
+      "loss": 0.7251,
+      "step": 7563
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.865530578826342,
+      "learning_rate": 1.8870167053256838e-05,
+      "loss": 0.9371,
+      "step": 7564
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7766150492747192,
+      "learning_rate": 1.8869814702890733e-05,
+      "loss": 0.7391,
+      "step": 7565
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7709297513381024,
+      "learning_rate": 1.8869462300881755e-05,
+      "loss": 0.8622,
+      "step": 7566
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6908261719383255,
+      "learning_rate": 1.8869109847231956e-05,
+      "loss": 0.8626,
+      "step": 7567
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6641391777375105,
+      "learning_rate": 1.8868757341943396e-05,
+      "loss": 0.9818,
+      "step": 7568
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6168358809697219,
+      "learning_rate": 1.886840478501812e-05,
+      "loss": 0.8921,
+      "step": 7569
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8032464424781978,
+      "learning_rate": 1.8868052176458188e-05,
+      "loss": 1.0289,
+      "step": 7570
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.872244774161907,
+      "learning_rate": 1.8867699516265646e-05,
+      "loss": 0.907,
+      "step": 7571
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7245659787287368,
+      "learning_rate": 1.886734680444255e-05,
+      "loss": 0.9085,
+      "step": 7572
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.43900869482779,
+      "learning_rate": 1.8866994040990954e-05,
+      "loss": 0.8994,
+      "step": 7573
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7546212175710503,
+      "learning_rate": 1.8866641225912908e-05,
+      "loss": 0.9642,
+      "step": 7574
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8858260448177497,
+      "learning_rate": 1.8866288359210474e-05,
+      "loss": 1.0874,
+      "step": 7575
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.57675116207097,
+      "learning_rate": 1.88659354408857e-05,
+      "loss": 0.8831,
+      "step": 7576
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.807484009217076,
+      "learning_rate": 1.8865582470940644e-05,
+      "loss": 0.9706,
+      "step": 7577
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8241186166017458,
+      "learning_rate": 1.886522944937736e-05,
+      "loss": 1.0535,
+      "step": 7578
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.74955063963441,
+      "learning_rate": 1.8864876376197904e-05,
+      "loss": 0.7848,
+      "step": 7579
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.725001736196315,
+      "learning_rate": 1.886452325140433e-05,
+      "loss": 0.8079,
+      "step": 7580
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7477290319566863,
+      "learning_rate": 1.8864170074998695e-05,
+      "loss": 0.8129,
+      "step": 7581
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7698116016261827,
+      "learning_rate": 1.8863816846983055e-05,
+      "loss": 0.8095,
+      "step": 7582
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.828412265959195,
+      "learning_rate": 1.886346356735947e-05,
+      "loss": 0.9431,
+      "step": 7583
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7303561983906344,
+      "learning_rate": 1.8863110236129993e-05,
+      "loss": 0.9122,
+      "step": 7584
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7001399921046607,
+      "learning_rate": 1.886275685329668e-05,
+      "loss": 0.8017,
+      "step": 7585
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6795606229708664,
+      "learning_rate": 1.8862403418861593e-05,
+      "loss": 0.863,
+      "step": 7586
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6490833470105475,
+      "learning_rate": 1.8862049932826787e-05,
+      "loss": 0.9398,
+      "step": 7587
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.613340651411344,
+      "learning_rate": 1.886169639519432e-05,
+      "loss": 0.8743,
+      "step": 7588
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7780789694527708,
+      "learning_rate": 1.886134280596625e-05,
+      "loss": 0.8099,
+      "step": 7589
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.709606460955251,
+      "learning_rate": 1.886098916514464e-05,
+      "loss": 0.8595,
+      "step": 7590
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7877078078729949,
+      "learning_rate": 1.886063547273154e-05,
+      "loss": 0.8596,
+      "step": 7591
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.916166191140713,
+      "learning_rate": 1.886028172872902e-05,
+      "loss": 0.802,
+      "step": 7592
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7106205359374065,
+      "learning_rate": 1.885992793313913e-05,
+      "loss": 0.8705,
+      "step": 7593
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.839551653740033,
+      "learning_rate": 1.885957408596394e-05,
+      "loss": 0.9573,
+      "step": 7594
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1196855703143664,
+      "learning_rate": 1.8859220187205505e-05,
+      "loss": 1.0065,
+      "step": 7595
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2081544601529446,
+      "learning_rate": 1.8858866236865883e-05,
+      "loss": 0.9905,
+      "step": 7596
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6302195114669364,
+      "learning_rate": 1.8858512234947138e-05,
+      "loss": 0.9376,
+      "step": 7597
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.568780228906368,
+      "learning_rate": 1.8858158181451327e-05,
+      "loss": 0.8972,
+      "step": 7598
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.751845026399113,
+      "learning_rate": 1.885780407638052e-05,
+      "loss": 1.0679,
+      "step": 7599
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6028055080175014,
+      "learning_rate": 1.885744991973677e-05,
+      "loss": 0.9075,
+      "step": 7600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7822235350223852,
+      "learning_rate": 1.885709571152214e-05,
+      "loss": 0.8836,
+      "step": 7601
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1316900717937353,
+      "learning_rate": 1.8856741451738697e-05,
+      "loss": 0.8687,
+      "step": 7602
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8641794269662222,
+      "learning_rate": 1.88563871403885e-05,
+      "loss": 0.9568,
+      "step": 7603
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7701593561027225,
+      "learning_rate": 1.8856032777473617e-05,
+      "loss": 0.9454,
+      "step": 7604
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1969261359131775,
+      "learning_rate": 1.8855678362996105e-05,
+      "loss": 0.7958,
+      "step": 7605
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6558651649427472,
+      "learning_rate": 1.885532389695803e-05,
+      "loss": 0.9562,
+      "step": 7606
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.00148512462567,
+      "learning_rate": 1.8854969379361456e-05,
+      "loss": 0.9063,
+      "step": 7607
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.917879127390996,
+      "learning_rate": 1.8854614810208446e-05,
+      "loss": 0.8412,
+      "step": 7608
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0507162599824347,
+      "learning_rate": 1.8854260189501068e-05,
+      "loss": 0.8893,
+      "step": 7609
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2793008482819963,
+      "learning_rate": 1.8853905517241384e-05,
+      "loss": 0.9655,
+      "step": 7610
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6868484146316078,
+      "learning_rate": 1.885355079343145e-05,
+      "loss": 1.0012,
+      "step": 7611
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.547563465650796,
+      "learning_rate": 1.8853196018073352e-05,
+      "loss": 0.9228,
+      "step": 7612
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6536590377197204,
+      "learning_rate": 1.8852841191169138e-05,
+      "loss": 0.8893,
+      "step": 7613
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9196352004901693,
+      "learning_rate": 1.885248631272088e-05,
+      "loss": 0.9022,
+      "step": 7614
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7554085039952587,
+      "learning_rate": 1.8852131382730645e-05,
+      "loss": 0.9344,
+      "step": 7615
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6697740629341036,
+      "learning_rate": 1.88517764012005e-05,
+      "loss": 0.9648,
+      "step": 7616
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.775760713653117,
+      "learning_rate": 1.8851421368132507e-05,
+      "loss": 0.7256,
+      "step": 7617
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7220039336028397,
+      "learning_rate": 1.8851066283528738e-05,
+      "loss": 0.7989,
+      "step": 7618
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6206802967480052,
+      "learning_rate": 1.8850711147391257e-05,
+      "loss": 0.9271,
+      "step": 7619
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8159406642493944,
+      "learning_rate": 1.8850355959722138e-05,
+      "loss": 0.7829,
+      "step": 7620
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9185050200802523,
+      "learning_rate": 1.885000072052344e-05,
+      "loss": 0.7765,
+      "step": 7621
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7849317308781507,
+      "learning_rate": 1.884964542979724e-05,
+      "loss": 0.8334,
+      "step": 7622
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8494985577369286,
+      "learning_rate": 1.88492900875456e-05,
+      "loss": 0.7401,
+      "step": 7623
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5802749914087533,
+      "learning_rate": 1.8848934693770594e-05,
+      "loss": 0.8908,
+      "step": 7624
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9738264393078528,
+      "learning_rate": 1.8848579248474286e-05,
+      "loss": 0.9502,
+      "step": 7625
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8300388741094398,
+      "learning_rate": 1.884822375165875e-05,
+      "loss": 1.0054,
+      "step": 7626
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.699823142309572,
+      "learning_rate": 1.8847868203326053e-05,
+      "loss": 0.8683,
+      "step": 7627
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.539598301452101,
+      "learning_rate": 1.8847512603478267e-05,
+      "loss": 1.047,
+      "step": 7628
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4604730504693948,
+      "learning_rate": 1.8847156952117462e-05,
+      "loss": 0.9188,
+      "step": 7629
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.694741703546731,
+      "learning_rate": 1.8846801249245706e-05,
+      "loss": 0.931,
+      "step": 7630
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5979994200224925,
+      "learning_rate": 1.8846445494865074e-05,
+      "loss": 0.9479,
+      "step": 7631
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5624788931262934,
+      "learning_rate": 1.8846089688977635e-05,
+      "loss": 0.9868,
+      "step": 7632
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.735949912467985,
+      "learning_rate": 1.884573383158546e-05,
+      "loss": 0.9257,
+      "step": 7633
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6969306006103109,
+      "learning_rate": 1.8845377922690623e-05,
+      "loss": 0.8764,
+      "step": 7634
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8826124050148596,
+      "learning_rate": 1.88450219622952e-05,
+      "loss": 0.9253,
+      "step": 7635
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7784429504070333,
+      "learning_rate": 1.884466595040125e-05,
+      "loss": 0.7366,
+      "step": 7636
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7304899858576135,
+      "learning_rate": 1.8844309887010862e-05,
+      "loss": 0.8748,
+      "step": 7637
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6877136497931429,
+      "learning_rate": 1.88439537721261e-05,
+      "loss": 0.9076,
+      "step": 7638
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.613834818368463,
+      "learning_rate": 1.8843597605749035e-05,
+      "loss": 0.9566,
+      "step": 7639
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8326076365495616,
+      "learning_rate": 1.884324138788175e-05,
+      "loss": 0.9539,
+      "step": 7640
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9026784813327917,
+      "learning_rate": 1.8842885118526314e-05,
+      "loss": 0.94,
+      "step": 7641
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.644142365138203,
+      "learning_rate": 1.8842528797684795e-05,
+      "loss": 0.8415,
+      "step": 7642
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.45796524633742,
+      "learning_rate": 1.884217242535928e-05,
+      "loss": 0.9054,
+      "step": 7643
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7274006751378028,
+      "learning_rate": 1.8841816001551836e-05,
+      "loss": 0.813,
+      "step": 7644
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7814457823944392,
+      "learning_rate": 1.8841459526264542e-05,
+      "loss": 0.92,
+      "step": 7645
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6368979445246132,
+      "learning_rate": 1.8841102999499468e-05,
+      "loss": 0.8611,
+      "step": 7646
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8769761953135589,
+      "learning_rate": 1.8840746421258694e-05,
+      "loss": 0.842,
+      "step": 7647
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.684206010154245,
+      "learning_rate": 1.8840389791544297e-05,
+      "loss": 1.0101,
+      "step": 7648
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.759093803236157,
+      "learning_rate": 1.884003311035835e-05,
+      "loss": 0.9275,
+      "step": 7649
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8248244280164854,
+      "learning_rate": 1.8839676377702934e-05,
+      "loss": 0.8762,
+      "step": 7650
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7770308879441712,
+      "learning_rate": 1.883931959358012e-05,
+      "loss": 0.9725,
+      "step": 7651
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.707342686190102,
+      "learning_rate": 1.8838962757991993e-05,
+      "loss": 0.9487,
+      "step": 7652
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8326903590617074,
+      "learning_rate": 1.883860587094062e-05,
+      "loss": 0.9224,
+      "step": 7653
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9658769545286476,
+      "learning_rate": 1.883824893242809e-05,
+      "loss": 0.8027,
+      "step": 7654
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7319853434871832,
+      "learning_rate": 1.8837891942456477e-05,
+      "loss": 0.888,
+      "step": 7655
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8284271433646954,
+      "learning_rate": 1.883753490102786e-05,
+      "loss": 0.7366,
+      "step": 7656
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.863353514528936,
+      "learning_rate": 1.8837177808144313e-05,
+      "loss": 0.9309,
+      "step": 7657
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8051216563169066,
+      "learning_rate": 1.8836820663807924e-05,
+      "loss": 0.7889,
+      "step": 7658
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7022664180869251,
+      "learning_rate": 1.8836463468020764e-05,
+      "loss": 0.8076,
+      "step": 7659
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.586893425587203,
+      "learning_rate": 1.8836106220784915e-05,
+      "loss": 0.831,
+      "step": 7660
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.93941054021011,
+      "learning_rate": 1.883574892210246e-05,
+      "loss": 0.7935,
+      "step": 7661
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9463984452244103,
+      "learning_rate": 1.8835391571975475e-05,
+      "loss": 0.9328,
+      "step": 7662
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.63487131689379,
+      "learning_rate": 1.8835034170406046e-05,
+      "loss": 0.9458,
+      "step": 7663
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.377003302129141,
+      "learning_rate": 1.883467671739625e-05,
+      "loss": 0.8732,
+      "step": 7664
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.630219776465528,
+      "learning_rate": 1.8834319212948168e-05,
+      "loss": 1.1315,
+      "step": 7665
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6507050775079335,
+      "learning_rate": 1.8833961657063887e-05,
+      "loss": 0.7941,
+      "step": 7666
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.81628854512108,
+      "learning_rate": 1.883360404974548e-05,
+      "loss": 0.7368,
+      "step": 7667
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7913216847226459,
+      "learning_rate": 1.8833246390995034e-05,
+      "loss": 0.928,
+      "step": 7668
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4088968430056383,
+      "learning_rate": 1.883288868081463e-05,
+      "loss": 0.9636,
+      "step": 7669
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7259335694510245,
+      "learning_rate": 1.8832530919206356e-05,
+      "loss": 0.7166,
+      "step": 7670
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6922512084271648,
+      "learning_rate": 1.883217310617229e-05,
+      "loss": 1.0484,
+      "step": 7671
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8078715429075407,
+      "learning_rate": 1.883181524171451e-05,
+      "loss": 0.9314,
+      "step": 7672
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5508213414688397,
+      "learning_rate": 1.8831457325835113e-05,
+      "loss": 0.9139,
+      "step": 7673
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5499812141604805,
+      "learning_rate": 1.8831099358536172e-05,
+      "loss": 0.8895,
+      "step": 7674
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.580296144983423,
+      "learning_rate": 1.8830741339819774e-05,
+      "loss": 0.8764,
+      "step": 7675
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.696228262081928,
+      "learning_rate": 1.8830383269688005e-05,
+      "loss": 0.8287,
+      "step": 7676
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7673259692815749,
+      "learning_rate": 1.8830025148142948e-05,
+      "loss": 0.7168,
+      "step": 7677
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5690267744372404,
+      "learning_rate": 1.882966697518669e-05,
+      "loss": 1.0086,
+      "step": 7678
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.675549279187098,
+      "learning_rate": 1.882930875082132e-05,
+      "loss": 0.9705,
+      "step": 7679
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8466259571233602,
+      "learning_rate": 1.8828950475048913e-05,
+      "loss": 0.7671,
+      "step": 7680
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4656412674589439,
+      "learning_rate": 1.8828592147871562e-05,
+      "loss": 0.9563,
+      "step": 7681
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9138756148361245,
+      "learning_rate": 1.8828233769291353e-05,
+      "loss": 0.9175,
+      "step": 7682
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.821483551275894,
+      "learning_rate": 1.8827875339310372e-05,
+      "loss": 0.9054,
+      "step": 7683
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5964453180508202,
+      "learning_rate": 1.8827516857930706e-05,
+      "loss": 0.8804,
+      "step": 7684
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.742852497673581,
+      "learning_rate": 1.882715832515444e-05,
+      "loss": 0.9254,
+      "step": 7685
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8778599274449026,
+      "learning_rate": 1.8826799740983665e-05,
+      "loss": 0.934,
+      "step": 7686
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8005159353045699,
+      "learning_rate": 1.882644110542047e-05,
+      "loss": 0.8491,
+      "step": 7687
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9052078862941,
+      "learning_rate": 1.8826082418466938e-05,
+      "loss": 0.8854,
+      "step": 7688
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7663413746993952,
+      "learning_rate": 1.882572368012516e-05,
+      "loss": 0.9518,
+      "step": 7689
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5563542419011693,
+      "learning_rate": 1.8825364890397225e-05,
+      "loss": 0.9417,
+      "step": 7690
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8324453343192032,
+      "learning_rate": 1.8825006049285225e-05,
+      "loss": 0.8055,
+      "step": 7691
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5822609581603273,
+      "learning_rate": 1.882464715679124e-05,
+      "loss": 0.8764,
+      "step": 7692
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4477761770074526,
+      "learning_rate": 1.8824288212917367e-05,
+      "loss": 0.8708,
+      "step": 7693
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8322482788486845,
+      "learning_rate": 1.8823929217665696e-05,
+      "loss": 1.0172,
+      "step": 7694
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7711045125784475,
+      "learning_rate": 1.8823570171038316e-05,
+      "loss": 0.8398,
+      "step": 7695
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5545848701438707,
+      "learning_rate": 1.8823211073037316e-05,
+      "loss": 0.9497,
+      "step": 7696
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8639280388072594,
+      "learning_rate": 1.8822851923664785e-05,
+      "loss": 1.0163,
+      "step": 7697
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7845344553565949,
+      "learning_rate": 1.882249272292282e-05,
+      "loss": 0.6959,
+      "step": 7698
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.052060852249362,
+      "learning_rate": 1.8822133470813507e-05,
+      "loss": 0.8309,
+      "step": 7699
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.612390413749389,
+      "learning_rate": 1.8821774167338944e-05,
+      "loss": 0.9025,
+      "step": 7700
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6152638088565805,
+      "learning_rate": 1.8821414812501214e-05,
+      "loss": 0.9513,
+      "step": 7701
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9302794784384838,
+      "learning_rate": 1.8821055406302418e-05,
+      "loss": 0.9544,
+      "step": 7702
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6885807922246472,
+      "learning_rate": 1.882069594874464e-05,
+      "loss": 0.9664,
+      "step": 7703
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8521364832033946,
+      "learning_rate": 1.882033643982998e-05,
+      "loss": 0.9558,
+      "step": 7704
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.802449004410757,
+      "learning_rate": 1.8819976879560534e-05,
+      "loss": 1.1021,
+      "step": 7705
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9432188334127212,
+      "learning_rate": 1.8819617267938383e-05,
+      "loss": 0.9714,
+      "step": 7706
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6153205071990901,
+      "learning_rate": 1.881925760496563e-05,
+      "loss": 0.8955,
+      "step": 7707
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6470555669983287,
+      "learning_rate": 1.881889789064437e-05,
+      "loss": 0.8177,
+      "step": 7708
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5616422571871371,
+      "learning_rate": 1.881853812497669e-05,
+      "loss": 1.0008,
+      "step": 7709
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8208858567010335,
+      "learning_rate": 1.881817830796469e-05,
+      "loss": 0.8357,
+      "step": 7710
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0961855842579236,
+      "learning_rate": 1.8817818439610466e-05,
+      "loss": 0.8413,
+      "step": 7711
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5735847990195655,
+      "learning_rate": 1.8817458519916113e-05,
+      "loss": 0.9244,
+      "step": 7712
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9096542068398645,
+      "learning_rate": 1.881709854888372e-05,
+      "loss": 0.8567,
+      "step": 7713
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.754507717500303,
+      "learning_rate": 1.881673852651539e-05,
+      "loss": 0.9239,
+      "step": 7714
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6660768709658798,
+      "learning_rate": 1.881637845281322e-05,
+      "loss": 0.9635,
+      "step": 7715
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5263990871683184,
+      "learning_rate": 1.88160183277793e-05,
+      "loss": 0.898,
+      "step": 7716
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8001697651122368,
+      "learning_rate": 1.8815658151415734e-05,
+      "loss": 0.9169,
+      "step": 7717
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.608612264395177,
+      "learning_rate": 1.8815297923724612e-05,
+      "loss": 0.9533,
+      "step": 7718
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5168614682365753,
+      "learning_rate": 1.8814937644708035e-05,
+      "loss": 0.893,
+      "step": 7719
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3344589738334154,
+      "learning_rate": 1.8814577314368098e-05,
+      "loss": 0.8292,
+      "step": 7720
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5843371747392239,
+      "learning_rate": 1.8814216932706905e-05,
+      "loss": 0.9443,
+      "step": 7721
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7086683829779898,
+      "learning_rate": 1.8813856499726548e-05,
+      "loss": 0.8389,
+      "step": 7722
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.590111430377693,
+      "learning_rate": 1.881349601542913e-05,
+      "loss": 0.9432,
+      "step": 7723
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5283124206913552,
+      "learning_rate": 1.8813135479816748e-05,
+      "loss": 0.9329,
+      "step": 7724
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1205443514788844,
+      "learning_rate": 1.88127748928915e-05,
+      "loss": 0.9143,
+      "step": 7725
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8974848981244878,
+      "learning_rate": 1.8812414254655483e-05,
+      "loss": 0.8304,
+      "step": 7726
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9234297468799426,
+      "learning_rate": 1.8812053565110804e-05,
+      "loss": 0.7813,
+      "step": 7727
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.88793653657333,
+      "learning_rate": 1.881169282425956e-05,
+      "loss": 0.7113,
+      "step": 7728
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.675936901152177,
+      "learning_rate": 1.881133203210385e-05,
+      "loss": 0.8219,
+      "step": 7729
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7130873920443523,
+      "learning_rate": 1.8810971188645775e-05,
+      "loss": 0.9708,
+      "step": 7730
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6309224276788759,
+      "learning_rate": 1.8810610293887435e-05,
+      "loss": 0.9548,
+      "step": 7731
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7418016226852808,
+      "learning_rate": 1.8810249347830935e-05,
+      "loss": 0.8636,
+      "step": 7732
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6169816400972101,
+      "learning_rate": 1.8809888350478372e-05,
+      "loss": 0.847,
+      "step": 7733
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.96318164965918,
+      "learning_rate": 1.880952730183185e-05,
+      "loss": 0.8125,
+      "step": 7734
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6319228097421405,
+      "learning_rate": 1.880916620189347e-05,
+      "loss": 0.9792,
+      "step": 7735
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5628768593084905,
+      "learning_rate": 1.880880505066534e-05,
+      "loss": 0.9607,
+      "step": 7736
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5755214838838252,
+      "learning_rate": 1.8808443848149556e-05,
+      "loss": 0.9972,
+      "step": 7737
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8134536956132188,
+      "learning_rate": 1.880808259434822e-05,
+      "loss": 1.0507,
+      "step": 7738
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.833932805212274,
+      "learning_rate": 1.8807721289263442e-05,
+      "loss": 0.8437,
+      "step": 7739
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.663197089763118,
+      "learning_rate": 1.880735993289732e-05,
+      "loss": 0.8892,
+      "step": 7740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8928763275496523,
+      "learning_rate": 1.8806998525251964e-05,
+      "loss": 1.0175,
+      "step": 7741
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.55177218574437,
+      "learning_rate": 1.880663706632947e-05,
+      "loss": 0.8573,
+      "step": 7742
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.004744830446127,
+      "learning_rate": 1.8806275556131948e-05,
+      "loss": 1.0821,
+      "step": 7743
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9142280600067576,
+      "learning_rate": 1.88059139946615e-05,
+      "loss": 0.7294,
+      "step": 7744
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8511436731782906,
+      "learning_rate": 1.8805552381920238e-05,
+      "loss": 0.8602,
+      "step": 7745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5654474169183175,
+      "learning_rate": 1.880519071791026e-05,
+      "loss": 0.8914,
+      "step": 7746
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6472832964710726,
+      "learning_rate": 1.8804829002633672e-05,
+      "loss": 0.986,
+      "step": 7747
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8100269108464775,
+      "learning_rate": 1.8804467236092584e-05,
+      "loss": 0.831,
+      "step": 7748
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7033849062838602,
+      "learning_rate": 1.8804105418289097e-05,
+      "loss": 0.9302,
+      "step": 7749
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5826296724053037,
+      "learning_rate": 1.8803743549225323e-05,
+      "loss": 0.887,
+      "step": 7750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5709606322549625,
+      "learning_rate": 1.880338162890337e-05,
+      "loss": 0.9348,
+      "step": 7751
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7740812050992847,
+      "learning_rate": 1.8803019657325338e-05,
+      "loss": 0.7694,
+      "step": 7752
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.208607510325065,
+      "learning_rate": 1.880265763449334e-05,
+      "loss": 1.0393,
+      "step": 7753
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6438536772931849,
+      "learning_rate": 1.880229556040948e-05,
+      "loss": 0.9528,
+      "step": 7754
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.825306315528404,
+      "learning_rate": 1.8801933435075872e-05,
+      "loss": 0.7605,
+      "step": 7755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7909649607415932,
+      "learning_rate": 1.880157125849462e-05,
+      "loss": 1.0282,
+      "step": 7756
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7552753093702205,
+      "learning_rate": 1.8801209030667834e-05,
+      "loss": 0.7386,
+      "step": 7757
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6334089700222887,
+      "learning_rate": 1.880084675159762e-05,
+      "loss": 1.0153,
+      "step": 7758
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.712271137631986,
+      "learning_rate": 1.8800484421286094e-05,
+      "loss": 1.043,
+      "step": 7759
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.823043619716354,
+      "learning_rate": 1.880012203973536e-05,
+      "loss": 0.9891,
+      "step": 7760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7178413564463098,
+      "learning_rate": 1.8799759606947527e-05,
+      "loss": 0.9386,
+      "step": 7761
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6007508341473262,
+      "learning_rate": 1.8799397122924708e-05,
+      "loss": 0.9169,
+      "step": 7762
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.525393853053641,
+      "learning_rate": 1.8799034587669016e-05,
+      "loss": 1.0138,
+      "step": 7763
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8206262751015584,
+      "learning_rate": 1.8798672001182553e-05,
+      "loss": 0.7742,
+      "step": 7764
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6539995788112873,
+      "learning_rate": 1.8798309363467446e-05,
+      "loss": 0.8978,
+      "step": 7765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4393748168837401,
+      "learning_rate": 1.8797946674525788e-05,
+      "loss": 0.9166,
+      "step": 7766
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6206897654480408,
+      "learning_rate": 1.87975839343597e-05,
+      "loss": 0.9102,
+      "step": 7767
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6399684889572086,
+      "learning_rate": 1.8797221142971295e-05,
+      "loss": 0.9735,
+      "step": 7768
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7832212732815802,
+      "learning_rate": 1.8796858300362682e-05,
+      "loss": 0.8702,
+      "step": 7769
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.403153393303939,
+      "learning_rate": 1.879649540653598e-05,
+      "loss": 0.8391,
+      "step": 7770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6235987925785593,
+      "learning_rate": 1.879613246149329e-05,
+      "loss": 1.0077,
+      "step": 7771
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9882540232842567,
+      "learning_rate": 1.8795769465236735e-05,
+      "loss": 0.9009,
+      "step": 7772
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.792132418327229,
+      "learning_rate": 1.8795406417768426e-05,
+      "loss": 1.0522,
+      "step": 7773
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7836031425838775,
+      "learning_rate": 1.8795043319090477e-05,
+      "loss": 0.7981,
+      "step": 7774
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7252404954404177,
+      "learning_rate": 1.8794680169205e-05,
+      "loss": 0.8656,
+      "step": 7775
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7741809132097595,
+      "learning_rate": 1.879431696811411e-05,
+      "loss": 0.8454,
+      "step": 7776
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.487966281157323,
+      "learning_rate": 1.8793953715819922e-05,
+      "loss": 0.8474,
+      "step": 7777
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0333973958274116,
+      "learning_rate": 1.8793590412324555e-05,
+      "loss": 0.8932,
+      "step": 7778
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5511762979097539,
+      "learning_rate": 1.8793227057630115e-05,
+      "loss": 0.8078,
+      "step": 7779
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6852168892044819,
+      "learning_rate": 1.8792863651738727e-05,
+      "loss": 1.0246,
+      "step": 7780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9769065215587054,
+      "learning_rate": 1.8792500194652505e-05,
+      "loss": 0.8658,
+      "step": 7781
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8480099811891029,
+      "learning_rate": 1.879213668637356e-05,
+      "loss": 0.9575,
+      "step": 7782
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7970685196411424,
+      "learning_rate": 1.879177312690401e-05,
+      "loss": 0.8249,
+      "step": 7783
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8984983436242213,
+      "learning_rate": 1.8791409516245974e-05,
+      "loss": 1.0446,
+      "step": 7784
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.72255793690541,
+      "learning_rate": 1.879104585440157e-05,
+      "loss": 0.8185,
+      "step": 7785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6414620272022964,
+      "learning_rate": 1.8790682141372913e-05,
+      "loss": 0.9561,
+      "step": 7786
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5464529709329564,
+      "learning_rate": 1.8790318377162126e-05,
+      "loss": 0.8332,
+      "step": 7787
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.725082410485334,
+      "learning_rate": 1.8789954561771314e-05,
+      "loss": 0.8121,
+      "step": 7788
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5968340927383426,
+      "learning_rate": 1.878959069520261e-05,
+      "loss": 0.8552,
+      "step": 7789
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7736796150996699,
+      "learning_rate": 1.8789226777458126e-05,
+      "loss": 0.8888,
+      "step": 7790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.111692749438293,
+      "learning_rate": 1.878886280853998e-05,
+      "loss": 0.8403,
+      "step": 7791
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1556953990640046,
+      "learning_rate": 1.878849878845029e-05,
+      "loss": 0.9033,
+      "step": 7792
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.049096737443023,
+      "learning_rate": 1.8788134717191182e-05,
+      "loss": 0.9068,
+      "step": 7793
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.698950490521332,
+      "learning_rate": 1.8787770594764768e-05,
+      "loss": 0.8443,
+      "step": 7794
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8368630226758196,
+      "learning_rate": 1.8787406421173173e-05,
+      "loss": 0.9166,
+      "step": 7795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7486254150266534,
+      "learning_rate": 1.8787042196418514e-05,
+      "loss": 0.9297,
+      "step": 7796
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7263955929768164,
+      "learning_rate": 1.8786677920502913e-05,
+      "loss": 0.9784,
+      "step": 7797
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6307086839804974,
+      "learning_rate": 1.8786313593428493e-05,
+      "loss": 0.9058,
+      "step": 7798
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8388879945470984,
+      "learning_rate": 1.8785949215197377e-05,
+      "loss": 1.0075,
+      "step": 7799
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5420573466646124,
+      "learning_rate": 1.8785584785811677e-05,
+      "loss": 0.8735,
+      "step": 7800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5602071258939898,
+      "learning_rate": 1.8785220305273526e-05,
+      "loss": 0.9417,
+      "step": 7801
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7713736136705525,
+      "learning_rate": 1.8784855773585037e-05,
+      "loss": 0.9488,
+      "step": 7802
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.646151656767299,
+      "learning_rate": 1.8784491190748338e-05,
+      "loss": 0.8176,
+      "step": 7803
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8436959252766123,
+      "learning_rate": 1.878412655676555e-05,
+      "loss": 0.8434,
+      "step": 7804
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4726617282416448,
+      "learning_rate": 1.8783761871638798e-05,
+      "loss": 0.979,
+      "step": 7805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7432371208030333,
+      "learning_rate": 1.87833971353702e-05,
+      "loss": 0.9415,
+      "step": 7806
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7792296658668217,
+      "learning_rate": 1.8783032347961884e-05,
+      "loss": 0.6903,
+      "step": 7807
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.732022714449328,
+      "learning_rate": 1.8782667509415974e-05,
+      "loss": 0.9611,
+      "step": 7808
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7992155961687337,
+      "learning_rate": 1.8782302619734594e-05,
+      "loss": 0.8855,
+      "step": 7809
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6419222195780563,
+      "learning_rate": 1.8781937678919867e-05,
+      "loss": 0.9322,
+      "step": 7810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5648381843144845,
+      "learning_rate": 1.878157268697392e-05,
+      "loss": 0.9144,
+      "step": 7811
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7572712419720186,
+      "learning_rate": 1.8781207643898873e-05,
+      "loss": 0.7692,
+      "step": 7812
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8471907501728007,
+      "learning_rate": 1.8780842549696857e-05,
+      "loss": 0.827,
+      "step": 7813
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9621558953107063,
+      "learning_rate": 1.8780477404369996e-05,
+      "loss": 0.8085,
+      "step": 7814
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5495643793352887,
+      "learning_rate": 1.8780112207920415e-05,
+      "loss": 0.9389,
+      "step": 7815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5319939906036848,
+      "learning_rate": 1.8779746960350243e-05,
+      "loss": 0.7375,
+      "step": 7816
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5308874831867232,
+      "learning_rate": 1.8779381661661603e-05,
+      "loss": 0.9844,
+      "step": 7817
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5890036132678644,
+      "learning_rate": 1.877901631185662e-05,
+      "loss": 0.8706,
+      "step": 7818
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1328958217904517,
+      "learning_rate": 1.877865091093743e-05,
+      "loss": 1.0791,
+      "step": 7819
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6230388813411543,
+      "learning_rate": 1.8778285458906152e-05,
+      "loss": 0.9412,
+      "step": 7820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.542006588447183,
+      "learning_rate": 1.8777919955764917e-05,
+      "loss": 0.9522,
+      "step": 7821
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0325356047523573,
+      "learning_rate": 1.8777554401515855e-05,
+      "loss": 0.9087,
+      "step": 7822
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6277774203138227,
+      "learning_rate": 1.877718879616109e-05,
+      "loss": 0.8544,
+      "step": 7823
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5755056069987392,
+      "learning_rate": 1.877682313970276e-05,
+      "loss": 0.9562,
+      "step": 7824
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6211533992382683,
+      "learning_rate": 1.877645743214298e-05,
+      "loss": 0.9439,
+      "step": 7825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6439200368055948,
+      "learning_rate": 1.8776091673483886e-05,
+      "loss": 0.9947,
+      "step": 7826
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8436145485205033,
+      "learning_rate": 1.877572586372761e-05,
+      "loss": 0.8648,
+      "step": 7827
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8047950374442585,
+      "learning_rate": 1.877536000287628e-05,
+      "loss": 0.8611,
+      "step": 7828
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.289797592903806,
+      "learning_rate": 1.8774994090932025e-05,
+      "loss": 0.9598,
+      "step": 7829
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7137104420394536,
+      "learning_rate": 1.8774628127896976e-05,
+      "loss": 0.8632,
+      "step": 7830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6734487421189523,
+      "learning_rate": 1.8774262113773262e-05,
+      "loss": 0.8031,
+      "step": 7831
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6977171734145837,
+      "learning_rate": 1.877389604856302e-05,
+      "loss": 0.9029,
+      "step": 7832
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5850691846524614,
+      "learning_rate": 1.8773529932268374e-05,
+      "loss": 0.8694,
+      "step": 7833
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6491023930331012,
+      "learning_rate": 1.8773163764891464e-05,
+      "loss": 0.9867,
+      "step": 7834
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3202160603647988,
+      "learning_rate": 1.8772797546434415e-05,
+      "loss": 0.9214,
+      "step": 7835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8536819344392335,
+      "learning_rate": 1.8772431276899358e-05,
+      "loss": 0.8429,
+      "step": 7836
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6968894127844045,
+      "learning_rate": 1.877206495628843e-05,
+      "loss": 0.9901,
+      "step": 7837
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6193354802357607,
+      "learning_rate": 1.8771698584603766e-05,
+      "loss": 0.9569,
+      "step": 7838
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4906941496509665,
+      "learning_rate": 1.8771332161847493e-05,
+      "loss": 0.8723,
+      "step": 7839
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7722849892473272,
+      "learning_rate": 1.8770965688021747e-05,
+      "loss": 0.7249,
+      "step": 7840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9078922015978268,
+      "learning_rate": 1.8770599163128663e-05,
+      "loss": 0.9206,
+      "step": 7841
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.981969822000119,
+      "learning_rate": 1.8770232587170374e-05,
+      "loss": 0.9709,
+      "step": 7842
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9717725588258361,
+      "learning_rate": 1.8769865960149012e-05,
+      "loss": 0.8776,
+      "step": 7843
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8253174408027601,
+      "learning_rate": 1.8769499282066716e-05,
+      "loss": 0.9314,
+      "step": 7844
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8033218942037434,
+      "learning_rate": 1.876913255292562e-05,
+      "loss": 0.859,
+      "step": 7845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7412767774412727,
+      "learning_rate": 1.8768765772727856e-05,
+      "loss": 0.9515,
+      "step": 7846
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6086905139284549,
+      "learning_rate": 1.876839894147556e-05,
+      "loss": 0.8757,
+      "step": 7847
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.449352033050925,
+      "learning_rate": 1.8768032059170874e-05,
+      "loss": 0.9779,
+      "step": 7848
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.74211690167898,
+      "learning_rate": 1.8767665125815927e-05,
+      "loss": 1.0102,
+      "step": 7849
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.638534439219203,
+      "learning_rate": 1.876729814141286e-05,
+      "loss": 0.8625,
+      "step": 7850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.818571645720796,
+      "learning_rate": 1.87669311059638e-05,
+      "loss": 0.8938,
+      "step": 7851
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2224571651859715,
+      "learning_rate": 1.8766564019470898e-05,
+      "loss": 1.0648,
+      "step": 7852
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.764506156417045,
+      "learning_rate": 1.8766196881936284e-05,
+      "loss": 0.7904,
+      "step": 7853
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8287071561877388,
+      "learning_rate": 1.8765829693362097e-05,
+      "loss": 0.7513,
+      "step": 7854
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4618092922946988,
+      "learning_rate": 1.8765462453750472e-05,
+      "loss": 0.9935,
+      "step": 7855
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7150910409884859,
+      "learning_rate": 1.876509516310355e-05,
+      "loss": 0.8969,
+      "step": 7856
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.754556594746707,
+      "learning_rate": 1.8764727821423473e-05,
+      "loss": 0.7353,
+      "step": 7857
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5068511916536218,
+      "learning_rate": 1.876436042871237e-05,
+      "loss": 0.9018,
+      "step": 7858
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6120003603157063,
+      "learning_rate": 1.8763992984972388e-05,
+      "loss": 0.8326,
+      "step": 7859
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5761604873326804,
+      "learning_rate": 1.8763625490205664e-05,
+      "loss": 0.7865,
+      "step": 7860
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5844935340674078,
+      "learning_rate": 1.876325794441434e-05,
+      "loss": 0.8923,
+      "step": 7861
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6468021300371156,
+      "learning_rate": 1.876289034760055e-05,
+      "loss": 0.8968,
+      "step": 7862
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7134635694901572,
+      "learning_rate": 1.8762522699766443e-05,
+      "loss": 0.8413,
+      "step": 7863
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.83391841503783,
+      "learning_rate": 1.876215500091415e-05,
+      "loss": 0.8566,
+      "step": 7864
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7283400360812213,
+      "learning_rate": 1.876178725104582e-05,
+      "loss": 0.7477,
+      "step": 7865
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.808038079341751,
+      "learning_rate": 1.8761419450163586e-05,
+      "loss": 0.785,
+      "step": 7866
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8525340115886035,
+      "learning_rate": 1.87610515982696e-05,
+      "loss": 0.8559,
+      "step": 7867
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5784489874398995,
+      "learning_rate": 1.8760683695365994e-05,
+      "loss": 0.8096,
+      "step": 7868
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.654175986540167,
+      "learning_rate": 1.8760315741454917e-05,
+      "loss": 0.9429,
+      "step": 7869
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6231229309144686,
+      "learning_rate": 1.8759947736538507e-05,
+      "loss": 0.8462,
+      "step": 7870
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.009245802781256,
+      "learning_rate": 1.8759579680618907e-05,
+      "loss": 1.0377,
+      "step": 7871
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7018480684076867,
+      "learning_rate": 1.8759211573698263e-05,
+      "loss": 0.9735,
+      "step": 7872
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.503917927568943,
+      "learning_rate": 1.875884341577871e-05,
+      "loss": 0.9745,
+      "step": 7873
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5345495339609168,
+      "learning_rate": 1.87584752068624e-05,
+      "loss": 0.9307,
+      "step": 7874
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5974099506523627,
+      "learning_rate": 1.8758106946951484e-05,
+      "loss": 0.98,
+      "step": 7875
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7881326778847126,
+      "learning_rate": 1.8757738636048088e-05,
+      "loss": 0.9573,
+      "step": 7876
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6635896911830892,
+      "learning_rate": 1.8757370274154362e-05,
+      "loss": 0.8645,
+      "step": 7877
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7936593235230753,
+      "learning_rate": 1.875700186127246e-05,
+      "loss": 0.9717,
+      "step": 7878
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.577083427920682,
+      "learning_rate": 1.8756633397404518e-05,
+      "loss": 0.8622,
+      "step": 7879
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9306249507238034,
+      "learning_rate": 1.8756264882552685e-05,
+      "loss": 0.87,
+      "step": 7880
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9562979287663531,
+      "learning_rate": 1.8755896316719104e-05,
+      "loss": 0.908,
+      "step": 7881
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5591614401244853,
+      "learning_rate": 1.8755527699905927e-05,
+      "loss": 0.9319,
+      "step": 7882
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8859721086322304,
+      "learning_rate": 1.8755159032115292e-05,
+      "loss": 0.7482,
+      "step": 7883
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.517422041789149,
+      "learning_rate": 1.8754790313349347e-05,
+      "loss": 0.8349,
+      "step": 7884
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6746852014002969,
+      "learning_rate": 1.8754421543610247e-05,
+      "loss": 0.8742,
+      "step": 7885
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6163285444938573,
+      "learning_rate": 1.875405272290013e-05,
+      "loss": 0.8454,
+      "step": 7886
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4635776207351454,
+      "learning_rate": 1.8753683851221146e-05,
+      "loss": 0.8661,
+      "step": 7887
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4382468473610774,
+      "learning_rate": 1.8753314928575444e-05,
+      "loss": 0.8238,
+      "step": 7888
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6793733936407178,
+      "learning_rate": 1.875294595496517e-05,
+      "loss": 0.9363,
+      "step": 7889
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8482515185455236,
+      "learning_rate": 1.8752576930392473e-05,
+      "loss": 0.696,
+      "step": 7890
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7552564837394504,
+      "learning_rate": 1.8752207854859507e-05,
+      "loss": 0.9773,
+      "step": 7891
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7407044508063598,
+      "learning_rate": 1.875183872836841e-05,
+      "loss": 0.8918,
+      "step": 7892
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.883404481513779,
+      "learning_rate": 1.8751469550921342e-05,
+      "loss": 0.9161,
+      "step": 7893
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.866056887064672,
+      "learning_rate": 1.8751100322520447e-05,
+      "loss": 0.8964,
+      "step": 7894
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.808871649358013,
+      "learning_rate": 1.8750731043167873e-05,
+      "loss": 0.7622,
+      "step": 7895
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7515282788260063,
+      "learning_rate": 1.8750361712865774e-05,
+      "loss": 0.8522,
+      "step": 7896
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5473166168408956,
+      "learning_rate": 1.8749992331616296e-05,
+      "loss": 0.906,
+      "step": 7897
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8934158650225843,
+      "learning_rate": 1.8749622899421597e-05,
+      "loss": 0.8287,
+      "step": 7898
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6776843715436451,
+      "learning_rate": 1.874925341628382e-05,
+      "loss": 0.9501,
+      "step": 7899
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6976507400922993,
+      "learning_rate": 1.874888388220512e-05,
+      "loss": 0.8916,
+      "step": 7900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9379094574976206,
+      "learning_rate": 1.874851429718765e-05,
+      "loss": 0.9967,
+      "step": 7901
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5954174556048286,
+      "learning_rate": 1.8748144661233558e-05,
+      "loss": 0.772,
+      "step": 7902
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6516433396588783,
+      "learning_rate": 1.8747774974345e-05,
+      "loss": 0.8623,
+      "step": 7903
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7162406836052893,
+      "learning_rate": 1.8747405236524125e-05,
+      "loss": 1.0677,
+      "step": 7904
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7361261200475453,
+      "learning_rate": 1.8747035447773086e-05,
+      "loss": 0.8129,
+      "step": 7905
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5840604204147142,
+      "learning_rate": 1.874666560809404e-05,
+      "loss": 1.0064,
+      "step": 7906
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6305034795558888,
+      "learning_rate": 1.8746295717489138e-05,
+      "loss": 0.8402,
+      "step": 7907
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5540978738235651,
+      "learning_rate": 1.874592577596053e-05,
+      "loss": 0.8855,
+      "step": 7908
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5738855997330974,
+      "learning_rate": 1.8745555783510378e-05,
+      "loss": 0.9118,
+      "step": 7909
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6575600540756132,
+      "learning_rate": 1.874518574014083e-05,
+      "loss": 0.9472,
+      "step": 7910
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4737068123957133,
+      "learning_rate": 1.874481564585404e-05,
+      "loss": 0.8323,
+      "step": 7911
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5803572626023705,
+      "learning_rate": 1.8744445500652165e-05,
+      "loss": 0.9409,
+      "step": 7912
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6463486737422177,
+      "learning_rate": 1.8744075304537358e-05,
+      "loss": 0.9288,
+      "step": 7913
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.534663973906103,
+      "learning_rate": 1.8743705057511777e-05,
+      "loss": 0.9365,
+      "step": 7914
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7115704225843875,
+      "learning_rate": 1.8743334759577582e-05,
+      "loss": 0.8992,
+      "step": 7915
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5641984749253854,
+      "learning_rate": 1.874296441073692e-05,
+      "loss": 0.8526,
+      "step": 7916
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5322504595974182,
+      "learning_rate": 1.8742594010991953e-05,
+      "loss": 0.939,
+      "step": 7917
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8103096893354433,
+      "learning_rate": 1.8742223560344835e-05,
+      "loss": 0.9745,
+      "step": 7918
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6494817313696957,
+      "learning_rate": 1.874185305879772e-05,
+      "loss": 0.9454,
+      "step": 7919
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7903992363430395,
+      "learning_rate": 1.8741482506352776e-05,
+      "loss": 0.7765,
+      "step": 7920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7650892679839416,
+      "learning_rate": 1.8741111903012148e-05,
+      "loss": 0.9119,
+      "step": 7921
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.155409959151394,
+      "learning_rate": 1.8740741248778002e-05,
+      "loss": 0.833,
+      "step": 7922
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0003335925161956,
+      "learning_rate": 1.8740370543652487e-05,
+      "loss": 0.7727,
+      "step": 7923
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7673628491307829,
+      "learning_rate": 1.8739999787637775e-05,
+      "loss": 0.724,
+      "step": 7924
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.526566761959507,
+      "learning_rate": 1.8739628980736016e-05,
+      "loss": 0.9343,
+      "step": 7925
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8413760064440197,
+      "learning_rate": 1.8739258122949367e-05,
+      "loss": 0.904,
+      "step": 7926
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8767779360751875,
+      "learning_rate": 1.873888721427999e-05,
+      "loss": 0.9416,
+      "step": 7927
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.514828908229341,
+      "learning_rate": 1.8738516254730052e-05,
+      "loss": 0.7627,
+      "step": 7928
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6766016542852902,
+      "learning_rate": 1.87381452443017e-05,
+      "loss": 0.9136,
+      "step": 7929
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.67038028361938,
+      "learning_rate": 1.87377741829971e-05,
+      "loss": 1.0444,
+      "step": 7930
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8624777295289052,
+      "learning_rate": 1.873740307081841e-05,
+      "loss": 0.9164,
+      "step": 7931
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4159553327138759,
+      "learning_rate": 1.8737031907767797e-05,
+      "loss": 0.9802,
+      "step": 7932
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6318462259493143,
+      "learning_rate": 1.873666069384742e-05,
+      "loss": 0.9072,
+      "step": 7933
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6994971538139991,
+      "learning_rate": 1.8736289429059436e-05,
+      "loss": 0.9866,
+      "step": 7934
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8079800901253021,
+      "learning_rate": 1.873591811340601e-05,
+      "loss": 0.7567,
+      "step": 7935
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7487839569484775,
+      "learning_rate": 1.87355467468893e-05,
+      "loss": 0.9128,
+      "step": 7936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2041660637115656,
+      "learning_rate": 1.873517532951147e-05,
+      "loss": 0.8379,
+      "step": 7937
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3254815121474164,
+      "learning_rate": 1.8734803861274692e-05,
+      "loss": 0.9285,
+      "step": 7938
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5687902694141462,
+      "learning_rate": 1.8734432342181113e-05,
+      "loss": 0.8628,
+      "step": 7939
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7905617050347578,
+      "learning_rate": 1.8734060772232906e-05,
+      "loss": 0.9038,
+      "step": 7940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.508865260735506,
+      "learning_rate": 1.873368915143223e-05,
+      "loss": 0.9769,
+      "step": 7941
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.008843594359375,
+      "learning_rate": 1.8733317479781256e-05,
+      "loss": 0.9652,
+      "step": 7942
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4177261013481681,
+      "learning_rate": 1.8732945757282137e-05,
+      "loss": 0.8881,
+      "step": 7943
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6011513582082166,
+      "learning_rate": 1.8732573983937046e-05,
+      "loss": 0.9621,
+      "step": 7944
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4876644843633002,
+      "learning_rate": 1.8732202159748146e-05,
+      "loss": 0.906,
+      "step": 7945
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.691193963063789,
+      "learning_rate": 1.8731830284717597e-05,
+      "loss": 0.8873,
+      "step": 7946
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.530996448987458,
+      "learning_rate": 1.873145835884757e-05,
+      "loss": 1.0394,
+      "step": 7947
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7817114948509059,
+      "learning_rate": 1.8731086382140228e-05,
+      "loss": 0.767,
+      "step": 7948
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5015636435427404,
+      "learning_rate": 1.8730714354597738e-05,
+      "loss": 1.0243,
+      "step": 7949
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7983700827624396,
+      "learning_rate": 1.873034227622226e-05,
+      "loss": 0.8964,
+      "step": 7950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.784272800151075,
+      "learning_rate": 1.8729970147015968e-05,
+      "loss": 0.8712,
+      "step": 7951
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8688831496744072,
+      "learning_rate": 1.8729597966981025e-05,
+      "loss": 0.7921,
+      "step": 7952
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5125430359778593,
+      "learning_rate": 1.8729225736119602e-05,
+      "loss": 0.9402,
+      "step": 7953
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.680977882064403,
+      "learning_rate": 1.872885345443386e-05,
+      "loss": 0.8882,
+      "step": 7954
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.776720206449011,
+      "learning_rate": 1.872848112192597e-05,
+      "loss": 0.7579,
+      "step": 7955
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.760548146697104,
+      "learning_rate": 1.8728108738598103e-05,
+      "loss": 0.8691,
+      "step": 7956
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8045001182815148,
+      "learning_rate": 1.872773630445242e-05,
+      "loss": 0.9424,
+      "step": 7957
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9593244858585561,
+      "learning_rate": 1.8727363819491097e-05,
+      "loss": 0.9128,
+      "step": 7958
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6250214111082135,
+      "learning_rate": 1.8726991283716295e-05,
+      "loss": 0.8602,
+      "step": 7959
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7388716414668963,
+      "learning_rate": 1.8726618697130186e-05,
+      "loss": 0.6543,
+      "step": 7960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1843202140138893,
+      "learning_rate": 1.872624605973494e-05,
+      "loss": 0.8731,
+      "step": 7961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.277610249751547,
+      "learning_rate": 1.8725873371532733e-05,
+      "loss": 0.8837,
+      "step": 7962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.581489576109475,
+      "learning_rate": 1.872550063252572e-05,
+      "loss": 0.8532,
+      "step": 7963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.715745905241587,
+      "learning_rate": 1.8725127842716086e-05,
+      "loss": 0.9858,
+      "step": 7964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.877709134604333,
+      "learning_rate": 1.8724755002105996e-05,
+      "loss": 0.9134,
+      "step": 7965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6421223866242272,
+      "learning_rate": 1.8724382110697615e-05,
+      "loss": 1.0171,
+      "step": 7966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5864247020918538,
+      "learning_rate": 1.8724009168493126e-05,
+      "loss": 0.87,
+      "step": 7967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.462031605619473,
+      "learning_rate": 1.872363617549469e-05,
+      "loss": 0.9027,
+      "step": 7968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7715483701546703,
+      "learning_rate": 1.872326313170448e-05,
+      "loss": 0.7513,
+      "step": 7969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2246432664844766,
+      "learning_rate": 1.8722890037124674e-05,
+      "loss": 1.0158,
+      "step": 7970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6886197275185038,
+      "learning_rate": 1.8722516891757435e-05,
+      "loss": 0.9415,
+      "step": 7971
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7729016827272848,
+      "learning_rate": 1.872214369560495e-05,
+      "loss": 0.9029,
+      "step": 7972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.735438282542121,
+      "learning_rate": 1.8721770448669377e-05,
+      "loss": 0.9209,
+      "step": 7973
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7503614274950654,
+      "learning_rate": 1.8721397150952894e-05,
+      "loss": 0.8924,
+      "step": 7974
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.454095937035569,
+      "learning_rate": 1.8721023802457678e-05,
+      "loss": 0.9504,
+      "step": 7975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.09622340721608,
+      "learning_rate": 1.87206504031859e-05,
+      "loss": 0.9187,
+      "step": 7976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9519926549327342,
+      "learning_rate": 1.8720276953139737e-05,
+      "loss": 0.8797,
+      "step": 7977
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9445459467166823,
+      "learning_rate": 1.871990345232136e-05,
+      "loss": 0.9337,
+      "step": 7978
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8946488394941403,
+      "learning_rate": 1.871952990073294e-05,
+      "loss": 0.9498,
+      "step": 7979
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4734567527717892,
+      "learning_rate": 1.871915629837666e-05,
+      "loss": 0.8536,
+      "step": 7980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.632195289443804,
+      "learning_rate": 1.8718782645254688e-05,
+      "loss": 0.8397,
+      "step": 7981
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4271739028486834,
+      "learning_rate": 1.8718408941369208e-05,
+      "loss": 0.9634,
+      "step": 7982
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1325812920766154,
+      "learning_rate": 1.871803518672239e-05,
+      "loss": 0.9107,
+      "step": 7983
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.162359480703867,
+      "learning_rate": 1.871766138131641e-05,
+      "loss": 0.7853,
+      "step": 7984
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.71186831858812,
+      "learning_rate": 1.8717287525153443e-05,
+      "loss": 1.0244,
+      "step": 7985
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.523014975624438,
+      "learning_rate": 1.8716913618235676e-05,
+      "loss": 0.8302,
+      "step": 7986
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6364282055896768,
+      "learning_rate": 1.871653966056527e-05,
+      "loss": 1.0287,
+      "step": 7987
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5596120035069652,
+      "learning_rate": 1.8716165652144415e-05,
+      "loss": 0.7954,
+      "step": 7988
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.644830535029658,
+      "learning_rate": 1.871579159297528e-05,
+      "loss": 0.9193,
+      "step": 7989
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.542651365972366,
+      "learning_rate": 1.871541748306005e-05,
+      "loss": 0.9677,
+      "step": 7990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8990598147827644,
+      "learning_rate": 1.87150433224009e-05,
+      "loss": 0.8816,
+      "step": 7991
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.150859870282456,
+      "learning_rate": 1.871466911100001e-05,
+      "loss": 0.8975,
+      "step": 7992
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5579840559331481,
+      "learning_rate": 1.8714294848859555e-05,
+      "loss": 0.8903,
+      "step": 7993
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7667558242248181,
+      "learning_rate": 1.8713920535981716e-05,
+      "loss": 0.7386,
+      "step": 7994
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8643328641339928,
+      "learning_rate": 1.8713546172368676e-05,
+      "loss": 0.7257,
+      "step": 7995
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6923986470894052,
+      "learning_rate": 1.8713171758022608e-05,
+      "loss": 0.8969,
+      "step": 7996
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5616713687738668,
+      "learning_rate": 1.8712797292945698e-05,
+      "loss": 0.9901,
+      "step": 7997
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4857045568873928,
+      "learning_rate": 1.8712422777140123e-05,
+      "loss": 0.8769,
+      "step": 7998
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4928081654650818,
+      "learning_rate": 1.8712048210608062e-05,
+      "loss": 0.8896,
+      "step": 7999
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7007204204138997,
+      "learning_rate": 1.87116735933517e-05,
+      "loss": 0.8423,
+      "step": 8000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7244121562547783,
+      "learning_rate": 1.8711298925373218e-05,
+      "loss": 0.7356,
+      "step": 8001
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7278468937209825,
+      "learning_rate": 1.8710924206674795e-05,
+      "loss": 0.825,
+      "step": 8002
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6152025698970685,
+      "learning_rate": 1.871054943725861e-05,
+      "loss": 0.9519,
+      "step": 8003
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6999538442804898,
+      "learning_rate": 1.871017461712685e-05,
+      "loss": 1.0017,
+      "step": 8004
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.03490964746927,
+      "learning_rate": 1.87097997462817e-05,
+      "loss": 0.868,
+      "step": 8005
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6448875438093984,
+      "learning_rate": 1.8709424824725335e-05,
+      "loss": 0.9374,
+      "step": 8006
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.490499294293943,
+      "learning_rate": 1.8709049852459942e-05,
+      "loss": 0.9032,
+      "step": 8007
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5830464390743417,
+      "learning_rate": 1.87086748294877e-05,
+      "loss": 0.8415,
+      "step": 8008
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6382929742506547,
+      "learning_rate": 1.87082997558108e-05,
+      "loss": 0.9298,
+      "step": 8009
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4467507184541728,
+      "learning_rate": 1.8707924631431424e-05,
+      "loss": 0.9681,
+      "step": 8010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8648042772452513,
+      "learning_rate": 1.870754945635175e-05,
+      "loss": 0.9278,
+      "step": 8011
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0877083128052303,
+      "learning_rate": 1.8707174230573967e-05,
+      "loss": 0.9711,
+      "step": 8012
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.673476940191639,
+      "learning_rate": 1.870679895410026e-05,
+      "loss": 0.9145,
+      "step": 8013
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.637510834690881,
+      "learning_rate": 1.870642362693281e-05,
+      "loss": 0.9079,
+      "step": 8014
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.732408552218697,
+      "learning_rate": 1.870604824907381e-05,
+      "loss": 0.9647,
+      "step": 8015
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.998693286810658,
+      "learning_rate": 1.8705672820525437e-05,
+      "loss": 0.9102,
+      "step": 8016
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6745235181032798,
+      "learning_rate": 1.870529734128988e-05,
+      "loss": 0.9644,
+      "step": 8017
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8052650797293572,
+      "learning_rate": 1.8704921811369328e-05,
+      "loss": 0.9688,
+      "step": 8018
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8280726262798956,
+      "learning_rate": 1.8704546230765964e-05,
+      "loss": 0.7896,
+      "step": 8019
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.509149012868513,
+      "learning_rate": 1.870417059948198e-05,
+      "loss": 0.8873,
+      "step": 8020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.168349249938789,
+      "learning_rate": 1.8703794917519555e-05,
+      "loss": 0.898,
+      "step": 8021
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3621133707236504,
+      "learning_rate": 1.8703419184880882e-05,
+      "loss": 0.8926,
+      "step": 8022
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9362749784005635,
+      "learning_rate": 1.8703043401568144e-05,
+      "loss": 1.0955,
+      "step": 8023
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8952957728082713,
+      "learning_rate": 1.8702667567583533e-05,
+      "loss": 1.0351,
+      "step": 8024
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7495891644133899,
+      "learning_rate": 1.870229168292924e-05,
+      "loss": 0.8713,
+      "step": 8025
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6911275163852841,
+      "learning_rate": 1.8701915747607446e-05,
+      "loss": 0.8488,
+      "step": 8026
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9614743496351792,
+      "learning_rate": 1.8701539761620348e-05,
+      "loss": 0.8886,
+      "step": 8027
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3347898388280814,
+      "learning_rate": 1.8701163724970125e-05,
+      "loss": 0.8526,
+      "step": 8028
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9239687342380136,
+      "learning_rate": 1.8700787637658976e-05,
+      "loss": 0.8646,
+      "step": 8029
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5885178909633908,
+      "learning_rate": 1.8700411499689088e-05,
+      "loss": 1.016,
+      "step": 8030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.825253626246258,
+      "learning_rate": 1.8700035311062644e-05,
+      "loss": 0.9105,
+      "step": 8031
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7564401370249796,
+      "learning_rate": 1.8699659071781845e-05,
+      "loss": 0.9703,
+      "step": 8032
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2022873864775407,
+      "learning_rate": 1.8699282781848873e-05,
+      "loss": 0.8547,
+      "step": 8033
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7851382788641966,
+      "learning_rate": 1.8698906441265925e-05,
+      "loss": 0.7584,
+      "step": 8034
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9384648787533123,
+      "learning_rate": 1.8698530050035188e-05,
+      "loss": 0.9687,
+      "step": 8035
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.044883290349114,
+      "learning_rate": 1.8698153608158856e-05,
+      "loss": 0.8727,
+      "step": 8036
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.858550952229815,
+      "learning_rate": 1.869777711563912e-05,
+      "loss": 0.9266,
+      "step": 8037
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6378399191216633,
+      "learning_rate": 1.8697400572478173e-05,
+      "loss": 1.0145,
+      "step": 8038
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.639621517954134,
+      "learning_rate": 1.8697023978678205e-05,
+      "loss": 0.9763,
+      "step": 8039
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7044697609522181,
+      "learning_rate": 1.869664733424141e-05,
+      "loss": 1.0152,
+      "step": 8040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7130804025062947,
+      "learning_rate": 1.869627063916998e-05,
+      "loss": 0.9778,
+      "step": 8041
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7927625782398324,
+      "learning_rate": 1.869589389346611e-05,
+      "loss": 0.7611,
+      "step": 8042
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.68114182843984,
+      "learning_rate": 1.8695517097131993e-05,
+      "loss": 1.0045,
+      "step": 8043
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9942405228013547,
+      "learning_rate": 1.8695140250169825e-05,
+      "loss": 0.9747,
+      "step": 8044
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4823423308208332,
+      "learning_rate": 1.8694763352581792e-05,
+      "loss": 1.0429,
+      "step": 8045
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3691375422880254,
+      "learning_rate": 1.86943864043701e-05,
+      "loss": 1.0178,
+      "step": 8046
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2129747026984408,
+      "learning_rate": 1.8694009405536934e-05,
+      "loss": 0.8889,
+      "step": 8047
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.738500162301038,
+      "learning_rate": 1.8693632356084493e-05,
+      "loss": 1.0002,
+      "step": 8048
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9503857442509027,
+      "learning_rate": 1.8693255256014976e-05,
+      "loss": 0.997,
+      "step": 8049
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6335392692675073,
+      "learning_rate": 1.8692878105330568e-05,
+      "loss": 1.0533,
+      "step": 8050
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1753862242719233,
+      "learning_rate": 1.8692500904033478e-05,
+      "loss": 0.9077,
+      "step": 8051
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7590762820270576,
+      "learning_rate": 1.869212365212589e-05,
+      "loss": 0.7653,
+      "step": 8052
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3267504020863266,
+      "learning_rate": 1.8691746349610008e-05,
+      "loss": 0.999,
+      "step": 8053
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.562508487138199,
+      "learning_rate": 1.869136899648803e-05,
+      "loss": 0.846,
+      "step": 8054
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6612632567512584,
+      "learning_rate": 1.8690991592762147e-05,
+      "loss": 0.9185,
+      "step": 8055
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5634641616877223,
+      "learning_rate": 1.869061413843456e-05,
+      "loss": 0.9798,
+      "step": 8056
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.654810087125079,
+      "learning_rate": 1.869023663350747e-05,
+      "loss": 0.8643,
+      "step": 8057
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9680887327547483,
+      "learning_rate": 1.8689859077983064e-05,
+      "loss": 0.8967,
+      "step": 8058
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9235345774133719,
+      "learning_rate": 1.868948147186355e-05,
+      "loss": 0.9238,
+      "step": 8059
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7330987133888176,
+      "learning_rate": 1.8689103815151127e-05,
+      "loss": 0.8992,
+      "step": 8060
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7960710191400329,
+      "learning_rate": 1.8688726107847986e-05,
+      "loss": 0.7519,
+      "step": 8061
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.636233523880956,
+      "learning_rate": 1.8688348349956334e-05,
+      "loss": 0.8718,
+      "step": 8062
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5668187768002628,
+      "learning_rate": 1.8687970541478367e-05,
+      "loss": 1.0862,
+      "step": 8063
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6488929397344376,
+      "learning_rate": 1.8687592682416284e-05,
+      "loss": 1.0099,
+      "step": 8064
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6522906232477406,
+      "learning_rate": 1.8687214772772283e-05,
+      "loss": 0.8284,
+      "step": 8065
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5818652310578079,
+      "learning_rate": 1.8686836812548573e-05,
+      "loss": 0.8167,
+      "step": 8066
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8045026162931486,
+      "learning_rate": 1.8686458801747347e-05,
+      "loss": 0.7212,
+      "step": 8067
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4344391598534467,
+      "learning_rate": 1.8686080740370807e-05,
+      "loss": 0.9834,
+      "step": 8068
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.775512837095171,
+      "learning_rate": 1.8685702628421155e-05,
+      "loss": 0.7463,
+      "step": 8069
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5312263394341468,
+      "learning_rate": 1.8685324465900594e-05,
+      "loss": 1.0372,
+      "step": 8070
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8886987717126296,
+      "learning_rate": 1.868494625281132e-05,
+      "loss": 0.8999,
+      "step": 8071
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7420641287382632,
+      "learning_rate": 1.8684567989155545e-05,
+      "loss": 0.6992,
+      "step": 8072
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8345677477052025,
+      "learning_rate": 1.868418967493546e-05,
+      "loss": 0.7892,
+      "step": 8073
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5537924926834767,
+      "learning_rate": 1.8683811310153276e-05,
+      "loss": 0.9351,
+      "step": 8074
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0019823028019874,
+      "learning_rate": 1.8683432894811195e-05,
+      "loss": 0.9213,
+      "step": 8075
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7331964668790474,
+      "learning_rate": 1.8683054428911414e-05,
+      "loss": 0.9763,
+      "step": 8076
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6330529056933512,
+      "learning_rate": 1.8682675912456147e-05,
+      "loss": 0.8438,
+      "step": 8077
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4351194315238405,
+      "learning_rate": 1.868229734544759e-05,
+      "loss": 0.8797,
+      "step": 8078
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1933626827501023,
+      "learning_rate": 1.8681918727887945e-05,
+      "loss": 1.037,
+      "step": 8079
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5655706952159887,
+      "learning_rate": 1.8681540059779425e-05,
+      "loss": 0.8878,
+      "step": 8080
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5564476938237841,
+      "learning_rate": 1.8681161341124227e-05,
+      "loss": 0.9804,
+      "step": 8081
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.585277595958703,
+      "learning_rate": 1.8680782571924564e-05,
+      "loss": 0.8122,
+      "step": 8082
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5003340813088635,
+      "learning_rate": 1.868040375218263e-05,
+      "loss": 0.8675,
+      "step": 8083
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8574583331612973,
+      "learning_rate": 1.868002488190064e-05,
+      "loss": 0.9324,
+      "step": 8084
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.584974438957643,
+      "learning_rate": 1.86796459610808e-05,
+      "loss": 0.9257,
+      "step": 8085
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5554748923842268,
+      "learning_rate": 1.8679266989725313e-05,
+      "loss": 0.8662,
+      "step": 8086
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.198955518705947,
+      "learning_rate": 1.8678887967836382e-05,
+      "loss": 1.0412,
+      "step": 8087
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3891987310234426,
+      "learning_rate": 1.867850889541622e-05,
+      "loss": 0.9184,
+      "step": 8088
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8210574788099054,
+      "learning_rate": 1.867812977246703e-05,
+      "loss": 0.9652,
+      "step": 8089
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7295025426233142,
+      "learning_rate": 1.8677750598991023e-05,
+      "loss": 0.9335,
+      "step": 8090
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8209818266669111,
+      "learning_rate": 1.8677371374990406e-05,
+      "loss": 0.7247,
+      "step": 8091
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6706166396767121,
+      "learning_rate": 1.8676992100467383e-05,
+      "loss": 1.0183,
+      "step": 8092
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2137123487064496,
+      "learning_rate": 1.8676612775424165e-05,
+      "loss": 0.913,
+      "step": 8093
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9346667186701154,
+      "learning_rate": 1.8676233399862963e-05,
+      "loss": 0.9875,
+      "step": 8094
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7673558856118545,
+      "learning_rate": 1.8675853973785976e-05,
+      "loss": 0.7062,
+      "step": 8095
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.029393746599631,
+      "learning_rate": 1.8675474497195428e-05,
+      "loss": 0.863,
+      "step": 8096
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8726630997373357,
+      "learning_rate": 1.867509497009352e-05,
+      "loss": 0.8298,
+      "step": 8097
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7555267954923708,
+      "learning_rate": 1.867471539248246e-05,
+      "loss": 0.7251,
+      "step": 8098
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7514716743445287,
+      "learning_rate": 1.8674335764364464e-05,
+      "loss": 0.9812,
+      "step": 8099
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.70614880635276,
+      "learning_rate": 1.8673956085741734e-05,
+      "loss": 1.019,
+      "step": 8100
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7338985836761276,
+      "learning_rate": 1.8673576356616488e-05,
+      "loss": 0.9125,
+      "step": 8101
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6456568350507927,
+      "learning_rate": 1.8673196576990938e-05,
+      "loss": 0.8867,
+      "step": 8102
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7628643261964667,
+      "learning_rate": 1.8672816746867285e-05,
+      "loss": 0.8825,
+      "step": 8103
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8510764359439773,
+      "learning_rate": 1.867243686624775e-05,
+      "loss": 0.8934,
+      "step": 8104
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.363853353212601,
+      "learning_rate": 1.8672056935134545e-05,
+      "loss": 0.9111,
+      "step": 8105
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7929423652060168,
+      "learning_rate": 1.8671676953529875e-05,
+      "loss": 0.9396,
+      "step": 8106
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7256449487492345,
+      "learning_rate": 1.8671296921435956e-05,
+      "loss": 0.946,
+      "step": 8107
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9055713881025165,
+      "learning_rate": 1.8670916838855005e-05,
+      "loss": 0.7326,
+      "step": 8108
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8200426626097458,
+      "learning_rate": 1.867053670578923e-05,
+      "loss": 0.7841,
+      "step": 8109
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8435080282218501,
+      "learning_rate": 1.8670156522240843e-05,
+      "loss": 1.0318,
+      "step": 8110
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5667196295546209,
+      "learning_rate": 1.866977628821206e-05,
+      "loss": 0.8682,
+      "step": 8111
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8458185562868104,
+      "learning_rate": 1.8669396003705095e-05,
+      "loss": 0.9158,
+      "step": 8112
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6466612781451826,
+      "learning_rate": 1.8669015668722162e-05,
+      "loss": 0.8757,
+      "step": 8113
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6256352591398542,
+      "learning_rate": 1.8668635283265475e-05,
+      "loss": 0.8634,
+      "step": 8114
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5041288395749315,
+      "learning_rate": 1.8668254847337252e-05,
+      "loss": 0.995,
+      "step": 8115
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8226517431349895,
+      "learning_rate": 1.86678743609397e-05,
+      "loss": 0.7299,
+      "step": 8116
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.735626988045814,
+      "learning_rate": 1.8667493824075042e-05,
+      "loss": 0.8171,
+      "step": 8117
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.796784359372091,
+      "learning_rate": 1.866711323674549e-05,
+      "loss": 0.7763,
+      "step": 8118
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.054100661090207,
+      "learning_rate": 1.8666732598953263e-05,
+      "loss": 0.8944,
+      "step": 8119
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9037838668799993,
+      "learning_rate": 1.8666351910700572e-05,
+      "loss": 0.9754,
+      "step": 8120
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6932452276736287,
+      "learning_rate": 1.8665971171989638e-05,
+      "loss": 0.8632,
+      "step": 8121
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8716648413403623,
+      "learning_rate": 1.8665590382822675e-05,
+      "loss": 0.8028,
+      "step": 8122
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7330547622700794,
+      "learning_rate": 1.8665209543201904e-05,
+      "loss": 0.976,
+      "step": 8123
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.96556314338623,
+      "learning_rate": 1.866482865312954e-05,
+      "loss": 0.8556,
+      "step": 8124
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.239338407433914,
+      "learning_rate": 1.8664447712607798e-05,
+      "loss": 0.9441,
+      "step": 8125
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5100552087751857,
+      "learning_rate": 1.8664066721638895e-05,
+      "loss": 0.8879,
+      "step": 8126
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7220491578244397,
+      "learning_rate": 1.8663685680225056e-05,
+      "loss": 0.9603,
+      "step": 8127
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8871438485903254,
+      "learning_rate": 1.8663304588368495e-05,
+      "loss": 1.0613,
+      "step": 8128
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6154940479965045,
+      "learning_rate": 1.8662923446071436e-05,
+      "loss": 0.9684,
+      "step": 8129
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6464506008447388,
+      "learning_rate": 1.866254225333609e-05,
+      "loss": 0.777,
+      "step": 8130
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.941612602921063,
+      "learning_rate": 1.8662161010164682e-05,
+      "loss": 1.0058,
+      "step": 8131
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6694441029275606,
+      "learning_rate": 1.8661779716559428e-05,
+      "loss": 0.9353,
+      "step": 8132
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.626443660575807,
+      "learning_rate": 1.866139837252255e-05,
+      "loss": 0.9609,
+      "step": 8133
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8967591059335833,
+      "learning_rate": 1.8661016978056273e-05,
+      "loss": 0.8965,
+      "step": 8134
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3518484433200395,
+      "learning_rate": 1.8660635533162813e-05,
+      "loss": 0.9739,
+      "step": 8135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.014295055380223,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 0.9239,
+      "step": 8136
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8400981439239243,
+      "learning_rate": 1.8659872492103222e-05,
+      "loss": 0.8491,
+      "step": 8137
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8165212501088106,
+      "learning_rate": 1.865949089594154e-05,
+      "loss": 0.8356,
+      "step": 8138
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.652975799394754,
+      "learning_rate": 1.8659109249361557e-05,
+      "loss": 0.9742,
+      "step": 8139
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.933695220452242,
+      "learning_rate": 1.8658727552365498e-05,
+      "loss": 1.0863,
+      "step": 8140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9427664150924098,
+      "learning_rate": 1.865834580495559e-05,
+      "loss": 0.8602,
+      "step": 8141
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9042586493017843,
+      "learning_rate": 1.865796400713405e-05,
+      "loss": 0.8523,
+      "step": 8142
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.942094301026198,
+      "learning_rate": 1.86575821589031e-05,
+      "loss": 0.9995,
+      "step": 8143
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6334689834539748,
+      "learning_rate": 1.8657200260264966e-05,
+      "loss": 0.9385,
+      "step": 8144
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9552680316120272,
+      "learning_rate": 1.865681831122187e-05,
+      "loss": 0.8433,
+      "step": 8145
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7534737545402113,
+      "learning_rate": 1.865643631177604e-05,
+      "loss": 0.7927,
+      "step": 8146
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8030101335577393,
+      "learning_rate": 1.86560542619297e-05,
+      "loss": 0.9455,
+      "step": 8147
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7109541474078418,
+      "learning_rate": 1.865567216168507e-05,
+      "loss": 0.9799,
+      "step": 8148
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.817173671756924,
+      "learning_rate": 1.865529001104437e-05,
+      "loss": 0.8757,
+      "step": 8149
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2736648510872453,
+      "learning_rate": 1.8654907810009838e-05,
+      "loss": 0.9932,
+      "step": 8150
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.175912974605035,
+      "learning_rate": 1.865452555858369e-05,
+      "loss": 0.9746,
+      "step": 8151
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.456027763461248,
+      "learning_rate": 1.8654143256768157e-05,
+      "loss": 0.8124,
+      "step": 8152
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1475632149272075,
+      "learning_rate": 1.865376090456546e-05,
+      "loss": 0.8673,
+      "step": 8153
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7435825287772523,
+      "learning_rate": 1.8653378501977827e-05,
+      "loss": 0.7585,
+      "step": 8154
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.56333902839405,
+      "learning_rate": 1.8652996049007485e-05,
+      "loss": 0.942,
+      "step": 8155
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6726265728047207,
+      "learning_rate": 1.865261354565666e-05,
+      "loss": 0.7797,
+      "step": 8156
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2752691436282584,
+      "learning_rate": 1.8652230991927582e-05,
+      "loss": 1.0319,
+      "step": 8157
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9047207574536262,
+      "learning_rate": 1.8651848387822473e-05,
+      "loss": 0.9018,
+      "step": 8158
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.049223793039282,
+      "learning_rate": 1.8651465733343562e-05,
+      "loss": 0.9054,
+      "step": 8159
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.876259239470517,
+      "learning_rate": 1.8651083028493082e-05,
+      "loss": 0.8458,
+      "step": 8160
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4413407369645204,
+      "learning_rate": 1.8650700273273257e-05,
+      "loss": 0.8928,
+      "step": 8161
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.796432651229092,
+      "learning_rate": 1.8650317467686318e-05,
+      "loss": 0.9732,
+      "step": 8162
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1435160881349264,
+      "learning_rate": 1.864993461173449e-05,
+      "loss": 0.8345,
+      "step": 8163
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9852846940952018,
+      "learning_rate": 1.8649551705420004e-05,
+      "loss": 0.9736,
+      "step": 8164
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7195158230355139,
+      "learning_rate": 1.864916874874509e-05,
+      "loss": 1.0207,
+      "step": 8165
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.755891570926106,
+      "learning_rate": 1.8648785741711975e-05,
+      "loss": 0.9019,
+      "step": 8166
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8305274991868592,
+      "learning_rate": 1.8648402684322893e-05,
+      "loss": 0.8891,
+      "step": 8167
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4971975060331364,
+      "learning_rate": 1.8648019576580074e-05,
+      "loss": 0.941,
+      "step": 8168
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.083427943112337,
+      "learning_rate": 1.8647636418485748e-05,
+      "loss": 0.9439,
+      "step": 8169
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4365084317800947,
+      "learning_rate": 1.8647253210042143e-05,
+      "loss": 0.8244,
+      "step": 8170
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.800061746773021,
+      "learning_rate": 1.8646869951251493e-05,
+      "loss": 0.8493,
+      "step": 8171
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5966620121095647,
+      "learning_rate": 1.8646486642116026e-05,
+      "loss": 0.9543,
+      "step": 8172
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.719754985286612,
+      "learning_rate": 1.864610328263798e-05,
+      "loss": 0.9064,
+      "step": 8173
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.785229234623815,
+      "learning_rate": 1.864571987281958e-05,
+      "loss": 0.8419,
+      "step": 8174
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.721120054544898,
+      "learning_rate": 1.8645336412663062e-05,
+      "loss": 0.9026,
+      "step": 8175
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7874917245701756,
+      "learning_rate": 1.864495290217066e-05,
+      "loss": 0.9269,
+      "step": 8176
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8839634556716716,
+      "learning_rate": 1.8644569341344604e-05,
+      "loss": 0.9373,
+      "step": 8177
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.225732371617913,
+      "learning_rate": 1.864418573018713e-05,
+      "loss": 0.8669,
+      "step": 8178
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6276408798805782,
+      "learning_rate": 1.864380206870047e-05,
+      "loss": 0.8426,
+      "step": 8179
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.006851674964322,
+      "learning_rate": 1.8643418356886854e-05,
+      "loss": 0.8889,
+      "step": 8180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5493551993713053,
+      "learning_rate": 1.8643034594748525e-05,
+      "loss": 0.9809,
+      "step": 8181
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7904382939030141,
+      "learning_rate": 1.864265078228771e-05,
+      "loss": 0.8882,
+      "step": 8182
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.53638861484836,
+      "learning_rate": 1.8642266919506644e-05,
+      "loss": 0.8623,
+      "step": 8183
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.011011717300178,
+      "learning_rate": 1.864188300640757e-05,
+      "loss": 0.8293,
+      "step": 8184
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.841334297113441,
+      "learning_rate": 1.864149904299271e-05,
+      "loss": 0.9013,
+      "step": 8185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8515470427157639,
+      "learning_rate": 1.8641115029264312e-05,
+      "loss": 0.9734,
+      "step": 8186
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5543649261839458,
+      "learning_rate": 1.8640730965224602e-05,
+      "loss": 0.9521,
+      "step": 8187
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5628536043297034,
+      "learning_rate": 1.8640346850875825e-05,
+      "loss": 0.8842,
+      "step": 8188
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7151199411642035,
+      "learning_rate": 1.863996268622021e-05,
+      "loss": 0.9058,
+      "step": 8189
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.75546093356854,
+      "learning_rate": 1.8639578471259997e-05,
+      "loss": 0.9043,
+      "step": 8190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.53548361542771,
+      "learning_rate": 1.8639194205997423e-05,
+      "loss": 0.9378,
+      "step": 8191
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6852745270416627,
+      "learning_rate": 1.863880989043473e-05,
+      "loss": 0.8741,
+      "step": 8192
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9368460601563235,
+      "learning_rate": 1.8638425524574148e-05,
+      "loss": 0.9485,
+      "step": 8193
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.818836816845028,
+      "learning_rate": 1.8638041108417916e-05,
+      "loss": 0.9273,
+      "step": 8194
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5832194120185015,
+      "learning_rate": 1.8637656641968272e-05,
+      "loss": 0.9154,
+      "step": 8195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8289505439202,
+      "learning_rate": 1.863727212522746e-05,
+      "loss": 0.9075,
+      "step": 8196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4860654484780595,
+      "learning_rate": 1.8636887558197713e-05,
+      "loss": 0.8277,
+      "step": 8197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8992605324276464,
+      "learning_rate": 1.8636502940881276e-05,
+      "loss": 1.0068,
+      "step": 8198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5601770793666776,
+      "learning_rate": 1.863611827328038e-05,
+      "loss": 1.0051,
+      "step": 8199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7148372048815732,
+      "learning_rate": 1.8635733555397273e-05,
+      "loss": 0.9816,
+      "step": 8200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.027894581563827,
+      "learning_rate": 1.863534878723419e-05,
+      "loss": 0.9697,
+      "step": 8201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.725817607041178,
+      "learning_rate": 1.863496396879337e-05,
+      "loss": 0.944,
+      "step": 8202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0217344101206183,
+      "learning_rate": 1.863457910007706e-05,
+      "loss": 0.903,
+      "step": 8203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6006579459644261,
+      "learning_rate": 1.8634194181087494e-05,
+      "loss": 0.9689,
+      "step": 8204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9744232807067814,
+      "learning_rate": 1.8633809211826917e-05,
+      "loss": 0.8685,
+      "step": 8205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7623630867405593,
+      "learning_rate": 1.8633424192297566e-05,
+      "loss": 0.9766,
+      "step": 8206
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6283917806355448,
+      "learning_rate": 1.8633039122501692e-05,
+      "loss": 0.9258,
+      "step": 8207
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8247051805395836,
+      "learning_rate": 1.8632654002441528e-05,
+      "loss": 0.7169,
+      "step": 8208
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8186151192656335,
+      "learning_rate": 1.863226883211932e-05,
+      "loss": 0.9981,
+      "step": 8209
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5563146547609614,
+      "learning_rate": 1.863188361153731e-05,
+      "loss": 0.8174,
+      "step": 8210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8434256422347567,
+      "learning_rate": 1.8631498340697736e-05,
+      "loss": 0.7432,
+      "step": 8211
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0509272387176822,
+      "learning_rate": 1.8631113019602854e-05,
+      "loss": 0.97,
+      "step": 8212
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5364135781277668,
+      "learning_rate": 1.8630727648254896e-05,
+      "loss": 0.8829,
+      "step": 8213
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.907071240845834,
+      "learning_rate": 1.8630342226656108e-05,
+      "loss": 0.8278,
+      "step": 8214
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7806082667022328,
+      "learning_rate": 1.8629956754808734e-05,
+      "loss": 0.7072,
+      "step": 8215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8322073001268893,
+      "learning_rate": 1.8629571232715024e-05,
+      "loss": 1.0042,
+      "step": 8216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5381799666494658,
+      "learning_rate": 1.8629185660377216e-05,
+      "loss": 0.9506,
+      "step": 8217
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6949938518800163,
+      "learning_rate": 1.8628800037797558e-05,
+      "loss": 0.9208,
+      "step": 8218
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6579022486471948,
+      "learning_rate": 1.8628414364978292e-05,
+      "loss": 1.0077,
+      "step": 8219
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4858051625305728,
+      "learning_rate": 1.8628028641921668e-05,
+      "loss": 0.8403,
+      "step": 8220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.61278700351045,
+      "learning_rate": 1.862764286862993e-05,
+      "loss": 0.9268,
+      "step": 8221
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8308307083788486,
+      "learning_rate": 1.862725704510532e-05,
+      "loss": 0.87,
+      "step": 8222
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9074202774325038,
+      "learning_rate": 1.8626871171350093e-05,
+      "loss": 0.7886,
+      "step": 8223
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5799415394000493,
+      "learning_rate": 1.8626485247366486e-05,
+      "loss": 1.0994,
+      "step": 8224
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7760661187175741,
+      "learning_rate": 1.8626099273156753e-05,
+      "loss": 0.9728,
+      "step": 8225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.576813109485943,
+      "learning_rate": 1.862571324872314e-05,
+      "loss": 0.8266,
+      "step": 8226
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7464165424038536,
+      "learning_rate": 1.8625327174067896e-05,
+      "loss": 0.7128,
+      "step": 8227
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.556518799500786,
+      "learning_rate": 1.8624941049193262e-05,
+      "loss": 0.9758,
+      "step": 8228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8272325985944162,
+      "learning_rate": 1.862455487410149e-05,
+      "loss": 1.0443,
+      "step": 8229
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9290018055184104,
+      "learning_rate": 1.8624168648794833e-05,
+      "loss": 0.8609,
+      "step": 8230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7893089841619627,
+      "learning_rate": 1.8623782373275535e-05,
+      "loss": 0.8256,
+      "step": 8231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5011952100727355,
+      "learning_rate": 1.8623396047545844e-05,
+      "loss": 0.8607,
+      "step": 8232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.208134897623774,
+      "learning_rate": 1.862300967160801e-05,
+      "loss": 0.7574,
+      "step": 8233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.92837126523383,
+      "learning_rate": 1.8622623245464286e-05,
+      "loss": 0.9337,
+      "step": 8234
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6201452382664658,
+      "learning_rate": 1.862223676911692e-05,
+      "loss": 0.9013,
+      "step": 8235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6486888954819807,
+      "learning_rate": 1.862185024256816e-05,
+      "loss": 0.9534,
+      "step": 8236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1378639104821766,
+      "learning_rate": 1.862146366582026e-05,
+      "loss": 0.8317,
+      "step": 8237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7161945702958166,
+      "learning_rate": 1.862107703887547e-05,
+      "loss": 0.9076,
+      "step": 8238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.960128460227797,
+      "learning_rate": 1.8620690361736037e-05,
+      "loss": 1.0144,
+      "step": 8239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7705210946203465,
+      "learning_rate": 1.8620303634404218e-05,
+      "loss": 0.9016,
+      "step": 8240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6949924909313443,
+      "learning_rate": 1.8619916856882258e-05,
+      "loss": 0.8346,
+      "step": 8241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8535410450918453,
+      "learning_rate": 1.8619530029172417e-05,
+      "loss": 0.9308,
+      "step": 8242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8173362543384566,
+      "learning_rate": 1.861914315127694e-05,
+      "loss": 0.9134,
+      "step": 8243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7280386587427592,
+      "learning_rate": 1.8618756223198086e-05,
+      "loss": 0.9296,
+      "step": 8244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5321356348898503,
+      "learning_rate": 1.8618369244938103e-05,
+      "loss": 0.7648,
+      "step": 8245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6993735611311818,
+      "learning_rate": 1.8617982216499247e-05,
+      "loss": 0.9368,
+      "step": 8246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.371537392657399,
+      "learning_rate": 1.861759513788377e-05,
+      "loss": 0.9471,
+      "step": 8247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.619414781106504,
+      "learning_rate": 1.8617208009093922e-05,
+      "loss": 1.0161,
+      "step": 8248
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8838407166786775,
+      "learning_rate": 1.8616820830131966e-05,
+      "loss": 0.9416,
+      "step": 8249
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6954318187377364,
+      "learning_rate": 1.861643360100015e-05,
+      "loss": 0.7961,
+      "step": 8250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6930191506802157,
+      "learning_rate": 1.861604632170073e-05,
+      "loss": 1.069,
+      "step": 8251
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8699093098367194,
+      "learning_rate": 1.861565899223596e-05,
+      "loss": 1.0058,
+      "step": 8252
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5762630946656482,
+      "learning_rate": 1.8615271612608094e-05,
+      "loss": 0.8143,
+      "step": 8253
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7990785483632824,
+      "learning_rate": 1.861488418281939e-05,
+      "loss": 0.7251,
+      "step": 8254
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5354561519043206,
+      "learning_rate": 1.8614496702872102e-05,
+      "loss": 0.8513,
+      "step": 8255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8325369023792859,
+      "learning_rate": 1.8614109172768487e-05,
+      "loss": 0.7286,
+      "step": 8256
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8286617452755695,
+      "learning_rate": 1.8613721592510804e-05,
+      "loss": 0.753,
+      "step": 8257
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.84115178719008,
+      "learning_rate": 1.8613333962101303e-05,
+      "loss": 0.8545,
+      "step": 8258
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5589438594326823,
+      "learning_rate": 1.8612946281542247e-05,
+      "loss": 0.8695,
+      "step": 8259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.777284700760753,
+      "learning_rate": 1.861255855083589e-05,
+      "loss": 0.9731,
+      "step": 8260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6375224715191994,
+      "learning_rate": 1.861217076998449e-05,
+      "loss": 0.9825,
+      "step": 8261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1740241360517594,
+      "learning_rate": 1.8611782938990308e-05,
+      "loss": 0.9409,
+      "step": 8262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5046005534336042,
+      "learning_rate": 1.86113950578556e-05,
+      "loss": 1.0002,
+      "step": 8263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5083369314101096,
+      "learning_rate": 1.8611007126582617e-05,
+      "loss": 0.9532,
+      "step": 8264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.148685477937495,
+      "learning_rate": 1.861061914517363e-05,
+      "loss": 0.8678,
+      "step": 8265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5242147846942669,
+      "learning_rate": 1.861023111363089e-05,
+      "loss": 0.8994,
+      "step": 8266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4884916972249151,
+      "learning_rate": 1.860984303195666e-05,
+      "loss": 1.0107,
+      "step": 8267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.677324827017625,
+      "learning_rate": 1.8609454900153195e-05,
+      "loss": 1.0088,
+      "step": 8268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8533075841171274,
+      "learning_rate": 1.860906671822276e-05,
+      "loss": 0.9591,
+      "step": 8269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.155779678573945,
+      "learning_rate": 1.8608678486167615e-05,
+      "loss": 1.0213,
+      "step": 8270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0787381497622515,
+      "learning_rate": 1.8608290203990015e-05,
+      "loss": 0.8664,
+      "step": 8271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0003079788362497,
+      "learning_rate": 1.8607901871692226e-05,
+      "loss": 0.9372,
+      "step": 8272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8259542190633202,
+      "learning_rate": 1.8607513489276507e-05,
+      "loss": 0.9703,
+      "step": 8273
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5743124792111425,
+      "learning_rate": 1.860712505674512e-05,
+      "loss": 0.8822,
+      "step": 8274
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1071219599319777,
+      "learning_rate": 1.8606736574100325e-05,
+      "loss": 0.9019,
+      "step": 8275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.727670425940431,
+      "learning_rate": 1.8606348041344383e-05,
+      "loss": 0.8659,
+      "step": 8276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5833325286316193,
+      "learning_rate": 1.860595945847956e-05,
+      "loss": 0.9356,
+      "step": 8277
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5153333726754261,
+      "learning_rate": 1.8605570825508117e-05,
+      "loss": 0.9782,
+      "step": 8278
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9117391960126489,
+      "learning_rate": 1.8605182142432317e-05,
+      "loss": 0.9248,
+      "step": 8279
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9550901966301069,
+      "learning_rate": 1.860479340925442e-05,
+      "loss": 0.9097,
+      "step": 8280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6144159215643532,
+      "learning_rate": 1.860440462597669e-05,
+      "loss": 0.9531,
+      "step": 8281
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4678915483734596,
+      "learning_rate": 1.8604015792601395e-05,
+      "loss": 0.9384,
+      "step": 8282
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6931970595965298,
+      "learning_rate": 1.8603626909130796e-05,
+      "loss": 0.841,
+      "step": 8283
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4462243598888267,
+      "learning_rate": 1.8603237975567157e-05,
+      "loss": 0.9702,
+      "step": 8284
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8186546799117194,
+      "learning_rate": 1.860284899191274e-05,
+      "loss": 0.7818,
+      "step": 8285
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8422185144135956,
+      "learning_rate": 1.8602459958169817e-05,
+      "loss": 0.9963,
+      "step": 8286
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7995759228184165,
+      "learning_rate": 1.8602070874340646e-05,
+      "loss": 0.8683,
+      "step": 8287
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6108240504833393,
+      "learning_rate": 1.8601681740427495e-05,
+      "loss": 0.9731,
+      "step": 8288
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8726706727106188,
+      "learning_rate": 1.8601292556432632e-05,
+      "loss": 0.933,
+      "step": 8289
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.736511706953884,
+      "learning_rate": 1.860090332235832e-05,
+      "loss": 0.8627,
+      "step": 8290
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.744105666024887,
+      "learning_rate": 1.860051403820682e-05,
+      "loss": 0.8506,
+      "step": 8291
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7908680909047029,
+      "learning_rate": 1.860012470398041e-05,
+      "loss": 0.7836,
+      "step": 8292
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.70778517038094,
+      "learning_rate": 1.859973531968135e-05,
+      "loss": 0.9668,
+      "step": 8293
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.805449777196085,
+      "learning_rate": 1.859934588531191e-05,
+      "loss": 0.7137,
+      "step": 8294
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7105491162121218,
+      "learning_rate": 1.8598956400874354e-05,
+      "loss": 0.8785,
+      "step": 8295
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8116736420081052,
+      "learning_rate": 1.8598566866370952e-05,
+      "loss": 0.7219,
+      "step": 8296
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7763759242172007,
+      "learning_rate": 1.859817728180397e-05,
+      "loss": 0.7358,
+      "step": 8297
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7477783648436622,
+      "learning_rate": 1.859778764717568e-05,
+      "loss": 0.7507,
+      "step": 8298
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9920859426378121,
+      "learning_rate": 1.8597397962488347e-05,
+      "loss": 0.8666,
+      "step": 8299
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7372233678409976,
+      "learning_rate": 1.8597008227744242e-05,
+      "loss": 0.8073,
+      "step": 8300
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6796605987009487,
+      "learning_rate": 1.8596618442945634e-05,
+      "loss": 0.9104,
+      "step": 8301
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.146194464191477,
+      "learning_rate": 1.8596228608094792e-05,
+      "loss": 0.8908,
+      "step": 8302
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5649639860746989,
+      "learning_rate": 1.8595838723193986e-05,
+      "loss": 0.8628,
+      "step": 8303
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.78162309064383,
+      "learning_rate": 1.859544878824548e-05,
+      "loss": 0.8894,
+      "step": 8304
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2114537161741796,
+      "learning_rate": 1.859505880325155e-05,
+      "loss": 0.9223,
+      "step": 8305
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9968665335987408,
+      "learning_rate": 1.8594668768214474e-05,
+      "loss": 1.024,
+      "step": 8306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7482220267775586,
+      "learning_rate": 1.859427868313651e-05,
+      "loss": 0.8497,
+      "step": 8307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9881266946528298,
+      "learning_rate": 1.8593888548019938e-05,
+      "loss": 0.8294,
+      "step": 8308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.503812454870375,
+      "learning_rate": 1.8593498362867023e-05,
+      "loss": 1.0968,
+      "step": 8309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8749100491901837,
+      "learning_rate": 1.859310812768004e-05,
+      "loss": 0.7699,
+      "step": 8310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5750327513074984,
+      "learning_rate": 1.8592717842461262e-05,
+      "loss": 0.8927,
+      "step": 8311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5906024162560186,
+      "learning_rate": 1.859232750721296e-05,
+      "loss": 0.9455,
+      "step": 8312
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.50131905332512,
+      "learning_rate": 1.8591937121937408e-05,
+      "loss": 0.9808,
+      "step": 8313
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9935058647498856,
+      "learning_rate": 1.859154668663688e-05,
+      "loss": 0.9164,
+      "step": 8314
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.673856322175922,
+      "learning_rate": 1.8591156201313642e-05,
+      "loss": 1.0399,
+      "step": 8315
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9256993250735326,
+      "learning_rate": 1.8590765665969974e-05,
+      "loss": 1.0039,
+      "step": 8316
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.798534994286735,
+      "learning_rate": 1.859037508060815e-05,
+      "loss": 0.7141,
+      "step": 8317
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7423871009579103,
+      "learning_rate": 1.8589984445230444e-05,
+      "loss": 0.7744,
+      "step": 8318
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.083784261930011,
+      "learning_rate": 1.8589593759839125e-05,
+      "loss": 1.0172,
+      "step": 8319
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9246754818703793,
+      "learning_rate": 1.8589203024436477e-05,
+      "loss": 0.9021,
+      "step": 8320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5777120512084635,
+      "learning_rate": 1.8588812239024766e-05,
+      "loss": 0.9215,
+      "step": 8321
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0492013373391393,
+      "learning_rate": 1.858842140360627e-05,
+      "loss": 0.9433,
+      "step": 8322
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.234432932593048,
+      "learning_rate": 1.8588030518183265e-05,
+      "loss": 0.8803,
+      "step": 8323
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7743691440745375,
+      "learning_rate": 1.8587639582758032e-05,
+      "loss": 0.9423,
+      "step": 8324
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6386309494176798,
+      "learning_rate": 1.8587248597332837e-05,
+      "loss": 0.9547,
+      "step": 8325
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7139458761871338,
+      "learning_rate": 1.8586857561909967e-05,
+      "loss": 0.8705,
+      "step": 8326
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.653114963946628,
+      "learning_rate": 1.8586466476491692e-05,
+      "loss": 0.8412,
+      "step": 8327
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.758289771123672,
+      "learning_rate": 1.8586075341080293e-05,
+      "loss": 1.0165,
+      "step": 8328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.967518027622778,
+      "learning_rate": 1.858568415567804e-05,
+      "loss": 0.9668,
+      "step": 8329
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7838140573292988,
+      "learning_rate": 1.8585292920287217e-05,
+      "loss": 0.747,
+      "step": 8330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7664572989540819,
+      "learning_rate": 1.8584901634910105e-05,
+      "loss": 0.8319,
+      "step": 8331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.51823488799677,
+      "learning_rate": 1.8584510299548975e-05,
+      "loss": 0.9238,
+      "step": 8332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.275295745524475,
+      "learning_rate": 1.8584118914206105e-05,
+      "loss": 0.9385,
+      "step": 8333
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8056669868972474,
+      "learning_rate": 1.8583727478883782e-05,
+      "loss": 0.7049,
+      "step": 8334
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6262070508032211,
+      "learning_rate": 1.8583335993584275e-05,
+      "loss": 0.8183,
+      "step": 8335
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1120376956217446,
+      "learning_rate": 1.858294445830987e-05,
+      "loss": 1.0273,
+      "step": 8336
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.783704992901075,
+      "learning_rate": 1.858255287306285e-05,
+      "loss": 0.7897,
+      "step": 8337
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.625863270295888,
+      "learning_rate": 1.8582161237845486e-05,
+      "loss": 0.9615,
+      "step": 8338
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4639714842024873,
+      "learning_rate": 1.8581769552660064e-05,
+      "loss": 0.8375,
+      "step": 8339
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6952663845707046,
+      "learning_rate": 1.858137781750886e-05,
+      "loss": 0.9856,
+      "step": 8340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7104711191565056,
+      "learning_rate": 1.8580986032394157e-05,
+      "loss": 0.9658,
+      "step": 8341
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.88920525729605,
+      "learning_rate": 1.858059419731824e-05,
+      "loss": 0.7854,
+      "step": 8342
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.633700563612029,
+      "learning_rate": 1.8580202312283383e-05,
+      "loss": 0.9171,
+      "step": 8343
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5944383409887626,
+      "learning_rate": 1.8579810377291875e-05,
+      "loss": 0.9117,
+      "step": 8344
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8213840019311298,
+      "learning_rate": 1.8579418392345993e-05,
+      "loss": 0.9459,
+      "step": 8345
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0345100687318602,
+      "learning_rate": 1.8579026357448023e-05,
+      "loss": 0.9737,
+      "step": 8346
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9878584686366665,
+      "learning_rate": 1.8578634272600246e-05,
+      "loss": 0.9715,
+      "step": 8347
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9356508367826533,
+      "learning_rate": 1.8578242137804938e-05,
+      "loss": 1.0511,
+      "step": 8348
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5844954906340605,
+      "learning_rate": 1.8577849953064394e-05,
+      "loss": 0.9095,
+      "step": 8349
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.835495730809335,
+      "learning_rate": 1.8577457718380893e-05,
+      "loss": 0.8347,
+      "step": 8350
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8413737639770489,
+      "learning_rate": 1.8577065433756712e-05,
+      "loss": 1.0592,
+      "step": 8351
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.538645363751425,
+      "learning_rate": 1.8576673099194145e-05,
+      "loss": 1.0688,
+      "step": 8352
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6678316104763482,
+      "learning_rate": 1.857628071469547e-05,
+      "loss": 0.9679,
+      "step": 8353
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7414193048764373,
+      "learning_rate": 1.8575888280262975e-05,
+      "loss": 0.978,
+      "step": 8354
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8199661305006298,
+      "learning_rate": 1.8575495795898942e-05,
+      "loss": 0.9297,
+      "step": 8355
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8015307767355595,
+      "learning_rate": 1.8575103261605657e-05,
+      "loss": 1.0654,
+      "step": 8356
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5907290073824536,
+      "learning_rate": 1.8574710677385412e-05,
+      "loss": 0.8966,
+      "step": 8357
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6601450327886962,
+      "learning_rate": 1.857431804324048e-05,
+      "loss": 1.0142,
+      "step": 8358
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.544565126441637,
+      "learning_rate": 1.8573925359173154e-05,
+      "loss": 0.8613,
+      "step": 8359
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.913399052347165,
+      "learning_rate": 1.8573532625185723e-05,
+      "loss": 0.9286,
+      "step": 8360
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5651609966905062,
+      "learning_rate": 1.857313984128047e-05,
+      "loss": 1.0476,
+      "step": 8361
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8841109979306676,
+      "learning_rate": 1.8572747007459678e-05,
+      "loss": 0.913,
+      "step": 8362
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.640131670713251,
+      "learning_rate": 1.8572354123725645e-05,
+      "loss": 1.0672,
+      "step": 8363
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6960831938884686,
+      "learning_rate": 1.857196119008065e-05,
+      "loss": 0.7699,
+      "step": 8364
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7225241636112503,
+      "learning_rate": 1.857156820652698e-05,
+      "loss": 1.0271,
+      "step": 8365
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.589152709179957,
+      "learning_rate": 1.8571175173066935e-05,
+      "loss": 0.8828,
+      "step": 8366
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4936687407438447,
+      "learning_rate": 1.8570782089702787e-05,
+      "loss": 0.8859,
+      "step": 8367
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7916638633644149,
+      "learning_rate": 1.8570388956436836e-05,
+      "loss": 0.748,
+      "step": 8368
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5501205847296884,
+      "learning_rate": 1.8569995773271366e-05,
+      "loss": 1.042,
+      "step": 8369
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5639229584116128,
+      "learning_rate": 1.8569602540208664e-05,
+      "loss": 0.9936,
+      "step": 8370
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9972925589024402,
+      "learning_rate": 1.8569209257251028e-05,
+      "loss": 0.9785,
+      "step": 8371
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5794790817418745,
+      "learning_rate": 1.8568815924400742e-05,
+      "loss": 0.8834,
+      "step": 8372
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.534373771925316,
+      "learning_rate": 1.856842254166009e-05,
+      "loss": 0.8171,
+      "step": 8373
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8139828112926615,
+      "learning_rate": 1.8568029109031377e-05,
+      "loss": 0.9236,
+      "step": 8374
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5006962110588866,
+      "learning_rate": 1.8567635626516885e-05,
+      "loss": 0.8969,
+      "step": 8375
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.579789389762876,
+      "learning_rate": 1.8567242094118902e-05,
+      "loss": 1.0552,
+      "step": 8376
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7192634466513088,
+      "learning_rate": 1.8566848511839724e-05,
+      "loss": 0.9281,
+      "step": 8377
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0344134061478423,
+      "learning_rate": 1.8566454879681646e-05,
+      "loss": 0.8645,
+      "step": 8378
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7962316510516785,
+      "learning_rate": 1.856606119764695e-05,
+      "loss": 0.9078,
+      "step": 8379
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5655925261336898,
+      "learning_rate": 1.856566746573794e-05,
+      "loss": 0.9192,
+      "step": 8380
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7027021427441884,
+      "learning_rate": 1.85652736839569e-05,
+      "loss": 0.9762,
+      "step": 8381
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6557905870966525,
+      "learning_rate": 1.8564879852306122e-05,
+      "loss": 1.1268,
+      "step": 8382
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0378872526539245,
+      "learning_rate": 1.8564485970787904e-05,
+      "loss": 0.8362,
+      "step": 8383
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8056631560353005,
+      "learning_rate": 1.8564092039404537e-05,
+      "loss": 0.9258,
+      "step": 8384
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8280516192841028,
+      "learning_rate": 1.8563698058158313e-05,
+      "loss": 0.7513,
+      "step": 8385
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.895624398685948,
+      "learning_rate": 1.8563304027051532e-05,
+      "loss": 0.8745,
+      "step": 8386
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1760384634699093,
+      "learning_rate": 1.856290994608648e-05,
+      "loss": 0.913,
+      "step": 8387
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8541095594658417,
+      "learning_rate": 1.8562515815265456e-05,
+      "loss": 0.9501,
+      "step": 8388
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.0178595961596555,
+      "learning_rate": 1.8562121634590756e-05,
+      "loss": 0.8655,
+      "step": 8389
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.592044755366472,
+      "learning_rate": 1.856172740406467e-05,
+      "loss": 0.91,
+      "step": 8390
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5828457274615295,
+      "learning_rate": 1.85613331236895e-05,
+      "loss": 0.9175,
+      "step": 8391
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0105406903288188,
+      "learning_rate": 1.8560938793467537e-05,
+      "loss": 0.8056,
+      "step": 8392
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.61460194529713,
+      "learning_rate": 1.8560544413401077e-05,
+      "loss": 0.9494,
+      "step": 8393
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8292561919726193,
+      "learning_rate": 1.8560149983492418e-05,
+      "loss": 0.9263,
+      "step": 8394
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5982463984420565,
+      "learning_rate": 1.855975550374385e-05,
+      "loss": 0.9514,
+      "step": 8395
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.784423776868343,
+      "learning_rate": 1.8559360974157683e-05,
+      "loss": 0.7738,
+      "step": 8396
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8363947757328845,
+      "learning_rate": 1.8558966394736205e-05,
+      "loss": 0.9159,
+      "step": 8397
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5322106906392172,
+      "learning_rate": 1.8558571765481712e-05,
+      "loss": 0.8402,
+      "step": 8398
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.45850875232878,
+      "learning_rate": 1.855817708639651e-05,
+      "loss": 0.9138,
+      "step": 8399
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6957112692555754,
+      "learning_rate": 1.8557782357482884e-05,
+      "loss": 0.9137,
+      "step": 8400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0208336690446895,
+      "learning_rate": 1.8557387578743146e-05,
+      "loss": 0.991,
+      "step": 8401
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5811605213005768,
+      "learning_rate": 1.855699275017959e-05,
+      "loss": 0.8596,
+      "step": 8402
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7599224367752353,
+      "learning_rate": 1.855659787179451e-05,
+      "loss": 0.7555,
+      "step": 8403
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.782027694045014,
+      "learning_rate": 1.8556202943590206e-05,
+      "loss": 0.9352,
+      "step": 8404
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7457687402276116,
+      "learning_rate": 1.855580796556898e-05,
+      "loss": 0.9,
+      "step": 8405
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7872076436811877,
+      "learning_rate": 1.855541293773313e-05,
+      "loss": 0.8745,
+      "step": 8406
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8683373335951199,
+      "learning_rate": 1.855501786008496e-05,
+      "loss": 0.9814,
+      "step": 8407
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9897386590842148,
+      "learning_rate": 1.855462273262677e-05,
+      "loss": 0.8325,
+      "step": 8408
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5931314578715092,
+      "learning_rate": 1.855422755536085e-05,
+      "loss": 0.8624,
+      "step": 8409
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7664340135423087,
+      "learning_rate": 1.855383232828952e-05,
+      "loss": 0.9086,
+      "step": 8410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7256610919401891,
+      "learning_rate": 1.855343705141506e-05,
+      "loss": 0.9227,
+      "step": 8411
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6067534831906802,
+      "learning_rate": 1.855304172473979e-05,
+      "loss": 0.982,
+      "step": 8412
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6478046856123516,
+      "learning_rate": 1.8552646348266e-05,
+      "loss": 0.8925,
+      "step": 8413
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.843784715069425,
+      "learning_rate": 1.8552250921995995e-05,
+      "loss": 0.8535,
+      "step": 8414
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7045127144142278,
+      "learning_rate": 1.8551855445932077e-05,
+      "loss": 1.0073,
+      "step": 8415
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.971269400153061,
+      "learning_rate": 1.8551459920076547e-05,
+      "loss": 0.9092,
+      "step": 8416
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6704429121744846,
+      "learning_rate": 1.8551064344431712e-05,
+      "loss": 0.8314,
+      "step": 8417
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6431967088107575,
+      "learning_rate": 1.8550668718999873e-05,
+      "loss": 0.9237,
+      "step": 8418
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5418629829796482,
+      "learning_rate": 1.8550273043783337e-05,
+      "loss": 0.9368,
+      "step": 8419
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6148162558864643,
+      "learning_rate": 1.85498773187844e-05,
+      "loss": 0.9432,
+      "step": 8420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5962058803630366,
+      "learning_rate": 1.8549481544005373e-05,
+      "loss": 0.9198,
+      "step": 8421
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7639835440767384,
+      "learning_rate": 1.8549085719448553e-05,
+      "loss": 0.932,
+      "step": 8422
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7906371370482792,
+      "learning_rate": 1.8548689845116255e-05,
+      "loss": 0.8359,
+      "step": 8423
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7748712933520371,
+      "learning_rate": 1.8548293921010776e-05,
+      "loss": 0.7454,
+      "step": 8424
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6506925074896575,
+      "learning_rate": 1.854789794713442e-05,
+      "loss": 0.9387,
+      "step": 8425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7443779097342924,
+      "learning_rate": 1.8547501923489503e-05,
+      "loss": 0.952,
+      "step": 8426
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8198907169925604,
+      "learning_rate": 1.854710585007832e-05,
+      "loss": 0.9344,
+      "step": 8427
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.334942680720376,
+      "learning_rate": 1.8546709726903178e-05,
+      "loss": 0.9111,
+      "step": 8428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6594930159156565,
+      "learning_rate": 1.854631355396639e-05,
+      "loss": 1.0477,
+      "step": 8429
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7498350899467813,
+      "learning_rate": 1.8545917331270257e-05,
+      "loss": 0.7241,
+      "step": 8430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9589164308258844,
+      "learning_rate": 1.8545521058817088e-05,
+      "loss": 0.8901,
+      "step": 8431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5627426534413809,
+      "learning_rate": 1.854512473660919e-05,
+      "loss": 0.754,
+      "step": 8432
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7411765227343166,
+      "learning_rate": 1.8544728364648875e-05,
+      "loss": 0.7384,
+      "step": 8433
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6274039043923525,
+      "learning_rate": 1.8544331942938442e-05,
+      "loss": 0.9596,
+      "step": 8434
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6679555083638133,
+      "learning_rate": 1.85439354714802e-05,
+      "loss": 0.9343,
+      "step": 8435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.93857796033571,
+      "learning_rate": 1.8543538950276465e-05,
+      "loss": 0.8906,
+      "step": 8436
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.728450873005321,
+      "learning_rate": 1.8543142379329543e-05,
+      "loss": 1.0122,
+      "step": 8437
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.387089910419164,
+      "learning_rate": 1.8542745758641736e-05,
+      "loss": 0.9803,
+      "step": 8438
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7183829141777052,
+      "learning_rate": 1.8542349088215362e-05,
+      "loss": 1.0553,
+      "step": 8439
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.471135353528829,
+      "learning_rate": 1.8541952368052727e-05,
+      "loss": 0.8407,
+      "step": 8440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.568397885929503,
+      "learning_rate": 1.8541555598156143e-05,
+      "loss": 0.9665,
+      "step": 8441
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7643781520238708,
+      "learning_rate": 1.8541158778527914e-05,
+      "loss": 0.9463,
+      "step": 8442
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5243805357596016,
+      "learning_rate": 1.8540761909170356e-05,
+      "loss": 0.8815,
+      "step": 8443
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6795090024316315,
+      "learning_rate": 1.8540364990085778e-05,
+      "loss": 0.8517,
+      "step": 8444
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8911941498502338,
+      "learning_rate": 1.8539968021276493e-05,
+      "loss": 0.8527,
+      "step": 8445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0320845318985667,
+      "learning_rate": 1.853957100274481e-05,
+      "loss": 0.8852,
+      "step": 8446
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5661792067369813,
+      "learning_rate": 1.853917393449304e-05,
+      "loss": 0.9229,
+      "step": 8447
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8324467149396506,
+      "learning_rate": 1.8538776816523495e-05,
+      "loss": 0.9877,
+      "step": 8448
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4250214605585598,
+      "learning_rate": 1.853837964883849e-05,
+      "loss": 0.8194,
+      "step": 8449
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.132422570301334,
+      "learning_rate": 1.8537982431440333e-05,
+      "loss": 0.9512,
+      "step": 8450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.572885977824651,
+      "learning_rate": 1.8537585164331343e-05,
+      "loss": 0.8559,
+      "step": 8451
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0769295364685676,
+      "learning_rate": 1.8537187847513825e-05,
+      "loss": 0.9518,
+      "step": 8452
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7612034385054238,
+      "learning_rate": 1.85367904809901e-05,
+      "loss": 0.8864,
+      "step": 8453
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9430742458377654,
+      "learning_rate": 1.8536393064762474e-05,
+      "loss": 0.987,
+      "step": 8454
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6307302497381633,
+      "learning_rate": 1.853599559883327e-05,
+      "loss": 0.9679,
+      "step": 8455
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.764626494129648,
+      "learning_rate": 1.8535598083204792e-05,
+      "loss": 0.9237,
+      "step": 8456
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0303797692011303,
+      "learning_rate": 1.8535200517879364e-05,
+      "loss": 0.9227,
+      "step": 8457
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.868553877740166,
+      "learning_rate": 1.8534802902859296e-05,
+      "loss": 0.901,
+      "step": 8458
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6851421224855956,
+      "learning_rate": 1.85344052381469e-05,
+      "loss": 0.9605,
+      "step": 8459
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.575072277611008,
+      "learning_rate": 1.8534007523744498e-05,
+      "loss": 0.8138,
+      "step": 8460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5613455354770787,
+      "learning_rate": 1.85336097596544e-05,
+      "loss": 0.9601,
+      "step": 8461
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.105659097896622,
+      "learning_rate": 1.8533211945878928e-05,
+      "loss": 0.8703,
+      "step": 8462
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6486047009282803,
+      "learning_rate": 1.8532814082420393e-05,
+      "loss": 0.7567,
+      "step": 8463
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.996049931400639,
+      "learning_rate": 1.8532416169281113e-05,
+      "loss": 0.927,
+      "step": 8464
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.706697237340503,
+      "learning_rate": 1.8532018206463404e-05,
+      "loss": 0.9667,
+      "step": 8465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.918629619939082,
+      "learning_rate": 1.853162019396958e-05,
+      "loss": 0.8675,
+      "step": 8466
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8360610239022143,
+      "learning_rate": 1.8531222131801965e-05,
+      "loss": 0.705,
+      "step": 8467
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8886582591111896,
+      "learning_rate": 1.8530824019962877e-05,
+      "loss": 0.8795,
+      "step": 8468
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3565013531679844,
+      "learning_rate": 1.8530425858454627e-05,
+      "loss": 0.8995,
+      "step": 8469
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.794714397692399,
+      "learning_rate": 1.853002764727954e-05,
+      "loss": 0.8128,
+      "step": 8470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4004503492197014,
+      "learning_rate": 1.8529629386439933e-05,
+      "loss": 0.9413,
+      "step": 8471
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.240404289492165,
+      "learning_rate": 1.8529231075938117e-05,
+      "loss": 0.866,
+      "step": 8472
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9363687666780967,
+      "learning_rate": 1.8528832715776422e-05,
+      "loss": 0.9123,
+      "step": 8473
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.817198883027449,
+      "learning_rate": 1.852843430595716e-05,
+      "loss": 0.7548,
+      "step": 8474
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8609026508459865,
+      "learning_rate": 1.8528035846482656e-05,
+      "loss": 0.8862,
+      "step": 8475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8053971331015473,
+      "learning_rate": 1.8527637337355225e-05,
+      "loss": 0.7388,
+      "step": 8476
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9753474896812986,
+      "learning_rate": 1.852723877857719e-05,
+      "loss": 1.0256,
+      "step": 8477
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7174424021022057,
+      "learning_rate": 1.8526840170150874e-05,
+      "loss": 0.9432,
+      "step": 8478
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8685987056867268,
+      "learning_rate": 1.852644151207859e-05,
+      "loss": 0.8133,
+      "step": 8479
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1612440830500605,
+      "learning_rate": 1.8526042804362666e-05,
+      "loss": 0.9538,
+      "step": 8480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6390894845335848,
+      "learning_rate": 1.852564404700542e-05,
+      "loss": 0.8637,
+      "step": 8481
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6453957655895917,
+      "learning_rate": 1.852524524000918e-05,
+      "loss": 0.8041,
+      "step": 8482
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1871658819611537,
+      "learning_rate": 1.852484638337626e-05,
+      "loss": 0.8828,
+      "step": 8483
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8242511446453522,
+      "learning_rate": 1.852444747710899e-05,
+      "loss": 0.9101,
+      "step": 8484
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0883690904275585,
+      "learning_rate": 1.852404852120968e-05,
+      "loss": 0.9346,
+      "step": 8485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6121165720108215,
+      "learning_rate": 1.852364951568067e-05,
+      "loss": 0.9024,
+      "step": 8486
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7775002962117965,
+      "learning_rate": 1.8523250460524264e-05,
+      "loss": 0.9171,
+      "step": 8487
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.539935006661114,
+      "learning_rate": 1.85228513557428e-05,
+      "loss": 0.9722,
+      "step": 8488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.057728556712492,
+      "learning_rate": 1.85224522013386e-05,
+      "loss": 0.9659,
+      "step": 8489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2605612684780856,
+      "learning_rate": 1.8522052997313982e-05,
+      "loss": 0.8696,
+      "step": 8490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5498246663636537,
+      "learning_rate": 1.8521653743671276e-05,
+      "loss": 0.8762,
+      "step": 8491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9679863982971697,
+      "learning_rate": 1.8521254440412798e-05,
+      "loss": 0.9755,
+      "step": 8492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9634919705301486,
+      "learning_rate": 1.8520855087540887e-05,
+      "loss": 0.8852,
+      "step": 8493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5454441686965754,
+      "learning_rate": 1.8520455685057856e-05,
+      "loss": 0.8163,
+      "step": 8494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8317514899825185,
+      "learning_rate": 1.8520056232966033e-05,
+      "loss": 0.8795,
+      "step": 8495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.787377926218575,
+      "learning_rate": 1.8519656731267746e-05,
+      "loss": 0.9423,
+      "step": 8496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6616529470916621,
+      "learning_rate": 1.8519257179965322e-05,
+      "loss": 0.8662,
+      "step": 8497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7526405043291478,
+      "learning_rate": 1.8518857579061083e-05,
+      "loss": 0.7212,
+      "step": 8498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8094338750909014,
+      "learning_rate": 1.851845792855736e-05,
+      "loss": 0.9188,
+      "step": 8499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.99187073202312,
+      "learning_rate": 1.8518058228456478e-05,
+      "loss": 0.9931,
+      "step": 8500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.535045687914837,
+      "learning_rate": 1.851765847876076e-05,
+      "loss": 0.8116,
+      "step": 8501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.937380496151916,
+      "learning_rate": 1.8517258679472546e-05,
+      "loss": 0.9455,
+      "step": 8502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6181966518476965,
+      "learning_rate": 1.851685883059415e-05,
+      "loss": 0.9334,
+      "step": 8503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7546706351655366,
+      "learning_rate": 1.8516458932127906e-05,
+      "loss": 0.9779,
+      "step": 8504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.667290964012016,
+      "learning_rate": 1.8516058984076145e-05,
+      "loss": 0.8695,
+      "step": 8505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6138477160058102,
+      "learning_rate": 1.851565898644119e-05,
+      "loss": 0.8523,
+      "step": 8506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6618926080839702,
+      "learning_rate": 1.851525893922537e-05,
+      "loss": 1.0215,
+      "step": 8507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6146601685837225,
+      "learning_rate": 1.8514858842431017e-05,
+      "loss": 0.7908,
+      "step": 8508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.249518876509936,
+      "learning_rate": 1.8514458696060463e-05,
+      "loss": 0.9511,
+      "step": 8509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6158594754679874,
+      "learning_rate": 1.8514058500116033e-05,
+      "loss": 0.7815,
+      "step": 8510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5409367684358848,
+      "learning_rate": 1.851365825460006e-05,
+      "loss": 0.8879,
+      "step": 8511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1697349749643573,
+      "learning_rate": 1.851325795951487e-05,
+      "loss": 0.8049,
+      "step": 8512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5960580393426071,
+      "learning_rate": 1.85128576148628e-05,
+      "loss": 0.8145,
+      "step": 8513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9827969165791595,
+      "learning_rate": 1.851245722064618e-05,
+      "loss": 0.7953,
+      "step": 8514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6682022407683916,
+      "learning_rate": 1.8512056776867335e-05,
+      "loss": 0.8842,
+      "step": 8515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5677347014568361,
+      "learning_rate": 1.8511656283528603e-05,
+      "loss": 0.7634,
+      "step": 8516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6065840386941106,
+      "learning_rate": 1.851125574063231e-05,
+      "loss": 0.8404,
+      "step": 8517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6815569815944067,
+      "learning_rate": 1.8510855148180794e-05,
+      "loss": 0.9745,
+      "step": 8518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.743000710437514,
+      "learning_rate": 1.851045450617639e-05,
+      "loss": 0.8907,
+      "step": 8519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2021486823548098,
+      "learning_rate": 1.8510053814621417e-05,
+      "loss": 0.8812,
+      "step": 8520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5296473446746053,
+      "learning_rate": 1.850965307351822e-05,
+      "loss": 0.9723,
+      "step": 8521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6942146581481234,
+      "learning_rate": 1.8509252282869127e-05,
+      "loss": 0.9978,
+      "step": 8522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6545101047914004,
+      "learning_rate": 1.8508851442676475e-05,
+      "loss": 0.9208,
+      "step": 8523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.014450653007087,
+      "learning_rate": 1.8508450552942597e-05,
+      "loss": 0.9178,
+      "step": 8524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6990550428899402,
+      "learning_rate": 1.8508049613669824e-05,
+      "loss": 0.9737,
+      "step": 8525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1096192457488008,
+      "learning_rate": 1.850764862486049e-05,
+      "loss": 0.9065,
+      "step": 8526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7661350834629004,
+      "learning_rate": 1.850724758651694e-05,
+      "loss": 1.0323,
+      "step": 8527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8139233357976647,
+      "learning_rate": 1.8506846498641496e-05,
+      "loss": 0.7282,
+      "step": 8528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2175542816521463,
+      "learning_rate": 1.8506445361236502e-05,
+      "loss": 0.962,
+      "step": 8529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.027066081797319,
+      "learning_rate": 1.8506044174304282e-05,
+      "loss": 0.8533,
+      "step": 8530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6241337945459107,
+      "learning_rate": 1.8505642937847188e-05,
+      "loss": 1.0138,
+      "step": 8531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.59776564198221,
+      "learning_rate": 1.8505241651867543e-05,
+      "loss": 0.7683,
+      "step": 8532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0683681980445576,
+      "learning_rate": 1.8504840316367692e-05,
+      "loss": 0.956,
+      "step": 8533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7964661731922056,
+      "learning_rate": 1.8504438931349967e-05,
+      "loss": 0.8988,
+      "step": 8534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.709366049976162,
+      "learning_rate": 1.8504037496816708e-05,
+      "loss": 0.9086,
+      "step": 8535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9101120818076478,
+      "learning_rate": 1.8503636012770247e-05,
+      "loss": 0.8495,
+      "step": 8536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7108145186101669,
+      "learning_rate": 1.8503234479212924e-05,
+      "loss": 0.8917,
+      "step": 8537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4899360837187816,
+      "learning_rate": 1.850283289614708e-05,
+      "loss": 0.9344,
+      "step": 8538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9894166866962084,
+      "learning_rate": 1.850243126357505e-05,
+      "loss": 0.8684,
+      "step": 8539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1156382734973027,
+      "learning_rate": 1.8502029581499173e-05,
+      "loss": 0.9265,
+      "step": 8540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7292380634258262,
+      "learning_rate": 1.850162784992179e-05,
+      "loss": 0.9365,
+      "step": 8541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8833082966899335,
+      "learning_rate": 1.8501226068845238e-05,
+      "loss": 1.0139,
+      "step": 8542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5573523253258177,
+      "learning_rate": 1.8500824238271855e-05,
+      "loss": 0.9441,
+      "step": 8543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.549032698870728,
+      "learning_rate": 1.8500422358203982e-05,
+      "loss": 0.802,
+      "step": 8544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6826808163481308,
+      "learning_rate": 1.850002042864396e-05,
+      "loss": 0.8737,
+      "step": 8545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7142441347780606,
+      "learning_rate": 1.849961844959413e-05,
+      "loss": 0.9322,
+      "step": 8546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5116872974414808,
+      "learning_rate": 1.8499216421056823e-05,
+      "loss": 0.8624,
+      "step": 8547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.839821453839946,
+      "learning_rate": 1.8498814343034396e-05,
+      "loss": 0.9695,
+      "step": 8548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8172716373730189,
+      "learning_rate": 1.8498412215529175e-05,
+      "loss": 0.8692,
+      "step": 8549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8025677161105773,
+      "learning_rate": 1.849801003854351e-05,
+      "loss": 0.7457,
+      "step": 8550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9506045722651408,
+      "learning_rate": 1.849760781207974e-05,
+      "loss": 0.8989,
+      "step": 8551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5758824462352072,
+      "learning_rate": 1.8497205536140207e-05,
+      "loss": 0.9542,
+      "step": 8552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.072715516024121,
+      "learning_rate": 1.8496803210727252e-05,
+      "loss": 0.8753,
+      "step": 8553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6065007002030396,
+      "learning_rate": 1.8496400835843223e-05,
+      "loss": 0.9206,
+      "step": 8554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5786061517034995,
+      "learning_rate": 1.849599841149045e-05,
+      "loss": 0.8963,
+      "step": 8555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6125663963562167,
+      "learning_rate": 1.8495595937671293e-05,
+      "loss": 1.0028,
+      "step": 8556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7473307797744644,
+      "learning_rate": 1.8495193414388084e-05,
+      "loss": 0.9022,
+      "step": 8557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0979686281560412,
+      "learning_rate": 1.849479084164317e-05,
+      "loss": 0.9059,
+      "step": 8558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0627934535422883,
+      "learning_rate": 1.8494388219438892e-05,
+      "loss": 0.8586,
+      "step": 8559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.270367685526462,
+      "learning_rate": 1.8493985547777598e-05,
+      "loss": 0.9881,
+      "step": 8560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5499238459372044,
+      "learning_rate": 1.8493582826661628e-05,
+      "loss": 0.985,
+      "step": 8561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7716640461624483,
+      "learning_rate": 1.8493180056093334e-05,
+      "loss": 0.7494,
+      "step": 8562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6807207373360855,
+      "learning_rate": 1.8492777236075056e-05,
+      "loss": 0.7725,
+      "step": 8563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6117713912402878,
+      "learning_rate": 1.849237436660914e-05,
+      "loss": 0.8378,
+      "step": 8564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.632873654539316,
+      "learning_rate": 1.849197144769793e-05,
+      "loss": 0.9057,
+      "step": 8565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7212984774813707,
+      "learning_rate": 1.8491568479343775e-05,
+      "loss": 0.8733,
+      "step": 8566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6566263458833554,
+      "learning_rate": 1.849116546154902e-05,
+      "loss": 1.0447,
+      "step": 8567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4470360683020318,
+      "learning_rate": 1.849076239431601e-05,
+      "loss": 0.8616,
+      "step": 8568
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.507131370211232,
+      "learning_rate": 1.8490359277647096e-05,
+      "loss": 0.8953,
+      "step": 8569
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7127486300103063,
+      "learning_rate": 1.8489956111544624e-05,
+      "loss": 0.8803,
+      "step": 8570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4557805739964864,
+      "learning_rate": 1.8489552896010932e-05,
+      "loss": 0.7982,
+      "step": 8571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9635993636865947,
+      "learning_rate": 1.8489149631048378e-05,
+      "loss": 0.9414,
+      "step": 8572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8755198372601003,
+      "learning_rate": 1.848874631665931e-05,
+      "loss": 0.9856,
+      "step": 8573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6928747857611672,
+      "learning_rate": 1.8488342952846074e-05,
+      "loss": 0.8769,
+      "step": 8574
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.786746828573641,
+      "learning_rate": 1.8487939539611016e-05,
+      "loss": 0.9355,
+      "step": 8575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6387873140303921,
+      "learning_rate": 1.8487536076956488e-05,
+      "loss": 0.9004,
+      "step": 8576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.610264467981291,
+      "learning_rate": 1.8487132564884835e-05,
+      "loss": 0.836,
+      "step": 8577
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0269280573064115,
+      "learning_rate": 1.848672900339841e-05,
+      "loss": 1.0369,
+      "step": 8578
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5469705227506703,
+      "learning_rate": 1.848632539249957e-05,
+      "loss": 0.8928,
+      "step": 8579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8261533412527079,
+      "learning_rate": 1.8485921732190644e-05,
+      "loss": 0.735,
+      "step": 8580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.18527784341083,
+      "learning_rate": 1.8485518022474004e-05,
+      "loss": 0.8952,
+      "step": 8581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9103762677074283,
+      "learning_rate": 1.8485114263351988e-05,
+      "loss": 0.971,
+      "step": 8582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8390993299470144,
+      "learning_rate": 1.8484710454826952e-05,
+      "loss": 0.9376,
+      "step": 8583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9187390792051777,
+      "learning_rate": 1.8484306596901243e-05,
+      "loss": 0.9989,
+      "step": 8584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6246531817695946,
+      "learning_rate": 1.8483902689577217e-05,
+      "loss": 0.8798,
+      "step": 8585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0948429126760324,
+      "learning_rate": 1.8483498732857224e-05,
+      "loss": 0.7424,
+      "step": 8586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7691338978667127,
+      "learning_rate": 1.8483094726743613e-05,
+      "loss": 0.7356,
+      "step": 8587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9329577294317484,
+      "learning_rate": 1.8482690671238738e-05,
+      "loss": 0.7533,
+      "step": 8588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.068104068708287,
+      "learning_rate": 1.8482286566344957e-05,
+      "loss": 0.8973,
+      "step": 8589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0941071795444204,
+      "learning_rate": 1.8481882412064616e-05,
+      "loss": 0.9331,
+      "step": 8590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7417406567863543,
+      "learning_rate": 1.848147820840007e-05,
+      "loss": 0.9236,
+      "step": 8591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.233786682132495,
+      "learning_rate": 1.8481073955353673e-05,
+      "loss": 0.8086,
+      "step": 8592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.138533006570326,
+      "learning_rate": 1.8480669652927778e-05,
+      "loss": 0.9629,
+      "step": 8593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.734978783003305,
+      "learning_rate": 1.848026530112474e-05,
+      "loss": 0.8002,
+      "step": 8594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5909867746423096,
+      "learning_rate": 1.8479860899946912e-05,
+      "loss": 0.8145,
+      "step": 8595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7630520895936856,
+      "learning_rate": 1.847945644939665e-05,
+      "loss": 0.7528,
+      "step": 8596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6197586078616397,
+      "learning_rate": 1.8479051949476307e-05,
+      "loss": 0.8971,
+      "step": 8597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.40625727063028,
+      "learning_rate": 1.847864740018824e-05,
+      "loss": 0.9698,
+      "step": 8598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7646931636924501,
+      "learning_rate": 1.8478242801534803e-05,
+      "loss": 0.7904,
+      "step": 8599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7809517316561204,
+      "learning_rate": 1.8477838153518354e-05,
+      "loss": 0.7442,
+      "step": 8600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5546340193742207,
+      "learning_rate": 1.8477433456141242e-05,
+      "loss": 0.9671,
+      "step": 8601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7214664729122142,
+      "learning_rate": 1.8477028709405834e-05,
+      "loss": 0.8984,
+      "step": 8602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7360432840761781,
+      "learning_rate": 1.8476623913314476e-05,
+      "loss": 0.8788,
+      "step": 8603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7560534721183783,
+      "learning_rate": 1.8476219067869534e-05,
+      "loss": 0.9427,
+      "step": 8604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5690711696928483,
+      "learning_rate": 1.847581417307336e-05,
+      "loss": 0.8101,
+      "step": 8605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6548587164902386,
+      "learning_rate": 1.8475409228928314e-05,
+      "loss": 0.8858,
+      "step": 8606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8209542793896,
+      "learning_rate": 1.847500423543675e-05,
+      "loss": 0.9396,
+      "step": 8607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7833520396253223,
+      "learning_rate": 1.847459919260103e-05,
+      "loss": 0.7567,
+      "step": 8608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7060311663794832,
+      "learning_rate": 1.8474194100423508e-05,
+      "loss": 0.8774,
+      "step": 8609
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6084929292211831,
+      "learning_rate": 1.8473788958906548e-05,
+      "loss": 0.8693,
+      "step": 8610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8381585501389714,
+      "learning_rate": 1.8473383768052504e-05,
+      "loss": 0.7536,
+      "step": 8611
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.864025386247005,
+      "learning_rate": 1.8472978527863737e-05,
+      "loss": 0.8647,
+      "step": 8612
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.713892218007599,
+      "learning_rate": 1.8472573238342606e-05,
+      "loss": 0.954,
+      "step": 8613
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2350819780796356,
+      "learning_rate": 1.8472167899491472e-05,
+      "loss": 1.0373,
+      "step": 8614
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5264569347746673,
+      "learning_rate": 1.8471762511312693e-05,
+      "loss": 0.9809,
+      "step": 8615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7568500529927398,
+      "learning_rate": 1.8471357073808635e-05,
+      "loss": 0.9656,
+      "step": 8616
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7320098041314984,
+      "learning_rate": 1.8470951586981648e-05,
+      "loss": 1.0898,
+      "step": 8617
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7854762114210376,
+      "learning_rate": 1.8470546050834104e-05,
+      "loss": 0.9249,
+      "step": 8618
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5440913657182778,
+      "learning_rate": 1.8470140465368355e-05,
+      "loss": 0.9792,
+      "step": 8619
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6360225981410705,
+      "learning_rate": 1.8469734830586773e-05,
+      "loss": 0.9749,
+      "step": 8620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.514550134913989,
+      "learning_rate": 1.8469329146491708e-05,
+      "loss": 0.8171,
+      "step": 8621
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8188042454051414,
+      "learning_rate": 1.846892341308553e-05,
+      "loss": 0.845,
+      "step": 8622
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9055713778237737,
+      "learning_rate": 1.8468517630370594e-05,
+      "loss": 0.8301,
+      "step": 8623
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6632844649535294,
+      "learning_rate": 1.846811179834927e-05,
+      "loss": 1.0765,
+      "step": 8624
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8443007211161917,
+      "learning_rate": 1.846770591702392e-05,
+      "loss": 0.943,
+      "step": 8625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9453977723000593,
+      "learning_rate": 1.8467299986396903e-05,
+      "loss": 0.9942,
+      "step": 8626
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6921202320530098,
+      "learning_rate": 1.8466894006470588e-05,
+      "loss": 0.7878,
+      "step": 8627
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6476908826025745,
+      "learning_rate": 1.846648797724733e-05,
+      "loss": 0.8202,
+      "step": 8628
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8044355081115757,
+      "learning_rate": 1.8466081898729507e-05,
+      "loss": 0.7629,
+      "step": 8629
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6724914997261517,
+      "learning_rate": 1.8465675770919468e-05,
+      "loss": 0.9942,
+      "step": 8630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.412288430984325,
+      "learning_rate": 1.8465269593819586e-05,
+      "loss": 0.8426,
+      "step": 8631
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8624186655961257,
+      "learning_rate": 1.8464863367432225e-05,
+      "loss": 0.8379,
+      "step": 8632
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8367948181276124,
+      "learning_rate": 1.8464457091759748e-05,
+      "loss": 0.8496,
+      "step": 8633
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.065663367071242,
+      "learning_rate": 1.8464050766804526e-05,
+      "loss": 0.9464,
+      "step": 8634
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6017992203064846,
+      "learning_rate": 1.8463644392568916e-05,
+      "loss": 0.7825,
+      "step": 8635
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9880240467432448,
+      "learning_rate": 1.846323796905529e-05,
+      "loss": 0.9619,
+      "step": 8636
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1953762513062003,
+      "learning_rate": 1.8462831496266013e-05,
+      "loss": 0.9619,
+      "step": 8637
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7930739111956484,
+      "learning_rate": 1.8462424974203452e-05,
+      "loss": 0.865,
+      "step": 8638
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7961762861189827,
+      "learning_rate": 1.8462018402869977e-05,
+      "loss": 0.9025,
+      "step": 8639
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6976777833264385,
+      "learning_rate": 1.8461611782267947e-05,
+      "loss": 0.8659,
+      "step": 8640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.543339744444445,
+      "learning_rate": 1.8461205112399736e-05,
+      "loss": 0.9059,
+      "step": 8641
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8509996472305668,
+      "learning_rate": 1.8460798393267713e-05,
+      "loss": 0.8045,
+      "step": 8642
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.176871552572001,
+      "learning_rate": 1.846039162487424e-05,
+      "loss": 0.8302,
+      "step": 8643
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5615835385508947,
+      "learning_rate": 1.8459984807221686e-05,
+      "loss": 0.9818,
+      "step": 8644
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5326339891984404,
+      "learning_rate": 1.845957794031243e-05,
+      "loss": 0.826,
+      "step": 8645
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5937817315818057,
+      "learning_rate": 1.8459171024148826e-05,
+      "loss": 0.8112,
+      "step": 8646
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9506305886408928,
+      "learning_rate": 1.845876405873325e-05,
+      "loss": 0.9696,
+      "step": 8647
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.771379497752999,
+      "learning_rate": 1.8458357044068073e-05,
+      "loss": 0.9242,
+      "step": 8648
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.8260476860600985,
+      "learning_rate": 1.8457949980155666e-05,
+      "loss": 0.9896,
+      "step": 8649
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5745924764372936,
+      "learning_rate": 1.8457542866998394e-05,
+      "loss": 0.9201,
+      "step": 8650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.556396494248334,
+      "learning_rate": 1.8457135704598634e-05,
+      "loss": 0.9316,
+      "step": 8651
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5632825100121932,
+      "learning_rate": 1.8456728492958748e-05,
+      "loss": 0.8923,
+      "step": 8652
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2038229999667442,
+      "learning_rate": 1.845632123208111e-05,
+      "loss": 0.9217,
+      "step": 8653
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.694996269496756,
+      "learning_rate": 1.8455913921968095e-05,
+      "loss": 0.9132,
+      "step": 8654
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7558508925556586,
+      "learning_rate": 1.845550656262207e-05,
+      "loss": 0.9265,
+      "step": 8655
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8183946056453737,
+      "learning_rate": 1.8455099154045416e-05,
+      "loss": 1.0041,
+      "step": 8656
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7700852022797172,
+      "learning_rate": 1.8454691696240493e-05,
+      "loss": 0.7228,
+      "step": 8657
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5911045268716943,
+      "learning_rate": 1.8454284189209674e-05,
+      "loss": 0.8768,
+      "step": 8658
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0584105921039053,
+      "learning_rate": 1.8453876632955342e-05,
+      "loss": 0.8738,
+      "step": 8659
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7612661673566286,
+      "learning_rate": 1.845346902747986e-05,
+      "loss": 0.8607,
+      "step": 8660
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7740357408304065,
+      "learning_rate": 1.8453061372785605e-05,
+      "loss": 0.7606,
+      "step": 8661
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.911855199639178,
+      "learning_rate": 1.8452653668874954e-05,
+      "loss": 0.9219,
+      "step": 8662
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7666503142916472,
+      "learning_rate": 1.8452245915750276e-05,
+      "loss": 0.9015,
+      "step": 8663
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8704108638537524,
+      "learning_rate": 1.8451838113413946e-05,
+      "loss": 0.9598,
+      "step": 8664
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0587739092290716,
+      "learning_rate": 1.8451430261868336e-05,
+      "loss": 0.918,
+      "step": 8665
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7957689434076228,
+      "learning_rate": 1.8451022361115828e-05,
+      "loss": 0.7473,
+      "step": 8666
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6142652774146062,
+      "learning_rate": 1.845061441115879e-05,
+      "loss": 0.9023,
+      "step": 8667
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9356326775028867,
+      "learning_rate": 1.84502064119996e-05,
+      "loss": 0.9929,
+      "step": 8668
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5286199315237228,
+      "learning_rate": 1.844979836364063e-05,
+      "loss": 0.9842,
+      "step": 8669
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.889064055342878,
+      "learning_rate": 1.8449390266084263e-05,
+      "loss": 0.8507,
+      "step": 8670
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7939909845284109,
+      "learning_rate": 1.8448982119332868e-05,
+      "loss": 0.7284,
+      "step": 8671
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5369832124830571,
+      "learning_rate": 1.8448573923388825e-05,
+      "loss": 1.033,
+      "step": 8672
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5086755488418715,
+      "learning_rate": 1.844816567825451e-05,
+      "loss": 0.9773,
+      "step": 8673
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1766050589267913,
+      "learning_rate": 1.8447757383932302e-05,
+      "loss": 0.8721,
+      "step": 8674
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.014110915911494,
+      "learning_rate": 1.8447349040424572e-05,
+      "loss": 0.8955,
+      "step": 8675
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.190163903614739,
+      "learning_rate": 1.8446940647733705e-05,
+      "loss": 0.9919,
+      "step": 8676
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.468296673779067,
+      "learning_rate": 1.8446532205862072e-05,
+      "loss": 0.836,
+      "step": 8677
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7803226693834249,
+      "learning_rate": 1.8446123714812055e-05,
+      "loss": 0.7414,
+      "step": 8678
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6374069822417985,
+      "learning_rate": 1.8445715174586033e-05,
+      "loss": 0.8691,
+      "step": 8679
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.750884750659187,
+      "learning_rate": 1.8445306585186384e-05,
+      "loss": 0.7228,
+      "step": 8680
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7584766523005914,
+      "learning_rate": 1.8444897946615486e-05,
+      "loss": 0.6897,
+      "step": 8681
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1674191642037677,
+      "learning_rate": 1.8444489258875718e-05,
+      "loss": 0.8688,
+      "step": 8682
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7290748554001172,
+      "learning_rate": 1.844408052196946e-05,
+      "loss": 0.7451,
+      "step": 8683
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1089531581273864,
+      "learning_rate": 1.8443671735899092e-05,
+      "loss": 0.9048,
+      "step": 8684
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5829558600037732,
+      "learning_rate": 1.8443262900666995e-05,
+      "loss": 0.9246,
+      "step": 8685
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8299027203373377,
+      "learning_rate": 1.8442854016275545e-05,
+      "loss": 0.9315,
+      "step": 8686
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6510231651006797,
+      "learning_rate": 1.8442445082727127e-05,
+      "loss": 0.9284,
+      "step": 8687
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8078237148091592,
+      "learning_rate": 1.8442036100024124e-05,
+      "loss": 0.7664,
+      "step": 8688
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.82788756512805,
+      "learning_rate": 1.8441627068168914e-05,
+      "loss": 0.8788,
+      "step": 8689
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.103215542426798,
+      "learning_rate": 1.8441217987163874e-05,
+      "loss": 0.7952,
+      "step": 8690
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9403406731356783,
+      "learning_rate": 1.8440808857011392e-05,
+      "loss": 1.0484,
+      "step": 8691
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7499368665660724,
+      "learning_rate": 1.844039967771385e-05,
+      "loss": 0.9217,
+      "step": 8692
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7972127311486283,
+      "learning_rate": 1.843999044927363e-05,
+      "loss": 0.9252,
+      "step": 8693
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6079803259657601,
+      "learning_rate": 1.843958117169311e-05,
+      "loss": 0.9017,
+      "step": 8694
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5075421014807944,
+      "learning_rate": 1.843917184497468e-05,
+      "loss": 0.9486,
+      "step": 8695
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6536990791278363,
+      "learning_rate": 1.843876246912072e-05,
+      "loss": 0.9472,
+      "step": 8696
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5767744789825073,
+      "learning_rate": 1.843835304413361e-05,
+      "loss": 0.9627,
+      "step": 8697
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8173478075519652,
+      "learning_rate": 1.8437943570015737e-05,
+      "loss": 0.7642,
+      "step": 8698
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5276352655520238,
+      "learning_rate": 1.843753404676949e-05,
+      "loss": 0.9349,
+      "step": 8699
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6879024429995648,
+      "learning_rate": 1.8437124474397247e-05,
+      "loss": 0.9005,
+      "step": 8700
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.601152075501491,
+      "learning_rate": 1.843671485290139e-05,
+      "loss": 1.0045,
+      "step": 8701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3355786389604094,
+      "learning_rate": 1.843630518228431e-05,
+      "loss": 0.9055,
+      "step": 8702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5270808281265775,
+      "learning_rate": 1.843589546254839e-05,
+      "loss": 0.9029,
+      "step": 8703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.579477624042867,
+      "learning_rate": 1.8435485693696015e-05,
+      "loss": 0.9258,
+      "step": 8704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6291759985728047,
+      "learning_rate": 1.8435075875729577e-05,
+      "loss": 0.9823,
+      "step": 8705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8260369720874339,
+      "learning_rate": 1.843466600865145e-05,
+      "loss": 0.7112,
+      "step": 8706
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.664297556911803,
+      "learning_rate": 1.843425609246403e-05,
+      "loss": 0.9082,
+      "step": 8707
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.790271013601674,
+      "learning_rate": 1.8433846127169704e-05,
+      "loss": 0.7925,
+      "step": 8708
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7888243291145487,
+      "learning_rate": 1.843343611277085e-05,
+      "loss": 0.7424,
+      "step": 8709
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.664280263581358,
+      "learning_rate": 1.8433026049269866e-05,
+      "loss": 0.8933,
+      "step": 8710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.946095974008104,
+      "learning_rate": 1.8432615936669133e-05,
+      "loss": 0.8957,
+      "step": 8711
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.613966458824085,
+      "learning_rate": 1.8432205774971036e-05,
+      "loss": 0.8139,
+      "step": 8712
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8214170602688621,
+      "learning_rate": 1.8431795564177975e-05,
+      "loss": 0.7435,
+      "step": 8713
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.859007123920347,
+      "learning_rate": 1.8431385304292324e-05,
+      "loss": 0.8914,
+      "step": 8714
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8901260178459418,
+      "learning_rate": 1.8430974995316482e-05,
+      "loss": 1.0191,
+      "step": 8715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8662449808299357,
+      "learning_rate": 1.8430564637252835e-05,
+      "loss": 0.8944,
+      "step": 8716
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.71946523544186,
+      "learning_rate": 1.843015423010377e-05,
+      "loss": 0.9414,
+      "step": 8717
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7459561344368961,
+      "learning_rate": 1.8429743773871678e-05,
+      "loss": 0.9056,
+      "step": 8718
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5273779113112655,
+      "learning_rate": 1.842933326855895e-05,
+      "loss": 0.8545,
+      "step": 8719
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.556835924092473,
+      "learning_rate": 1.842892271416797e-05,
+      "loss": 0.9693,
+      "step": 8720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5799568796933035,
+      "learning_rate": 1.842851211070114e-05,
+      "loss": 1.0202,
+      "step": 8721
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6237849809131009,
+      "learning_rate": 1.8428101458160843e-05,
+      "loss": 0.9891,
+      "step": 8722
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6586927523634185,
+      "learning_rate": 1.842769075654947e-05,
+      "loss": 0.9347,
+      "step": 8723
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2128328327549234,
+      "learning_rate": 1.842728000586941e-05,
+      "loss": 0.8692,
+      "step": 8724
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.543025447537126,
+      "learning_rate": 1.8426869206123062e-05,
+      "loss": 0.8119,
+      "step": 8725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0038927637501787,
+      "learning_rate": 1.8426458357312814e-05,
+      "loss": 0.8678,
+      "step": 8726
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.000136428050431,
+      "learning_rate": 1.8426047459441053e-05,
+      "loss": 0.8031,
+      "step": 8727
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5176651809359438,
+      "learning_rate": 1.842563651251018e-05,
+      "loss": 0.9419,
+      "step": 8728
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8134676897475637,
+      "learning_rate": 1.842522551652258e-05,
+      "loss": 0.9023,
+      "step": 8729
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.699183445819275,
+      "learning_rate": 1.8424814471480658e-05,
+      "loss": 0.9109,
+      "step": 8730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2917382169323526,
+      "learning_rate": 1.842440337738679e-05,
+      "loss": 0.9069,
+      "step": 8731
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7205742905217556,
+      "learning_rate": 1.842399223424338e-05,
+      "loss": 0.8484,
+      "step": 8732
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6638262464283202,
+      "learning_rate": 1.8423581042052823e-05,
+      "loss": 0.8581,
+      "step": 8733
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6949820341813044,
+      "learning_rate": 1.842316980081751e-05,
+      "loss": 0.9103,
+      "step": 8734
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.69131740212549,
+      "learning_rate": 1.8422758510539837e-05,
+      "loss": 0.825,
+      "step": 8735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.612800480920809,
+      "learning_rate": 1.8422347171222194e-05,
+      "loss": 0.7503,
+      "step": 8736
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9240631640248882,
+      "learning_rate": 1.8421935782866982e-05,
+      "loss": 0.9691,
+      "step": 8737
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6360968267438154,
+      "learning_rate": 1.842152434547659e-05,
+      "loss": 0.9679,
+      "step": 8738
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9435919205291328,
+      "learning_rate": 1.8421112859053417e-05,
+      "loss": 0.7978,
+      "step": 8739
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8057840433084786,
+      "learning_rate": 1.8420701323599865e-05,
+      "loss": 0.9486,
+      "step": 8740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8048204845694881,
+      "learning_rate": 1.8420289739118318e-05,
+      "loss": 0.7116,
+      "step": 8741
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9021975041072126,
+      "learning_rate": 1.841987810561118e-05,
+      "loss": 0.8654,
+      "step": 8742
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7235374433417552,
+      "learning_rate": 1.8419466423080847e-05,
+      "loss": 0.8916,
+      "step": 8743
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.280820576333117,
+      "learning_rate": 1.8419054691529715e-05,
+      "loss": 0.956,
+      "step": 8744
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5449493190216763,
+      "learning_rate": 1.841864291096018e-05,
+      "loss": 0.9763,
+      "step": 8745
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.752678966705807,
+      "learning_rate": 1.841823108137464e-05,
+      "loss": 0.98,
+      "step": 8746
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7924943801977427,
+      "learning_rate": 1.8417819202775497e-05,
+      "loss": 0.7407,
+      "step": 8747
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.912343970692861,
+      "learning_rate": 1.841740727516514e-05,
+      "loss": 1.0063,
+      "step": 8748
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7509980742426394,
+      "learning_rate": 1.8416995298545978e-05,
+      "loss": 0.9953,
+      "step": 8749
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6296568118382362,
+      "learning_rate": 1.84165832729204e-05,
+      "loss": 0.9085,
+      "step": 8750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7245164620189348,
+      "learning_rate": 1.8416171198290812e-05,
+      "loss": 0.8562,
+      "step": 8751
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9343679819562072,
+      "learning_rate": 1.8415759074659614e-05,
+      "loss": 1.0302,
+      "step": 8752
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7164925032443785,
+      "learning_rate": 1.8415346902029196e-05,
+      "loss": 0.7767,
+      "step": 8753
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9589460439060122,
+      "learning_rate": 1.8414934680401968e-05,
+      "loss": 0.8477,
+      "step": 8754
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0281044670436175,
+      "learning_rate": 1.8414522409780328e-05,
+      "loss": 0.8654,
+      "step": 8755
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6900401712397104,
+      "learning_rate": 1.8414110090166672e-05,
+      "loss": 0.9934,
+      "step": 8756
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.073835231084113,
+      "learning_rate": 1.84136977215634e-05,
+      "loss": 0.9707,
+      "step": 8757
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.325351677324342,
+      "learning_rate": 1.8413285303972922e-05,
+      "loss": 0.8918,
+      "step": 8758
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8559822181264944,
+      "learning_rate": 1.8412872837397627e-05,
+      "loss": 0.9272,
+      "step": 8759
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6391420073608158,
+      "learning_rate": 1.841246032183993e-05,
+      "loss": 0.8463,
+      "step": 8760
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5807856917912748,
+      "learning_rate": 1.8412047757302222e-05,
+      "loss": 0.9184,
+      "step": 8761
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6318401412271941,
+      "learning_rate": 1.8411635143786906e-05,
+      "loss": 0.8499,
+      "step": 8762
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2575363361189895,
+      "learning_rate": 1.841122248129639e-05,
+      "loss": 0.8863,
+      "step": 8763
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8247762848548768,
+      "learning_rate": 1.8410809769833075e-05,
+      "loss": 0.8903,
+      "step": 8764
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7018917970228413,
+      "learning_rate": 1.8410397009399363e-05,
+      "loss": 0.9226,
+      "step": 8765
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5900609954603242,
+      "learning_rate": 1.8409984199997653e-05,
+      "loss": 0.844,
+      "step": 8766
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6918867531786175,
+      "learning_rate": 1.8409571341630358e-05,
+      "loss": 0.9296,
+      "step": 8767
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4545994292017335,
+      "learning_rate": 1.8409158434299872e-05,
+      "loss": 0.8482,
+      "step": 8768
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7940554328383427,
+      "learning_rate": 1.8408745478008602e-05,
+      "loss": 0.8949,
+      "step": 8769
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8228517813189073,
+      "learning_rate": 1.840833247275896e-05,
+      "loss": 0.7366,
+      "step": 8770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4919146007881967,
+      "learning_rate": 1.8407919418553337e-05,
+      "loss": 0.9684,
+      "step": 8771
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.452755174234879,
+      "learning_rate": 1.840750631539415e-05,
+      "loss": 0.8862,
+      "step": 8772
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6900434844824241,
+      "learning_rate": 1.84070931632838e-05,
+      "loss": 0.9048,
+      "step": 8773
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6849831282619248,
+      "learning_rate": 1.840667996222469e-05,
+      "loss": 0.8074,
+      "step": 8774
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7415791491613244,
+      "learning_rate": 1.8406266712219225e-05,
+      "loss": 0.8415,
+      "step": 8775
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8721349703590444,
+      "learning_rate": 1.840585341326982e-05,
+      "loss": 0.9977,
+      "step": 8776
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6287227497489898,
+      "learning_rate": 1.840544006537887e-05,
+      "loss": 0.9203,
+      "step": 8777
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4847586445199286,
+      "learning_rate": 1.840502666854879e-05,
+      "loss": 0.9162,
+      "step": 8778
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6777760063849785,
+      "learning_rate": 1.840461322278198e-05,
+      "loss": 0.787,
+      "step": 8779
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7936305741505552,
+      "learning_rate": 1.8404199728080854e-05,
+      "loss": 1.1114,
+      "step": 8780
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5849037931333316,
+      "learning_rate": 1.8403786184447818e-05,
+      "loss": 0.9439,
+      "step": 8781
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7656869019008944,
+      "learning_rate": 1.8403372591885272e-05,
+      "loss": 0.7263,
+      "step": 8782
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7815081110765898,
+      "learning_rate": 1.8402958950395637e-05,
+      "loss": 0.7885,
+      "step": 8783
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.766783734127397,
+      "learning_rate": 1.8402545259981312e-05,
+      "loss": 0.9195,
+      "step": 8784
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5675523580448794,
+      "learning_rate": 1.8402131520644708e-05,
+      "loss": 0.9571,
+      "step": 8785
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6146513046950035,
+      "learning_rate": 1.8401717732388234e-05,
+      "loss": 0.8921,
+      "step": 8786
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8646993372866402,
+      "learning_rate": 1.8401303895214303e-05,
+      "loss": 0.7727,
+      "step": 8787
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6416460625969926,
+      "learning_rate": 1.8400890009125314e-05,
+      "loss": 0.8749,
+      "step": 8788
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.279070218537133,
+      "learning_rate": 1.8400476074123686e-05,
+      "loss": 0.9101,
+      "step": 8789
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3761964082392164,
+      "learning_rate": 1.840006209021183e-05,
+      "loss": 0.7733,
+      "step": 8790
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.026250704281268,
+      "learning_rate": 1.839964805739215e-05,
+      "loss": 0.8493,
+      "step": 8791
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0283281311105776,
+      "learning_rate": 1.8399233975667062e-05,
+      "loss": 0.9216,
+      "step": 8792
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9421866329787993,
+      "learning_rate": 1.8398819845038973e-05,
+      "loss": 1.0096,
+      "step": 8793
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9469141472554328,
+      "learning_rate": 1.83984056655103e-05,
+      "loss": 0.9556,
+      "step": 8794
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.159737728910353,
+      "learning_rate": 1.8397991437083445e-05,
+      "loss": 0.9319,
+      "step": 8795
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1225637293414095,
+      "learning_rate": 1.8397577159760828e-05,
+      "loss": 0.8804,
+      "step": 8796
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.698814537950394,
+      "learning_rate": 1.8397162833544856e-05,
+      "loss": 0.7782,
+      "step": 8797
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.878144075432478,
+      "learning_rate": 1.8396748458437945e-05,
+      "loss": 0.8409,
+      "step": 8798
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6692096105147662,
+      "learning_rate": 1.8396334034442506e-05,
+      "loss": 0.8485,
+      "step": 8799
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.84042390651973,
+      "learning_rate": 1.839591956156095e-05,
+      "loss": 0.9444,
+      "step": 8800
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7595931573723744,
+      "learning_rate": 1.8395505039795696e-05,
+      "loss": 0.8633,
+      "step": 8801
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8586669978771517,
+      "learning_rate": 1.8395090469149153e-05,
+      "loss": 0.8659,
+      "step": 8802
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5481143066050236,
+      "learning_rate": 1.8394675849623733e-05,
+      "loss": 0.8931,
+      "step": 8803
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.723834868136047,
+      "learning_rate": 1.8394261181221853e-05,
+      "loss": 0.9649,
+      "step": 8804
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5357688683281685,
+      "learning_rate": 1.8393846463945925e-05,
+      "loss": 0.9378,
+      "step": 8805
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4954801278899426,
+      "learning_rate": 1.839343169779837e-05,
+      "loss": 0.9168,
+      "step": 8806
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9379799829561506,
+      "learning_rate": 1.83930168827816e-05,
+      "loss": 0.9591,
+      "step": 8807
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6263902229279648,
+      "learning_rate": 1.8392602018898023e-05,
+      "loss": 0.9316,
+      "step": 8808
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0730679798089704,
+      "learning_rate": 1.8392187106150058e-05,
+      "loss": 0.9021,
+      "step": 8809
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6393749069681227,
+      "learning_rate": 1.8391772144540127e-05,
+      "loss": 0.7911,
+      "step": 8810
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4811631063683781,
+      "learning_rate": 1.8391357134070642e-05,
+      "loss": 0.9569,
+      "step": 8811
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8758288910078244,
+      "learning_rate": 1.839094207474402e-05,
+      "loss": 0.786,
+      "step": 8812
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7849686254471071,
+      "learning_rate": 1.8390526966562673e-05,
+      "loss": 0.7616,
+      "step": 8813
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.807400598986517,
+      "learning_rate": 1.8390111809529024e-05,
+      "loss": 0.8361,
+      "step": 8814
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9636838858896635,
+      "learning_rate": 1.8389696603645487e-05,
+      "loss": 0.969,
+      "step": 8815
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5315892491065726,
+      "learning_rate": 1.8389281348914482e-05,
+      "loss": 0.9265,
+      "step": 8816
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.456861662059818,
+      "learning_rate": 1.8388866045338424e-05,
+      "loss": 0.8439,
+      "step": 8817
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5786449852745588,
+      "learning_rate": 1.8388450692919733e-05,
+      "loss": 0.8905,
+      "step": 8818
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.780602485479189,
+      "learning_rate": 1.8388035291660823e-05,
+      "loss": 0.8148,
+      "step": 8819
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.755031457336164,
+      "learning_rate": 1.8387619841564118e-05,
+      "loss": 0.9671,
+      "step": 8820
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9715872949179645,
+      "learning_rate": 1.8387204342632034e-05,
+      "loss": 0.9798,
+      "step": 8821
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.582804097282771,
+      "learning_rate": 1.8386788794866995e-05,
+      "loss": 1.0247,
+      "step": 8822
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.625998197732358,
+      "learning_rate": 1.8386373198271412e-05,
+      "loss": 0.9535,
+      "step": 8823
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.792569928949052,
+      "learning_rate": 1.8385957552847707e-05,
+      "loss": 0.9191,
+      "step": 8824
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7403323296839188,
+      "learning_rate": 1.8385541858598306e-05,
+      "loss": 0.8614,
+      "step": 8825
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5872414899086662,
+      "learning_rate": 1.8385126115525627e-05,
+      "loss": 0.9749,
+      "step": 8826
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.594548048204447,
+      "learning_rate": 1.8384710323632083e-05,
+      "loss": 0.9164,
+      "step": 8827
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7173786109090174,
+      "learning_rate": 1.838429448292011e-05,
+      "loss": 0.7767,
+      "step": 8828
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7300650549762802,
+      "learning_rate": 1.838387859339211e-05,
+      "loss": 1.0439,
+      "step": 8829
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8053580469681445,
+      "learning_rate": 1.8383462655050525e-05,
+      "loss": 0.7066,
+      "step": 8830
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8389939774972723,
+      "learning_rate": 1.838304666789776e-05,
+      "loss": 0.8967,
+      "step": 8831
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.855961246286226,
+      "learning_rate": 1.8382630631936243e-05,
+      "loss": 0.9791,
+      "step": 8832
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.564767030464751,
+      "learning_rate": 1.8382214547168396e-05,
+      "loss": 0.9872,
+      "step": 8833
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.301572065716818,
+      "learning_rate": 1.8381798413596646e-05,
+      "loss": 0.8754,
+      "step": 8834
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.937233772224033,
+      "learning_rate": 1.838138223122341e-05,
+      "loss": 0.9087,
+      "step": 8835
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7412127944862308,
+      "learning_rate": 1.8380966000051113e-05,
+      "loss": 0.9224,
+      "step": 8836
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9703425879911085,
+      "learning_rate": 1.8380549720082175e-05,
+      "loss": 0.9696,
+      "step": 8837
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6077126801437471,
+      "learning_rate": 1.838013339131903e-05,
+      "loss": 0.8684,
+      "step": 8838
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3512465847069066,
+      "learning_rate": 1.837971701376409e-05,
+      "loss": 0.8272,
+      "step": 8839
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.63268424681803,
+      "learning_rate": 1.837930058741979e-05,
+      "loss": 0.8873,
+      "step": 8840
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7407724679832732,
+      "learning_rate": 1.8378884112288543e-05,
+      "loss": 0.9423,
+      "step": 8841
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.797513459022841,
+      "learning_rate": 1.8378467588372785e-05,
+      "loss": 0.7253,
+      "step": 8842
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4737446226614224,
+      "learning_rate": 1.8378051015674933e-05,
+      "loss": 0.9171,
+      "step": 8843
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.828143952188794,
+      "learning_rate": 1.837763439419742e-05,
+      "loss": 0.9035,
+      "step": 8844
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.656855999877573,
+      "learning_rate": 1.8377217723942662e-05,
+      "loss": 0.9068,
+      "step": 8845
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5865290407258885,
+      "learning_rate": 1.837680100491309e-05,
+      "loss": 0.9342,
+      "step": 8846
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1304786909414273,
+      "learning_rate": 1.8376384237111134e-05,
+      "loss": 0.8958,
+      "step": 8847
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5831208729029818,
+      "learning_rate": 1.8375967420539215e-05,
+      "loss": 0.8889,
+      "step": 8848
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.740706093536807,
+      "learning_rate": 1.8375550555199764e-05,
+      "loss": 0.8918,
+      "step": 8849
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.237425834875246,
+      "learning_rate": 1.8375133641095206e-05,
+      "loss": 0.862,
+      "step": 8850
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6031123329274117,
+      "learning_rate": 1.8374716678227964e-05,
+      "loss": 0.925,
+      "step": 8851
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2325685430286835,
+      "learning_rate": 1.8374299666600474e-05,
+      "loss": 0.8679,
+      "step": 8852
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.229866980469793,
+      "learning_rate": 1.837388260621516e-05,
+      "loss": 0.8186,
+      "step": 8853
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5906974491541588,
+      "learning_rate": 1.8373465497074448e-05,
+      "loss": 0.9673,
+      "step": 8854
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.250408253097248,
+      "learning_rate": 1.837304833918077e-05,
+      "loss": 0.7992,
+      "step": 8855
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7805184994701977,
+      "learning_rate": 1.8372631132536552e-05,
+      "loss": 0.736,
+      "step": 8856
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.897657833713032,
+      "learning_rate": 1.837221387714423e-05,
+      "loss": 0.9558,
+      "step": 8857
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7865573917496744,
+      "learning_rate": 1.8371796573006223e-05,
+      "loss": 0.7624,
+      "step": 8858
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.260689957251731,
+      "learning_rate": 1.8371379220124968e-05,
+      "loss": 0.9126,
+      "step": 8859
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.840718913115611,
+      "learning_rate": 1.8370961818502893e-05,
+      "loss": 0.9743,
+      "step": 8860
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6457460063550446,
+      "learning_rate": 1.837054436814243e-05,
+      "loss": 0.8691,
+      "step": 8861
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9845826848605235,
+      "learning_rate": 1.8370126869046002e-05,
+      "loss": 0.9867,
+      "step": 8862
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4297450331621628,
+      "learning_rate": 1.836970932121605e-05,
+      "loss": 0.8722,
+      "step": 8863
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7308302893125933,
+      "learning_rate": 1.8369291724655e-05,
+      "loss": 0.9647,
+      "step": 8864
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.773333532666767,
+      "learning_rate": 1.836887407936528e-05,
+      "loss": 0.9252,
+      "step": 8865
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5123979921282034,
+      "learning_rate": 1.8368456385349333e-05,
+      "loss": 0.9739,
+      "step": 8866
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5397681387150144,
+      "learning_rate": 1.8368038642609577e-05,
+      "loss": 0.8329,
+      "step": 8867
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.300172924646575,
+      "learning_rate": 1.8367620851148455e-05,
+      "loss": 0.8887,
+      "step": 8868
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5133449925971572,
+      "learning_rate": 1.8367203010968393e-05,
+      "loss": 1.015,
+      "step": 8869
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.239376346624852,
+      "learning_rate": 1.8366785122071828e-05,
+      "loss": 0.8051,
+      "step": 8870
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8397494248135735,
+      "learning_rate": 1.8366367184461188e-05,
+      "loss": 0.8855,
+      "step": 8871
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0323668847331904,
+      "learning_rate": 1.8365949198138913e-05,
+      "loss": 1.0577,
+      "step": 8872
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6012778960793674,
+      "learning_rate": 1.836553116310743e-05,
+      "loss": 1.0133,
+      "step": 8873
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6099220166600479,
+      "learning_rate": 1.8365113079369176e-05,
+      "loss": 0.9866,
+      "step": 8874
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.575006081181709,
+      "learning_rate": 1.8364694946926586e-05,
+      "loss": 0.7987,
+      "step": 8875
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7508144930286764,
+      "learning_rate": 1.8364276765782095e-05,
+      "loss": 0.8592,
+      "step": 8876
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6283326632076025,
+      "learning_rate": 1.8363858535938136e-05,
+      "loss": 0.962,
+      "step": 8877
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.473065392691009,
+      "learning_rate": 1.8363440257397145e-05,
+      "loss": 0.8619,
+      "step": 8878
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2171193201609363,
+      "learning_rate": 1.8363021930161555e-05,
+      "loss": 0.8487,
+      "step": 8879
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7293234388210361,
+      "learning_rate": 1.8362603554233807e-05,
+      "loss": 0.9648,
+      "step": 8880
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4095519819902236,
+      "learning_rate": 1.8362185129616333e-05,
+      "loss": 0.8952,
+      "step": 8881
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7982390474802907,
+      "learning_rate": 1.8361766656311568e-05,
+      "loss": 0.9902,
+      "step": 8882
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7634816912091567,
+      "learning_rate": 1.836134813432195e-05,
+      "loss": 0.9433,
+      "step": 8883
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6757467405352218,
+      "learning_rate": 1.8360929563649918e-05,
+      "loss": 1.0347,
+      "step": 8884
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8011622062988787,
+      "learning_rate": 1.8360510944297905e-05,
+      "loss": 0.7767,
+      "step": 8885
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.977758092700832,
+      "learning_rate": 1.836009227626835e-05,
+      "loss": 0.9704,
+      "step": 8886
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.830399874585732,
+      "learning_rate": 1.8359673559563696e-05,
+      "loss": 0.9123,
+      "step": 8887
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.61733367338927,
+      "learning_rate": 1.8359254794186368e-05,
+      "loss": 0.9569,
+      "step": 8888
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.675265120693435,
+      "learning_rate": 1.8358835980138822e-05,
+      "loss": 0.9027,
+      "step": 8889
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.525687926910339,
+      "learning_rate": 1.8358417117423476e-05,
+      "loss": 0.8671,
+      "step": 8890
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5604477330638884,
+      "learning_rate": 1.8357998206042788e-05,
+      "loss": 1.0113,
+      "step": 8891
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7943816833059074,
+      "learning_rate": 1.8357579245999182e-05,
+      "loss": 0.9648,
+      "step": 8892
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5495431124511925,
+      "learning_rate": 1.8357160237295106e-05,
+      "loss": 0.9292,
+      "step": 8893
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7451074454988778,
+      "learning_rate": 1.8356741179933e-05,
+      "loss": 0.9321,
+      "step": 8894
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7218988869203573,
+      "learning_rate": 1.83563220739153e-05,
+      "loss": 0.9445,
+      "step": 8895
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6539529587504853,
+      "learning_rate": 1.835590291924444e-05,
+      "loss": 0.9714,
+      "step": 8896
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6351082437193776,
+      "learning_rate": 1.8355483715922876e-05,
+      "loss": 0.9717,
+      "step": 8897
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5534300918954982,
+      "learning_rate": 1.835506446395304e-05,
+      "loss": 0.8466,
+      "step": 8898
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6768200833397258,
+      "learning_rate": 1.835464516333737e-05,
+      "loss": 0.8598,
+      "step": 8899
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9793851163279959,
+      "learning_rate": 1.8354225814078313e-05,
+      "loss": 0.8416,
+      "step": 8900
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7985758683513582,
+      "learning_rate": 1.835380641617831e-05,
+      "loss": 0.9861,
+      "step": 8901
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4906406964780123,
+      "learning_rate": 1.83533869696398e-05,
+      "loss": 0.9888,
+      "step": 8902
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3624396794900617,
+      "learning_rate": 1.8352967474465223e-05,
+      "loss": 0.9442,
+      "step": 8903
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4057505449157888,
+      "learning_rate": 1.8352547930657027e-05,
+      "loss": 1.0393,
+      "step": 8904
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8275121784855645,
+      "learning_rate": 1.8352128338217654e-05,
+      "loss": 0.7029,
+      "step": 8905
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.467193762895837,
+      "learning_rate": 1.8351708697149543e-05,
+      "loss": 0.8409,
+      "step": 8906
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0369193997240265,
+      "learning_rate": 1.8351289007455145e-05,
+      "loss": 0.9277,
+      "step": 8907
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8206461850936628,
+      "learning_rate": 1.835086926913689e-05,
+      "loss": 0.9483,
+      "step": 8908
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9010064770072659,
+      "learning_rate": 1.8350449482197236e-05,
+      "loss": 0.9093,
+      "step": 8909
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6080906372752561,
+      "learning_rate": 1.8350029646638625e-05,
+      "loss": 0.9048,
+      "step": 8910
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5408082912343684,
+      "learning_rate": 1.834960976246349e-05,
+      "loss": 0.8655,
+      "step": 8911
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8195502299726025,
+      "learning_rate": 1.8349189829674288e-05,
+      "loss": 0.8231,
+      "step": 8912
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.158162315485008,
+      "learning_rate": 1.834876984827346e-05,
+      "loss": 0.9316,
+      "step": 8913
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5528719360705536,
+      "learning_rate": 1.8348349818263445e-05,
+      "loss": 0.8844,
+      "step": 8914
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.827833730528368,
+      "learning_rate": 1.83479297396467e-05,
+      "loss": 0.8603,
+      "step": 8915
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.666277517531575,
+      "learning_rate": 1.8347509612425666e-05,
+      "loss": 0.9205,
+      "step": 8916
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.864920612540786,
+      "learning_rate": 1.8347089436602786e-05,
+      "loss": 0.9553,
+      "step": 8917
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7716895101215236,
+      "learning_rate": 1.8346669212180507e-05,
+      "loss": 0.7349,
+      "step": 8918
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5312044275985404,
+      "learning_rate": 1.834624893916128e-05,
+      "loss": 0.9557,
+      "step": 8919
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7833198649467467,
+      "learning_rate": 1.834582861754755e-05,
+      "loss": 0.7508,
+      "step": 8920
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7225755210169438,
+      "learning_rate": 1.834540824734176e-05,
+      "loss": 0.9904,
+      "step": 8921
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6506899059738378,
+      "learning_rate": 1.8344987828546364e-05,
+      "loss": 0.9492,
+      "step": 8922
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.17056358881768,
+      "learning_rate": 1.8344567361163804e-05,
+      "loss": 0.9055,
+      "step": 8923
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7434209059044294,
+      "learning_rate": 1.8344146845196534e-05,
+      "loss": 0.9496,
+      "step": 8924
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6463463313076732,
+      "learning_rate": 1.8343726280646996e-05,
+      "loss": 1.0063,
+      "step": 8925
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2842308624197463,
+      "learning_rate": 1.8343305667517644e-05,
+      "loss": 0.8342,
+      "step": 8926
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8513435391531792,
+      "learning_rate": 1.8342885005810926e-05,
+      "loss": 0.6808,
+      "step": 8927
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5271451707335568,
+      "learning_rate": 1.834246429552929e-05,
+      "loss": 0.8469,
+      "step": 8928
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7092036339740484,
+      "learning_rate": 1.8342043536675185e-05,
+      "loss": 0.9204,
+      "step": 8929
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6438842679987091,
+      "learning_rate": 1.8341622729251062e-05,
+      "loss": 0.9322,
+      "step": 8930
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6893970760484256,
+      "learning_rate": 1.834120187325937e-05,
+      "loss": 0.9812,
+      "step": 8931
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.65087068704696,
+      "learning_rate": 1.834078096870256e-05,
+      "loss": 0.8295,
+      "step": 8932
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7343530078553646,
+      "learning_rate": 1.834036001558308e-05,
+      "loss": 0.7656,
+      "step": 8933
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5662074325556556,
+      "learning_rate": 1.833993901390339e-05,
+      "loss": 0.85,
+      "step": 8934
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7623008643325215,
+      "learning_rate": 1.8339517963665928e-05,
+      "loss": 0.8489,
+      "step": 8935
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7418346175608102,
+      "learning_rate": 1.8339096864873153e-05,
+      "loss": 0.9051,
+      "step": 8936
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7864495706506958,
+      "learning_rate": 1.833867571752752e-05,
+      "loss": 0.8989,
+      "step": 8937
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7048166728209442,
+      "learning_rate": 1.8338254521631476e-05,
+      "loss": 0.7768,
+      "step": 8938
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8077678594826286,
+      "learning_rate": 1.8337833277187472e-05,
+      "loss": 0.7493,
+      "step": 8939
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8043901225073179,
+      "learning_rate": 1.8337411984197965e-05,
+      "loss": 0.7527,
+      "step": 8940
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7997870059038739,
+      "learning_rate": 1.8336990642665405e-05,
+      "loss": 1.031,
+      "step": 8941
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5958194953055354,
+      "learning_rate": 1.8336569252592243e-05,
+      "loss": 0.9106,
+      "step": 8942
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6380795737867595,
+      "learning_rate": 1.8336147813980936e-05,
+      "loss": 1.012,
+      "step": 8943
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.188547631599865,
+      "learning_rate": 1.8335726326833943e-05,
+      "loss": 0.682,
+      "step": 8944
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.237214919646581,
+      "learning_rate": 1.8335304791153707e-05,
+      "loss": 0.9435,
+      "step": 8945
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5846014980323997,
+      "learning_rate": 1.833488320694269e-05,
+      "loss": 1.0795,
+      "step": 8946
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4964565796840752,
+      "learning_rate": 1.8334461574203346e-05,
+      "loss": 0.9606,
+      "step": 8947
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5451020309319128,
+      "learning_rate": 1.8334039892938126e-05,
+      "loss": 0.8975,
+      "step": 8948
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6905257090607522,
+      "learning_rate": 1.8333618163149485e-05,
+      "loss": 0.8967,
+      "step": 8949
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.330254369444004,
+      "learning_rate": 1.8333196384839884e-05,
+      "loss": 0.8832,
+      "step": 8950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5075192223197547,
+      "learning_rate": 1.833277455801177e-05,
+      "loss": 0.9934,
+      "step": 8951
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5845716736820115,
+      "learning_rate": 1.8332352682667613e-05,
+      "loss": 0.9785,
+      "step": 8952
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4778519357556354,
+      "learning_rate": 1.8331930758809856e-05,
+      "loss": 0.8536,
+      "step": 8953
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7815960888185314,
+      "learning_rate": 1.833150878644096e-05,
+      "loss": 0.9813,
+      "step": 8954
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7675346528550704,
+      "learning_rate": 1.8331086765563384e-05,
+      "loss": 0.8342,
+      "step": 8955
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6407028971994386,
+      "learning_rate": 1.833066469617958e-05,
+      "loss": 0.8806,
+      "step": 8956
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.022226069215082,
+      "learning_rate": 1.8330242578292013e-05,
+      "loss": 0.7012,
+      "step": 8957
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6388944722711203,
+      "learning_rate": 1.832982041190313e-05,
+      "loss": 0.9152,
+      "step": 8958
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6083407705849186,
+      "learning_rate": 1.83293981970154e-05,
+      "loss": 0.8508,
+      "step": 8959
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5615750583508783,
+      "learning_rate": 1.8328975933631277e-05,
+      "loss": 0.9592,
+      "step": 8960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.769917497094597,
+      "learning_rate": 1.832855362175322e-05,
+      "loss": 0.9181,
+      "step": 8961
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.263513615870314,
+      "learning_rate": 1.832813126138368e-05,
+      "loss": 0.8931,
+      "step": 8962
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.27424789348539,
+      "learning_rate": 1.8327708852525132e-05,
+      "loss": 0.8506,
+      "step": 8963
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6311002389096683,
+      "learning_rate": 1.832728639518002e-05,
+      "loss": 0.7731,
+      "step": 8964
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4812937231388004,
+      "learning_rate": 1.8326863889350816e-05,
+      "loss": 0.9036,
+      "step": 8965
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.558341868502518,
+      "learning_rate": 1.832644133503997e-05,
+      "loss": 0.9265,
+      "step": 8966
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4880241488533046,
+      "learning_rate": 1.832601873224995e-05,
+      "loss": 0.8969,
+      "step": 8967
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8476095666712633,
+      "learning_rate": 1.832559608098321e-05,
+      "loss": 0.9006,
+      "step": 8968
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5451959385244571,
+      "learning_rate": 1.8325173381242213e-05,
+      "loss": 0.8692,
+      "step": 8969
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7908373307517451,
+      "learning_rate": 1.8324750633029424e-05,
+      "loss": 0.9296,
+      "step": 8970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3079809354879024,
+      "learning_rate": 1.83243278363473e-05,
+      "loss": 0.9676,
+      "step": 8971
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.745036281441738,
+      "learning_rate": 1.8323904991198304e-05,
+      "loss": 0.9519,
+      "step": 8972
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8780092403637403,
+      "learning_rate": 1.8323482097584896e-05,
+      "loss": 1.0161,
+      "step": 8973
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7664669226843162,
+      "learning_rate": 1.8323059155509542e-05,
+      "loss": 0.961,
+      "step": 8974
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8412415108591045,
+      "learning_rate": 1.8322636164974705e-05,
+      "loss": 0.7747,
+      "step": 8975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7245252063781271,
+      "learning_rate": 1.832221312598284e-05,
+      "loss": 0.9096,
+      "step": 8976
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5239436476423138,
+      "learning_rate": 1.8321790038536415e-05,
+      "loss": 1.0164,
+      "step": 8977
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.821311669362772,
+      "learning_rate": 1.83213669026379e-05,
+      "loss": 0.9683,
+      "step": 8978
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.24838433757825,
+      "learning_rate": 1.8320943718289747e-05,
+      "loss": 0.9203,
+      "step": 8979
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4260838442685073,
+      "learning_rate": 1.832052048549443e-05,
+      "loss": 0.9097,
+      "step": 8980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9679313645398018,
+      "learning_rate": 1.8320097204254402e-05,
+      "loss": 0.9299,
+      "step": 8981
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.90571321079457,
+      "learning_rate": 1.831967387457214e-05,
+      "loss": 0.9866,
+      "step": 8982
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4430997902862244,
+      "learning_rate": 1.83192504964501e-05,
+      "loss": 0.8904,
+      "step": 8983
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7634784878442462,
+      "learning_rate": 1.831882706989075e-05,
+      "loss": 0.7119,
+      "step": 8984
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.922928172618884,
+      "learning_rate": 1.8318403594896557e-05,
+      "loss": 0.9006,
+      "step": 8985
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.753718135203128,
+      "learning_rate": 1.8317980071469982e-05,
+      "loss": 0.8999,
+      "step": 8986
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4339179531280613,
+      "learning_rate": 1.8317556499613492e-05,
+      "loss": 0.9951,
+      "step": 8987
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9523169235104056,
+      "learning_rate": 1.831713287932956e-05,
+      "loss": 0.956,
+      "step": 8988
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.962143557489,
+      "learning_rate": 1.8316709210620646e-05,
+      "loss": 0.9194,
+      "step": 8989
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.005005508869834,
+      "learning_rate": 1.8316285493489213e-05,
+      "loss": 0.9752,
+      "step": 8990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.501150913992381,
+      "learning_rate": 1.831586172793774e-05,
+      "loss": 0.912,
+      "step": 8991
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1474741515976694,
+      "learning_rate": 1.831543791396868e-05,
+      "loss": 0.8393,
+      "step": 8992
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7982405179870296,
+      "learning_rate": 1.831501405158451e-05,
+      "loss": 0.7428,
+      "step": 8993
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8324523358263245,
+      "learning_rate": 1.83145901407877e-05,
+      "loss": 0.9329,
+      "step": 8994
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.692143493231175,
+      "learning_rate": 1.831416618158071e-05,
+      "loss": 0.9432,
+      "step": 8995
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.680251325693409,
+      "learning_rate": 1.8313742173966017e-05,
+      "loss": 0.8342,
+      "step": 8996
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4828996981984308,
+      "learning_rate": 1.831331811794608e-05,
+      "loss": 0.8883,
+      "step": 8997
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.68321555747044,
+      "learning_rate": 1.831289401352337e-05,
+      "loss": 0.8591,
+      "step": 8998
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4819166628508296,
+      "learning_rate": 1.8312469860700366e-05,
+      "loss": 1.0183,
+      "step": 8999
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.152141828225651,
+      "learning_rate": 1.831204565947953e-05,
+      "loss": 0.8313,
+      "step": 9000
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8261591253944796,
+      "learning_rate": 1.831162140986333e-05,
+      "loss": 0.8851,
+      "step": 9001
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.649252746833095,
+      "learning_rate": 1.831119711185424e-05,
+      "loss": 0.8392,
+      "step": 9002
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8879228029376467,
+      "learning_rate": 1.831077276545473e-05,
+      "loss": 0.8987,
+      "step": 9003
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7536004764751971,
+      "learning_rate": 1.831034837066727e-05,
+      "loss": 0.8684,
+      "step": 9004
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6036128290618994,
+      "learning_rate": 1.830992392749433e-05,
+      "loss": 0.9612,
+      "step": 9005
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8447641265446815,
+      "learning_rate": 1.8309499435938377e-05,
+      "loss": 0.7956,
+      "step": 9006
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5293482517422032,
+      "learning_rate": 1.8309074896001893e-05,
+      "loss": 0.8661,
+      "step": 9007
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6363319822006919,
+      "learning_rate": 1.8308650307687347e-05,
+      "loss": 0.7568,
+      "step": 9008
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.523201095947987,
+      "learning_rate": 1.83082256709972e-05,
+      "loss": 0.7766,
+      "step": 9009
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4400648571139394,
+      "learning_rate": 1.8307800985933938e-05,
+      "loss": 0.9742,
+      "step": 9010
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.821259519061108,
+      "learning_rate": 1.830737625250003e-05,
+      "loss": 0.7601,
+      "step": 9011
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.096711030133602,
+      "learning_rate": 1.8306951470697946e-05,
+      "loss": 0.9376,
+      "step": 9012
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5621815426964805,
+      "learning_rate": 1.830652664053016e-05,
+      "loss": 0.8769,
+      "step": 9013
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.677237415915949,
+      "learning_rate": 1.8306101761999147e-05,
+      "loss": 0.9689,
+      "step": 9014
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7707331202581227,
+      "learning_rate": 1.8305676835107378e-05,
+      "loss": 0.879,
+      "step": 9015
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6734724163949106,
+      "learning_rate": 1.830525185985733e-05,
+      "loss": 0.8458,
+      "step": 9016
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.98047691664244,
+      "learning_rate": 1.8304826836251477e-05,
+      "loss": 0.8394,
+      "step": 9017
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.964158284569616,
+      "learning_rate": 1.8304401764292292e-05,
+      "loss": 0.9773,
+      "step": 9018
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.191981973542911,
+      "learning_rate": 1.830397664398225e-05,
+      "loss": 0.8364,
+      "step": 9019
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8100088297959782,
+      "learning_rate": 1.830355147532383e-05,
+      "loss": 0.9644,
+      "step": 9020
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.515376634036461,
+      "learning_rate": 1.83031262583195e-05,
+      "loss": 0.8184,
+      "step": 9021
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.429895466235139,
+      "learning_rate": 1.8302700992971744e-05,
+      "loss": 0.9022,
+      "step": 9022
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8614008309424699,
+      "learning_rate": 1.8302275679283033e-05,
+      "loss": 0.8693,
+      "step": 9023
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7062424101210136,
+      "learning_rate": 1.8301850317255843e-05,
+      "loss": 0.9009,
+      "step": 9024
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.042307380556395,
+      "learning_rate": 1.8301424906892653e-05,
+      "loss": 0.911,
+      "step": 9025
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4811534196993155,
+      "learning_rate": 1.830099944819594e-05,
+      "loss": 0.8415,
+      "step": 9026
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1910625439791476,
+      "learning_rate": 1.830057394116818e-05,
+      "loss": 0.9189,
+      "step": 9027
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7707347743160793,
+      "learning_rate": 1.8300148385811848e-05,
+      "loss": 0.7167,
+      "step": 9028
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7812329680383783,
+      "learning_rate": 1.8299722782129428e-05,
+      "loss": 1.0398,
+      "step": 9029
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6572999737057281,
+      "learning_rate": 1.8299297130123395e-05,
+      "loss": 0.9901,
+      "step": 9030
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6210784400438172,
+      "learning_rate": 1.8298871429796223e-05,
+      "loss": 0.9494,
+      "step": 9031
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4540392749926538,
+      "learning_rate": 1.8298445681150395e-05,
+      "loss": 0.9251,
+      "step": 9032
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9522891426109412,
+      "learning_rate": 1.829801988418839e-05,
+      "loss": 0.8748,
+      "step": 9033
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5664830104663579,
+      "learning_rate": 1.8297594038912686e-05,
+      "loss": 0.9636,
+      "step": 9034
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6806623703298633,
+      "learning_rate": 1.8297168145325762e-05,
+      "loss": 0.9353,
+      "step": 9035
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.613379221402358,
+      "learning_rate": 1.82967422034301e-05,
+      "loss": 0.9349,
+      "step": 9036
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.457466789630013,
+      "learning_rate": 1.8296316213228174e-05,
+      "loss": 0.8667,
+      "step": 9037
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5769103326296605,
+      "learning_rate": 1.829589017472247e-05,
+      "loss": 0.867,
+      "step": 9038
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8851376656587453,
+      "learning_rate": 1.829546408791547e-05,
+      "loss": 0.9476,
+      "step": 9039
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9752599391058392,
+      "learning_rate": 1.829503795280965e-05,
+      "loss": 1.0088,
+      "step": 9040
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7907339312998125,
+      "learning_rate": 1.8294611769407493e-05,
+      "loss": 0.8904,
+      "step": 9041
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.870493714482072,
+      "learning_rate": 1.829418553771148e-05,
+      "loss": 0.8213,
+      "step": 9042
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5598337170518453,
+      "learning_rate": 1.8293759257724095e-05,
+      "loss": 1.0146,
+      "step": 9043
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.619380114802591,
+      "learning_rate": 1.8293332929447817e-05,
+      "loss": 0.9519,
+      "step": 9044
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.06661167191593,
+      "learning_rate": 1.8292906552885122e-05,
+      "loss": 0.9557,
+      "step": 9045
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1481477388407955,
+      "learning_rate": 1.8292480128038507e-05,
+      "loss": 0.958,
+      "step": 9046
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6681281100619456,
+      "learning_rate": 1.8292053654910447e-05,
+      "loss": 0.9064,
+      "step": 9047
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6882795418654442,
+      "learning_rate": 1.8291627133503424e-05,
+      "loss": 0.8571,
+      "step": 9048
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.786131461176392,
+      "learning_rate": 1.829120056381992e-05,
+      "loss": 0.758,
+      "step": 9049
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6019631653128117,
+      "learning_rate": 1.8290773945862428e-05,
+      "loss": 1.0086,
+      "step": 9050
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5902553137381341,
+      "learning_rate": 1.829034727963342e-05,
+      "loss": 0.9241,
+      "step": 9051
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8223078631678711,
+      "learning_rate": 1.828992056513539e-05,
+      "loss": 0.7529,
+      "step": 9052
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2954413293436176,
+      "learning_rate": 1.8289493802370813e-05,
+      "loss": 0.8657,
+      "step": 9053
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7342902472057518,
+      "learning_rate": 1.8289066991342182e-05,
+      "loss": 0.909,
+      "step": 9054
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4828485172111125,
+      "learning_rate": 1.8288640132051975e-05,
+      "loss": 0.9296,
+      "step": 9055
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.097958833465737,
+      "learning_rate": 1.8288213224502685e-05,
+      "loss": 0.956,
+      "step": 9056
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.181934184005303,
+      "learning_rate": 1.828778626869679e-05,
+      "loss": 0.8618,
+      "step": 9057
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.374058716330083,
+      "learning_rate": 1.828735926463678e-05,
+      "loss": 0.9055,
+      "step": 9058
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4930414856234686,
+      "learning_rate": 1.8286932212325143e-05,
+      "loss": 0.8908,
+      "step": 9059
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6466246375446933,
+      "learning_rate": 1.828650511176436e-05,
+      "loss": 0.9673,
+      "step": 9060
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4257906838084264,
+      "learning_rate": 1.8286077962956926e-05,
+      "loss": 0.8627,
+      "step": 9061
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5516482752713296,
+      "learning_rate": 1.828565076590532e-05,
+      "loss": 0.8354,
+      "step": 9062
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.778987035739668,
+      "learning_rate": 1.828522352061203e-05,
+      "loss": 0.9398,
+      "step": 9063
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9332208591907494,
+      "learning_rate": 1.828479622707955e-05,
+      "loss": 0.9705,
+      "step": 9064
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.153449097418492,
+      "learning_rate": 1.828436888531036e-05,
+      "loss": 0.8209,
+      "step": 9065
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.712594532875397,
+      "learning_rate": 1.828394149530695e-05,
+      "loss": 0.8778,
+      "step": 9066
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6817105590393253,
+      "learning_rate": 1.8283514057071813e-05,
+      "loss": 0.8944,
+      "step": 9067
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.632697745369181,
+      "learning_rate": 1.8283086570607434e-05,
+      "loss": 0.9386,
+      "step": 9068
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7146857136291813,
+      "learning_rate": 1.8282659035916302e-05,
+      "loss": 0.8624,
+      "step": 9069
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.259233719865446,
+      "learning_rate": 1.828223145300091e-05,
+      "loss": 0.8977,
+      "step": 9070
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3246633660120257,
+      "learning_rate": 1.828180382186374e-05,
+      "loss": 1.0287,
+      "step": 9071
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6950605075630427,
+      "learning_rate": 1.8281376142507288e-05,
+      "loss": 0.9324,
+      "step": 9072
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6133446468992187,
+      "learning_rate": 1.8280948414934042e-05,
+      "loss": 0.8386,
+      "step": 9073
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.23385592475887,
+      "learning_rate": 1.828052063914649e-05,
+      "loss": 0.8465,
+      "step": 9074
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6745120131483233,
+      "learning_rate": 1.828009281514713e-05,
+      "loss": 0.7907,
+      "step": 9075
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5878002542355212,
+      "learning_rate": 1.8279664942938448e-05,
+      "loss": 0.9039,
+      "step": 9076
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5201994997982258,
+      "learning_rate": 1.8279237022522932e-05,
+      "loss": 0.8548,
+      "step": 9077
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7802892953382987,
+      "learning_rate": 1.8278809053903076e-05,
+      "loss": 0.933,
+      "step": 9078
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6186116367719106,
+      "learning_rate": 1.8278381037081377e-05,
+      "loss": 0.8941,
+      "step": 9079
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7421252160087395,
+      "learning_rate": 1.827795297206032e-05,
+      "loss": 0.831,
+      "step": 9080
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4618775620829922,
+      "learning_rate": 1.8277524858842404e-05,
+      "loss": 0.9342,
+      "step": 9081
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4576000550205463,
+      "learning_rate": 1.827709669743011e-05,
+      "loss": 0.9572,
+      "step": 9082
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4492054592078036,
+      "learning_rate": 1.8276668487825946e-05,
+      "loss": 0.9113,
+      "step": 9083
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.666699119954434,
+      "learning_rate": 1.8276240230032396e-05,
+      "loss": 0.8304,
+      "step": 9084
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6461343853744956,
+      "learning_rate": 1.8275811924051955e-05,
+      "loss": 0.8137,
+      "step": 9085
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6927443929149366,
+      "learning_rate": 1.8275383569887112e-05,
+      "loss": 0.8917,
+      "step": 9086
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8340391054134985,
+      "learning_rate": 1.827495516754037e-05,
+      "loss": 0.7711,
+      "step": 9087
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8235002388621786,
+      "learning_rate": 1.827452671701422e-05,
+      "loss": 0.7412,
+      "step": 9088
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.119056785013822,
+      "learning_rate": 1.8274098218311154e-05,
+      "loss": 1.06,
+      "step": 9089
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8920486129759915,
+      "learning_rate": 1.827366967143367e-05,
+      "loss": 0.8813,
+      "step": 9090
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.317944721419519,
+      "learning_rate": 1.8273241076384264e-05,
+      "loss": 0.8652,
+      "step": 9091
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7480477246323711,
+      "learning_rate": 1.8272812433165424e-05,
+      "loss": 0.8346,
+      "step": 9092
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7395937262301853,
+      "learning_rate": 1.8272383741779652e-05,
+      "loss": 0.9094,
+      "step": 9093
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9691394585553197,
+      "learning_rate": 1.8271955002229446e-05,
+      "loss": 0.8204,
+      "step": 9094
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6480385400795734,
+      "learning_rate": 1.8271526214517298e-05,
+      "loss": 0.9692,
+      "step": 9095
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.751349768425558,
+      "learning_rate": 1.8271097378645703e-05,
+      "loss": 0.9584,
+      "step": 9096
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.864268128571111,
+      "learning_rate": 1.827066849461716e-05,
+      "loss": 0.7704,
+      "step": 9097
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.999256490698664,
+      "learning_rate": 1.827023956243417e-05,
+      "loss": 0.8561,
+      "step": 9098
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5050186925782483,
+      "learning_rate": 1.826981058209923e-05,
+      "loss": 0.903,
+      "step": 9099
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7955503787656457,
+      "learning_rate": 1.826938155361483e-05,
+      "loss": 0.8863,
+      "step": 9100
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5873862211525733,
+      "learning_rate": 1.8268952476983473e-05,
+      "loss": 0.8417,
+      "step": 9101
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.830789419573194,
+      "learning_rate": 1.8268523352207656e-05,
+      "loss": 0.9236,
+      "step": 9102
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7032050076500915,
+      "learning_rate": 1.826809417928988e-05,
+      "loss": 0.851,
+      "step": 9103
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7593566136932963,
+      "learning_rate": 1.826766495823264e-05,
+      "loss": 0.9507,
+      "step": 9104
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9728599853872095,
+      "learning_rate": 1.826723568903844e-05,
+      "loss": 0.8708,
+      "step": 9105
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7373380618437189,
+      "learning_rate": 1.8266806371709773e-05,
+      "loss": 1.056,
+      "step": 9106
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7085943835472452,
+      "learning_rate": 1.8266377006249144e-05,
+      "loss": 0.9269,
+      "step": 9107
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6349969107652411,
+      "learning_rate": 1.826594759265905e-05,
+      "loss": 1.006,
+      "step": 9108
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7287646725087589,
+      "learning_rate": 1.8265518130941992e-05,
+      "loss": 0.9554,
+      "step": 9109
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9394674169616704,
+      "learning_rate": 1.8265088621100474e-05,
+      "loss": 0.9044,
+      "step": 9110
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7001207747956326,
+      "learning_rate": 1.8264659063136988e-05,
+      "loss": 0.9728,
+      "step": 9111
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.787353780861004,
+      "learning_rate": 1.8264229457054045e-05,
+      "loss": 0.7327,
+      "step": 9112
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5617765223554252,
+      "learning_rate": 1.826379980285414e-05,
+      "loss": 0.953,
+      "step": 9113
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8608666852280416,
+      "learning_rate": 1.826337010053978e-05,
+      "loss": 0.9191,
+      "step": 9114
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9329238033382665,
+      "learning_rate": 1.8262940350113456e-05,
+      "loss": 0.9116,
+      "step": 9115
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5574961144379629,
+      "learning_rate": 1.8262510551577678e-05,
+      "loss": 0.8027,
+      "step": 9116
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5813421140139428,
+      "learning_rate": 1.8262080704934955e-05,
+      "loss": 0.878,
+      "step": 9117
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7959419815608149,
+      "learning_rate": 1.8261650810187778e-05,
+      "loss": 0.749,
+      "step": 9118
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7678754374531538,
+      "learning_rate": 1.8261220867338653e-05,
+      "loss": 0.7987,
+      "step": 9119
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6392124180016219,
+      "learning_rate": 1.8260790876390085e-05,
+      "loss": 0.9597,
+      "step": 9120
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8640852889422017,
+      "learning_rate": 1.826036083734458e-05,
+      "loss": 0.8813,
+      "step": 9121
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5867602819978797,
+      "learning_rate": 1.8259930750204637e-05,
+      "loss": 0.993,
+      "step": 9122
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5203566747510822,
+      "learning_rate": 1.825950061497276e-05,
+      "loss": 0.8673,
+      "step": 9123
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4979850516429694,
+      "learning_rate": 1.8259070431651457e-05,
+      "loss": 1.0712,
+      "step": 9124
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4773313306541955,
+      "learning_rate": 1.8258640200243235e-05,
+      "loss": 0.7496,
+      "step": 9125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.718741121509633,
+      "learning_rate": 1.825820992075059e-05,
+      "loss": 0.9918,
+      "step": 9126
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0265268853752687,
+      "learning_rate": 1.8257779593176034e-05,
+      "loss": 0.9638,
+      "step": 9127
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5897447641704296,
+      "learning_rate": 1.8257349217522072e-05,
+      "loss": 0.8704,
+      "step": 9128
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9371182051477758,
+      "learning_rate": 1.8256918793791208e-05,
+      "loss": 0.9309,
+      "step": 9129
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6001074514439906,
+      "learning_rate": 1.825648832198595e-05,
+      "loss": 0.9238,
+      "step": 9130
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9156406864003177,
+      "learning_rate": 1.82560578021088e-05,
+      "loss": 0.9616,
+      "step": 9131
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.248142573618854,
+      "learning_rate": 1.8255627234162267e-05,
+      "loss": 0.8642,
+      "step": 9132
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6736281623720362,
+      "learning_rate": 1.8255196618148863e-05,
+      "loss": 0.9682,
+      "step": 9133
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5831337960968668,
+      "learning_rate": 1.825476595407109e-05,
+      "loss": 0.8722,
+      "step": 9134
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6141706699962126,
+      "learning_rate": 1.8254335241931453e-05,
+      "loss": 0.9633,
+      "step": 9135
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6715048353802695,
+      "learning_rate": 1.8253904481732464e-05,
+      "loss": 0.7908,
+      "step": 9136
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4688870663544324,
+      "learning_rate": 1.825347367347663e-05,
+      "loss": 0.9207,
+      "step": 9137
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1243947546057047,
+      "learning_rate": 1.8253042817166458e-05,
+      "loss": 0.7598,
+      "step": 9138
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8620064486103258,
+      "learning_rate": 1.825261191280446e-05,
+      "loss": 0.8935,
+      "step": 9139
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7868353503319068,
+      "learning_rate": 1.8252180960393145e-05,
+      "loss": 0.937,
+      "step": 9140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5283100970295516,
+      "learning_rate": 1.8251749959935017e-05,
+      "loss": 0.9431,
+      "step": 9141
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8157426370278148,
+      "learning_rate": 1.8251318911432586e-05,
+      "loss": 0.7897,
+      "step": 9142
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0673494809591904,
+      "learning_rate": 1.8250887814888368e-05,
+      "loss": 0.8757,
+      "step": 9143
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5455169207251305,
+      "learning_rate": 1.8250456670304866e-05,
+      "loss": 0.9395,
+      "step": 9144
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.514004562391708,
+      "learning_rate": 1.8250025477684598e-05,
+      "loss": 0.9256,
+      "step": 9145
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.638219531456722,
+      "learning_rate": 1.8249594237030064e-05,
+      "loss": 0.8436,
+      "step": 9146
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8040456804919762,
+      "learning_rate": 1.8249162948343784e-05,
+      "loss": 0.9204,
+      "step": 9147
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6577804998365235,
+      "learning_rate": 1.8248731611628266e-05,
+      "loss": 0.9071,
+      "step": 9148
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8747744038026446,
+      "learning_rate": 1.8248300226886018e-05,
+      "loss": 0.9445,
+      "step": 9149
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8174036479512246,
+      "learning_rate": 1.8247868794119556e-05,
+      "loss": 0.9138,
+      "step": 9150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7995626977676862,
+      "learning_rate": 1.8247437313331394e-05,
+      "loss": 0.9904,
+      "step": 9151
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8092602151825657,
+      "learning_rate": 1.8247005784524034e-05,
+      "loss": 0.7112,
+      "step": 9152
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7653604565187095,
+      "learning_rate": 1.82465742077e-05,
+      "loss": 0.6862,
+      "step": 9153
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4533066700301904,
+      "learning_rate": 1.8246142582861802e-05,
+      "loss": 0.9133,
+      "step": 9154
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.91765915472108,
+      "learning_rate": 1.8245710910011945e-05,
+      "loss": 0.9521,
+      "step": 9155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4591218984231586,
+      "learning_rate": 1.8245279189152952e-05,
+      "loss": 0.8906,
+      "step": 9156
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6307499469982627,
+      "learning_rate": 1.8244847420287334e-05,
+      "loss": 0.9437,
+      "step": 9157
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0933407404485815,
+      "learning_rate": 1.8244415603417603e-05,
+      "loss": 0.9715,
+      "step": 9158
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.55372300229867,
+      "learning_rate": 1.8243983738546276e-05,
+      "loss": 0.8731,
+      "step": 9159
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5947077405707093,
+      "learning_rate": 1.8243551825675864e-05,
+      "loss": 0.881,
+      "step": 9160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.781046854740911,
+      "learning_rate": 1.8243119864808886e-05,
+      "loss": 0.7674,
+      "step": 9161
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7964661644762827,
+      "learning_rate": 1.824268785594785e-05,
+      "loss": 0.86,
+      "step": 9162
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8844868473875838,
+      "learning_rate": 1.824225579909528e-05,
+      "loss": 0.9017,
+      "step": 9163
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5985698272641042,
+      "learning_rate": 1.824182369425368e-05,
+      "loss": 0.9373,
+      "step": 9164
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5582231780789437,
+      "learning_rate": 1.824139154142558e-05,
+      "loss": 0.8335,
+      "step": 9165
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7920916735307844,
+      "learning_rate": 1.8240959340613488e-05,
+      "loss": 0.9113,
+      "step": 9166
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.749848832416412,
+      "learning_rate": 1.8240527091819923e-05,
+      "loss": 0.892,
+      "step": 9167
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.639268835422674,
+      "learning_rate": 1.82400947950474e-05,
+      "loss": 1.0419,
+      "step": 9168
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.00244118264291,
+      "learning_rate": 1.8239662450298433e-05,
+      "loss": 0.9947,
+      "step": 9169
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.70511906084891,
+      "learning_rate": 1.8239230057575542e-05,
+      "loss": 0.8727,
+      "step": 9170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7426228679889786,
+      "learning_rate": 1.823879761688125e-05,
+      "loss": 1.1311,
+      "step": 9171
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9537435754890715,
+      "learning_rate": 1.8238365128218066e-05,
+      "loss": 0.736,
+      "step": 9172
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7265274731842237,
+      "learning_rate": 1.8237932591588513e-05,
+      "loss": 0.8713,
+      "step": 9173
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6716709533465297,
+      "learning_rate": 1.823750000699511e-05,
+      "loss": 0.9949,
+      "step": 9174
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.590769771330842,
+      "learning_rate": 1.8237067374440372e-05,
+      "loss": 0.8907,
+      "step": 9175
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4882473315004283,
+      "learning_rate": 1.823663469392682e-05,
+      "loss": 0.9291,
+      "step": 9176
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.054209736555574,
+      "learning_rate": 1.8236201965456975e-05,
+      "loss": 0.8646,
+      "step": 9177
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3180454605106764,
+      "learning_rate": 1.8235769189033354e-05,
+      "loss": 0.947,
+      "step": 9178
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7451588427778681,
+      "learning_rate": 1.8235336364658475e-05,
+      "loss": 0.9334,
+      "step": 9179
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8660263125713168,
+      "learning_rate": 1.8234903492334864e-05,
+      "loss": 0.9334,
+      "step": 9180
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5859626618241134,
+      "learning_rate": 1.8234470572065034e-05,
+      "loss": 0.9007,
+      "step": 9181
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5713739618736657,
+      "learning_rate": 1.8234037603851512e-05,
+      "loss": 0.8743,
+      "step": 9182
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3254312291838435,
+      "learning_rate": 1.8233604587696818e-05,
+      "loss": 0.9388,
+      "step": 9183
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5289783503849013,
+      "learning_rate": 1.823317152360347e-05,
+      "loss": 0.9875,
+      "step": 9184
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0608341401263957,
+      "learning_rate": 1.823273841157399e-05,
+      "loss": 0.8405,
+      "step": 9185
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6322835347833609,
+      "learning_rate": 1.82323052516109e-05,
+      "loss": 0.8988,
+      "step": 9186
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7083920039368388,
+      "learning_rate": 1.823187204371672e-05,
+      "loss": 0.9996,
+      "step": 9187
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.580340745111553,
+      "learning_rate": 1.8231438787893978e-05,
+      "loss": 0.863,
+      "step": 9188
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6213467571798459,
+      "learning_rate": 1.8231005484145194e-05,
+      "loss": 0.8691,
+      "step": 9189
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.625066732172763,
+      "learning_rate": 1.8230572132472888e-05,
+      "loss": 1.0248,
+      "step": 9190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7996231945625191,
+      "learning_rate": 1.8230138732879588e-05,
+      "loss": 0.7428,
+      "step": 9191
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8384252046711447,
+      "learning_rate": 1.8229705285367812e-05,
+      "loss": 0.9301,
+      "step": 9192
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6113705911964755,
+      "learning_rate": 1.822927178994009e-05,
+      "loss": 0.8044,
+      "step": 9193
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6410304339164508,
+      "learning_rate": 1.822883824659894e-05,
+      "loss": 0.794,
+      "step": 9194
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6830415144831918,
+      "learning_rate": 1.8228404655346885e-05,
+      "loss": 0.8734,
+      "step": 9195
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2954122538046975,
+      "learning_rate": 1.8227971016186457e-05,
+      "loss": 0.8655,
+      "step": 9196
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3698158329546346,
+      "learning_rate": 1.8227537329120177e-05,
+      "loss": 0.861,
+      "step": 9197
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6806638844999489,
+      "learning_rate": 1.8227103594150568e-05,
+      "loss": 0.9682,
+      "step": 9198
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.769548553814819,
+      "learning_rate": 1.822666981128016e-05,
+      "loss": 0.8006,
+      "step": 9199
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8488409797308765,
+      "learning_rate": 1.8226235980511476e-05,
+      "loss": 0.8577,
+      "step": 9200
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2359649392312817,
+      "learning_rate": 1.8225802101847038e-05,
+      "loss": 0.9453,
+      "step": 9201
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4838930232467828,
+      "learning_rate": 1.8225368175289378e-05,
+      "loss": 0.9679,
+      "step": 9202
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7521252062386914,
+      "learning_rate": 1.822493420084102e-05,
+      "loss": 0.9688,
+      "step": 9203
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4499941428852443,
+      "learning_rate": 1.822450017850449e-05,
+      "loss": 0.8438,
+      "step": 9204
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6427229847425318,
+      "learning_rate": 1.8224066108282318e-05,
+      "loss": 0.8797,
+      "step": 9205
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5829429873709226,
+      "learning_rate": 1.8223631990177027e-05,
+      "loss": 0.8591,
+      "step": 9206
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6084122200107849,
+      "learning_rate": 1.822319782419115e-05,
+      "loss": 0.7885,
+      "step": 9207
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0907375504808745,
+      "learning_rate": 1.822276361032721e-05,
+      "loss": 0.9134,
+      "step": 9208
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6190911302972406,
+      "learning_rate": 1.822232934858774e-05,
+      "loss": 0.8804,
+      "step": 9209
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.764252755824297,
+      "learning_rate": 1.822189503897526e-05,
+      "loss": 0.7697,
+      "step": 9210
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3955772692011306,
+      "learning_rate": 1.822146068149231e-05,
+      "loss": 0.8339,
+      "step": 9211
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5406882618598345,
+      "learning_rate": 1.822102627614141e-05,
+      "loss": 0.9355,
+      "step": 9212
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7630823689881185,
+      "learning_rate": 1.822059182292509e-05,
+      "loss": 0.6915,
+      "step": 9213
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5571524126566936,
+      "learning_rate": 1.8220157321845887e-05,
+      "loss": 1.0378,
+      "step": 9214
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.682934420290578,
+      "learning_rate": 1.821972277290632e-05,
+      "loss": 0.9619,
+      "step": 9215
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2822507218552959,
+      "learning_rate": 1.8219288176108928e-05,
+      "loss": 0.848,
+      "step": 9216
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9875476203490021,
+      "learning_rate": 1.821885353145624e-05,
+      "loss": 0.8271,
+      "step": 9217
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8772297181472082,
+      "learning_rate": 1.821841883895078e-05,
+      "loss": 1.0519,
+      "step": 9218
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5084044474344702,
+      "learning_rate": 1.8217984098595086e-05,
+      "loss": 0.8949,
+      "step": 9219
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7330186842436102,
+      "learning_rate": 1.8217549310391687e-05,
+      "loss": 0.8924,
+      "step": 9220
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4626267853400305,
+      "learning_rate": 1.8217114474343113e-05,
+      "loss": 1.0275,
+      "step": 9221
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.626988529648551,
+      "learning_rate": 1.82166795904519e-05,
+      "loss": 0.8873,
+      "step": 9222
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1072128147761506,
+      "learning_rate": 1.8216244658720573e-05,
+      "loss": 0.9808,
+      "step": 9223
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7548663659192787,
+      "learning_rate": 1.821580967915167e-05,
+      "loss": 0.901,
+      "step": 9224
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6495528746767145,
+      "learning_rate": 1.8215374651747723e-05,
+      "loss": 0.8557,
+      "step": 9225
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2973136196180772,
+      "learning_rate": 1.8214939576511265e-05,
+      "loss": 0.8876,
+      "step": 9226
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5653645825546025,
+      "learning_rate": 1.8214504453444825e-05,
+      "loss": 0.8051,
+      "step": 9227
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.557260943387838,
+      "learning_rate": 1.821406928255094e-05,
+      "loss": 1.0936,
+      "step": 9228
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7469449063900244,
+      "learning_rate": 1.8213634063832146e-05,
+      "loss": 0.9897,
+      "step": 9229
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7735992639370545,
+      "learning_rate": 1.8213198797290973e-05,
+      "loss": 0.7288,
+      "step": 9230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7375457318732495,
+      "learning_rate": 1.821276348292995e-05,
+      "loss": 0.6883,
+      "step": 9231
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5875213451605161,
+      "learning_rate": 1.8212328120751626e-05,
+      "loss": 1.0144,
+      "step": 9232
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5896191230341359,
+      "learning_rate": 1.8211892710758525e-05,
+      "loss": 0.8997,
+      "step": 9233
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8216435918322875,
+      "learning_rate": 1.8211457252953185e-05,
+      "loss": 0.8756,
+      "step": 9234
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5673863727315238,
+      "learning_rate": 1.821102174733814e-05,
+      "loss": 0.9951,
+      "step": 9235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.656729559923906,
+      "learning_rate": 1.8210586193915928e-05,
+      "loss": 0.8944,
+      "step": 9236
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7591943346657544,
+      "learning_rate": 1.8210150592689083e-05,
+      "loss": 0.7595,
+      "step": 9237
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6456251212606425,
+      "learning_rate": 1.8209714943660143e-05,
+      "loss": 0.9255,
+      "step": 9238
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4156440454506067,
+      "learning_rate": 1.8209279246831643e-05,
+      "loss": 0.8867,
+      "step": 9239
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1917618886713366,
+      "learning_rate": 1.8208843502206118e-05,
+      "loss": 0.8847,
+      "step": 9240
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.255405752752306,
+      "learning_rate": 1.8208407709786107e-05,
+      "loss": 0.9532,
+      "step": 9241
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5537438328840378,
+      "learning_rate": 1.8207971869574154e-05,
+      "loss": 0.9695,
+      "step": 9242
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5141442366133429,
+      "learning_rate": 1.8207535981572785e-05,
+      "loss": 0.9302,
+      "step": 9243
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8102453761779027,
+      "learning_rate": 1.8207100045784544e-05,
+      "loss": 0.8788,
+      "step": 9244
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7364995696652494,
+      "learning_rate": 1.8206664062211966e-05,
+      "loss": 0.959,
+      "step": 9245
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.061873025086393,
+      "learning_rate": 1.8206228030857595e-05,
+      "loss": 0.7699,
+      "step": 9246
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2427495338358634,
+      "learning_rate": 1.8205791951723966e-05,
+      "loss": 1.0133,
+      "step": 9247
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.637725899171308,
+      "learning_rate": 1.8205355824813617e-05,
+      "loss": 0.9858,
+      "step": 9248
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7764725373059957,
+      "learning_rate": 1.8204919650129088e-05,
+      "loss": 0.898,
+      "step": 9249
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7725454936698826,
+      "learning_rate": 1.8204483427672924e-05,
+      "loss": 0.8415,
+      "step": 9250
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6052354373494164,
+      "learning_rate": 1.8204047157447652e-05,
+      "loss": 0.8003,
+      "step": 9251
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7223285854521584,
+      "learning_rate": 1.8203610839455828e-05,
+      "loss": 0.9823,
+      "step": 9252
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4046911899789065,
+      "learning_rate": 1.820317447369998e-05,
+      "loss": 0.8685,
+      "step": 9253
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7571585287269161,
+      "learning_rate": 1.8202738060182655e-05,
+      "loss": 0.7631,
+      "step": 9254
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8346399448831918,
+      "learning_rate": 1.8202301598906388e-05,
+      "loss": 0.928,
+      "step": 9255
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7722156924451805,
+      "learning_rate": 1.820186508987373e-05,
+      "loss": 0.7344,
+      "step": 9256
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0662286085513415,
+      "learning_rate": 1.8201428533087214e-05,
+      "loss": 0.9091,
+      "step": 9257
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4434455014647045,
+      "learning_rate": 1.820099192854938e-05,
+      "loss": 0.9882,
+      "step": 9258
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6684963840854017,
+      "learning_rate": 1.8200555276262784e-05,
+      "loss": 1.0022,
+      "step": 9259
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7547783016653118,
+      "learning_rate": 1.8200118576229953e-05,
+      "loss": 0.9332,
+      "step": 9260
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8861526948425387,
+      "learning_rate": 1.8199681828453438e-05,
+      "loss": 1.0515,
+      "step": 9261
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.617815697327686,
+      "learning_rate": 1.8199245032935775e-05,
+      "loss": 0.8297,
+      "step": 9262
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.675620457193548,
+      "learning_rate": 1.8198808189679518e-05,
+      "loss": 1.0173,
+      "step": 9263
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.496854453297212,
+      "learning_rate": 1.81983712986872e-05,
+      "loss": 0.9366,
+      "step": 9264
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4429698127190413,
+      "learning_rate": 1.819793435996137e-05,
+      "loss": 0.8982,
+      "step": 9265
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.169876449025356,
+      "learning_rate": 1.819749737350457e-05,
+      "loss": 0.8962,
+      "step": 9266
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7558942568893757,
+      "learning_rate": 1.8197060339319345e-05,
+      "loss": 0.7775,
+      "step": 9267
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9849462636238262,
+      "learning_rate": 1.8196623257408238e-05,
+      "loss": 0.9421,
+      "step": 9268
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.618172567578441,
+      "learning_rate": 1.8196186127773797e-05,
+      "loss": 0.9568,
+      "step": 9269
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.459708627117925,
+      "learning_rate": 1.8195748950418567e-05,
+      "loss": 0.9783,
+      "step": 9270
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7423863443478986,
+      "learning_rate": 1.819531172534509e-05,
+      "loss": 0.9391,
+      "step": 9271
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.091243435074845,
+      "learning_rate": 1.8194874452555914e-05,
+      "loss": 1.0691,
+      "step": 9272
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7566653579708411,
+      "learning_rate": 1.8194437132053586e-05,
+      "loss": 0.7283,
+      "step": 9273
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7795655677598526,
+      "learning_rate": 1.8193999763840647e-05,
+      "loss": 0.7617,
+      "step": 9274
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6862839148263369,
+      "learning_rate": 1.819356234791965e-05,
+      "loss": 1.0391,
+      "step": 9275
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9643520427467034,
+      "learning_rate": 1.819312488429314e-05,
+      "loss": 0.9045,
+      "step": 9276
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.108679991042721,
+      "learning_rate": 1.819268737296366e-05,
+      "loss": 0.8632,
+      "step": 9277
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9010958229646933,
+      "learning_rate": 1.8192249813933762e-05,
+      "loss": 0.9384,
+      "step": 9278
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.532382293298193,
+      "learning_rate": 1.819181220720599e-05,
+      "loss": 0.8279,
+      "step": 9279
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6353792768688404,
+      "learning_rate": 1.81913745527829e-05,
+      "loss": 0.8929,
+      "step": 9280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6106443741493077,
+      "learning_rate": 1.8190936850667028e-05,
+      "loss": 1.0236,
+      "step": 9281
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9455054172193653,
+      "learning_rate": 1.819049910086093e-05,
+      "loss": 0.8725,
+      "step": 9282
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8026973044419438,
+      "learning_rate": 1.819006130336715e-05,
+      "loss": 0.7643,
+      "step": 9283
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8647628142059733,
+      "learning_rate": 1.818962345818825e-05,
+      "loss": 0.9856,
+      "step": 9284
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5483360380705202,
+      "learning_rate": 1.818918556532676e-05,
+      "loss": 0.9503,
+      "step": 9285
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8597621971469904,
+      "learning_rate": 1.818874762478524e-05,
+      "loss": 0.8872,
+      "step": 9286
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.597195517848466,
+      "learning_rate": 1.818830963656624e-05,
+      "loss": 0.9233,
+      "step": 9287
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6254003625584428,
+      "learning_rate": 1.818787160067231e-05,
+      "loss": 0.8403,
+      "step": 9288
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7062667894534869,
+      "learning_rate": 1.8187433517106e-05,
+      "loss": 0.8494,
+      "step": 9289
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7286771538862074,
+      "learning_rate": 1.8186995385869857e-05,
+      "loss": 0.7904,
+      "step": 9290
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.35543441688131,
+      "learning_rate": 1.818655720696644e-05,
+      "loss": 0.9227,
+      "step": 9291
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8229180165429685,
+      "learning_rate": 1.818611898039829e-05,
+      "loss": 0.7424,
+      "step": 9292
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.865054670572262,
+      "learning_rate": 1.8185680706167966e-05,
+      "loss": 0.8194,
+      "step": 9293
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.058311918547822,
+      "learning_rate": 1.8185242384278016e-05,
+      "loss": 0.9412,
+      "step": 9294
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2609639698558786,
+      "learning_rate": 1.8184804014730994e-05,
+      "loss": 0.8668,
+      "step": 9295
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5893879240541349,
+      "learning_rate": 1.8184365597529454e-05,
+      "loss": 0.8971,
+      "step": 9296
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5469790007764748,
+      "learning_rate": 1.8183927132675945e-05,
+      "loss": 0.9289,
+      "step": 9297
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5719316272911228,
+      "learning_rate": 1.818348862017302e-05,
+      "loss": 1.0299,
+      "step": 9298
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7952466052943599,
+      "learning_rate": 1.8183050060023233e-05,
+      "loss": 0.8974,
+      "step": 9299
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.705527363088517,
+      "learning_rate": 1.818261145222914e-05,
+      "loss": 1.0,
+      "step": 9300
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.79667589529934,
+      "learning_rate": 1.8182172796793292e-05,
+      "loss": 0.9379,
+      "step": 9301
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.743357509059103,
+      "learning_rate": 1.8181734093718245e-05,
+      "loss": 0.8734,
+      "step": 9302
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5587404526463984,
+      "learning_rate": 1.818129534300655e-05,
+      "loss": 0.9672,
+      "step": 9303
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7324870811870468,
+      "learning_rate": 1.8180856544660763e-05,
+      "loss": 0.9311,
+      "step": 9304
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5857189340489248,
+      "learning_rate": 1.8180417698683437e-05,
+      "loss": 0.9716,
+      "step": 9305
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7778299242302882,
+      "learning_rate": 1.817997880507713e-05,
+      "loss": 0.8847,
+      "step": 9306
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.553214597103997,
+      "learning_rate": 1.81795398638444e-05,
+      "loss": 0.9546,
+      "step": 9307
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5377790002875242,
+      "learning_rate": 1.81791008749878e-05,
+      "loss": 0.9987,
+      "step": 9308
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4828928994567974,
+      "learning_rate": 1.817866183850988e-05,
+      "loss": 0.9762,
+      "step": 9309
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.20749393531671,
+      "learning_rate": 1.8178222754413206e-05,
+      "loss": 0.9794,
+      "step": 9310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6072631220304854,
+      "learning_rate": 1.8177783622700328e-05,
+      "loss": 0.8442,
+      "step": 9311
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7840434636648574,
+      "learning_rate": 1.8177344443373806e-05,
+      "loss": 0.9572,
+      "step": 9312
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7579401068908377,
+      "learning_rate": 1.8176905216436192e-05,
+      "loss": 0.7269,
+      "step": 9313
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8067083846253041,
+      "learning_rate": 1.817646594189005e-05,
+      "loss": 0.7696,
+      "step": 9314
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.894030785561376,
+      "learning_rate": 1.8176026619737935e-05,
+      "loss": 0.9453,
+      "step": 9315
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.203628902590722,
+      "learning_rate": 1.8175587249982405e-05,
+      "loss": 1.0296,
+      "step": 9316
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6747872164707502,
+      "learning_rate": 1.8175147832626018e-05,
+      "loss": 0.9508,
+      "step": 9317
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4985338102215326,
+      "learning_rate": 1.8174708367671326e-05,
+      "loss": 0.8756,
+      "step": 9318
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.663104316733908,
+      "learning_rate": 1.81742688551209e-05,
+      "loss": 0.8681,
+      "step": 9319
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5801946837910013,
+      "learning_rate": 1.8173829294977294e-05,
+      "loss": 0.8001,
+      "step": 9320
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6469474268755615,
+      "learning_rate": 1.8173389687243063e-05,
+      "loss": 1.0519,
+      "step": 9321
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7300729343138501,
+      "learning_rate": 1.8172950031920772e-05,
+      "loss": 0.9368,
+      "step": 9322
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5470226269250513,
+      "learning_rate": 1.8172510329012974e-05,
+      "loss": 0.8495,
+      "step": 9323
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.349253367067115,
+      "learning_rate": 1.8172070578522235e-05,
+      "loss": 0.8323,
+      "step": 9324
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7946114490074028,
+      "learning_rate": 1.8171630780451115e-05,
+      "loss": 0.9522,
+      "step": 9325
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.777119094591636,
+      "learning_rate": 1.8171190934802175e-05,
+      "loss": 0.9415,
+      "step": 9326
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8832746158345635,
+      "learning_rate": 1.8170751041577972e-05,
+      "loss": 0.8988,
+      "step": 9327
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8580735616975187,
+      "learning_rate": 1.817031110078107e-05,
+      "loss": 0.7341,
+      "step": 9328
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8678486747438858,
+      "learning_rate": 1.816987111241403e-05,
+      "loss": 0.9277,
+      "step": 9329
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7976242236867823,
+      "learning_rate": 1.8169431076479415e-05,
+      "loss": 0.9463,
+      "step": 9330
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7881021623575881,
+      "learning_rate": 1.8168990992979785e-05,
+      "loss": 0.8997,
+      "step": 9331
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0204588991956265,
+      "learning_rate": 1.8168550861917702e-05,
+      "loss": 0.8671,
+      "step": 9332
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.198120991710964,
+      "learning_rate": 1.8168110683295733e-05,
+      "loss": 0.9821,
+      "step": 9333
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0511460408686735,
+      "learning_rate": 1.8167670457116434e-05,
+      "loss": 0.8344,
+      "step": 9334
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7487656650306703,
+      "learning_rate": 1.8167230183382375e-05,
+      "loss": 0.7136,
+      "step": 9335
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0838335075017684,
+      "learning_rate": 1.8166789862096114e-05,
+      "loss": 0.947,
+      "step": 9336
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7271944315696772,
+      "learning_rate": 1.8166349493260218e-05,
+      "loss": 0.9096,
+      "step": 9337
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9470048324925215,
+      "learning_rate": 1.816590907687725e-05,
+      "loss": 0.9117,
+      "step": 9338
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5820107385658206,
+      "learning_rate": 1.816546861294977e-05,
+      "loss": 0.884,
+      "step": 9339
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7062659324061358,
+      "learning_rate": 1.8165028101480354e-05,
+      "loss": 1.0458,
+      "step": 9340
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8160571049896537,
+      "learning_rate": 1.8164587542471555e-05,
+      "loss": 0.9397,
+      "step": 9341
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7000561079934147,
+      "learning_rate": 1.816414693592594e-05,
+      "loss": 0.9586,
+      "step": 9342
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2046404537241684,
+      "learning_rate": 1.816370628184608e-05,
+      "loss": 0.8286,
+      "step": 9343
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4942984158357084,
+      "learning_rate": 1.816326558023454e-05,
+      "loss": 0.8381,
+      "step": 9344
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.618443756073384,
+      "learning_rate": 1.816282483109388e-05,
+      "loss": 0.9941,
+      "step": 9345
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7136633914670294,
+      "learning_rate": 1.816238403442667e-05,
+      "loss": 0.9243,
+      "step": 9346
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.640325976373535,
+      "learning_rate": 1.8161943190235476e-05,
+      "loss": 0.9202,
+      "step": 9347
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7984546168387102,
+      "learning_rate": 1.8161502298522865e-05,
+      "loss": 0.6869,
+      "step": 9348
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6280783529133722,
+      "learning_rate": 1.81610613592914e-05,
+      "loss": 1.0291,
+      "step": 9349
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.297481137069097,
+      "learning_rate": 1.816062037254366e-05,
+      "loss": 0.878,
+      "step": 9350
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7320014213490829,
+      "learning_rate": 1.8160179338282196e-05,
+      "loss": 0.7365,
+      "step": 9351
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6759612603570369,
+      "learning_rate": 1.815973825650959e-05,
+      "loss": 0.8524,
+      "step": 9352
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5882703506273537,
+      "learning_rate": 1.8159297127228403e-05,
+      "loss": 0.9999,
+      "step": 9353
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5543799348072027,
+      "learning_rate": 1.8158855950441206e-05,
+      "loss": 1.1072,
+      "step": 9354
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9627499359898308,
+      "learning_rate": 1.8158414726150565e-05,
+      "loss": 0.8789,
+      "step": 9355
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9596359278985425,
+      "learning_rate": 1.815797345435905e-05,
+      "loss": 0.9173,
+      "step": 9356
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.693185862661194,
+      "learning_rate": 1.815753213506923e-05,
+      "loss": 0.9612,
+      "step": 9357
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7959927510331867,
+      "learning_rate": 1.815709076828368e-05,
+      "loss": 0.7135,
+      "step": 9358
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7833937398344502,
+      "learning_rate": 1.815664935400496e-05,
+      "loss": 0.9287,
+      "step": 9359
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.572265297211441,
+      "learning_rate": 1.815620789223565e-05,
+      "loss": 0.9328,
+      "step": 9360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4870541212240254,
+      "learning_rate": 1.8155766382978314e-05,
+      "loss": 0.8093,
+      "step": 9361
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7112640323810382,
+      "learning_rate": 1.815532482623552e-05,
+      "loss": 1.0894,
+      "step": 9362
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0007050439324128,
+      "learning_rate": 1.8154883222009844e-05,
+      "loss": 0.9283,
+      "step": 9363
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.750595391692833,
+      "learning_rate": 1.815444157030386e-05,
+      "loss": 0.8977,
+      "step": 9364
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.193231673965686,
+      "learning_rate": 1.8153999871120135e-05,
+      "loss": 0.7699,
+      "step": 9365
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9491339897400906,
+      "learning_rate": 1.8153558124461236e-05,
+      "loss": 0.9149,
+      "step": 9366
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7931164681700046,
+      "learning_rate": 1.8153116330329745e-05,
+      "loss": 0.7725,
+      "step": 9367
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5210451925276305,
+      "learning_rate": 1.815267448872823e-05,
+      "loss": 0.8796,
+      "step": 9368
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.880800361205127,
+      "learning_rate": 1.815223259965926e-05,
+      "loss": 1.0427,
+      "step": 9369
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1071493754944277,
+      "learning_rate": 1.8151790663125412e-05,
+      "loss": 1.0717,
+      "step": 9370
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8946227720271769,
+      "learning_rate": 1.8151348679129257e-05,
+      "loss": 0.8687,
+      "step": 9371
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4259007222408298,
+      "learning_rate": 1.815090664767337e-05,
+      "loss": 0.9452,
+      "step": 9372
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0180809523556404,
+      "learning_rate": 1.8150464568760327e-05,
+      "loss": 0.9211,
+      "step": 9373
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1590637406824023,
+      "learning_rate": 1.8150022442392695e-05,
+      "loss": 1.0148,
+      "step": 9374
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.801442530973802,
+      "learning_rate": 1.814958026857305e-05,
+      "loss": 0.8805,
+      "step": 9375
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5411803089708889,
+      "learning_rate": 1.8149138047303977e-05,
+      "loss": 0.9414,
+      "step": 9376
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5510684411874305,
+      "learning_rate": 1.8148695778588034e-05,
+      "loss": 0.8308,
+      "step": 9377
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7969366005388423,
+      "learning_rate": 1.814825346242781e-05,
+      "loss": 0.8789,
+      "step": 9378
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6033546419798264,
+      "learning_rate": 1.814781109882587e-05,
+      "loss": 0.986,
+      "step": 9379
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.793828003689812,
+      "learning_rate": 1.8147368687784798e-05,
+      "loss": 1.083,
+      "step": 9380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7523046713314743,
+      "learning_rate": 1.8146926229307166e-05,
+      "loss": 0.9278,
+      "step": 9381
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5746003343231059,
+      "learning_rate": 1.814648372339555e-05,
+      "loss": 0.8861,
+      "step": 9382
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1603479275410096,
+      "learning_rate": 1.8146041170052524e-05,
+      "loss": 0.9398,
+      "step": 9383
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5396922465734268,
+      "learning_rate": 1.814559856928067e-05,
+      "loss": 0.8032,
+      "step": 9384
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.269142575570653,
+      "learning_rate": 1.814515592108256e-05,
+      "loss": 0.8715,
+      "step": 9385
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7724430271332754,
+      "learning_rate": 1.814471322546077e-05,
+      "loss": 0.9202,
+      "step": 9386
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7831188356749452,
+      "learning_rate": 1.8144270482417887e-05,
+      "loss": 0.8997,
+      "step": 9387
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7592410457680958,
+      "learning_rate": 1.8143827691956483e-05,
+      "loss": 0.9909,
+      "step": 9388
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6315816068837679,
+      "learning_rate": 1.8143384854079132e-05,
+      "loss": 0.8855,
+      "step": 9389
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1347520132025677,
+      "learning_rate": 1.8142941968788422e-05,
+      "loss": 0.815,
+      "step": 9390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.595806522322404,
+      "learning_rate": 1.8142499036086922e-05,
+      "loss": 0.9509,
+      "step": 9391
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7554412439810703,
+      "learning_rate": 1.8142056055977212e-05,
+      "loss": 1.0255,
+      "step": 9392
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.403543299734387,
+      "learning_rate": 1.8141613028461877e-05,
+      "loss": 0.9713,
+      "step": 9393
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8049070151523614,
+      "learning_rate": 1.8141169953543495e-05,
+      "loss": 0.8983,
+      "step": 9394
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5505099401554785,
+      "learning_rate": 1.8140726831224644e-05,
+      "loss": 0.8672,
+      "step": 9395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.771450694654712,
+      "learning_rate": 1.8140283661507904e-05,
+      "loss": 0.8923,
+      "step": 9396
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6282527333246428,
+      "learning_rate": 1.813984044439585e-05,
+      "loss": 0.8895,
+      "step": 9397
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7278693304396289,
+      "learning_rate": 1.8139397179891075e-05,
+      "loss": 0.9722,
+      "step": 9398
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.713518061885266,
+      "learning_rate": 1.8138953867996147e-05,
+      "loss": 0.968,
+      "step": 9399
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.848888199534125,
+      "learning_rate": 1.8138510508713656e-05,
+      "loss": 0.8779,
+      "step": 9400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5269184259302775,
+      "learning_rate": 1.813806710204618e-05,
+      "loss": 0.9456,
+      "step": 9401
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.853213321348467,
+      "learning_rate": 1.81376236479963e-05,
+      "loss": 0.9823,
+      "step": 9402
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7232880603897773,
+      "learning_rate": 1.81371801465666e-05,
+      "loss": 0.8902,
+      "step": 9403
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.957344664137291,
+      "learning_rate": 1.813673659775966e-05,
+      "loss": 0.8903,
+      "step": 9404
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8168174977017388,
+      "learning_rate": 1.813629300157806e-05,
+      "loss": 0.8918,
+      "step": 9405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9564770460923084,
+      "learning_rate": 1.8135849358024392e-05,
+      "loss": 0.9459,
+      "step": 9406
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.565788441910842,
+      "learning_rate": 1.813540566710123e-05,
+      "loss": 0.9375,
+      "step": 9407
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8376050265205749,
+      "learning_rate": 1.8134961928811164e-05,
+      "loss": 0.7741,
+      "step": 9408
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6903597944784783,
+      "learning_rate": 1.813451814315677e-05,
+      "loss": 1.0021,
+      "step": 9409
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6403896139512815,
+      "learning_rate": 1.8134074310140638e-05,
+      "loss": 0.9933,
+      "step": 9410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6075258318979664,
+      "learning_rate": 1.813363042976535e-05,
+      "loss": 0.8362,
+      "step": 9411
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.86707356860509,
+      "learning_rate": 1.8133186502033493e-05,
+      "loss": 1.0599,
+      "step": 9412
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.834652932055509,
+      "learning_rate": 1.8132742526947648e-05,
+      "loss": 1.0357,
+      "step": 9413
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9516567357575891,
+      "learning_rate": 1.8132298504510398e-05,
+      "loss": 1.0538,
+      "step": 9414
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.829061135004015,
+      "learning_rate": 1.8131854434724335e-05,
+      "loss": 0.9454,
+      "step": 9415
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5490830886070441,
+      "learning_rate": 1.8131410317592037e-05,
+      "loss": 0.8541,
+      "step": 9416
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6333875728221552,
+      "learning_rate": 1.8130966153116096e-05,
+      "loss": 0.8283,
+      "step": 9417
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6070011330685585,
+      "learning_rate": 1.81305219412991e-05,
+      "loss": 0.8004,
+      "step": 9418
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6185034284948308,
+      "learning_rate": 1.8130077682143627e-05,
+      "loss": 0.9453,
+      "step": 9419
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8101129123022768,
+      "learning_rate": 1.812963337565227e-05,
+      "loss": 0.9323,
+      "step": 9420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0161804078679624,
+      "learning_rate": 1.8129189021827608e-05,
+      "loss": 0.8913,
+      "step": 9421
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7855916805272094,
+      "learning_rate": 1.8128744620672237e-05,
+      "loss": 0.8724,
+      "step": 9422
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.528595739978762,
+      "learning_rate": 1.8128300172188744e-05,
+      "loss": 0.9038,
+      "step": 9423
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.808766335942967,
+      "learning_rate": 1.812785567637971e-05,
+      "loss": 0.7621,
+      "step": 9424
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9822677110758669,
+      "learning_rate": 1.812741113324773e-05,
+      "loss": 0.8802,
+      "step": 9425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0749287383018875,
+      "learning_rate": 1.8126966542795387e-05,
+      "loss": 0.8581,
+      "step": 9426
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.892958713198179,
+      "learning_rate": 1.8126521905025272e-05,
+      "loss": 0.9014,
+      "step": 9427
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9216804280968072,
+      "learning_rate": 1.812607721993998e-05,
+      "loss": 0.8533,
+      "step": 9428
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6946814718237395,
+      "learning_rate": 1.8125632487542084e-05,
+      "loss": 0.8372,
+      "step": 9429
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8024372726583263,
+      "learning_rate": 1.812518770783419e-05,
+      "loss": 0.7662,
+      "step": 9430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5044294090230252,
+      "learning_rate": 1.8124742880818877e-05,
+      "loss": 0.8536,
+      "step": 9431
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4933519160252546,
+      "learning_rate": 1.812429800649874e-05,
+      "loss": 1.0211,
+      "step": 9432
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.555972341222278,
+      "learning_rate": 1.812385308487637e-05,
+      "loss": 1.0062,
+      "step": 9433
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9169651613702414,
+      "learning_rate": 1.8123408115954354e-05,
+      "loss": 0.9631,
+      "step": 9434
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7143705225151573,
+      "learning_rate": 1.812296309973528e-05,
+      "loss": 0.8333,
+      "step": 9435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7184311789497768,
+      "learning_rate": 1.812251803622175e-05,
+      "loss": 0.8776,
+      "step": 9436
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4213887205104891,
+      "learning_rate": 1.8122072925416347e-05,
+      "loss": 0.8942,
+      "step": 9437
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5695445280066154,
+      "learning_rate": 1.8121627767321658e-05,
+      "loss": 0.8153,
+      "step": 9438
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.6941316875260615,
+      "learning_rate": 1.8121182561940288e-05,
+      "loss": 0.7949,
+      "step": 9439
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4619413748489916,
+      "learning_rate": 1.8120737309274818e-05,
+      "loss": 0.9045,
+      "step": 9440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5662637411860256,
+      "learning_rate": 1.8120292009327843e-05,
+      "loss": 0.9281,
+      "step": 9441
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6528117502093767,
+      "learning_rate": 1.811984666210196e-05,
+      "loss": 0.8976,
+      "step": 9442
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7983934581945926,
+      "learning_rate": 1.811940126759976e-05,
+      "loss": 0.9929,
+      "step": 9443
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.685925070524137,
+      "learning_rate": 1.8118955825823833e-05,
+      "loss": 0.9615,
+      "step": 9444
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7761512950251629,
+      "learning_rate": 1.8118510336776776e-05,
+      "loss": 0.7511,
+      "step": 9445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6654448982108454,
+      "learning_rate": 1.8118064800461182e-05,
+      "loss": 0.9104,
+      "step": 9446
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.937163716169845,
+      "learning_rate": 1.8117619216879645e-05,
+      "loss": 0.8787,
+      "step": 9447
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.470802581485677,
+      "learning_rate": 1.8117173586034757e-05,
+      "loss": 0.883,
+      "step": 9448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7732653658387019,
+      "learning_rate": 1.8116727907929113e-05,
+      "loss": 0.8588,
+      "step": 9449
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5565943405386717,
+      "learning_rate": 1.8116282182565313e-05,
+      "loss": 0.814,
+      "step": 9450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7218495962541789,
+      "learning_rate": 1.8115836409945947e-05,
+      "loss": 0.9042,
+      "step": 9451
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6041380540210095,
+      "learning_rate": 1.8115390590073612e-05,
+      "loss": 0.8777,
+      "step": 9452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7051462520893337,
+      "learning_rate": 1.8114944722950903e-05,
+      "loss": 0.851,
+      "step": 9453
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5895454448299084,
+      "learning_rate": 1.8114498808580418e-05,
+      "loss": 0.8636,
+      "step": 9454
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8358283534699429,
+      "learning_rate": 1.8114052846964753e-05,
+      "loss": 0.8734,
+      "step": 9455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.86748316667626,
+      "learning_rate": 1.8113606838106503e-05,
+      "loss": 0.8543,
+      "step": 9456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7072237078681334,
+      "learning_rate": 1.8113160782008263e-05,
+      "loss": 0.7947,
+      "step": 9457
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4764639010067775,
+      "learning_rate": 1.8112714678672632e-05,
+      "loss": 0.9512,
+      "step": 9458
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5709679365376557,
+      "learning_rate": 1.811226852810221e-05,
+      "loss": 0.9635,
+      "step": 9459
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8085288443978889,
+      "learning_rate": 1.811182233029959e-05,
+      "loss": 1.0135,
+      "step": 9460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7999392635147737,
+      "learning_rate": 1.8111376085267376e-05,
+      "loss": 0.8937,
+      "step": 9461
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.659917485689472,
+      "learning_rate": 1.811092979300816e-05,
+      "loss": 0.9752,
+      "step": 9462
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7928499273524126,
+      "learning_rate": 1.8110483453524543e-05,
+      "loss": 0.8145,
+      "step": 9463
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7788099495056404,
+      "learning_rate": 1.811003706681912e-05,
+      "loss": 0.8337,
+      "step": 9464
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5865022642736422,
+      "learning_rate": 1.81095906328945e-05,
+      "loss": 0.8733,
+      "step": 9465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7775059464827395,
+      "learning_rate": 1.8109144151753274e-05,
+      "loss": 0.7526,
+      "step": 9466
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7464144260114725,
+      "learning_rate": 1.810869762339804e-05,
+      "loss": 0.8574,
+      "step": 9467
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5271392959283185,
+      "learning_rate": 1.8108251047831405e-05,
+      "loss": 0.8903,
+      "step": 9468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.000047663488231,
+      "learning_rate": 1.8107804425055967e-05,
+      "loss": 0.8566,
+      "step": 9469
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5840848352960555,
+      "learning_rate": 1.810735775507432e-05,
+      "loss": 0.9648,
+      "step": 9470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.019571596926389,
+      "learning_rate": 1.8106911037889072e-05,
+      "loss": 0.9285,
+      "step": 9471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5649785339612505,
+      "learning_rate": 1.810646427350282e-05,
+      "loss": 0.9873,
+      "step": 9472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5361136395403145,
+      "learning_rate": 1.8106017461918165e-05,
+      "loss": 0.9765,
+      "step": 9473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9287123578706171,
+      "learning_rate": 1.8105570603137714e-05,
+      "loss": 0.9562,
+      "step": 9474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.066719512563691,
+      "learning_rate": 1.810512369716406e-05,
+      "loss": 0.9198,
+      "step": 9475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5414446411772151,
+      "learning_rate": 1.8104676743999816e-05,
+      "loss": 0.8201,
+      "step": 9476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.706133348481525,
+      "learning_rate": 1.8104229743647573e-05,
+      "loss": 0.8229,
+      "step": 9477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6092961446856413,
+      "learning_rate": 1.8103782696109943e-05,
+      "loss": 0.9377,
+      "step": 9478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.794069405451606,
+      "learning_rate": 1.8103335601389522e-05,
+      "loss": 1.0438,
+      "step": 9479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7913874759161239,
+      "learning_rate": 1.8102888459488916e-05,
+      "loss": 0.6768,
+      "step": 9480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7316597786483459,
+      "learning_rate": 1.810244127041073e-05,
+      "loss": 1.0097,
+      "step": 9481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4821414761443195,
+      "learning_rate": 1.810199403415756e-05,
+      "loss": 0.88,
+      "step": 9482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.587188506108511,
+      "learning_rate": 1.8101546750732022e-05,
+      "loss": 0.9855,
+      "step": 9483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7275754127969352,
+      "learning_rate": 1.810109942013671e-05,
+      "loss": 0.8509,
+      "step": 9484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7763373173592523,
+      "learning_rate": 1.8100652042374232e-05,
+      "loss": 0.8839,
+      "step": 9485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9917838452604633,
+      "learning_rate": 1.8100204617447195e-05,
+      "loss": 0.8433,
+      "step": 9486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7705176411570003,
+      "learning_rate": 1.80997571453582e-05,
+      "loss": 0.9614,
+      "step": 9487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7704401509929675,
+      "learning_rate": 1.8099309626109857e-05,
+      "loss": 0.9057,
+      "step": 9488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8680268658594743,
+      "learning_rate": 1.809886205970477e-05,
+      "loss": 0.9057,
+      "step": 9489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6870889713634996,
+      "learning_rate": 1.8098414446145543e-05,
+      "loss": 0.9502,
+      "step": 9490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.685736094052493,
+      "learning_rate": 1.8097966785434786e-05,
+      "loss": 0.9821,
+      "step": 9491
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7231237797063195,
+      "learning_rate": 1.8097519077575095e-05,
+      "loss": 0.9304,
+      "step": 9492
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7310952636136971,
+      "learning_rate": 1.809707132256909e-05,
+      "loss": 0.8996,
+      "step": 9493
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.193892684768822,
+      "learning_rate": 1.8096623520419372e-05,
+      "loss": 0.9124,
+      "step": 9494
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5504793144667945,
+      "learning_rate": 1.8096175671128547e-05,
+      "loss": 1.0425,
+      "step": 9495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9888848653957996,
+      "learning_rate": 1.8095727774699224e-05,
+      "loss": 0.9311,
+      "step": 9496
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4689331254995956,
+      "learning_rate": 1.8095279831134012e-05,
+      "loss": 0.9972,
+      "step": 9497
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8392614516894124,
+      "learning_rate": 1.809483184043552e-05,
+      "loss": 0.8206,
+      "step": 9498
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.605217381667774,
+      "learning_rate": 1.809438380260635e-05,
+      "loss": 0.8427,
+      "step": 9499
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8542861654008677,
+      "learning_rate": 1.8093935717649117e-05,
+      "loss": 0.7518,
+      "step": 9500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7541054690694171,
+      "learning_rate": 1.8093487585566426e-05,
+      "loss": 0.8982,
+      "step": 9501
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6080553470076786,
+      "learning_rate": 1.809303940636089e-05,
+      "loss": 0.9826,
+      "step": 9502
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4576154326672797,
+      "learning_rate": 1.8092591180035114e-05,
+      "loss": 0.966,
+      "step": 9503
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.765813022762524,
+      "learning_rate": 1.809214290659171e-05,
+      "loss": 0.7582,
+      "step": 9504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7119644502602909,
+      "learning_rate": 1.809169458603329e-05,
+      "loss": 0.8708,
+      "step": 9505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7627419532769864,
+      "learning_rate": 1.8091246218362464e-05,
+      "loss": 0.9133,
+      "step": 9506
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6476549493704615,
+      "learning_rate": 1.809079780358184e-05,
+      "loss": 0.8832,
+      "step": 9507
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7965948528032223,
+      "learning_rate": 1.809034934169403e-05,
+      "loss": 0.7771,
+      "step": 9508
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7806259845014034,
+      "learning_rate": 1.808990083270164e-05,
+      "loss": 0.9153,
+      "step": 9509
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.45260257738928,
+      "learning_rate": 1.808945227660729e-05,
+      "loss": 0.7112,
+      "step": 9510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8577876175911738,
+      "learning_rate": 1.808900367341359e-05,
+      "loss": 0.9988,
+      "step": 9511
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6170248243717131,
+      "learning_rate": 1.8088555023123144e-05,
+      "loss": 0.8942,
+      "step": 9512
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5031086890307552,
+      "learning_rate": 1.8088106325738572e-05,
+      "loss": 0.8262,
+      "step": 9513
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7476127269357982,
+      "learning_rate": 1.8087657581262488e-05,
+      "loss": 0.6813,
+      "step": 9514
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6168838400313634,
+      "learning_rate": 1.8087208789697498e-05,
+      "loss": 0.9844,
+      "step": 9515
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8252768064308067,
+      "learning_rate": 1.8086759951046217e-05,
+      "loss": 0.9171,
+      "step": 9516
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7595991740681893,
+      "learning_rate": 1.8086311065311258e-05,
+      "loss": 0.7634,
+      "step": 9517
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.980757343862393,
+      "learning_rate": 1.808586213249524e-05,
+      "loss": 0.9803,
+      "step": 9518
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6304534559704678,
+      "learning_rate": 1.808541315260077e-05,
+      "loss": 0.9537,
+      "step": 9519
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8294644882311117,
+      "learning_rate": 1.808496412563047e-05,
+      "loss": 0.735,
+      "step": 9520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6413704478186582,
+      "learning_rate": 1.8084515051586943e-05,
+      "loss": 0.9262,
+      "step": 9521
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5744755314910888,
+      "learning_rate": 1.808406593047281e-05,
+      "loss": 0.9647,
+      "step": 9522
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5933628130576658,
+      "learning_rate": 1.808361676229069e-05,
+      "loss": 0.924,
+      "step": 9523
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.106166900618155,
+      "learning_rate": 1.808316754704319e-05,
+      "loss": 0.8717,
+      "step": 9524
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8208136331904372,
+      "learning_rate": 1.8082718284732932e-05,
+      "loss": 0.9882,
+      "step": 9525
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8849962492272396,
+      "learning_rate": 1.808226897536253e-05,
+      "loss": 0.7721,
+      "step": 9526
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7201509640478652,
+      "learning_rate": 1.8081819618934597e-05,
+      "loss": 0.9321,
+      "step": 9527
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7425427920839507,
+      "learning_rate": 1.8081370215451752e-05,
+      "loss": 0.9739,
+      "step": 9528
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.426250068606066,
+      "learning_rate": 1.8080920764916613e-05,
+      "loss": 0.9371,
+      "step": 9529
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7671471630541304,
+      "learning_rate": 1.8080471267331792e-05,
+      "loss": 0.9119,
+      "step": 9530
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5375742234773093,
+      "learning_rate": 1.808002172269991e-05,
+      "loss": 0.9493,
+      "step": 9531
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7717470141917349,
+      "learning_rate": 1.8079572131023585e-05,
+      "loss": 0.99,
+      "step": 9532
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6564138515771996,
+      "learning_rate": 1.807912249230543e-05,
+      "loss": 0.925,
+      "step": 9533
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5075959541328143,
+      "learning_rate": 1.807867280654807e-05,
+      "loss": 1.0099,
+      "step": 9534
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5101436816962683,
+      "learning_rate": 1.807822307375412e-05,
+      "loss": 0.957,
+      "step": 9535
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.697873289384662,
+      "learning_rate": 1.807777329392619e-05,
+      "loss": 0.8516,
+      "step": 9536
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7492543579894424,
+      "learning_rate": 1.8077323467066913e-05,
+      "loss": 1.005,
+      "step": 9537
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7736051230193142,
+      "learning_rate": 1.80768735931789e-05,
+      "loss": 0.8262,
+      "step": 9538
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2668566466269153,
+      "learning_rate": 1.807642367226477e-05,
+      "loss": 0.878,
+      "step": 9539
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5650047676988854,
+      "learning_rate": 1.8075973704327148e-05,
+      "loss": 0.9792,
+      "step": 9540
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.613295508435998,
+      "learning_rate": 1.8075523689368646e-05,
+      "loss": 0.9199,
+      "step": 9541
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.622318850390041,
+      "learning_rate": 1.8075073627391892e-05,
+      "loss": 0.8714,
+      "step": 9542
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1416890524222016,
+      "learning_rate": 1.8074623518399503e-05,
+      "loss": 0.9244,
+      "step": 9543
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5772047334882293,
+      "learning_rate": 1.8074173362394095e-05,
+      "loss": 0.8941,
+      "step": 9544
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7423987591289223,
+      "learning_rate": 1.8073723159378298e-05,
+      "loss": 0.9916,
+      "step": 9545
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7216447842146618,
+      "learning_rate": 1.8073272909354727e-05,
+      "loss": 0.8516,
+      "step": 9546
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.849066835713059,
+      "learning_rate": 1.8072822612326007e-05,
+      "loss": 0.9171,
+      "step": 9547
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6895940624694241,
+      "learning_rate": 1.8072372268294755e-05,
+      "loss": 0.9909,
+      "step": 9548
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7988335949825908,
+      "learning_rate": 1.8071921877263597e-05,
+      "loss": 0.7221,
+      "step": 9549
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6299700879697245,
+      "learning_rate": 1.8071471439235154e-05,
+      "loss": 0.9943,
+      "step": 9550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4081611420782394,
+      "learning_rate": 1.807102095421205e-05,
+      "loss": 0.8932,
+      "step": 9551
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8481159120962384,
+      "learning_rate": 1.807057042219691e-05,
+      "loss": 0.6981,
+      "step": 9552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6824331390818295,
+      "learning_rate": 1.8070119843192348e-05,
+      "loss": 0.9558,
+      "step": 9553
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6910896868471563,
+      "learning_rate": 1.8069669217200995e-05,
+      "loss": 0.9142,
+      "step": 9554
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5846774427387331,
+      "learning_rate": 1.8069218544225474e-05,
+      "loss": 1.0254,
+      "step": 9555
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6949036098652384,
+      "learning_rate": 1.8068767824268407e-05,
+      "loss": 0.863,
+      "step": 9556
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3212514416340575,
+      "learning_rate": 1.8068317057332422e-05,
+      "loss": 0.934,
+      "step": 9557
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6637633456610965,
+      "learning_rate": 1.8067866243420136e-05,
+      "loss": 0.8645,
+      "step": 9558
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.504869140853529,
+      "learning_rate": 1.806741538253418e-05,
+      "loss": 1.0028,
+      "step": 9559
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4399539573286715,
+      "learning_rate": 1.806696447467718e-05,
+      "loss": 0.998,
+      "step": 9560
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7507927381701075,
+      "learning_rate": 1.806651351985176e-05,
+      "loss": 0.7892,
+      "step": 9561
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7004475967173884,
+      "learning_rate": 1.806606251806054e-05,
+      "loss": 0.9526,
+      "step": 9562
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4783758499497108,
+      "learning_rate": 1.8065611469306152e-05,
+      "loss": 0.8871,
+      "step": 9563
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9252648561191053,
+      "learning_rate": 1.806516037359122e-05,
+      "loss": 0.927,
+      "step": 9564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5398217369083613,
+      "learning_rate": 1.806470923091837e-05,
+      "loss": 0.9689,
+      "step": 9565
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6604707466639297,
+      "learning_rate": 1.806425804129023e-05,
+      "loss": 0.9045,
+      "step": 9566
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9373978412695696,
+      "learning_rate": 1.8063806804709433e-05,
+      "loss": 0.9842,
+      "step": 9567
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4923424894801467,
+      "learning_rate": 1.8063355521178594e-05,
+      "loss": 0.8981,
+      "step": 9568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8477980481514917,
+      "learning_rate": 1.8062904190700344e-05,
+      "loss": 0.878,
+      "step": 9569
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4919328952362103,
+      "learning_rate": 1.8062452813277316e-05,
+      "loss": 0.7809,
+      "step": 9570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7919023753507275,
+      "learning_rate": 1.8062001388912135e-05,
+      "loss": 0.7566,
+      "step": 9571
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.817224696629718,
+      "learning_rate": 1.806154991760743e-05,
+      "loss": 0.9588,
+      "step": 9572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.740225682400123,
+      "learning_rate": 1.806109839936583e-05,
+      "loss": 0.8229,
+      "step": 9573
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8478133830096641,
+      "learning_rate": 1.806064683418996e-05,
+      "loss": 0.831,
+      "step": 9574
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6903611373940928,
+      "learning_rate": 1.8060195222082457e-05,
+      "loss": 0.9795,
+      "step": 9575
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.332021672152894,
+      "learning_rate": 1.805974356304594e-05,
+      "loss": 0.9119,
+      "step": 9576
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6181383282687665,
+      "learning_rate": 1.8059291857083048e-05,
+      "loss": 0.9931,
+      "step": 9577
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8910471114880076,
+      "learning_rate": 1.8058840104196406e-05,
+      "loss": 0.9382,
+      "step": 9578
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9232195371288614,
+      "learning_rate": 1.8058388304388646e-05,
+      "loss": 0.9604,
+      "step": 9579
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6894286880903304,
+      "learning_rate": 1.80579364576624e-05,
+      "loss": 0.8989,
+      "step": 9580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.224428858358421,
+      "learning_rate": 1.8057484564020294e-05,
+      "loss": 0.8793,
+      "step": 9581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.606106947422406,
+      "learning_rate": 1.805703262346496e-05,
+      "loss": 0.8232,
+      "step": 9582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8559015636026246,
+      "learning_rate": 1.8056580635999033e-05,
+      "loss": 0.8525,
+      "step": 9583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5566421690175465,
+      "learning_rate": 1.8056128601625146e-05,
+      "loss": 0.88,
+      "step": 9584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5096826662022067,
+      "learning_rate": 1.8055676520345928e-05,
+      "loss": 0.8371,
+      "step": 9585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8506215219800561,
+      "learning_rate": 1.8055224392164005e-05,
+      "loss": 0.9243,
+      "step": 9586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6475395631598335,
+      "learning_rate": 1.805477221708202e-05,
+      "loss": 0.9097,
+      "step": 9587
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6618293934800659,
+      "learning_rate": 1.80543199951026e-05,
+      "loss": 1.056,
+      "step": 9588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8635692703091615,
+      "learning_rate": 1.8053867726228376e-05,
+      "loss": 0.9522,
+      "step": 9589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0797460841167097,
+      "learning_rate": 1.8053415410461988e-05,
+      "loss": 0.9989,
+      "step": 9590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6390658879324624,
+      "learning_rate": 1.8052963047806065e-05,
+      "loss": 0.9152,
+      "step": 9591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5701700180848628,
+      "learning_rate": 1.8052510638263244e-05,
+      "loss": 0.9294,
+      "step": 9592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7880304362505548,
+      "learning_rate": 1.8052058181836152e-05,
+      "loss": 0.7982,
+      "step": 9593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7116917030640517,
+      "learning_rate": 1.805160567852743e-05,
+      "loss": 0.9531,
+      "step": 9594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7257459206722767,
+      "learning_rate": 1.805115312833971e-05,
+      "loss": 0.8756,
+      "step": 9595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7833895864076181,
+      "learning_rate": 1.8050700531275632e-05,
+      "loss": 0.8688,
+      "step": 9596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.09842911856607,
+      "learning_rate": 1.8050247887337825e-05,
+      "loss": 0.8058,
+      "step": 9597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4310439235122798,
+      "learning_rate": 1.8049795196528924e-05,
+      "loss": 0.8411,
+      "step": 9598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7755108714790984,
+      "learning_rate": 1.8049342458851563e-05,
+      "loss": 0.9752,
+      "step": 9599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5894605738265661,
+      "learning_rate": 1.8048889674308387e-05,
+      "loss": 0.8719,
+      "step": 9600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7639179428710443,
+      "learning_rate": 1.8048436842902025e-05,
+      "loss": 0.8348,
+      "step": 9601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7114336851987746,
+      "learning_rate": 1.8047983964635118e-05,
+      "loss": 0.8848,
+      "step": 9602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1301229550261263,
+      "learning_rate": 1.80475310395103e-05,
+      "loss": 0.8577,
+      "step": 9603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.995062294527225,
+      "learning_rate": 1.8047078067530205e-05,
+      "loss": 0.9651,
+      "step": 9604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7272467635150677,
+      "learning_rate": 1.804662504869748e-05,
+      "loss": 0.8589,
+      "step": 9605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5513926977735575,
+      "learning_rate": 1.804617198301475e-05,
+      "loss": 0.8742,
+      "step": 9606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9069559567407028,
+      "learning_rate": 1.8045718870484662e-05,
+      "loss": 0.9225,
+      "step": 9607
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9165807531828352,
+      "learning_rate": 1.804526571110985e-05,
+      "loss": 0.9124,
+      "step": 9608
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.14785121009652,
+      "learning_rate": 1.8044812504892958e-05,
+      "loss": 0.7856,
+      "step": 9609
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9582587253928805,
+      "learning_rate": 1.8044359251836615e-05,
+      "loss": 0.8205,
+      "step": 9610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4530817699667435,
+      "learning_rate": 1.8043905951943472e-05,
+      "loss": 0.8684,
+      "step": 9611
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8350328120773174,
+      "learning_rate": 1.8043452605216157e-05,
+      "loss": 0.9212,
+      "step": 9612
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6216006649440562,
+      "learning_rate": 1.804299921165732e-05,
+      "loss": 1.0037,
+      "step": 9613
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7442952973677872,
+      "learning_rate": 1.804254577126959e-05,
+      "loss": 0.8828,
+      "step": 9614
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9021026334725069,
+      "learning_rate": 1.8042092284055615e-05,
+      "loss": 0.8813,
+      "step": 9615
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6119115380351237,
+      "learning_rate": 1.804163875001803e-05,
+      "loss": 0.9652,
+      "step": 9616
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.861296582718943,
+      "learning_rate": 1.8041185169159478e-05,
+      "loss": 0.7066,
+      "step": 9617
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8178082221861849,
+      "learning_rate": 1.8040731541482603e-05,
+      "loss": 0.8783,
+      "step": 9618
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8300234323092754,
+      "learning_rate": 1.804027786699004e-05,
+      "loss": 0.9714,
+      "step": 9619
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6362468650144915,
+      "learning_rate": 1.8039824145684438e-05,
+      "loss": 0.9463,
+      "step": 9620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7954870571090744,
+      "learning_rate": 1.803937037756843e-05,
+      "loss": 0.7583,
+      "step": 9621
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6030041528910133,
+      "learning_rate": 1.8038916562644666e-05,
+      "loss": 0.9395,
+      "step": 9622
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7261974514981,
+      "learning_rate": 1.8038462700915782e-05,
+      "loss": 0.9172,
+      "step": 9623
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7696525141509852,
+      "learning_rate": 1.8038008792384426e-05,
+      "loss": 0.7577,
+      "step": 9624
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6561346924174793,
+      "learning_rate": 1.8037554837053238e-05,
+      "loss": 0.8487,
+      "step": 9625
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6353254196169587,
+      "learning_rate": 1.803710083492486e-05,
+      "loss": 0.9548,
+      "step": 9626
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8555168891766778,
+      "learning_rate": 1.8036646786001935e-05,
+      "loss": 0.9052,
+      "step": 9627
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8140338137890684,
+      "learning_rate": 1.803619269028711e-05,
+      "loss": 0.9129,
+      "step": 9628
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3841746558554726,
+      "learning_rate": 1.8035738547783025e-05,
+      "loss": 0.8768,
+      "step": 9629
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8320167116831745,
+      "learning_rate": 1.803528435849233e-05,
+      "loss": 0.8909,
+      "step": 9630
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6374017197485813,
+      "learning_rate": 1.8034830122417658e-05,
+      "loss": 0.7883,
+      "step": 9631
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6841960257423478,
+      "learning_rate": 1.803437583956167e-05,
+      "loss": 0.8233,
+      "step": 9632
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6882803477251398,
+      "learning_rate": 1.8033921509926998e-05,
+      "loss": 0.9123,
+      "step": 9633
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7903480605918917,
+      "learning_rate": 1.8033467133516288e-05,
+      "loss": 0.7788,
+      "step": 9634
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.866886331754283,
+      "learning_rate": 1.8033012710332195e-05,
+      "loss": 0.8472,
+      "step": 9635
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5453463061718495,
+      "learning_rate": 1.8032558240377355e-05,
+      "loss": 0.8346,
+      "step": 9636
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.594078481093178,
+      "learning_rate": 1.803210372365442e-05,
+      "loss": 0.8761,
+      "step": 9637
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9846276369193372,
+      "learning_rate": 1.8031649160166035e-05,
+      "loss": 1.0367,
+      "step": 9638
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.01884625683406,
+      "learning_rate": 1.8031194549914845e-05,
+      "loss": 0.81,
+      "step": 9639
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8418689522919611,
+      "learning_rate": 1.8030739892903498e-05,
+      "loss": 0.9845,
+      "step": 9640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5026261514302783,
+      "learning_rate": 1.803028518913464e-05,
+      "loss": 0.8416,
+      "step": 9641
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6940239444713006,
+      "learning_rate": 1.8029830438610916e-05,
+      "loss": 0.859,
+      "step": 9642
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0013203894360125,
+      "learning_rate": 1.802937564133498e-05,
+      "loss": 0.9874,
+      "step": 9643
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0574485619923877,
+      "learning_rate": 1.8028920797309478e-05,
+      "loss": 0.8417,
+      "step": 9644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3229915529602336,
+      "learning_rate": 1.8028465906537055e-05,
+      "loss": 0.876,
+      "step": 9645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5426986117963752,
+      "learning_rate": 1.8028010969020364e-05,
+      "loss": 0.8614,
+      "step": 9646
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.554688255151548,
+      "learning_rate": 1.802755598476205e-05,
+      "loss": 0.7504,
+      "step": 9647
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6130443558292717,
+      "learning_rate": 1.8027100953764766e-05,
+      "loss": 0.9566,
+      "step": 9648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.371235515630516,
+      "learning_rate": 1.802664587603116e-05,
+      "loss": 0.9766,
+      "step": 9649
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6755215113322477,
+      "learning_rate": 1.8026190751563874e-05,
+      "loss": 0.8054,
+      "step": 9650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.111535238238948,
+      "learning_rate": 1.8025735580365568e-05,
+      "loss": 0.8827,
+      "step": 9651
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8197073552058505,
+      "learning_rate": 1.802528036243889e-05,
+      "loss": 0.8909,
+      "step": 9652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7405405635043942,
+      "learning_rate": 1.8024825097786487e-05,
+      "loss": 0.7209,
+      "step": 9653
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5692737883570942,
+      "learning_rate": 1.802436978641101e-05,
+      "loss": 0.8729,
+      "step": 9654
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.528854304213289,
+      "learning_rate": 1.8023914428315115e-05,
+      "loss": 0.9327,
+      "step": 9655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0233020063533242,
+      "learning_rate": 1.802345902350145e-05,
+      "loss": 0.8056,
+      "step": 9656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0974047108374183,
+      "learning_rate": 1.8023003571972664e-05,
+      "loss": 0.9163,
+      "step": 9657
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.933302109482013,
+      "learning_rate": 1.8022548073731413e-05,
+      "loss": 0.8131,
+      "step": 9658
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9532799264139749,
+      "learning_rate": 1.8022092528780348e-05,
+      "loss": 0.8765,
+      "step": 9659
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9448543672878993,
+      "learning_rate": 1.8021636937122117e-05,
+      "loss": 0.9503,
+      "step": 9660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5398333054458493,
+      "learning_rate": 1.8021181298759377e-05,
+      "loss": 0.9679,
+      "step": 9661
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8529291537762038,
+      "learning_rate": 1.802072561369478e-05,
+      "loss": 0.8418,
+      "step": 9662
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.858255340188077,
+      "learning_rate": 1.8020269881930982e-05,
+      "loss": 0.8986,
+      "step": 9663
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4837473253190587,
+      "learning_rate": 1.8019814103470633e-05,
+      "loss": 0.9515,
+      "step": 9664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8369340141345784,
+      "learning_rate": 1.8019358278316386e-05,
+      "loss": 0.7865,
+      "step": 9665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7839723001707157,
+      "learning_rate": 1.8018902406470894e-05,
+      "loss": 1.0294,
+      "step": 9666
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5622410363495471,
+      "learning_rate": 1.8018446487936815e-05,
+      "loss": 0.8402,
+      "step": 9667
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.744716495437285,
+      "learning_rate": 1.8017990522716804e-05,
+      "loss": 0.8727,
+      "step": 9668
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5323602788501303,
+      "learning_rate": 1.8017534510813513e-05,
+      "loss": 0.8167,
+      "step": 9669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.907226953378858,
+      "learning_rate": 1.8017078452229597e-05,
+      "loss": 0.9327,
+      "step": 9670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5318591655108307,
+      "learning_rate": 1.8016622346967714e-05,
+      "loss": 0.9806,
+      "step": 9671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.777988253460645,
+      "learning_rate": 1.8016166195030514e-05,
+      "loss": 0.741,
+      "step": 9672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7603639703373187,
+      "learning_rate": 1.801570999642066e-05,
+      "loss": 0.7014,
+      "step": 9673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6980510471664378,
+      "learning_rate": 1.8015253751140803e-05,
+      "loss": 0.9496,
+      "step": 9674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7816789756832632,
+      "learning_rate": 1.80147974591936e-05,
+      "loss": 0.8798,
+      "step": 9675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2321970070035406,
+      "learning_rate": 1.801434112058171e-05,
+      "loss": 1.0975,
+      "step": 9676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.570056421847422,
+      "learning_rate": 1.801388473530779e-05,
+      "loss": 0.8665,
+      "step": 9677
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8239489140473062,
+      "learning_rate": 1.8013428303374495e-05,
+      "loss": 1.0187,
+      "step": 9678
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7943189295296986,
+      "learning_rate": 1.801297182478448e-05,
+      "loss": 0.7365,
+      "step": 9679
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6650672347728237,
+      "learning_rate": 1.8012515299540413e-05,
+      "loss": 0.9862,
+      "step": 9680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5229622656138602,
+      "learning_rate": 1.8012058727644943e-05,
+      "loss": 0.9222,
+      "step": 9681
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9041524698604508,
+      "learning_rate": 1.8011602109100724e-05,
+      "loss": 0.8646,
+      "step": 9682
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.715183943709363,
+      "learning_rate": 1.8011145443910426e-05,
+      "loss": 0.9467,
+      "step": 9683
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9025636372310817,
+      "learning_rate": 1.8010688732076706e-05,
+      "loss": 0.8137,
+      "step": 9684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.859820373360182,
+      "learning_rate": 1.8010231973602218e-05,
+      "loss": 0.851,
+      "step": 9685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6353346292360418,
+      "learning_rate": 1.8009775168489622e-05,
+      "loss": 0.903,
+      "step": 9686
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.570036828907694,
+      "learning_rate": 1.8009318316741577e-05,
+      "loss": 0.9273,
+      "step": 9687
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0545309435409442,
+      "learning_rate": 1.8008861418360746e-05,
+      "loss": 0.8158,
+      "step": 9688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7358270672671332,
+      "learning_rate": 1.8008404473349793e-05,
+      "loss": 0.9416,
+      "step": 9689
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7832243480164327,
+      "learning_rate": 1.800794748171137e-05,
+      "loss": 0.8001,
+      "step": 9690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.559008085370411,
+      "learning_rate": 1.8007490443448144e-05,
+      "loss": 0.9345,
+      "step": 9691
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9315960232153182,
+      "learning_rate": 1.800703335856277e-05,
+      "loss": 0.908,
+      "step": 9692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5828085951194144,
+      "learning_rate": 1.8006576227057914e-05,
+      "loss": 0.9531,
+      "step": 9693
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.50728881974015,
+      "learning_rate": 1.8006119048936236e-05,
+      "loss": 0.977,
+      "step": 9694
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.862931683966974,
+      "learning_rate": 1.80056618242004e-05,
+      "loss": 0.9061,
+      "step": 9695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6573529922651105,
+      "learning_rate": 1.8005204552853068e-05,
+      "loss": 1.0141,
+      "step": 9696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5111642863089931,
+      "learning_rate": 1.8004747234896896e-05,
+      "loss": 0.9492,
+      "step": 9697
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5621879430121153,
+      "learning_rate": 1.8004289870334552e-05,
+      "loss": 0.8849,
+      "step": 9698
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7519539640664108,
+      "learning_rate": 1.8003832459168703e-05,
+      "loss": 0.9236,
+      "step": 9699
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0018022657605465,
+      "learning_rate": 1.8003375001402005e-05,
+      "loss": 0.9908,
+      "step": 9700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8261013446092944,
+      "learning_rate": 1.800291749703712e-05,
+      "loss": 0.9136,
+      "step": 9701
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6472330115554763,
+      "learning_rate": 1.8002459946076717e-05,
+      "loss": 0.7501,
+      "step": 9702
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7674090115514705,
+      "learning_rate": 1.8002002348523463e-05,
+      "loss": 0.8851,
+      "step": 9703
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6101162987580753,
+      "learning_rate": 1.8001544704380013e-05,
+      "loss": 0.9629,
+      "step": 9704
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.798814251204998,
+      "learning_rate": 1.800108701364904e-05,
+      "loss": 0.8958,
+      "step": 9705
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.77977547633125,
+      "learning_rate": 1.8000629276333204e-05,
+      "loss": 0.9817,
+      "step": 9706
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9966379048592664,
+      "learning_rate": 1.8000171492435172e-05,
+      "loss": 0.9275,
+      "step": 9707
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6252928688628836,
+      "learning_rate": 1.799971366195761e-05,
+      "loss": 0.9584,
+      "step": 9708
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6278162500090336,
+      "learning_rate": 1.7999255784903178e-05,
+      "loss": 0.962,
+      "step": 9709
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7975555512105712,
+      "learning_rate": 1.799879786127455e-05,
+      "loss": 0.8932,
+      "step": 9710
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8182668865168489,
+      "learning_rate": 1.7998339891074385e-05,
+      "loss": 0.7851,
+      "step": 9711
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5938287612457547,
+      "learning_rate": 1.7997881874305356e-05,
+      "loss": 0.8986,
+      "step": 9712
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7869784661750734,
+      "learning_rate": 1.7997423810970128e-05,
+      "loss": 0.995,
+      "step": 9713
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8056201321496679,
+      "learning_rate": 1.7996965701071364e-05,
+      "loss": 0.8174,
+      "step": 9714
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5092356976742982,
+      "learning_rate": 1.7996507544611735e-05,
+      "loss": 0.8991,
+      "step": 9715
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7421985220299858,
+      "learning_rate": 1.7996049341593903e-05,
+      "loss": 0.8673,
+      "step": 9716
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5546326311591296,
+      "learning_rate": 1.7995591092020544e-05,
+      "loss": 0.9754,
+      "step": 9717
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6522159388716882,
+      "learning_rate": 1.7995132795894323e-05,
+      "loss": 0.9165,
+      "step": 9718
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0067460584991026,
+      "learning_rate": 1.799467445321791e-05,
+      "loss": 0.9895,
+      "step": 9719
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9608430450971541,
+      "learning_rate": 1.7994216063993968e-05,
+      "loss": 0.8287,
+      "step": 9720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6347521898844026,
+      "learning_rate": 1.7993757628225166e-05,
+      "loss": 0.8955,
+      "step": 9721
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.039489584441452,
+      "learning_rate": 1.799329914591418e-05,
+      "loss": 0.882,
+      "step": 9722
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7984203250829884,
+      "learning_rate": 1.799284061706368e-05,
+      "loss": 0.8923,
+      "step": 9723
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9631275183213357,
+      "learning_rate": 1.7992382041676326e-05,
+      "loss": 0.8734,
+      "step": 9724
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7059797590289902,
+      "learning_rate": 1.7991923419754793e-05,
+      "loss": 0.7938,
+      "step": 9725
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.821638669184535,
+      "learning_rate": 1.799146475130175e-05,
+      "loss": 0.8687,
+      "step": 9726
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.181940317048502,
+      "learning_rate": 1.7991006036319876e-05,
+      "loss": 0.8409,
+      "step": 9727
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4980277310819592,
+      "learning_rate": 1.799054727481183e-05,
+      "loss": 0.7481,
+      "step": 9728
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4094395377073086,
+      "learning_rate": 1.799008846678029e-05,
+      "loss": 0.7783,
+      "step": 9729
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.877997457827619,
+      "learning_rate": 1.7989629612227924e-05,
+      "loss": 0.9836,
+      "step": 9730
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5097639752689886,
+      "learning_rate": 1.7989170711157406e-05,
+      "loss": 0.8716,
+      "step": 9731
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8339306024590467,
+      "learning_rate": 1.7988711763571406e-05,
+      "loss": 0.9322,
+      "step": 9732
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3401855476327027,
+      "learning_rate": 1.7988252769472597e-05,
+      "loss": 0.8343,
+      "step": 9733
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7485580203086415,
+      "learning_rate": 1.798779372886365e-05,
+      "loss": 0.7329,
+      "step": 9734
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1209642219000684,
+      "learning_rate": 1.798733464174724e-05,
+      "loss": 0.93,
+      "step": 9735
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9578766358966069,
+      "learning_rate": 1.798687550812604e-05,
+      "loss": 0.8349,
+      "step": 9736
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.502093960781598,
+      "learning_rate": 1.7986416328002723e-05,
+      "loss": 0.8466,
+      "step": 9737
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4675275327902546,
+      "learning_rate": 1.798595710137996e-05,
+      "loss": 0.9018,
+      "step": 9738
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7330174647840004,
+      "learning_rate": 1.7985497828260425e-05,
+      "loss": 0.8997,
+      "step": 9739
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9480295181371594,
+      "learning_rate": 1.7985038508646794e-05,
+      "loss": 0.8961,
+      "step": 9740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7181414981750776,
+      "learning_rate": 1.7984579142541743e-05,
+      "loss": 0.8611,
+      "step": 9741
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.569362358269655,
+      "learning_rate": 1.7984119729947944e-05,
+      "loss": 0.9462,
+      "step": 9742
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.814233231825221,
+      "learning_rate": 1.7983660270868074e-05,
+      "loss": 0.9715,
+      "step": 9743
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.382969470428308,
+      "learning_rate": 1.7983200765304802e-05,
+      "loss": 0.9439,
+      "step": 9744
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6320686661272077,
+      "learning_rate": 1.798274121326081e-05,
+      "loss": 0.904,
+      "step": 9745
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7919766530979913,
+      "learning_rate": 1.798228161473877e-05,
+      "loss": 0.9276,
+      "step": 9746
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8132080215418969,
+      "learning_rate": 1.798182196974136e-05,
+      "loss": 0.8112,
+      "step": 9747
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5408645391584406,
+      "learning_rate": 1.7981362278271258e-05,
+      "loss": 0.874,
+      "step": 9748
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5759573669218365,
+      "learning_rate": 1.7980902540331134e-05,
+      "loss": 0.9491,
+      "step": 9749
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1848531159162032,
+      "learning_rate": 1.798044275592367e-05,
+      "loss": 1.0246,
+      "step": 9750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8154754504332355,
+      "learning_rate": 1.797998292505154e-05,
+      "loss": 0.9645,
+      "step": 9751
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.808716559076863,
+      "learning_rate": 1.7979523047717427e-05,
+      "loss": 0.913,
+      "step": 9752
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.623336805115776,
+      "learning_rate": 1.7979063123924004e-05,
+      "loss": 0.9429,
+      "step": 9753
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8662199990994721,
+      "learning_rate": 1.7978603153673944e-05,
+      "loss": 0.9087,
+      "step": 9754
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7759451481519859,
+      "learning_rate": 1.7978143136969938e-05,
+      "loss": 0.7625,
+      "step": 9755
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4785520447806886,
+      "learning_rate": 1.7977683073814655e-05,
+      "loss": 0.8746,
+      "step": 9756
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6462945683759815,
+      "learning_rate": 1.797722296421077e-05,
+      "loss": 0.9929,
+      "step": 9757
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8065192665437917,
+      "learning_rate": 1.7976762808160972e-05,
+      "loss": 0.8981,
+      "step": 9758
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9698281631923833,
+      "learning_rate": 1.7976302605667936e-05,
+      "loss": 0.8598,
+      "step": 9759
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.557944792175995,
+      "learning_rate": 1.797584235673434e-05,
+      "loss": 0.7801,
+      "step": 9760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.954158998773744,
+      "learning_rate": 1.7975382061362863e-05,
+      "loss": 0.9809,
+      "step": 9761
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6320062732985174,
+      "learning_rate": 1.797492171955619e-05,
+      "loss": 0.9364,
+      "step": 9762
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2908293794904115,
+      "learning_rate": 1.7974461331316994e-05,
+      "loss": 0.997,
+      "step": 9763
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0358171940700232,
+      "learning_rate": 1.797400089664796e-05,
+      "loss": 0.8281,
+      "step": 9764
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4802039998663157,
+      "learning_rate": 1.797354041555177e-05,
+      "loss": 0.8857,
+      "step": 9765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6297389357865673,
+      "learning_rate": 1.7973079888031102e-05,
+      "loss": 0.8522,
+      "step": 9766
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6411366890737802,
+      "learning_rate": 1.7972619314088643e-05,
+      "loss": 0.9431,
+      "step": 9767
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1160571024699903,
+      "learning_rate": 1.7972158693727064e-05,
+      "loss": 1.0624,
+      "step": 9768
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6115884860973058,
+      "learning_rate": 1.7971698026949055e-05,
+      "loss": 0.9582,
+      "step": 9769
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6429674977344204,
+      "learning_rate": 1.79712373137573e-05,
+      "loss": 0.8154,
+      "step": 9770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.894963350543833,
+      "learning_rate": 1.7970776554154472e-05,
+      "loss": 0.849,
+      "step": 9771
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8922594472179919,
+      "learning_rate": 1.7970315748143264e-05,
+      "loss": 0.9525,
+      "step": 9772
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.637344186867348,
+      "learning_rate": 1.7969854895726354e-05,
+      "loss": 0.9633,
+      "step": 9773
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.871175572166971,
+      "learning_rate": 1.796939399690642e-05,
+      "loss": 0.8725,
+      "step": 9774
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7061334026929804,
+      "learning_rate": 1.7968933051686157e-05,
+      "loss": 0.9072,
+      "step": 9775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.600579624504693,
+      "learning_rate": 1.7968472060068242e-05,
+      "loss": 0.951,
+      "step": 9776
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6356534682241013,
+      "learning_rate": 1.796801102205536e-05,
+      "loss": 0.8805,
+      "step": 9777
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4528284336006423,
+      "learning_rate": 1.7967549937650193e-05,
+      "loss": 0.8285,
+      "step": 9778
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6874300957217383,
+      "learning_rate": 1.796708880685543e-05,
+      "loss": 0.9575,
+      "step": 9779
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.814790536053634,
+      "learning_rate": 1.796662762967375e-05,
+      "loss": 0.7943,
+      "step": 9780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6068363062807849,
+      "learning_rate": 1.7966166406107847e-05,
+      "loss": 0.9171,
+      "step": 9781
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.642704491075686,
+      "learning_rate": 1.7965705136160397e-05,
+      "loss": 0.8065,
+      "step": 9782
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4761698685854752,
+      "learning_rate": 1.7965243819834092e-05,
+      "loss": 0.9047,
+      "step": 9783
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7549795282388012,
+      "learning_rate": 1.7964782457131615e-05,
+      "loss": 0.8962,
+      "step": 9784
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.718064898650604,
+      "learning_rate": 1.7964321048055656e-05,
+      "loss": 0.986,
+      "step": 9785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5226594835493101,
+      "learning_rate": 1.7963859592608896e-05,
+      "loss": 0.9326,
+      "step": 9786
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4885658753636188,
+      "learning_rate": 1.796339809079402e-05,
+      "loss": 0.8247,
+      "step": 9787
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0418951815582647,
+      "learning_rate": 1.7962936542613723e-05,
+      "loss": 0.9527,
+      "step": 9788
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6278400494255705,
+      "learning_rate": 1.7962474948070688e-05,
+      "loss": 0.9406,
+      "step": 9789
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.825155677564066,
+      "learning_rate": 1.7962013307167602e-05,
+      "loss": 1.0516,
+      "step": 9790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.891925858098905,
+      "learning_rate": 1.7961551619907158e-05,
+      "loss": 0.9144,
+      "step": 9791
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8490108315021596,
+      "learning_rate": 1.7961089886292036e-05,
+      "loss": 0.8215,
+      "step": 9792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2158828596649243,
+      "learning_rate": 1.7960628106324927e-05,
+      "loss": 0.9276,
+      "step": 9793
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2573929474289405,
+      "learning_rate": 1.7960166280008528e-05,
+      "loss": 0.8783,
+      "step": 9794
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.605381126170576,
+      "learning_rate": 1.795970440734551e-05,
+      "loss": 0.9071,
+      "step": 9795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8109799885166127,
+      "learning_rate": 1.795924248833858e-05,
+      "loss": 0.7787,
+      "step": 9796
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.768925301278777,
+      "learning_rate": 1.795878052299042e-05,
+      "loss": 0.7757,
+      "step": 9797
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5563395126424668,
+      "learning_rate": 1.7958318511303717e-05,
+      "loss": 0.8095,
+      "step": 9798
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3909562692826634,
+      "learning_rate": 1.795785645328117e-05,
+      "loss": 0.9113,
+      "step": 9799
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5903699013064976,
+      "learning_rate": 1.795739434892546e-05,
+      "loss": 0.9873,
+      "step": 9800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7593564569879439,
+      "learning_rate": 1.7956932198239278e-05,
+      "loss": 0.8388,
+      "step": 9801
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6885189922692827,
+      "learning_rate": 1.7956470001225322e-05,
+      "loss": 0.8799,
+      "step": 9802
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4968622978813273,
+      "learning_rate": 1.7956007757886275e-05,
+      "loss": 0.8443,
+      "step": 9803
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8442193981114936,
+      "learning_rate": 1.7955545468224832e-05,
+      "loss": 0.9371,
+      "step": 9804
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5626949277933535,
+      "learning_rate": 1.7955083132243687e-05,
+      "loss": 0.8751,
+      "step": 9805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7790168745473626,
+      "learning_rate": 1.795462074994553e-05,
+      "loss": 0.7462,
+      "step": 9806
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5619588235485962,
+      "learning_rate": 1.795415832133305e-05,
+      "loss": 0.9239,
+      "step": 9807
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5987264962773517,
+      "learning_rate": 1.7953695846408945e-05,
+      "loss": 0.9586,
+      "step": 9808
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5114351158284554,
+      "learning_rate": 1.7953233325175904e-05,
+      "loss": 0.8915,
+      "step": 9809
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5271351141579572,
+      "learning_rate": 1.795277075763662e-05,
+      "loss": 0.919,
+      "step": 9810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.719922061764369,
+      "learning_rate": 1.7952308143793783e-05,
+      "loss": 0.9159,
+      "step": 9811
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7575430247491387,
+      "learning_rate": 1.7951845483650093e-05,
+      "loss": 0.7377,
+      "step": 9812
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5487846636221412,
+      "learning_rate": 1.7951382777208243e-05,
+      "loss": 0.9105,
+      "step": 9813
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.191923037420851,
+      "learning_rate": 1.7950920024470923e-05,
+      "loss": 0.9412,
+      "step": 9814
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5227282376360727,
+      "learning_rate": 1.795045722544083e-05,
+      "loss": 0.9726,
+      "step": 9815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6216020116771122,
+      "learning_rate": 1.794999438012066e-05,
+      "loss": 0.8319,
+      "step": 9816
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.625267259362579,
+      "learning_rate": 1.7949531488513106e-05,
+      "loss": 0.967,
+      "step": 9817
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.700079123294486,
+      "learning_rate": 1.794906855062086e-05,
+      "loss": 0.8949,
+      "step": 9818
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5038852677801309,
+      "learning_rate": 1.794860556644662e-05,
+      "loss": 0.9363,
+      "step": 9819
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.465369520324321,
+      "learning_rate": 1.7948142535993084e-05,
+      "loss": 0.9335,
+      "step": 9820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.650585737771178,
+      "learning_rate": 1.7947679459262945e-05,
+      "loss": 1.0445,
+      "step": 9821
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8081028477211687,
+      "learning_rate": 1.79472163362589e-05,
+      "loss": 0.763,
+      "step": 9822
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5989028729171049,
+      "learning_rate": 1.7946753166983642e-05,
+      "loss": 0.8714,
+      "step": 9823
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.114173078366993,
+      "learning_rate": 1.7946289951439876e-05,
+      "loss": 0.8645,
+      "step": 9824
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4848087593453543,
+      "learning_rate": 1.7945826689630293e-05,
+      "loss": 0.8878,
+      "step": 9825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7656017805744328,
+      "learning_rate": 1.7945363381557594e-05,
+      "loss": 0.9943,
+      "step": 9826
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6004810436164445,
+      "learning_rate": 1.794490002722447e-05,
+      "loss": 0.8811,
+      "step": 9827
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.915426860020582,
+      "learning_rate": 1.7944436626633625e-05,
+      "loss": 0.8688,
+      "step": 9828
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9125896475419497,
+      "learning_rate": 1.7943973179787753e-05,
+      "loss": 0.9054,
+      "step": 9829
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8901875499576917,
+      "learning_rate": 1.7943509686689556e-05,
+      "loss": 0.8366,
+      "step": 9830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8078399471806208,
+      "learning_rate": 1.7943046147341727e-05,
+      "loss": 0.8641,
+      "step": 9831
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.628568739341852,
+      "learning_rate": 1.7942582561746973e-05,
+      "loss": 1.0438,
+      "step": 9832
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.087811541367086,
+      "learning_rate": 1.7942118929907987e-05,
+      "loss": 0.9519,
+      "step": 9833
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7064760980288742,
+      "learning_rate": 1.7941655251827467e-05,
+      "loss": 1.0111,
+      "step": 9834
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.614676141506988,
+      "learning_rate": 1.794119152750812e-05,
+      "loss": 0.9693,
+      "step": 9835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5236513232292648,
+      "learning_rate": 1.794072775695264e-05,
+      "loss": 0.9586,
+      "step": 9836
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5740921279291176,
+      "learning_rate": 1.7940263940163732e-05,
+      "loss": 0.9818,
+      "step": 9837
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5861040147076693,
+      "learning_rate": 1.793980007714409e-05,
+      "loss": 1.0072,
+      "step": 9838
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7083515631182928,
+      "learning_rate": 1.7939336167896416e-05,
+      "loss": 0.9882,
+      "step": 9839
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6756062344858214,
+      "learning_rate": 1.7938872212423417e-05,
+      "loss": 0.8391,
+      "step": 9840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6416272203793565,
+      "learning_rate": 1.793840821072779e-05,
+      "loss": 1.0538,
+      "step": 9841
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8090197739397031,
+      "learning_rate": 1.7937944162812238e-05,
+      "loss": 0.8721,
+      "step": 9842
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.208329254627245,
+      "learning_rate": 1.7937480068679462e-05,
+      "loss": 0.9359,
+      "step": 9843
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6529611185650468,
+      "learning_rate": 1.7937015928332165e-05,
+      "loss": 0.8443,
+      "step": 9844
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6678525463133131,
+      "learning_rate": 1.7936551741773043e-05,
+      "loss": 0.893,
+      "step": 9845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.729797834515621,
+      "learning_rate": 1.7936087509004812e-05,
+      "loss": 0.9205,
+      "step": 9846
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.736639381588157,
+      "learning_rate": 1.793562323003016e-05,
+      "loss": 0.7499,
+      "step": 9847
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.6749870785722023,
+      "learning_rate": 1.79351589048518e-05,
+      "loss": 0.8958,
+      "step": 9848
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.724147834151157,
+      "learning_rate": 1.7934694533472433e-05,
+      "loss": 0.9167,
+      "step": 9849
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7418297526926347,
+      "learning_rate": 1.7934230115894765e-05,
+      "loss": 0.9503,
+      "step": 9850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6524577090766042,
+      "learning_rate": 1.7933765652121493e-05,
+      "loss": 0.9337,
+      "step": 9851
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8124962880013331,
+      "learning_rate": 1.793330114215533e-05,
+      "loss": 0.9778,
+      "step": 9852
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5442705435061397,
+      "learning_rate": 1.7932836585998973e-05,
+      "loss": 1.0508,
+      "step": 9853
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8115508070874191,
+      "learning_rate": 1.7932371983655125e-05,
+      "loss": 0.8837,
+      "step": 9854
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.139545716358563,
+      "learning_rate": 1.7931907335126505e-05,
+      "loss": 0.7774,
+      "step": 9855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5217379301982885,
+      "learning_rate": 1.7931442640415805e-05,
+      "loss": 0.918,
+      "step": 9856
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7511154191617648,
+      "learning_rate": 1.7930977899525736e-05,
+      "loss": 0.7951,
+      "step": 9857
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6872718198036818,
+      "learning_rate": 1.7930513112459002e-05,
+      "loss": 0.9589,
+      "step": 9858
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.071152917813331,
+      "learning_rate": 1.7930048279218306e-05,
+      "loss": 0.8726,
+      "step": 9859
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8908445284025388,
+      "learning_rate": 1.7929583399806363e-05,
+      "loss": 0.8087,
+      "step": 9860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6424482751000773,
+      "learning_rate": 1.7929118474225876e-05,
+      "loss": 0.9094,
+      "step": 9861
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5403677873609702,
+      "learning_rate": 1.7928653502479546e-05,
+      "loss": 0.8768,
+      "step": 9862
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5453042036763804,
+      "learning_rate": 1.792818848457009e-05,
+      "loss": 0.887,
+      "step": 9863
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6549697024644943,
+      "learning_rate": 1.792772342050021e-05,
+      "loss": 1.0145,
+      "step": 9864
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8268841061754137,
+      "learning_rate": 1.7927258310272613e-05,
+      "loss": 0.7563,
+      "step": 9865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8804449156948266,
+      "learning_rate": 1.7926793153890006e-05,
+      "loss": 0.9433,
+      "step": 9866
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4312664313591923,
+      "learning_rate": 1.7926327951355106e-05,
+      "loss": 0.9077,
+      "step": 9867
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0956930602760555,
+      "learning_rate": 1.7925862702670608e-05,
+      "loss": 0.937,
+      "step": 9868
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7301649476377363,
+      "learning_rate": 1.7925397407839233e-05,
+      "loss": 0.8433,
+      "step": 9869
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6074636629075356,
+      "learning_rate": 1.7924932066863682e-05,
+      "loss": 1.0662,
+      "step": 9870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6844243285081388,
+      "learning_rate": 1.792446667974667e-05,
+      "loss": 0.9733,
+      "step": 9871
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.600367938813124,
+      "learning_rate": 1.7924001246490907e-05,
+      "loss": 0.9236,
+      "step": 9872
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7348013289918613,
+      "learning_rate": 1.7923535767099093e-05,
+      "loss": 0.9107,
+      "step": 9873
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.526275081740061,
+      "learning_rate": 1.792307024157395e-05,
+      "loss": 0.9699,
+      "step": 9874
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.63488516333478,
+      "learning_rate": 1.792260466991818e-05,
+      "loss": 1.0048,
+      "step": 9875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.437296392776579,
+      "learning_rate": 1.79221390521345e-05,
+      "loss": 0.9365,
+      "step": 9876
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8089410128116878,
+      "learning_rate": 1.792167338822562e-05,
+      "loss": 0.8665,
+      "step": 9877
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.671933393288656,
+      "learning_rate": 1.792120767819425e-05,
+      "loss": 1.024,
+      "step": 9878
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.689581055428095,
+      "learning_rate": 1.7920741922043093e-05,
+      "loss": 0.876,
+      "step": 9879
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6963745983274487,
+      "learning_rate": 1.7920276119774876e-05,
+      "loss": 0.8106,
+      "step": 9880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7944998776429253,
+      "learning_rate": 1.7919810271392305e-05,
+      "loss": 0.9697,
+      "step": 9881
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8865105508217324,
+      "learning_rate": 1.7919344376898088e-05,
+      "loss": 0.8246,
+      "step": 9882
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6082258828580127,
+      "learning_rate": 1.791887843629494e-05,
+      "loss": 0.8947,
+      "step": 9883
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7142261469184474,
+      "learning_rate": 1.7918412449585577e-05,
+      "loss": 0.8946,
+      "step": 9884
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8791487900572377,
+      "learning_rate": 1.7917946416772708e-05,
+      "loss": 1.0111,
+      "step": 9885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.663529408942093,
+      "learning_rate": 1.791748033785905e-05,
+      "loss": 0.9692,
+      "step": 9886
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8697987678174601,
+      "learning_rate": 1.7917014212847314e-05,
+      "loss": 0.801,
+      "step": 9887
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8792082727912633,
+      "learning_rate": 1.7916548041740213e-05,
+      "loss": 0.7667,
+      "step": 9888
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9768305728234439,
+      "learning_rate": 1.7916081824540466e-05,
+      "loss": 0.8251,
+      "step": 9889
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8328970427010738,
+      "learning_rate": 1.7915615561250783e-05,
+      "loss": 1.0216,
+      "step": 9890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7340627101817923,
+      "learning_rate": 1.791514925187388e-05,
+      "loss": 0.8828,
+      "step": 9891
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6550073934167226,
+      "learning_rate": 1.791468289641247e-05,
+      "loss": 0.8747,
+      "step": 9892
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7978276759499012,
+      "learning_rate": 1.7914216494869278e-05,
+      "loss": 0.8214,
+      "step": 9893
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0158948264900087,
+      "learning_rate": 1.7913750047247004e-05,
+      "loss": 0.9299,
+      "step": 9894
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5982935120614332,
+      "learning_rate": 1.7913283553548374e-05,
+      "loss": 0.9971,
+      "step": 9895
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7668812245452352,
+      "learning_rate": 1.7912817013776102e-05,
+      "loss": 0.7589,
+      "step": 9896
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6828148046774554,
+      "learning_rate": 1.7912350427932905e-05,
+      "loss": 1.0159,
+      "step": 9897
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7846602599855501,
+      "learning_rate": 1.79118837960215e-05,
+      "loss": 0.8043,
+      "step": 9898
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7222855960794858,
+      "learning_rate": 1.79114171180446e-05,
+      "loss": 0.8253,
+      "step": 9899
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5160407915154495,
+      "learning_rate": 1.7910950394004926e-05,
+      "loss": 0.9552,
+      "step": 9900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7951798046383154,
+      "learning_rate": 1.7910483623905195e-05,
+      "loss": 0.7014,
+      "step": 9901
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7200539630368974,
+      "learning_rate": 1.791001680774812e-05,
+      "loss": 0.9824,
+      "step": 9902
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6850437876735662,
+      "learning_rate": 1.7909549945536427e-05,
+      "loss": 0.9581,
+      "step": 9903
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6369384949777206,
+      "learning_rate": 1.790908303727283e-05,
+      "loss": 0.9208,
+      "step": 9904
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5662037067508652,
+      "learning_rate": 1.7908616082960047e-05,
+      "loss": 0.986,
+      "step": 9905
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6956454526870413,
+      "learning_rate": 1.7908149082600796e-05,
+      "loss": 0.7804,
+      "step": 9906
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7561093357400215,
+      "learning_rate": 1.79076820361978e-05,
+      "loss": 1.0492,
+      "step": 9907
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2103334551904688,
+      "learning_rate": 1.790721494375377e-05,
+      "loss": 0.9555,
+      "step": 9908
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7343061005452471,
+      "learning_rate": 1.790674780527144e-05,
+      "loss": 0.9319,
+      "step": 9909
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5059700260686415,
+      "learning_rate": 1.7906280620753517e-05,
+      "loss": 0.8805,
+      "step": 9910
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8085079142171685,
+      "learning_rate": 1.790581339020272e-05,
+      "loss": 0.7653,
+      "step": 9911
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.607418175234772,
+      "learning_rate": 1.790534611362178e-05,
+      "loss": 1.0127,
+      "step": 9912
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9337651865210705,
+      "learning_rate": 1.7904878791013412e-05,
+      "loss": 0.829,
+      "step": 9913
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.645978526464702,
+      "learning_rate": 1.7904411422380336e-05,
+      "loss": 0.906,
+      "step": 9914
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3418749977633144,
+      "learning_rate": 1.7903944007725273e-05,
+      "loss": 0.9006,
+      "step": 9915
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.720689873266827,
+      "learning_rate": 1.790347654705095e-05,
+      "loss": 0.8597,
+      "step": 9916
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7871349775517089,
+      "learning_rate": 1.790300904036008e-05,
+      "loss": 0.769,
+      "step": 9917
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.779102670833269,
+      "learning_rate": 1.790254148765539e-05,
+      "loss": 0.9889,
+      "step": 9918
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8526289917550112,
+      "learning_rate": 1.7902073888939603e-05,
+      "loss": 0.9929,
+      "step": 9919
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7954587045575037,
+      "learning_rate": 1.7901606244215436e-05,
+      "loss": 0.8204,
+      "step": 9920
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.600928051345805,
+      "learning_rate": 1.7901138553485623e-05,
+      "loss": 0.9529,
+      "step": 9921
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.511393430367098,
+      "learning_rate": 1.7900670816752875e-05,
+      "loss": 0.935,
+      "step": 9922
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5834267559112483,
+      "learning_rate": 1.790020303401992e-05,
+      "loss": 0.8124,
+      "step": 9923
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.494440761274893,
+      "learning_rate": 1.7899735205289483e-05,
+      "loss": 0.9817,
+      "step": 9924
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8264736623288202,
+      "learning_rate": 1.7899267330564287e-05,
+      "loss": 0.7043,
+      "step": 9925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9429645993235511,
+      "learning_rate": 1.7898799409847055e-05,
+      "loss": 0.9564,
+      "step": 9926
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8479416867247158,
+      "learning_rate": 1.7898331443140512e-05,
+      "loss": 0.9437,
+      "step": 9927
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6761853829707778,
+      "learning_rate": 1.7897863430447384e-05,
+      "loss": 0.9874,
+      "step": 9928
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.615213450372831,
+      "learning_rate": 1.7897395371770393e-05,
+      "loss": 0.8733,
+      "step": 9929
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.665932232381268,
+      "learning_rate": 1.789692726711227e-05,
+      "loss": 0.9442,
+      "step": 9930
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6262048093940844,
+      "learning_rate": 1.789645911647573e-05,
+      "loss": 0.9114,
+      "step": 9931
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6674626119405935,
+      "learning_rate": 1.789599091986351e-05,
+      "loss": 1.0746,
+      "step": 9932
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6791069962369898,
+      "learning_rate": 1.7895522677278327e-05,
+      "loss": 0.8124,
+      "step": 9933
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6785190190947747,
+      "learning_rate": 1.7895054388722913e-05,
+      "loss": 0.8257,
+      "step": 9934
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5285582715367823,
+      "learning_rate": 1.7894586054199993e-05,
+      "loss": 0.9601,
+      "step": 9935
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6154004722443844,
+      "learning_rate": 1.7894117673712295e-05,
+      "loss": 0.9834,
+      "step": 9936
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7152588882931126,
+      "learning_rate": 1.7893649247262543e-05,
+      "loss": 0.9751,
+      "step": 9937
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.883027879483452,
+      "learning_rate": 1.7893180774853467e-05,
+      "loss": 1.0326,
+      "step": 9938
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6386939936442053,
+      "learning_rate": 1.7892712256487792e-05,
+      "loss": 0.9478,
+      "step": 9939
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7483527996709507,
+      "learning_rate": 1.789224369216825e-05,
+      "loss": 0.9197,
+      "step": 9940
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7397388166362387,
+      "learning_rate": 1.7891775081897564e-05,
+      "loss": 0.7124,
+      "step": 9941
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5493574087813844,
+      "learning_rate": 1.7891306425678467e-05,
+      "loss": 0.9317,
+      "step": 9942
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.452960204482011,
+      "learning_rate": 1.7890837723513686e-05,
+      "loss": 0.9374,
+      "step": 9943
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9433165377778996,
+      "learning_rate": 1.7890368975405947e-05,
+      "loss": 0.9042,
+      "step": 9944
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6086275907209966,
+      "learning_rate": 1.7889900181357983e-05,
+      "loss": 0.858,
+      "step": 9945
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.075767628695061,
+      "learning_rate": 1.7889431341372523e-05,
+      "loss": 0.863,
+      "step": 9946
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9727445500476868,
+      "learning_rate": 1.7888962455452296e-05,
+      "loss": 0.8878,
+      "step": 9947
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0508449557037145,
+      "learning_rate": 1.788849352360003e-05,
+      "loss": 1.0051,
+      "step": 9948
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6427936587604517,
+      "learning_rate": 1.7888024545818458e-05,
+      "loss": 0.9741,
+      "step": 9949
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.15257159622046,
+      "learning_rate": 1.788755552211031e-05,
+      "loss": 0.9217,
+      "step": 9950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7726326864653852,
+      "learning_rate": 1.7887086452478318e-05,
+      "loss": 0.767,
+      "step": 9951
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5419796964637438,
+      "learning_rate": 1.788661733692521e-05,
+      "loss": 0.9236,
+      "step": 9952
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6483062464414378,
+      "learning_rate": 1.788614817545372e-05,
+      "loss": 1.0015,
+      "step": 9953
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6958427615877445,
+      "learning_rate": 1.7885678968066576e-05,
+      "loss": 0.9698,
+      "step": 9954
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6487555867856312,
+      "learning_rate": 1.7885209714766517e-05,
+      "loss": 0.845,
+      "step": 9955
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7355735763334263,
+      "learning_rate": 1.788474041555627e-05,
+      "loss": 0.7295,
+      "step": 9956
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4967133284944658,
+      "learning_rate": 1.7884271070438563e-05,
+      "loss": 0.8266,
+      "step": 9957
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.749300707811484,
+      "learning_rate": 1.7883801679416138e-05,
+      "loss": 0.9364,
+      "step": 9958
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.71678173786245,
+      "learning_rate": 1.7883332242491726e-05,
+      "loss": 0.9119,
+      "step": 9959
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6834621505240792,
+      "learning_rate": 1.7882862759668052e-05,
+      "loss": 0.9084,
+      "step": 9960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7333540393684967,
+      "learning_rate": 1.7882393230947857e-05,
+      "loss": 0.9132,
+      "step": 9961
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.754999389541667,
+      "learning_rate": 1.7881923656333876e-05,
+      "loss": 0.817,
+      "step": 9962
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2734400188709225,
+      "learning_rate": 1.788145403582884e-05,
+      "loss": 0.754,
+      "step": 9963
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7369703726675307,
+      "learning_rate": 1.788098436943548e-05,
+      "loss": 0.9626,
+      "step": 9964
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5641338230751063,
+      "learning_rate": 1.7880514657156534e-05,
+      "loss": 0.8924,
+      "step": 9965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.78666239639182,
+      "learning_rate": 1.788004489899474e-05,
+      "loss": 0.8051,
+      "step": 9966
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7927636895465129,
+      "learning_rate": 1.7879575094952826e-05,
+      "loss": 1.0417,
+      "step": 9967
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7836145360887156,
+      "learning_rate": 1.7879105245033533e-05,
+      "loss": 0.7831,
+      "step": 9968
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8043704291062157,
+      "learning_rate": 1.7878635349239593e-05,
+      "loss": 0.9648,
+      "step": 9969
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5939612416802866,
+      "learning_rate": 1.7878165407573747e-05,
+      "loss": 0.834,
+      "step": 9970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6033822451302628,
+      "learning_rate": 1.7877695420038725e-05,
+      "loss": 0.956,
+      "step": 9971
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.100687186697513,
+      "learning_rate": 1.7877225386637266e-05,
+      "loss": 0.8391,
+      "step": 9972
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.703737738257727,
+      "learning_rate": 1.787675530737211e-05,
+      "loss": 0.9382,
+      "step": 9973
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4552641387865275,
+      "learning_rate": 1.7876285182245992e-05,
+      "loss": 0.8754,
+      "step": 9974
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6240091929668359,
+      "learning_rate": 1.7875815011261645e-05,
+      "loss": 0.9141,
+      "step": 9975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0089734025586625,
+      "learning_rate": 1.7875344794421808e-05,
+      "loss": 0.8146,
+      "step": 9976
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9193335542158667,
+      "learning_rate": 1.7874874531729225e-05,
+      "loss": 0.8501,
+      "step": 9977
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6542896018099194,
+      "learning_rate": 1.7874404223186626e-05,
+      "loss": 0.8703,
+      "step": 9978
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.621593824652844,
+      "learning_rate": 1.7873933868796753e-05,
+      "loss": 0.838,
+      "step": 9979
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6669466989330062,
+      "learning_rate": 1.7873463468562343e-05,
+      "loss": 0.9913,
+      "step": 9980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.206581002806844,
+      "learning_rate": 1.7872993022486137e-05,
+      "loss": 1.0253,
+      "step": 9981
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7629185525634923,
+      "learning_rate": 1.7872522530570876e-05,
+      "loss": 0.9224,
+      "step": 9982
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5858753285126652,
+      "learning_rate": 1.7872051992819295e-05,
+      "loss": 0.891,
+      "step": 9983
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8356584213824354,
+      "learning_rate": 1.7871581409234134e-05,
+      "loss": 0.8625,
+      "step": 9984
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7076386640467403,
+      "learning_rate": 1.787111077981813e-05,
+      "loss": 0.8056,
+      "step": 9985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.247455579024133,
+      "learning_rate": 1.787064010457403e-05,
+      "loss": 0.9205,
+      "step": 9986
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7674364980644108,
+      "learning_rate": 1.7870169383504574e-05,
+      "loss": 0.8308,
+      "step": 9987
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6636053138894116,
+      "learning_rate": 1.78696986166125e-05,
+      "loss": 0.8824,
+      "step": 9988
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6253322805844677,
+      "learning_rate": 1.7869227803900548e-05,
+      "loss": 1.0076,
+      "step": 9989
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.570172750932819,
+      "learning_rate": 1.786875694537146e-05,
+      "loss": 0.8793,
+      "step": 9990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.518861068489803,
+      "learning_rate": 1.7868286041027977e-05,
+      "loss": 0.8179,
+      "step": 9991
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6623438726177138,
+      "learning_rate": 1.7867815090872845e-05,
+      "loss": 0.8444,
+      "step": 9992
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5104887018551265,
+      "learning_rate": 1.78673440949088e-05,
+      "loss": 0.8652,
+      "step": 9993
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8443362705133863,
+      "learning_rate": 1.7866873053138588e-05,
+      "loss": 0.9245,
+      "step": 9994
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.068161859443834,
+      "learning_rate": 1.786640196556495e-05,
+      "loss": 0.962,
+      "step": 9995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7481607842066392,
+      "learning_rate": 1.786593083219063e-05,
+      "loss": 0.7485,
+      "step": 9996
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9640628517248673,
+      "learning_rate": 1.7865459653018368e-05,
+      "loss": 0.9395,
+      "step": 9997
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.106683699269206,
+      "learning_rate": 1.7864988428050912e-05,
+      "loss": 0.8278,
+      "step": 9998
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1045970281034307,
+      "learning_rate": 1.7864517157291e-05,
+      "loss": 0.8415,
+      "step": 9999
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5923121407310985,
+      "learning_rate": 1.7864045840741383e-05,
+      "loss": 0.9475,
+      "step": 10000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7681649526462822,
+      "learning_rate": 1.78635744784048e-05,
+      "loss": 0.7648,
+      "step": 10001
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5470446926230357,
+      "learning_rate": 1.7863103070283995e-05,
+      "loss": 0.9085,
+      "step": 10002
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.696113466758017,
+      "learning_rate": 1.7862631616381717e-05,
+      "loss": 0.8819,
+      "step": 10003
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9030712433635568,
+      "learning_rate": 1.786216011670071e-05,
+      "loss": 0.7873,
+      "step": 10004
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.974440640896007,
+      "learning_rate": 1.7861688571243715e-05,
+      "loss": 0.8947,
+      "step": 10005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.028504689266288,
+      "learning_rate": 1.786121698001348e-05,
+      "loss": 0.8321,
+      "step": 10006
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7841023331116321,
+      "learning_rate": 1.786074534301275e-05,
+      "loss": 0.7826,
+      "step": 10007
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.352877787739958,
+      "learning_rate": 1.7860273660244275e-05,
+      "loss": 0.8976,
+      "step": 10008
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0363603861820545,
+      "learning_rate": 1.7859801931710798e-05,
+      "loss": 0.9001,
+      "step": 10009
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5012225191414297,
+      "learning_rate": 1.7859330157415065e-05,
+      "loss": 0.8473,
+      "step": 10010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8068708408600291,
+      "learning_rate": 1.785885833735982e-05,
+      "loss": 0.8981,
+      "step": 10011
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5017535479874253,
+      "learning_rate": 1.785838647154782e-05,
+      "loss": 0.9806,
+      "step": 10012
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0134560670162447,
+      "learning_rate": 1.7857914559981805e-05,
+      "loss": 1.0275,
+      "step": 10013
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8498990912434774,
+      "learning_rate": 1.785744260266452e-05,
+      "loss": 0.9103,
+      "step": 10014
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5853617555253126,
+      "learning_rate": 1.7856970599598716e-05,
+      "loss": 0.8626,
+      "step": 10015
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5346639698667066,
+      "learning_rate": 1.7856498550787144e-05,
+      "loss": 0.9305,
+      "step": 10016
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7505202565471658,
+      "learning_rate": 1.7856026456232553e-05,
+      "loss": 0.9303,
+      "step": 10017
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.121540466528076,
+      "learning_rate": 1.7855554315937685e-05,
+      "loss": 0.8876,
+      "step": 10018
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.544488806485582,
+      "learning_rate": 1.785508212990529e-05,
+      "loss": 1.0171,
+      "step": 10019
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4647944761140685,
+      "learning_rate": 1.7854609898138126e-05,
+      "loss": 1.064,
+      "step": 10020
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5550053153490282,
+      "learning_rate": 1.7854137620638933e-05,
+      "loss": 1.0674,
+      "step": 10021
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.729856347996126,
+      "learning_rate": 1.7853665297410465e-05,
+      "loss": 0.9441,
+      "step": 10022
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.292814221502987,
+      "learning_rate": 1.7853192928455475e-05,
+      "loss": 0.9125,
+      "step": 10023
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0902748367965334,
+      "learning_rate": 1.78527205137767e-05,
+      "loss": 0.9122,
+      "step": 10024
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6107737329680054,
+      "learning_rate": 1.785224805337691e-05,
+      "loss": 0.8488,
+      "step": 10025
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7911138190754885,
+      "learning_rate": 1.785177554725884e-05,
+      "loss": 0.8403,
+      "step": 10026
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5559755265036226,
+      "learning_rate": 1.7851302995425247e-05,
+      "loss": 0.9303,
+      "step": 10027
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.89236352883873,
+      "learning_rate": 1.7850830397878888e-05,
+      "loss": 0.9713,
+      "step": 10028
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7433499853705793,
+      "learning_rate": 1.7850357754622503e-05,
+      "loss": 0.7484,
+      "step": 10029
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.702638575398631,
+      "learning_rate": 1.7849885065658852e-05,
+      "loss": 0.9434,
+      "step": 10030
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7795453693279965,
+      "learning_rate": 1.7849412330990686e-05,
+      "loss": 0.7563,
+      "step": 10031
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7129060592202037,
+      "learning_rate": 1.7848939550620755e-05,
+      "loss": 0.7578,
+      "step": 10032
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.085591527485076,
+      "learning_rate": 1.784846672455181e-05,
+      "loss": 0.907,
+      "step": 10033
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5966451910758948,
+      "learning_rate": 1.7847993852786612e-05,
+      "loss": 1.0053,
+      "step": 10034
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8911677548939647,
+      "learning_rate": 1.7847520935327905e-05,
+      "loss": 0.9726,
+      "step": 10035
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5788171725859412,
+      "learning_rate": 1.784704797217845e-05,
+      "loss": 0.8616,
+      "step": 10036
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5850127414630073,
+      "learning_rate": 1.7846574963340995e-05,
+      "loss": 0.8905,
+      "step": 10037
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.584952577531729,
+      "learning_rate": 1.78461019088183e-05,
+      "loss": 0.9561,
+      "step": 10038
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6546049594343484,
+      "learning_rate": 1.7845628808613112e-05,
+      "loss": 0.8618,
+      "step": 10039
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.600205363165436,
+      "learning_rate": 1.784515566272819e-05,
+      "loss": 0.9172,
+      "step": 10040
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5964979649623197,
+      "learning_rate": 1.7844682471166288e-05,
+      "loss": 0.9604,
+      "step": 10041
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0067819923235724,
+      "learning_rate": 1.784420923393016e-05,
+      "loss": 0.8769,
+      "step": 10042
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7364132399054095,
+      "learning_rate": 1.7843735951022565e-05,
+      "loss": 0.959,
+      "step": 10043
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6202201762383823,
+      "learning_rate": 1.7843262622446253e-05,
+      "loss": 0.8959,
+      "step": 10044
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6198663333568286,
+      "learning_rate": 1.7842789248203985e-05,
+      "loss": 1.0043,
+      "step": 10045
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6645637033959098,
+      "learning_rate": 1.7842315828298514e-05,
+      "loss": 0.7374,
+      "step": 10046
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9016611819443914,
+      "learning_rate": 1.78418423627326e-05,
+      "loss": 0.8931,
+      "step": 10047
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1064593526215787,
+      "learning_rate": 1.7841368851508995e-05,
+      "loss": 0.9142,
+      "step": 10048
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8933701998547352,
+      "learning_rate": 1.7840895294630457e-05,
+      "loss": 0.9001,
+      "step": 10049
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8829682276359196,
+      "learning_rate": 1.7840421692099746e-05,
+      "loss": 0.8454,
+      "step": 10050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7438377479692149,
+      "learning_rate": 1.7839948043919617e-05,
+      "loss": 0.8601,
+      "step": 10051
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0196000554152813,
+      "learning_rate": 1.783947435009283e-05,
+      "loss": 0.872,
+      "step": 10052
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7066856751307236,
+      "learning_rate": 1.7839000610622138e-05,
+      "loss": 0.9253,
+      "step": 10053
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.446318693955384,
+      "learning_rate": 1.7838526825510304e-05,
+      "loss": 0.9345,
+      "step": 10054
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4515940307183404,
+      "learning_rate": 1.7838052994760087e-05,
+      "loss": 0.8013,
+      "step": 10055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7058939163888516,
+      "learning_rate": 1.783757911837424e-05,
+      "loss": 0.8951,
+      "step": 10056
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5310232639083854,
+      "learning_rate": 1.783710519635553e-05,
+      "loss": 0.8384,
+      "step": 10057
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6823350777255925,
+      "learning_rate": 1.783663122870671e-05,
+      "loss": 0.8843,
+      "step": 10058
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8640917563524468,
+      "learning_rate": 1.7836157215430544e-05,
+      "loss": 0.9208,
+      "step": 10059
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6218436566413394,
+      "learning_rate": 1.7835683156529787e-05,
+      "loss": 0.8031,
+      "step": 10060
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.727804999168879,
+      "learning_rate": 1.7835209052007206e-05,
+      "loss": 1.1091,
+      "step": 10061
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6310139751609276,
+      "learning_rate": 1.7834734901865554e-05,
+      "loss": 1.0055,
+      "step": 10062
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8597216832031178,
+      "learning_rate": 1.7834260706107597e-05,
+      "loss": 0.8502,
+      "step": 10063
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8482737868114998,
+      "learning_rate": 1.7833786464736092e-05,
+      "loss": 0.908,
+      "step": 10064
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7161980313222476,
+      "learning_rate": 1.7833312177753805e-05,
+      "loss": 0.9266,
+      "step": 10065
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6615631862067668,
+      "learning_rate": 1.7832837845163492e-05,
+      "loss": 0.973,
+      "step": 10066
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3936934088527066,
+      "learning_rate": 1.7832363466967913e-05,
+      "loss": 0.9335,
+      "step": 10067
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7246165888682423,
+      "learning_rate": 1.783188904316984e-05,
+      "loss": 0.9308,
+      "step": 10068
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.547280634804794,
+      "learning_rate": 1.783141457377203e-05,
+      "loss": 0.8935,
+      "step": 10069
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4208207276730587,
+      "learning_rate": 1.7830940058777244e-05,
+      "loss": 0.9818,
+      "step": 10070
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1004098729642537,
+      "learning_rate": 1.7830465498188244e-05,
+      "loss": 0.8443,
+      "step": 10071
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.997805615120836,
+      "learning_rate": 1.7829990892007794e-05,
+      "loss": 0.8497,
+      "step": 10072
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6358559415890674,
+      "learning_rate": 1.7829516240238662e-05,
+      "loss": 0.9702,
+      "step": 10073
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5271063992770926,
+      "learning_rate": 1.7829041542883606e-05,
+      "loss": 0.9733,
+      "step": 10074
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4138451616345085,
+      "learning_rate": 1.7828566799945394e-05,
+      "loss": 0.9334,
+      "step": 10075
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7651781322803592,
+      "learning_rate": 1.782809201142678e-05,
+      "loss": 0.9692,
+      "step": 10076
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8599421707761994,
+      "learning_rate": 1.7827617177330544e-05,
+      "loss": 0.8435,
+      "step": 10077
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6788881835838028,
+      "learning_rate": 1.7827142297659435e-05,
+      "loss": 0.8483,
+      "step": 10078
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6655404509712495,
+      "learning_rate": 1.7826667372416233e-05,
+      "loss": 0.9053,
+      "step": 10079
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4665792395976627,
+      "learning_rate": 1.782619240160369e-05,
+      "loss": 0.9141,
+      "step": 10080
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.951216463537964,
+      "learning_rate": 1.782571738522458e-05,
+      "loss": 0.813,
+      "step": 10081
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7198299699700055,
+      "learning_rate": 1.7825242323281666e-05,
+      "loss": 0.9499,
+      "step": 10082
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9605638793980817,
+      "learning_rate": 1.7824767215777715e-05,
+      "loss": 0.863,
+      "step": 10083
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6915603432188135,
+      "learning_rate": 1.7824292062715488e-05,
+      "loss": 0.925,
+      "step": 10084
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.639199103738421,
+      "learning_rate": 1.7823816864097758e-05,
+      "loss": 0.7144,
+      "step": 10085
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1788644717181964,
+      "learning_rate": 1.7823341619927288e-05,
+      "loss": 0.7947,
+      "step": 10086
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7574610808552127,
+      "learning_rate": 1.782286633020685e-05,
+      "loss": 0.8795,
+      "step": 10087
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.023788261573193,
+      "learning_rate": 1.7822390994939205e-05,
+      "loss": 0.8911,
+      "step": 10088
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8798620079487027,
+      "learning_rate": 1.782191561412712e-05,
+      "loss": 0.9026,
+      "step": 10089
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5026474759231931,
+      "learning_rate": 1.782144018777337e-05,
+      "loss": 0.9563,
+      "step": 10090
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.890454462257044,
+      "learning_rate": 1.7820964715880716e-05,
+      "loss": 1.0038,
+      "step": 10091
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6711079778495823,
+      "learning_rate": 1.7820489198451932e-05,
+      "loss": 0.886,
+      "step": 10092
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.772001215106189,
+      "learning_rate": 1.7820013635489783e-05,
+      "loss": 0.8942,
+      "step": 10093
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6910741688720714,
+      "learning_rate": 1.781953802699704e-05,
+      "loss": 0.8276,
+      "step": 10094
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.569873496649646,
+      "learning_rate": 1.7819062372976467e-05,
+      "loss": 0.8416,
+      "step": 10095
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.259746203332904,
+      "learning_rate": 1.7818586673430842e-05,
+      "loss": 0.9211,
+      "step": 10096
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.880521899031333,
+      "learning_rate": 1.781811092836293e-05,
+      "loss": 0.9921,
+      "step": 10097
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.581791022301474,
+      "learning_rate": 1.78176351377755e-05,
+      "loss": 0.996,
+      "step": 10098
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8415909007999873,
+      "learning_rate": 1.781715930167132e-05,
+      "loss": 0.7161,
+      "step": 10099
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.948502545569472,
+      "learning_rate": 1.7816683420053164e-05,
+      "loss": 0.8874,
+      "step": 10100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4997770280631044,
+      "learning_rate": 1.7816207492923807e-05,
+      "loss": 0.9444,
+      "step": 10101
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5502215660728085,
+      "learning_rate": 1.781573152028601e-05,
+      "loss": 0.9068,
+      "step": 10102
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.247834699949738,
+      "learning_rate": 1.7815255502142553e-05,
+      "loss": 1.0594,
+      "step": 10103
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6989117537568037,
+      "learning_rate": 1.7814779438496207e-05,
+      "loss": 0.9659,
+      "step": 10104
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7349449982328844,
+      "learning_rate": 1.781430332934974e-05,
+      "loss": 0.7758,
+      "step": 10105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5876906709206349,
+      "learning_rate": 1.781382717470592e-05,
+      "loss": 0.7234,
+      "step": 10106
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8643071507392799,
+      "learning_rate": 1.7813350974567528e-05,
+      "loss": 0.8831,
+      "step": 10107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.398960725841143,
+      "learning_rate": 1.781287472893733e-05,
+      "loss": 0.9053,
+      "step": 10108
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.601221995421083,
+      "learning_rate": 1.7812398437818107e-05,
+      "loss": 1.0033,
+      "step": 10109
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7927540076091018,
+      "learning_rate": 1.7811922101212622e-05,
+      "loss": 0.9327,
+      "step": 10110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6388189952150567,
+      "learning_rate": 1.7811445719123654e-05,
+      "loss": 0.9363,
+      "step": 10111
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.488722328040009,
+      "learning_rate": 1.781096929155398e-05,
+      "loss": 0.7493,
+      "step": 10112
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5982996745552829,
+      "learning_rate": 1.7810492818506365e-05,
+      "loss": 0.8895,
+      "step": 10113
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6251072196486784,
+      "learning_rate": 1.7810016299983594e-05,
+      "loss": 0.815,
+      "step": 10114
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6142125826503329,
+      "learning_rate": 1.780953973598843e-05,
+      "loss": 0.9669,
+      "step": 10115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7631883180431394,
+      "learning_rate": 1.7809063126523653e-05,
+      "loss": 0.8705,
+      "step": 10116
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0102040834028765,
+      "learning_rate": 1.780858647159204e-05,
+      "loss": 0.8719,
+      "step": 10117
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7702308238388821,
+      "learning_rate": 1.7808109771196364e-05,
+      "loss": 0.7286,
+      "step": 10118
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4376893204927375,
+      "learning_rate": 1.7807633025339403e-05,
+      "loss": 0.938,
+      "step": 10119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0724524729103266,
+      "learning_rate": 1.780715623402393e-05,
+      "loss": 0.9642,
+      "step": 10120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5376198994674255,
+      "learning_rate": 1.780667939725272e-05,
+      "loss": 0.829,
+      "step": 10121
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6652217005518695,
+      "learning_rate": 1.780620251502855e-05,
+      "loss": 0.7468,
+      "step": 10122
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8843554350283966,
+      "learning_rate": 1.78057255873542e-05,
+      "loss": 0.8264,
+      "step": 10123
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8293668408638224,
+      "learning_rate": 1.780524861423244e-05,
+      "loss": 0.9077,
+      "step": 10124
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7811357604730418,
+      "learning_rate": 1.7804771595666057e-05,
+      "loss": 0.9713,
+      "step": 10125
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8190277418809049,
+      "learning_rate": 1.7804294531657822e-05,
+      "loss": 0.9302,
+      "step": 10126
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5341275763721633,
+      "learning_rate": 1.7803817422210512e-05,
+      "loss": 0.9155,
+      "step": 10127
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7116810726766434,
+      "learning_rate": 1.7803340267326907e-05,
+      "loss": 0.9196,
+      "step": 10128
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4639757158210847,
+      "learning_rate": 1.7802863067009783e-05,
+      "loss": 0.8442,
+      "step": 10129
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8263624708905966,
+      "learning_rate": 1.7802385821261922e-05,
+      "loss": 0.7944,
+      "step": 10130
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6791668251243008,
+      "learning_rate": 1.78019085300861e-05,
+      "loss": 0.9951,
+      "step": 10131
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.595515599902624,
+      "learning_rate": 1.7801431193485097e-05,
+      "loss": 0.9683,
+      "step": 10132
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6517037286871308,
+      "learning_rate": 1.780095381146169e-05,
+      "loss": 0.9731,
+      "step": 10133
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7578687733013891,
+      "learning_rate": 1.7800476384018658e-05,
+      "loss": 0.9744,
+      "step": 10134
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5016770649587388,
+      "learning_rate": 1.7799998911158786e-05,
+      "loss": 0.8105,
+      "step": 10135
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8143709392728327,
+      "learning_rate": 1.7799521392884854e-05,
+      "loss": 0.8591,
+      "step": 10136
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6429423460561101,
+      "learning_rate": 1.7799043829199635e-05,
+      "loss": 0.9561,
+      "step": 10137
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.859710589680409,
+      "learning_rate": 1.7798566220105912e-05,
+      "loss": 0.8847,
+      "step": 10138
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6251649019959153,
+      "learning_rate": 1.7798088565606473e-05,
+      "loss": 0.9355,
+      "step": 10139
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0195356586060464,
+      "learning_rate": 1.779761086570409e-05,
+      "loss": 0.8105,
+      "step": 10140
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6874259546434933,
+      "learning_rate": 1.779713312040155e-05,
+      "loss": 0.8712,
+      "step": 10141
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.757056785882247,
+      "learning_rate": 1.7796655329701627e-05,
+      "loss": 0.8008,
+      "step": 10142
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.678610773257777,
+      "learning_rate": 1.7796177493607113e-05,
+      "loss": 1.0555,
+      "step": 10143
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8216941110513827,
+      "learning_rate": 1.7795699612120784e-05,
+      "loss": 0.7555,
+      "step": 10144
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7670266120878282,
+      "learning_rate": 1.7795221685245425e-05,
+      "loss": 0.9233,
+      "step": 10145
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.624885853472311,
+      "learning_rate": 1.779474371298382e-05,
+      "loss": 0.9248,
+      "step": 10146
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.811977667393501,
+      "learning_rate": 1.7794265695338743e-05,
+      "loss": 0.8294,
+      "step": 10147
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8468530603214959,
+      "learning_rate": 1.779378763231299e-05,
+      "loss": 0.8469,
+      "step": 10148
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9713771490771916,
+      "learning_rate": 1.779330952390933e-05,
+      "loss": 0.9448,
+      "step": 10149
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8149280225766018,
+      "learning_rate": 1.779283137013056e-05,
+      "loss": 0.9449,
+      "step": 10150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8389422009002747,
+      "learning_rate": 1.779235317097946e-05,
+      "loss": 0.7914,
+      "step": 10151
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6454763832543258,
+      "learning_rate": 1.779187492645881e-05,
+      "loss": 0.7539,
+      "step": 10152
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8992105766654628,
+      "learning_rate": 1.77913966365714e-05,
+      "loss": 0.9435,
+      "step": 10153
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5716578344676069,
+      "learning_rate": 1.779091830132001e-05,
+      "loss": 0.8782,
+      "step": 10154
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6176005825126063,
+      "learning_rate": 1.7790439920707424e-05,
+      "loss": 0.9311,
+      "step": 10155
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7502342425677224,
+      "learning_rate": 1.7789961494736438e-05,
+      "loss": 0.9186,
+      "step": 10156
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9005370976932676,
+      "learning_rate": 1.7789483023409822e-05,
+      "loss": 0.871,
+      "step": 10157
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4670014162999208,
+      "learning_rate": 1.7789004506730376e-05,
+      "loss": 0.8335,
+      "step": 10158
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.747328349888087,
+      "learning_rate": 1.7788525944700873e-05,
+      "loss": 0.9156,
+      "step": 10159
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7803331523257266,
+      "learning_rate": 1.778804733732411e-05,
+      "loss": 0.8215,
+      "step": 10160
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7916744333890948,
+      "learning_rate": 1.778756868460287e-05,
+      "loss": 0.7666,
+      "step": 10161
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5424671121573277,
+      "learning_rate": 1.778708998653994e-05,
+      "loss": 0.7357,
+      "step": 10162
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5973525908192823,
+      "learning_rate": 1.7786611243138104e-05,
+      "loss": 0.9344,
+      "step": 10163
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7050602992433892,
+      "learning_rate": 1.7786132454400155e-05,
+      "loss": 0.8516,
+      "step": 10164
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8730114476592126,
+      "learning_rate": 1.778565362032888e-05,
+      "loss": 0.8443,
+      "step": 10165
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9444032996555178,
+      "learning_rate": 1.778517474092706e-05,
+      "loss": 0.8795,
+      "step": 10166
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5226048390219447,
+      "learning_rate": 1.778469581619749e-05,
+      "loss": 0.8962,
+      "step": 10167
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.705999749471544,
+      "learning_rate": 1.7784216846142955e-05,
+      "loss": 0.8208,
+      "step": 10168
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.051721555647313,
+      "learning_rate": 1.7783737830766247e-05,
+      "loss": 0.8868,
+      "step": 10169
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.010188085255848,
+      "learning_rate": 1.7783258770070153e-05,
+      "loss": 0.9282,
+      "step": 10170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.903919826831854,
+      "learning_rate": 1.778277966405746e-05,
+      "loss": 0.9303,
+      "step": 10171
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8272118747913415,
+      "learning_rate": 1.7782300512730965e-05,
+      "loss": 1.0412,
+      "step": 10172
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7808147393629877,
+      "learning_rate": 1.7781821316093445e-05,
+      "loss": 0.7918,
+      "step": 10173
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.389743635967409,
+      "learning_rate": 1.7781342074147706e-05,
+      "loss": 0.9816,
+      "step": 10174
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5897162742240178,
+      "learning_rate": 1.7780862786896526e-05,
+      "loss": 0.9618,
+      "step": 10175
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5666908549963412,
+      "learning_rate": 1.7780383454342696e-05,
+      "loss": 0.9015,
+      "step": 10176
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.043451252169168,
+      "learning_rate": 1.7779904076489013e-05,
+      "loss": 0.9372,
+      "step": 10177
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8500564790657765,
+      "learning_rate": 1.7779424653338268e-05,
+      "loss": 0.9648,
+      "step": 10178
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.613237641162368,
+      "learning_rate": 1.7778945184893245e-05,
+      "loss": 0.8712,
+      "step": 10179
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7433943211053948,
+      "learning_rate": 1.7778465671156744e-05,
+      "loss": 0.9283,
+      "step": 10180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7525643904418223,
+      "learning_rate": 1.777798611213155e-05,
+      "loss": 0.696,
+      "step": 10181
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6421554980614474,
+      "learning_rate": 1.777750650782046e-05,
+      "loss": 0.9092,
+      "step": 10182
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.516680443895265,
+      "learning_rate": 1.7777026858226265e-05,
+      "loss": 1.0029,
+      "step": 10183
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6739582925880379,
+      "learning_rate": 1.7776547163351758e-05,
+      "loss": 0.9407,
+      "step": 10184
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7891808031908785,
+      "learning_rate": 1.777606742319973e-05,
+      "loss": 0.9949,
+      "step": 10185
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6823868410856662,
+      "learning_rate": 1.777558763777298e-05,
+      "loss": 0.9061,
+      "step": 10186
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.628308520971308,
+      "learning_rate": 1.777510780707429e-05,
+      "loss": 0.9794,
+      "step": 10187
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.393848903938817,
+      "learning_rate": 1.7774627931106464e-05,
+      "loss": 0.9851,
+      "step": 10188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7320050593369023,
+      "learning_rate": 1.7774148009872295e-05,
+      "loss": 0.9162,
+      "step": 10189
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.513980452536214,
+      "learning_rate": 1.7773668043374573e-05,
+      "loss": 0.8433,
+      "step": 10190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.601357993274315,
+      "learning_rate": 1.7773188031616096e-05,
+      "loss": 0.8704,
+      "step": 10191
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6061273904657416,
+      "learning_rate": 1.7772707974599656e-05,
+      "loss": 0.9053,
+      "step": 10192
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.221501001702175,
+      "learning_rate": 1.777222787232805e-05,
+      "loss": 0.8941,
+      "step": 10193
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6271264949412751,
+      "learning_rate": 1.7771747724804073e-05,
+      "loss": 0.9592,
+      "step": 10194
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.056732130952754,
+      "learning_rate": 1.777126753203052e-05,
+      "loss": 0.7954,
+      "step": 10195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5799323132543663,
+      "learning_rate": 1.7770787294010185e-05,
+      "loss": 0.9002,
+      "step": 10196
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.70351161332933,
+      "learning_rate": 1.7770307010745867e-05,
+      "loss": 0.8455,
+      "step": 10197
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6429715089827526,
+      "learning_rate": 1.7769826682240365e-05,
+      "loss": 0.9635,
+      "step": 10198
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8327675821641944,
+      "learning_rate": 1.7769346308496466e-05,
+      "loss": 0.9434,
+      "step": 10199
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.765795613777686,
+      "learning_rate": 1.776886588951698e-05,
+      "loss": 0.9372,
+      "step": 10200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7290009421072963,
+      "learning_rate": 1.7768385425304692e-05,
+      "loss": 0.9273,
+      "step": 10201
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6592533323858787,
+      "learning_rate": 1.7767904915862405e-05,
+      "loss": 0.8062,
+      "step": 10202
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8453412564235325,
+      "learning_rate": 1.776742436119292e-05,
+      "loss": 0.9009,
+      "step": 10203
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8082075244013833,
+      "learning_rate": 1.776694376129903e-05,
+      "loss": 0.7703,
+      "step": 10204
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6550172234393223,
+      "learning_rate": 1.776646311618353e-05,
+      "loss": 0.8656,
+      "step": 10205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1906157665461548,
+      "learning_rate": 1.776598242584923e-05,
+      "loss": 0.8708,
+      "step": 10206
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6520355484219824,
+      "learning_rate": 1.7765501690298913e-05,
+      "loss": 0.9707,
+      "step": 10207
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8015528583807534,
+      "learning_rate": 1.7765020909535395e-05,
+      "loss": 0.9921,
+      "step": 10208
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8906752433078489,
+      "learning_rate": 1.7764540083561462e-05,
+      "loss": 0.8562,
+      "step": 10209
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7558470311264023,
+      "learning_rate": 1.776405921237992e-05,
+      "loss": 1.1027,
+      "step": 10210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6570842592398685,
+      "learning_rate": 1.776357829599357e-05,
+      "loss": 0.8385,
+      "step": 10211
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8266759900811718,
+      "learning_rate": 1.7763097334405204e-05,
+      "loss": 0.9848,
+      "step": 10212
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6515215739724953,
+      "learning_rate": 1.776261632761763e-05,
+      "loss": 0.8219,
+      "step": 10213
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.765360704012878,
+      "learning_rate": 1.7762135275633645e-05,
+      "loss": 0.8047,
+      "step": 10214
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9417762014091975,
+      "learning_rate": 1.7761654178456053e-05,
+      "loss": 0.8164,
+      "step": 10215
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.811534289593593,
+      "learning_rate": 1.776117303608765e-05,
+      "loss": 0.9547,
+      "step": 10216
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6513368871250593,
+      "learning_rate": 1.7760691848531246e-05,
+      "loss": 1.0282,
+      "step": 10217
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9423796745481599,
+      "learning_rate": 1.7760210615789633e-05,
+      "loss": 0.9558,
+      "step": 10218
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0310278950761544,
+      "learning_rate": 1.7759729337865618e-05,
+      "loss": 0.8663,
+      "step": 10219
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8232622394461475,
+      "learning_rate": 1.7759248014762e-05,
+      "loss": 0.7163,
+      "step": 10220
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8548753941288592,
+      "learning_rate": 1.7758766646481587e-05,
+      "loss": 0.9002,
+      "step": 10221
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5289470555803082,
+      "learning_rate": 1.7758285233027174e-05,
+      "loss": 0.958,
+      "step": 10222
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.640234742410995,
+      "learning_rate": 1.7757803774401572e-05,
+      "loss": 0.833,
+      "step": 10223
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6734361345876771,
+      "learning_rate": 1.775732227060758e-05,
+      "loss": 0.8398,
+      "step": 10224
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7988350874006022,
+      "learning_rate": 1.7756840721648e-05,
+      "loss": 0.7116,
+      "step": 10225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4827292581118683,
+      "learning_rate": 1.7756359127525638e-05,
+      "loss": 0.9823,
+      "step": 10226
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5350987823687245,
+      "learning_rate": 1.7755877488243297e-05,
+      "loss": 0.9599,
+      "step": 10227
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5910473712402502,
+      "learning_rate": 1.7755395803803783e-05,
+      "loss": 0.8692,
+      "step": 10228
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8676980910495857,
+      "learning_rate": 1.7754914074209898e-05,
+      "loss": 0.8268,
+      "step": 10229
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7968653984111318,
+      "learning_rate": 1.7754432299464448e-05,
+      "loss": 0.851,
+      "step": 10230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4989067404790732,
+      "learning_rate": 1.775395047957024e-05,
+      "loss": 0.929,
+      "step": 10231
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5707692374586104,
+      "learning_rate": 1.7753468614530075e-05,
+      "loss": 0.7857,
+      "step": 10232
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8314232714533318,
+      "learning_rate": 1.7752986704346764e-05,
+      "loss": 0.7442,
+      "step": 10233
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7663710212214853,
+      "learning_rate": 1.7752504749023108e-05,
+      "loss": 0.6831,
+      "step": 10234
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5380289852765503,
+      "learning_rate": 1.7752022748561917e-05,
+      "loss": 0.9032,
+      "step": 10235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.733620446135317,
+      "learning_rate": 1.7751540702965993e-05,
+      "loss": 0.9037,
+      "step": 10236
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7736885218932713,
+      "learning_rate": 1.7751058612238147e-05,
+      "loss": 0.758,
+      "step": 10237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6891043687555267,
+      "learning_rate": 1.775057647638118e-05,
+      "loss": 1.0216,
+      "step": 10238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7389368332549806,
+      "learning_rate": 1.7750094295397907e-05,
+      "loss": 0.9291,
+      "step": 10239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8051894220527345,
+      "learning_rate": 1.774961206929113e-05,
+      "loss": 1.0299,
+      "step": 10240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.529120075614959,
+      "learning_rate": 1.7749129798063657e-05,
+      "loss": 0.8088,
+      "step": 10241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.188378026729015,
+      "learning_rate": 1.77486474817183e-05,
+      "loss": 0.9192,
+      "step": 10242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9952794505208116,
+      "learning_rate": 1.774816512025786e-05,
+      "loss": 0.9254,
+      "step": 10243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.655941571595956,
+      "learning_rate": 1.7747682713685154e-05,
+      "loss": 0.8212,
+      "step": 10244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7861770254496907,
+      "learning_rate": 1.7747200262002983e-05,
+      "loss": 0.8264,
+      "step": 10245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.58850864045638,
+      "learning_rate": 1.774671776521416e-05,
+      "loss": 0.8646,
+      "step": 10246
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4596046265927325,
+      "learning_rate": 1.7746235223321492e-05,
+      "loss": 0.9553,
+      "step": 10247
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.810156099145191,
+      "learning_rate": 1.7745752636327793e-05,
+      "loss": 0.7831,
+      "step": 10248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5606502765878636,
+      "learning_rate": 1.774527000423587e-05,
+      "loss": 0.8169,
+      "step": 10249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5628698668958205,
+      "learning_rate": 1.7744787327048533e-05,
+      "loss": 0.7788,
+      "step": 10250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7122334135426962,
+      "learning_rate": 1.7744304604768588e-05,
+      "loss": 0.9538,
+      "step": 10251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7085988889605703,
+      "learning_rate": 1.7743821837398854e-05,
+      "loss": 0.8981,
+      "step": 10252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6976674110215064,
+      "learning_rate": 1.7743339024942135e-05,
+      "loss": 0.957,
+      "step": 10253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6097633735880919,
+      "learning_rate": 1.7742856167401248e-05,
+      "loss": 1.0284,
+      "step": 10254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8441670831769237,
+      "learning_rate": 1.7742373264778997e-05,
+      "loss": 0.8683,
+      "step": 10255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8229754500920858,
+      "learning_rate": 1.7741890317078205e-05,
+      "loss": 0.8056,
+      "step": 10256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.472517167693316,
+      "learning_rate": 1.774140732430167e-05,
+      "loss": 0.7977,
+      "step": 10257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8813060793000158,
+      "learning_rate": 1.774092428645221e-05,
+      "loss": 0.9535,
+      "step": 10258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.417511389241043,
+      "learning_rate": 1.7740441203532638e-05,
+      "loss": 0.8183,
+      "step": 10259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7114875306670458,
+      "learning_rate": 1.7739958075545772e-05,
+      "loss": 0.946,
+      "step": 10260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5984049158835347,
+      "learning_rate": 1.7739474902494416e-05,
+      "loss": 0.771,
+      "step": 10261
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7029621751915547,
+      "learning_rate": 1.7738991684381387e-05,
+      "loss": 0.8406,
+      "step": 10262
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1548618644551722,
+      "learning_rate": 1.77385084212095e-05,
+      "loss": 0.8132,
+      "step": 10263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6731396450090386,
+      "learning_rate": 1.7738025112981565e-05,
+      "loss": 0.9741,
+      "step": 10264
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8524912031160171,
+      "learning_rate": 1.7737541759700398e-05,
+      "loss": 0.76,
+      "step": 10265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6816746103458493,
+      "learning_rate": 1.7737058361368814e-05,
+      "loss": 0.743,
+      "step": 10266
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5616642983933517,
+      "learning_rate": 1.7736574917989627e-05,
+      "loss": 0.8528,
+      "step": 10267
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.774542071027543,
+      "learning_rate": 1.773609142956565e-05,
+      "loss": 0.8561,
+      "step": 10268
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7231616790274433,
+      "learning_rate": 1.77356078960997e-05,
+      "loss": 0.9791,
+      "step": 10269
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8974681088051574,
+      "learning_rate": 1.773512431759459e-05,
+      "loss": 0.7675,
+      "step": 10270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4427836312775568,
+      "learning_rate": 1.7734640694053143e-05,
+      "loss": 1.0007,
+      "step": 10271
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7134485724716915,
+      "learning_rate": 1.7734157025478164e-05,
+      "loss": 0.9238,
+      "step": 10272
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7562688569404027,
+      "learning_rate": 1.7733673311872478e-05,
+      "loss": 0.7535,
+      "step": 10273
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7799811278221798,
+      "learning_rate": 1.7733189553238893e-05,
+      "loss": 0.9229,
+      "step": 10274
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9240374003205156,
+      "learning_rate": 1.7732705749580235e-05,
+      "loss": 0.9088,
+      "step": 10275
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8007975851768313,
+      "learning_rate": 1.7732221900899313e-05,
+      "loss": 0.7527,
+      "step": 10276
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5999106175474698,
+      "learning_rate": 1.773173800719895e-05,
+      "loss": 0.7869,
+      "step": 10277
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6894433942380889,
+      "learning_rate": 1.7731254068481955e-05,
+      "loss": 0.8809,
+      "step": 10278
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7849382535946762,
+      "learning_rate": 1.7730770084751156e-05,
+      "loss": 0.7438,
+      "step": 10279
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8328355524760467,
+      "learning_rate": 1.7730286056009363e-05,
+      "loss": 0.8736,
+      "step": 10280
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8139735719026695,
+      "learning_rate": 1.77298019822594e-05,
+      "loss": 1.05,
+      "step": 10281
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5196606072010514,
+      "learning_rate": 1.7729317863504082e-05,
+      "loss": 0.8194,
+      "step": 10282
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.238084823753203,
+      "learning_rate": 1.7728833699746226e-05,
+      "loss": 0.8877,
+      "step": 10283
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6457367306242816,
+      "learning_rate": 1.7728349490988653e-05,
+      "loss": 0.9787,
+      "step": 10284
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.571046234973575,
+      "learning_rate": 1.7727865237234187e-05,
+      "loss": 0.9299,
+      "step": 10285
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.496981183056459,
+      "learning_rate": 1.7727380938485642e-05,
+      "loss": 0.7797,
+      "step": 10286
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7150221990242924,
+      "learning_rate": 1.7726896594745834e-05,
+      "loss": 0.8711,
+      "step": 10287
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8233754605941743,
+      "learning_rate": 1.7726412206017593e-05,
+      "loss": 0.8147,
+      "step": 10288
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6800217421024188,
+      "learning_rate": 1.772592777230373e-05,
+      "loss": 0.8672,
+      "step": 10289
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4936057310213584,
+      "learning_rate": 1.7725443293607074e-05,
+      "loss": 0.8448,
+      "step": 10290
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7879435316486025,
+      "learning_rate": 1.7724958769930437e-05,
+      "loss": 0.7733,
+      "step": 10291
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.807593941039939,
+      "learning_rate": 1.7724474201276646e-05,
+      "loss": 0.82,
+      "step": 10292
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5475610425417228,
+      "learning_rate": 1.772398958764852e-05,
+      "loss": 0.9434,
+      "step": 10293
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7390265645602901,
+      "learning_rate": 1.7723504929048886e-05,
+      "loss": 0.7614,
+      "step": 10294
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7089595652380225,
+      "learning_rate": 1.7723020225480554e-05,
+      "loss": 0.9076,
+      "step": 10295
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8087520609572683,
+      "learning_rate": 1.7722535476946357e-05,
+      "loss": 0.8516,
+      "step": 10296
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.70908195428198,
+      "learning_rate": 1.7722050683449114e-05,
+      "loss": 0.9311,
+      "step": 10297
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9878524296358524,
+      "learning_rate": 1.7721565844991643e-05,
+      "loss": 0.8637,
+      "step": 10298
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9791440995561325,
+      "learning_rate": 1.7721080961576776e-05,
+      "loss": 1.0061,
+      "step": 10299
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7781050697161983,
+      "learning_rate": 1.7720596033207327e-05,
+      "loss": 0.7765,
+      "step": 10300
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.572229506391693,
+      "learning_rate": 1.7720111059886127e-05,
+      "loss": 0.9846,
+      "step": 10301
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4717801197334524,
+      "learning_rate": 1.7719626041615992e-05,
+      "loss": 0.8205,
+      "step": 10302
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4347740783360916,
+      "learning_rate": 1.7719140978399755e-05,
+      "loss": 0.8419,
+      "step": 10303
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0240194080448504,
+      "learning_rate": 1.7718655870240234e-05,
+      "loss": 0.8154,
+      "step": 10304
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.625331012041727,
+      "learning_rate": 1.7718170717140256e-05,
+      "loss": 0.9553,
+      "step": 10305
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5059696256027328,
+      "learning_rate": 1.771768551910264e-05,
+      "loss": 0.9213,
+      "step": 10306
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.822228379909634,
+      "learning_rate": 1.771720027613022e-05,
+      "loss": 0.8928,
+      "step": 10307
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.724261805020058,
+      "learning_rate": 1.7716714988225815e-05,
+      "loss": 1.0919,
+      "step": 10308
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.053930280657707,
+      "learning_rate": 1.7716229655392255e-05,
+      "loss": 0.9611,
+      "step": 10309
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7227874101298848,
+      "learning_rate": 1.771574427763236e-05,
+      "loss": 1.021,
+      "step": 10310
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5432921295759487,
+      "learning_rate": 1.771525885494896e-05,
+      "loss": 0.991,
+      "step": 10311
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6607194514768242,
+      "learning_rate": 1.7714773387344878e-05,
+      "loss": 0.7902,
+      "step": 10312
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6891529484978967,
+      "learning_rate": 1.7714287874822945e-05,
+      "loss": 0.8564,
+      "step": 10313
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6642565544826748,
+      "learning_rate": 1.7713802317385985e-05,
+      "loss": 0.8765,
+      "step": 10314
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.780542618470705,
+      "learning_rate": 1.7713316715036827e-05,
+      "loss": 0.8663,
+      "step": 10315
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.603295593659883,
+      "learning_rate": 1.7712831067778298e-05,
+      "loss": 0.9266,
+      "step": 10316
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8039856607716205,
+      "learning_rate": 1.771234537561322e-05,
+      "loss": 0.7201,
+      "step": 10317
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8164240211340428,
+      "learning_rate": 1.771185963854443e-05,
+      "loss": 0.9241,
+      "step": 10318
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6405426724957208,
+      "learning_rate": 1.771137385657475e-05,
+      "loss": 0.8252,
+      "step": 10319
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.545203232075912,
+      "learning_rate": 1.771088802970701e-05,
+      "loss": 0.8388,
+      "step": 10320
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7829415155498096,
+      "learning_rate": 1.771040215794404e-05,
+      "loss": 0.9078,
+      "step": 10321
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4812728643397464,
+      "learning_rate": 1.7709916241288667e-05,
+      "loss": 0.9048,
+      "step": 10322
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6527179787135122,
+      "learning_rate": 1.770943027974372e-05,
+      "loss": 0.9755,
+      "step": 10323
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.757878254283148,
+      "learning_rate": 1.770894427331203e-05,
+      "loss": 0.9377,
+      "step": 10324
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9108244566340051,
+      "learning_rate": 1.7708458221996426e-05,
+      "loss": 0.85,
+      "step": 10325
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.086887507025372,
+      "learning_rate": 1.7707972125799738e-05,
+      "loss": 0.7635,
+      "step": 10326
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5832150446715498,
+      "learning_rate": 1.7707485984724795e-05,
+      "loss": 1.0101,
+      "step": 10327
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.028536868260782,
+      "learning_rate": 1.7706999798774426e-05,
+      "loss": 0.9373,
+      "step": 10328
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8227240005676846,
+      "learning_rate": 1.7706513567951467e-05,
+      "loss": 0.8371,
+      "step": 10329
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7778162282220673,
+      "learning_rate": 1.7706027292258747e-05,
+      "loss": 0.9587,
+      "step": 10330
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7246508834316119,
+      "learning_rate": 1.7705540971699095e-05,
+      "loss": 0.8814,
+      "step": 10331
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8976408012677528,
+      "learning_rate": 1.7705054606275343e-05,
+      "loss": 0.9002,
+      "step": 10332
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7586216322485875,
+      "learning_rate": 1.7704568195990324e-05,
+      "loss": 0.7836,
+      "step": 10333
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5360839250951017,
+      "learning_rate": 1.770408174084687e-05,
+      "loss": 0.8674,
+      "step": 10334
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6992992668251816,
+      "learning_rate": 1.7703595240847815e-05,
+      "loss": 0.9248,
+      "step": 10335
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7618532025825888,
+      "learning_rate": 1.770310869599599e-05,
+      "loss": 0.6802,
+      "step": 10336
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6709459129353492,
+      "learning_rate": 1.770262210629422e-05,
+      "loss": 0.9031,
+      "step": 10337
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5835423355823046,
+      "learning_rate": 1.7702135471745358e-05,
+      "loss": 0.9637,
+      "step": 10338
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6266753154264872,
+      "learning_rate": 1.7701648792352215e-05,
+      "loss": 0.9134,
+      "step": 10339
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8193409150765801,
+      "learning_rate": 1.7701162068117636e-05,
+      "loss": 0.7525,
+      "step": 10340
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7540219579602387,
+      "learning_rate": 1.7700675299044457e-05,
+      "loss": 0.7179,
+      "step": 10341
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5798422141605768,
+      "learning_rate": 1.7700188485135506e-05,
+      "loss": 0.8379,
+      "step": 10342
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9928111563701372,
+      "learning_rate": 1.769970162639362e-05,
+      "loss": 0.8534,
+      "step": 10343
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6241348198514287,
+      "learning_rate": 1.7699214722821634e-05,
+      "loss": 0.7424,
+      "step": 10344
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4298040233401177,
+      "learning_rate": 1.769872777442238e-05,
+      "loss": 0.8382,
+      "step": 10345
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.619500768625262,
+      "learning_rate": 1.7698240781198696e-05,
+      "loss": 0.9247,
+      "step": 10346
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6018513535866952,
+      "learning_rate": 1.769775374315342e-05,
+      "loss": 0.9349,
+      "step": 10347
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8258358742217409,
+      "learning_rate": 1.7697266660289383e-05,
+      "loss": 0.9318,
+      "step": 10348
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.578180941951075,
+      "learning_rate": 1.769677953260942e-05,
+      "loss": 0.9769,
+      "step": 10349
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7998897459429484,
+      "learning_rate": 1.7696292360116373e-05,
+      "loss": 0.7948,
+      "step": 10350
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.719633566098783,
+      "learning_rate": 1.7695805142813073e-05,
+      "loss": 0.6987,
+      "step": 10351
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7360845627184953,
+      "learning_rate": 1.769531788070236e-05,
+      "loss": 0.8419,
+      "step": 10352
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7552518075386401,
+      "learning_rate": 1.7694830573787075e-05,
+      "loss": 0.7293,
+      "step": 10353
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5594363111535996,
+      "learning_rate": 1.769434322207004e-05,
+      "loss": 0.8624,
+      "step": 10354
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9911773685127816,
+      "learning_rate": 1.7693855825554113e-05,
+      "loss": 0.8336,
+      "step": 10355
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.852243381814298,
+      "learning_rate": 1.7693368384242115e-05,
+      "loss": 0.8701,
+      "step": 10356
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7848670522170094,
+      "learning_rate": 1.7692880898136894e-05,
+      "loss": 0.7602,
+      "step": 10357
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.439178140910759,
+      "learning_rate": 1.7692393367241283e-05,
+      "loss": 0.8467,
+      "step": 10358
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7116159751231659,
+      "learning_rate": 1.7691905791558123e-05,
+      "loss": 1.0978,
+      "step": 10359
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6859875355228187,
+      "learning_rate": 1.7691418171090254e-05,
+      "loss": 0.8442,
+      "step": 10360
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6706888321268938,
+      "learning_rate": 1.7690930505840513e-05,
+      "loss": 0.9661,
+      "step": 10361
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1335495816040315,
+      "learning_rate": 1.7690442795811734e-05,
+      "loss": 0.9044,
+      "step": 10362
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6491126626385615,
+      "learning_rate": 1.7689955041006768e-05,
+      "loss": 0.9535,
+      "step": 10363
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7455707960741793,
+      "learning_rate": 1.7689467241428446e-05,
+      "loss": 0.7619,
+      "step": 10364
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5254705307179133,
+      "learning_rate": 1.7688979397079612e-05,
+      "loss": 0.7893,
+      "step": 10365
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4669517059251884,
+      "learning_rate": 1.7688491507963108e-05,
+      "loss": 0.9621,
+      "step": 10366
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.685899340535082,
+      "learning_rate": 1.7688003574081773e-05,
+      "loss": 0.8623,
+      "step": 10367
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5847213478130286,
+      "learning_rate": 1.7687515595438443e-05,
+      "loss": 0.9574,
+      "step": 10368
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7030025489970217,
+      "learning_rate": 1.7687027572035965e-05,
+      "loss": 0.8138,
+      "step": 10369
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8219852756057503,
+      "learning_rate": 1.768653950387718e-05,
+      "loss": 0.7551,
+      "step": 10370
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8682651356897686,
+      "learning_rate": 1.768605139096493e-05,
+      "loss": 0.8797,
+      "step": 10371
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6937873494683235,
+      "learning_rate": 1.7685563233302053e-05,
+      "loss": 0.9212,
+      "step": 10372
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4872816621609881,
+      "learning_rate": 1.7685075030891393e-05,
+      "loss": 0.8927,
+      "step": 10373
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.683435350223454,
+      "learning_rate": 1.7684586783735793e-05,
+      "loss": 0.9019,
+      "step": 10374
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6520378000430442,
+      "learning_rate": 1.7684098491838097e-05,
+      "loss": 0.9032,
+      "step": 10375
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7819500227390668,
+      "learning_rate": 1.7683610155201148e-05,
+      "loss": 0.7523,
+      "step": 10376
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.746606522471857,
+      "learning_rate": 1.7683121773827787e-05,
+      "loss": 0.9871,
+      "step": 10377
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8009841958165406,
+      "learning_rate": 1.7682633347720858e-05,
+      "loss": 0.8176,
+      "step": 10378
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6473187628046082,
+      "learning_rate": 1.7682144876883207e-05,
+      "loss": 0.9673,
+      "step": 10379
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.523422631376866,
+      "learning_rate": 1.7681656361317676e-05,
+      "loss": 0.9444,
+      "step": 10380
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7214537372948098,
+      "learning_rate": 1.7681167801027107e-05,
+      "loss": 0.9491,
+      "step": 10381
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6423025204057597,
+      "learning_rate": 1.7680679196014348e-05,
+      "loss": 0.9014,
+      "step": 10382
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8351178092044425,
+      "learning_rate": 1.7680190546282244e-05,
+      "loss": 0.9541,
+      "step": 10383
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.531402404856524,
+      "learning_rate": 1.767970185183364e-05,
+      "loss": 1.1012,
+      "step": 10384
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7559845586237988,
+      "learning_rate": 1.7679213112671378e-05,
+      "loss": 0.8105,
+      "step": 10385
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.504050237769019,
+      "learning_rate": 1.7678724328798308e-05,
+      "loss": 0.8272,
+      "step": 10386
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.737067276984505,
+      "learning_rate": 1.7678235500217274e-05,
+      "loss": 0.8748,
+      "step": 10387
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4776750564707033,
+      "learning_rate": 1.7677746626931122e-05,
+      "loss": 0.9344,
+      "step": 10388
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7847623350847466,
+      "learning_rate": 1.7677257708942697e-05,
+      "loss": 0.8676,
+      "step": 10389
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9387043750915065,
+      "learning_rate": 1.767676874625485e-05,
+      "loss": 0.9626,
+      "step": 10390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8866784444274034,
+      "learning_rate": 1.7676279738870423e-05,
+      "loss": 0.9465,
+      "step": 10391
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6892054674302435,
+      "learning_rate": 1.7675790686792264e-05,
+      "loss": 0.9013,
+      "step": 10392
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7725759822905098,
+      "learning_rate": 1.7675301590023225e-05,
+      "loss": 0.8739,
+      "step": 10393
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.76486465725936,
+      "learning_rate": 1.7674812448566147e-05,
+      "loss": 1.0183,
+      "step": 10394
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.301578848982092,
+      "learning_rate": 1.7674323262423883e-05,
+      "loss": 1.0772,
+      "step": 10395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6563315224750155,
+      "learning_rate": 1.767383403159928e-05,
+      "loss": 0.912,
+      "step": 10396
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7617862123279304,
+      "learning_rate": 1.767334475609518e-05,
+      "loss": 0.9656,
+      "step": 10397
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6786116077030375,
+      "learning_rate": 1.7672855435914442e-05,
+      "loss": 0.9141,
+      "step": 10398
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.627042113803595,
+      "learning_rate": 1.767236607105991e-05,
+      "loss": 0.9814,
+      "step": 10399
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0152481535945825,
+      "learning_rate": 1.7671876661534436e-05,
+      "loss": 0.7432,
+      "step": 10400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6911137883082927,
+      "learning_rate": 1.7671387207340864e-05,
+      "loss": 0.8724,
+      "step": 10401
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7639228531606563,
+      "learning_rate": 1.767089770848205e-05,
+      "loss": 0.8839,
+      "step": 10402
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8702812966932318,
+      "learning_rate": 1.7670408164960833e-05,
+      "loss": 0.9215,
+      "step": 10403
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5291043101285253,
+      "learning_rate": 1.766991857678008e-05,
+      "loss": 0.8897,
+      "step": 10404
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8089562579564398,
+      "learning_rate": 1.766942894394263e-05,
+      "loss": 0.8588,
+      "step": 10405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.906009902525376,
+      "learning_rate": 1.7668939266451337e-05,
+      "loss": 0.8853,
+      "step": 10406
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.469139036263563,
+      "learning_rate": 1.7668449544309053e-05,
+      "loss": 0.9503,
+      "step": 10407
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7933266804618545,
+      "learning_rate": 1.7667959777518622e-05,
+      "loss": 0.7262,
+      "step": 10408
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8851126559815217,
+      "learning_rate": 1.7667469966082908e-05,
+      "loss": 0.8857,
+      "step": 10409
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.516152693632411,
+      "learning_rate": 1.7666980110004756e-05,
+      "loss": 0.9588,
+      "step": 10410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.742063741260133,
+      "learning_rate": 1.7666490209287015e-05,
+      "loss": 0.9482,
+      "step": 10411
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.76876264404159,
+      "learning_rate": 1.7666000263932542e-05,
+      "loss": 0.854,
+      "step": 10412
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2579670648927466,
+      "learning_rate": 1.766551027394419e-05,
+      "loss": 0.9379,
+      "step": 10413
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.320473026092424,
+      "learning_rate": 1.7665020239324808e-05,
+      "loss": 0.8311,
+      "step": 10414
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7437656674250828,
+      "learning_rate": 1.7664530160077252e-05,
+      "loss": 0.8824,
+      "step": 10415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1468367124854186,
+      "learning_rate": 1.7664040036204375e-05,
+      "loss": 0.9128,
+      "step": 10416
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5910568242060643,
+      "learning_rate": 1.766354986770903e-05,
+      "loss": 0.9572,
+      "step": 10417
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3756679873959954,
+      "learning_rate": 1.7663059654594073e-05,
+      "loss": 0.8828,
+      "step": 10418
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3651728293273204,
+      "learning_rate": 1.766256939686236e-05,
+      "loss": 1.0459,
+      "step": 10419
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.562622239725458,
+      "learning_rate": 1.7662079094516735e-05,
+      "loss": 0.8403,
+      "step": 10420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.288133551092626,
+      "learning_rate": 1.7661588747560062e-05,
+      "loss": 0.8437,
+      "step": 10421
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6474068809594098,
+      "learning_rate": 1.7661098355995195e-05,
+      "loss": 0.9548,
+      "step": 10422
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7481240359051002,
+      "learning_rate": 1.7660607919824988e-05,
+      "loss": 0.93,
+      "step": 10423
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.716834273489727,
+      "learning_rate": 1.76601174390523e-05,
+      "loss": 0.9336,
+      "step": 10424
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5997255941405515,
+      "learning_rate": 1.7659626913679977e-05,
+      "loss": 0.8406,
+      "step": 10425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7988950412784703,
+      "learning_rate": 1.765913634371088e-05,
+      "loss": 0.8586,
+      "step": 10426
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9835331290788025,
+      "learning_rate": 1.765864572914787e-05,
+      "loss": 0.7841,
+      "step": 10427
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8242258323996783,
+      "learning_rate": 1.7658155069993802e-05,
+      "loss": 0.7664,
+      "step": 10428
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8542204313360756,
+      "learning_rate": 1.7657664366251527e-05,
+      "loss": 0.7416,
+      "step": 10429
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4248878319602445,
+      "learning_rate": 1.7657173617923905e-05,
+      "loss": 0.8449,
+      "step": 10430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.951418333292033,
+      "learning_rate": 1.7656682825013794e-05,
+      "loss": 0.9585,
+      "step": 10431
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7424888482523282,
+      "learning_rate": 1.7656191987524055e-05,
+      "loss": 0.9063,
+      "step": 10432
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9363630626018065,
+      "learning_rate": 1.765570110545754e-05,
+      "loss": 1.009,
+      "step": 10433
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7496727798716031,
+      "learning_rate": 1.765521017881711e-05,
+      "loss": 0.7709,
+      "step": 10434
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.64088471707785,
+      "learning_rate": 1.7654719207605625e-05,
+      "loss": 0.7535,
+      "step": 10435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8064833348717755,
+      "learning_rate": 1.7654228191825937e-05,
+      "loss": 0.7194,
+      "step": 10436
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4470718394376136,
+      "learning_rate": 1.7653737131480914e-05,
+      "loss": 0.9048,
+      "step": 10437
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.529849826150956,
+      "learning_rate": 1.7653246026573404e-05,
+      "loss": 0.8469,
+      "step": 10438
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7448809361946501,
+      "learning_rate": 1.7652754877106275e-05,
+      "loss": 0.7329,
+      "step": 10439
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8489367205768121,
+      "learning_rate": 1.7652263683082388e-05,
+      "loss": 0.9819,
+      "step": 10440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.68462093028439,
+      "learning_rate": 1.7651772444504594e-05,
+      "loss": 0.8386,
+      "step": 10441
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2223297972345653,
+      "learning_rate": 1.765128116137576e-05,
+      "loss": 0.8375,
+      "step": 10442
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5467193009185465,
+      "learning_rate": 1.765078983369875e-05,
+      "loss": 0.8508,
+      "step": 10443
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7963514723421552,
+      "learning_rate": 1.765029846147641e-05,
+      "loss": 0.8647,
+      "step": 10444
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.971203085802386,
+      "learning_rate": 1.764980704471162e-05,
+      "loss": 0.8406,
+      "step": 10445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6709869470430136,
+      "learning_rate": 1.7649315583407226e-05,
+      "loss": 0.9953,
+      "step": 10446
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5566119060655168,
+      "learning_rate": 1.7648824077566095e-05,
+      "loss": 0.9503,
+      "step": 10447
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3760815368956454,
+      "learning_rate": 1.7648332527191086e-05,
+      "loss": 0.8852,
+      "step": 10448
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8842703756004369,
+      "learning_rate": 1.764784093228507e-05,
+      "loss": 0.9142,
+      "step": 10449
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9494611332090621,
+      "learning_rate": 1.7647349292850898e-05,
+      "loss": 0.7306,
+      "step": 10450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7590285484304,
+      "learning_rate": 1.7646857608891438e-05,
+      "loss": 1.0645,
+      "step": 10451
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5808661639434163,
+      "learning_rate": 1.7646365880409554e-05,
+      "loss": 0.962,
+      "step": 10452
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.776542639887663,
+      "learning_rate": 1.7645874107408107e-05,
+      "loss": 0.7465,
+      "step": 10453
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7572986746737506,
+      "learning_rate": 1.764538228988996e-05,
+      "loss": 0.7814,
+      "step": 10454
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4640589973331537,
+      "learning_rate": 1.7644890427857975e-05,
+      "loss": 0.9392,
+      "step": 10455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.669237137327922,
+      "learning_rate": 1.764439852131502e-05,
+      "loss": 1.0479,
+      "step": 10456
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1454460877061483,
+      "learning_rate": 1.7643906570263958e-05,
+      "loss": 0.9182,
+      "step": 10457
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6498025848024545,
+      "learning_rate": 1.764341457470765e-05,
+      "loss": 0.7995,
+      "step": 10458
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.720246719995,
+      "learning_rate": 1.7642922534648962e-05,
+      "loss": 0.8397,
+      "step": 10459
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6813478553344177,
+      "learning_rate": 1.7642430450090758e-05,
+      "loss": 0.8893,
+      "step": 10460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4748141844905502,
+      "learning_rate": 1.7641938321035904e-05,
+      "loss": 0.8783,
+      "step": 10461
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1475078856464376,
+      "learning_rate": 1.7641446147487267e-05,
+      "loss": 0.9412,
+      "step": 10462
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5019245565772636,
+      "learning_rate": 1.7640953929447712e-05,
+      "loss": 0.9007,
+      "step": 10463
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6759370368651776,
+      "learning_rate": 1.7640461666920104e-05,
+      "loss": 0.9321,
+      "step": 10464
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6147503875497113,
+      "learning_rate": 1.763996935990731e-05,
+      "loss": 0.8682,
+      "step": 10465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8080928549917614,
+      "learning_rate": 1.7639477008412192e-05,
+      "loss": 0.8324,
+      "step": 10466
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6403570959602893,
+      "learning_rate": 1.7638984612437623e-05,
+      "loss": 0.9568,
+      "step": 10467
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5511868722644138,
+      "learning_rate": 1.7638492171986468e-05,
+      "loss": 0.91,
+      "step": 10468
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8087853587578062,
+      "learning_rate": 1.763799968706159e-05,
+      "loss": 0.943,
+      "step": 10469
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4881146508362955,
+      "learning_rate": 1.7637507157665867e-05,
+      "loss": 0.8499,
+      "step": 10470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8771133030430955,
+      "learning_rate": 1.763701458380215e-05,
+      "loss": 0.8508,
+      "step": 10471
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5968222074378093,
+      "learning_rate": 1.7636521965473324e-05,
+      "loss": 0.8482,
+      "step": 10472
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4030555366719852,
+      "learning_rate": 1.7636029302682244e-05,
+      "loss": 0.9205,
+      "step": 10473
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7491874847785042,
+      "learning_rate": 1.7635536595431784e-05,
+      "loss": 0.7884,
+      "step": 10474
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3074634883113245,
+      "learning_rate": 1.7635043843724817e-05,
+      "loss": 0.865,
+      "step": 10475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.591707435406985,
+      "learning_rate": 1.7634551047564205e-05,
+      "loss": 0.9153,
+      "step": 10476
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5754590118064002,
+      "learning_rate": 1.763405820695282e-05,
+      "loss": 0.9498,
+      "step": 10477
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5003127433902583,
+      "learning_rate": 1.7633565321893527e-05,
+      "loss": 0.8603,
+      "step": 10478
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6218448443065017,
+      "learning_rate": 1.7633072392389204e-05,
+      "loss": 0.9863,
+      "step": 10479
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9475006566510715,
+      "learning_rate": 1.763257941844272e-05,
+      "loss": 0.9293,
+      "step": 10480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.649132721998107,
+      "learning_rate": 1.7632086400056936e-05,
+      "loss": 0.9937,
+      "step": 10481
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5739222425274702,
+      "learning_rate": 1.7631593337234733e-05,
+      "loss": 0.9504,
+      "step": 10482
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1711463607009613,
+      "learning_rate": 1.7631100229978977e-05,
+      "loss": 0.8448,
+      "step": 10483
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.597021135625009,
+      "learning_rate": 1.763060707829254e-05,
+      "loss": 0.783,
+      "step": 10484
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.06914826785656,
+      "learning_rate": 1.7630113882178287e-05,
+      "loss": 0.9646,
+      "step": 10485
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.112599370523303,
+      "learning_rate": 1.7629620641639102e-05,
+      "loss": 0.8618,
+      "step": 10486
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6520089824442006,
+      "learning_rate": 1.7629127356677847e-05,
+      "loss": 0.9278,
+      "step": 10487
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6415535468879277,
+      "learning_rate": 1.7628634027297397e-05,
+      "loss": 0.9199,
+      "step": 10488
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7358529707724801,
+      "learning_rate": 1.7628140653500628e-05,
+      "loss": 0.829,
+      "step": 10489
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6156116412079609,
+      "learning_rate": 1.7627647235290407e-05,
+      "loss": 0.8929,
+      "step": 10490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.53090591723032,
+      "learning_rate": 1.762715377266961e-05,
+      "loss": 0.8034,
+      "step": 10491
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9263462192502212,
+      "learning_rate": 1.7626660265641105e-05,
+      "loss": 0.8761,
+      "step": 10492
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6076662937256754,
+      "learning_rate": 1.7626166714207774e-05,
+      "loss": 0.8712,
+      "step": 10493
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1279298148334616,
+      "learning_rate": 1.762567311837248e-05,
+      "loss": 0.9272,
+      "step": 10494
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7184815681639949,
+      "learning_rate": 1.762517947813811e-05,
+      "loss": 0.8854,
+      "step": 10495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9422623868869007,
+      "learning_rate": 1.7624685793507526e-05,
+      "loss": 0.9911,
+      "step": 10496
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.666953042466854,
+      "learning_rate": 1.762419206448361e-05,
+      "loss": 0.9188,
+      "step": 10497
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8361742472724847,
+      "learning_rate": 1.7623698291069235e-05,
+      "loss": 1.0423,
+      "step": 10498
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1460192100019824,
+      "learning_rate": 1.762320447326727e-05,
+      "loss": 1.0029,
+      "step": 10499
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8633432485766697,
+      "learning_rate": 1.7622710611080597e-05,
+      "loss": 0.8564,
+      "step": 10500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7455355570204745,
+      "learning_rate": 1.762221670451209e-05,
+      "loss": 0.8998,
+      "step": 10501
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7413164464087458,
+      "learning_rate": 1.7621722753564628e-05,
+      "loss": 0.7384,
+      "step": 10502
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.375944650881002,
+      "learning_rate": 1.762122875824108e-05,
+      "loss": 0.9057,
+      "step": 10503
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7764425011070661,
+      "learning_rate": 1.7620734718544326e-05,
+      "loss": 0.8296,
+      "step": 10504
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.635364149431277,
+      "learning_rate": 1.762024063447724e-05,
+      "loss": 0.8893,
+      "step": 10505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5132581741327769,
+      "learning_rate": 1.76197465060427e-05,
+      "loss": 1.0055,
+      "step": 10506
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.826768080614853,
+      "learning_rate": 1.7619252333243586e-05,
+      "loss": 0.9335,
+      "step": 10507
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6099665627528212,
+      "learning_rate": 1.7618758116082772e-05,
+      "loss": 0.8956,
+      "step": 10508
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9138131965063176,
+      "learning_rate": 1.7618263854563137e-05,
+      "loss": 0.9612,
+      "step": 10509
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.730344801250831,
+      "learning_rate": 1.7617769548687555e-05,
+      "loss": 0.8271,
+      "step": 10510
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.435862343211079,
+      "learning_rate": 1.761727519845891e-05,
+      "loss": 0.97,
+      "step": 10511
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4606099908095358,
+      "learning_rate": 1.7616780803880077e-05,
+      "loss": 0.9121,
+      "step": 10512
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8395596022707534,
+      "learning_rate": 1.7616286364953935e-05,
+      "loss": 0.8471,
+      "step": 10513
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8285064185098397,
+      "learning_rate": 1.7615791881683358e-05,
+      "loss": 0.7578,
+      "step": 10514
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.931907791636577,
+      "learning_rate": 1.7615297354071233e-05,
+      "loss": 0.9591,
+      "step": 10515
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1628541398656984,
+      "learning_rate": 1.7614802782120433e-05,
+      "loss": 0.926,
+      "step": 10516
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8464205416225152,
+      "learning_rate": 1.7614308165833845e-05,
+      "loss": 0.8328,
+      "step": 10517
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.524989045354445,
+      "learning_rate": 1.7613813505214343e-05,
+      "loss": 0.8915,
+      "step": 10518
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1122502682543356,
+      "learning_rate": 1.7613318800264808e-05,
+      "loss": 0.9116,
+      "step": 10519
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6884808177284887,
+      "learning_rate": 1.761282405098812e-05,
+      "loss": 1.0934,
+      "step": 10520
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2090373497187903,
+      "learning_rate": 1.7612329257387163e-05,
+      "loss": 0.9079,
+      "step": 10521
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7611287854002515,
+      "learning_rate": 1.761183441946481e-05,
+      "loss": 0.7365,
+      "step": 10522
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.526630961746136,
+      "learning_rate": 1.761133953722395e-05,
+      "loss": 0.805,
+      "step": 10523
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5066917922623357,
+      "learning_rate": 1.7610844610667463e-05,
+      "loss": 0.7856,
+      "step": 10524
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5728076577660959,
+      "learning_rate": 1.7610349639798226e-05,
+      "loss": 0.8798,
+      "step": 10525
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7630854127310378,
+      "learning_rate": 1.7609854624619124e-05,
+      "loss": 0.7591,
+      "step": 10526
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.852896385189145,
+      "learning_rate": 1.7609359565133043e-05,
+      "loss": 0.826,
+      "step": 10527
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7700481106654162,
+      "learning_rate": 1.7608864461342856e-05,
+      "loss": 0.9951,
+      "step": 10528
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.55661445892062,
+      "learning_rate": 1.7608369313251453e-05,
+      "loss": 0.9134,
+      "step": 10529
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5503462265658774,
+      "learning_rate": 1.7607874120861715e-05,
+      "loss": 0.8384,
+      "step": 10530
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8015624237921257,
+      "learning_rate": 1.7607378884176528e-05,
+      "loss": 0.9502,
+      "step": 10531
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7644165850056126,
+      "learning_rate": 1.760688360319877e-05,
+      "loss": 0.8821,
+      "step": 10532
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7264970252531066,
+      "learning_rate": 1.760638827793133e-05,
+      "loss": 0.7374,
+      "step": 10533
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3657506206525887,
+      "learning_rate": 1.7605892908377088e-05,
+      "loss": 0.8772,
+      "step": 10534
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9182484424747792,
+      "learning_rate": 1.760539749453893e-05,
+      "loss": 1.0041,
+      "step": 10535
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7610156823869947,
+      "learning_rate": 1.7604902036419737e-05,
+      "loss": 0.8766,
+      "step": 10536
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3112609452579513,
+      "learning_rate": 1.76044065340224e-05,
+      "loss": 0.8457,
+      "step": 10537
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5970387906213253,
+      "learning_rate": 1.7603910987349803e-05,
+      "loss": 0.8378,
+      "step": 10538
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6188648778743577,
+      "learning_rate": 1.7603415396404822e-05,
+      "loss": 0.8188,
+      "step": 10539
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6549308078860174,
+      "learning_rate": 1.7602919761190355e-05,
+      "loss": 0.8492,
+      "step": 10540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1579210780369884,
+      "learning_rate": 1.760242408170928e-05,
+      "loss": 1.028,
+      "step": 10541
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7155084870270263,
+      "learning_rate": 1.760192835796448e-05,
+      "loss": 0.8772,
+      "step": 10542
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6494354943065423,
+      "learning_rate": 1.7601432589958856e-05,
+      "loss": 1.0114,
+      "step": 10543
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5864732865046587,
+      "learning_rate": 1.760093677769528e-05,
+      "loss": 0.8959,
+      "step": 10544
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7583270433578148,
+      "learning_rate": 1.7600440921176643e-05,
+      "loss": 0.7342,
+      "step": 10545
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6918031612786784,
+      "learning_rate": 1.7599945020405834e-05,
+      "loss": 0.9631,
+      "step": 10546
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7827806829792117,
+      "learning_rate": 1.7599449075385736e-05,
+      "loss": 0.9314,
+      "step": 10547
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8414583004691036,
+      "learning_rate": 1.7598953086119243e-05,
+      "loss": 0.7694,
+      "step": 10548
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6461068009778883,
+      "learning_rate": 1.759845705260924e-05,
+      "loss": 1.0194,
+      "step": 10549
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6727510548454345,
+      "learning_rate": 1.759796097485861e-05,
+      "loss": 0.9102,
+      "step": 10550
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.20938368623441,
+      "learning_rate": 1.759746485287025e-05,
+      "loss": 0.9384,
+      "step": 10551
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.532193685921436,
+      "learning_rate": 1.7596968686647042e-05,
+      "loss": 0.983,
+      "step": 10552
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2705729252697697,
+      "learning_rate": 1.759647247619188e-05,
+      "loss": 0.8407,
+      "step": 10553
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7567878914075126,
+      "learning_rate": 1.7595976221507647e-05,
+      "loss": 0.8926,
+      "step": 10554
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5356786983749322,
+      "learning_rate": 1.7595479922597238e-05,
+      "loss": 0.8666,
+      "step": 10555
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5744556311710474,
+      "learning_rate": 1.7594983579463537e-05,
+      "loss": 0.9809,
+      "step": 10556
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7138636749148712,
+      "learning_rate": 1.759448719210944e-05,
+      "loss": 0.8597,
+      "step": 10557
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7258484360218616,
+      "learning_rate": 1.7593990760537836e-05,
+      "loss": 0.8537,
+      "step": 10558
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5932434657413563,
+      "learning_rate": 1.759349428475161e-05,
+      "loss": 0.8934,
+      "step": 10559
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7048065381228754,
+      "learning_rate": 1.7592997764753655e-05,
+      "loss": 0.8894,
+      "step": 10560
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7339110939409488,
+      "learning_rate": 1.7592501200546865e-05,
+      "loss": 0.8819,
+      "step": 10561
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.012687746657968,
+      "learning_rate": 1.759200459213413e-05,
+      "loss": 1.0018,
+      "step": 10562
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.558923041819284,
+      "learning_rate": 1.7591507939518342e-05,
+      "loss": 0.8879,
+      "step": 10563
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.001591443210658,
+      "learning_rate": 1.7591011242702388e-05,
+      "loss": 1.0756,
+      "step": 10564
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4803476693985314,
+      "learning_rate": 1.7590514501689163e-05,
+      "loss": 0.76,
+      "step": 10565
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.511726051243319,
+      "learning_rate": 1.759001771648156e-05,
+      "loss": 0.824,
+      "step": 10566
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7249060715425473,
+      "learning_rate": 1.7589520887082474e-05,
+      "loss": 0.9546,
+      "step": 10567
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7894835027184204,
+      "learning_rate": 1.7589024013494796e-05,
+      "loss": 0.8295,
+      "step": 10568
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9046885297041876,
+      "learning_rate": 1.758852709572141e-05,
+      "loss": 0.9417,
+      "step": 10569
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.538120988245873,
+      "learning_rate": 1.7588030133765222e-05,
+      "loss": 0.9276,
+      "step": 10570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.737257213376762,
+      "learning_rate": 1.7587533127629123e-05,
+      "loss": 0.8451,
+      "step": 10571
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6456847592895647,
+      "learning_rate": 1.7587036077315996e-05,
+      "loss": 0.9808,
+      "step": 10572
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1258379929006583,
+      "learning_rate": 1.758653898282875e-05,
+      "loss": 0.9211,
+      "step": 10573
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6824440410913635,
+      "learning_rate": 1.758604184417027e-05,
+      "loss": 1.0531,
+      "step": 10574
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.613487423561785,
+      "learning_rate": 1.7585544661343454e-05,
+      "loss": 0.9674,
+      "step": 10575
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8193615336719503,
+      "learning_rate": 1.758504743435119e-05,
+      "loss": 0.7221,
+      "step": 10576
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5419504195174025,
+      "learning_rate": 1.7584550163196382e-05,
+      "loss": 0.9926,
+      "step": 10577
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.609997459710214,
+      "learning_rate": 1.758405284788192e-05,
+      "loss": 0.7669,
+      "step": 10578
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8196131971309233,
+      "learning_rate": 1.7583555488410703e-05,
+      "loss": 0.7519,
+      "step": 10579
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7832969674325585,
+      "learning_rate": 1.7583058084785626e-05,
+      "loss": 0.834,
+      "step": 10580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8183932250081963,
+      "learning_rate": 1.7582560637009584e-05,
+      "loss": 0.7056,
+      "step": 10581
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5649569865104316,
+      "learning_rate": 1.7582063145085473e-05,
+      "loss": 0.9187,
+      "step": 10582
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6681006118634991,
+      "learning_rate": 1.7581565609016186e-05,
+      "loss": 0.7772,
+      "step": 10583
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4365710616118865,
+      "learning_rate": 1.758106802880463e-05,
+      "loss": 0.8813,
+      "step": 10584
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7849435011510051,
+      "learning_rate": 1.7580570404453693e-05,
+      "loss": 0.7119,
+      "step": 10585
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7472541227554816,
+      "learning_rate": 1.7580072735966275e-05,
+      "loss": 0.9482,
+      "step": 10586
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7710357415533824,
+      "learning_rate": 1.7579575023345273e-05,
+      "loss": 0.8302,
+      "step": 10587
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7382210902799831,
+      "learning_rate": 1.7579077266593587e-05,
+      "loss": 0.7644,
+      "step": 10588
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9554754707241033,
+      "learning_rate": 1.757857946571411e-05,
+      "loss": 0.9802,
+      "step": 10589
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8842740566722216,
+      "learning_rate": 1.7578081620709747e-05,
+      "loss": 0.9372,
+      "step": 10590
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.799322686643781,
+      "learning_rate": 1.7577583731583396e-05,
+      "loss": 0.8991,
+      "step": 10591
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.301373242057289,
+      "learning_rate": 1.757708579833795e-05,
+      "loss": 0.8812,
+      "step": 10592
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5565617118159973,
+      "learning_rate": 1.7576587820976315e-05,
+      "loss": 0.898,
+      "step": 10593
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7782600958962862,
+      "learning_rate": 1.7576089799501384e-05,
+      "loss": 0.8405,
+      "step": 10594
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8081273574293557,
+      "learning_rate": 1.757559173391606e-05,
+      "loss": 0.8854,
+      "step": 10595
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8192677458369186,
+      "learning_rate": 1.7575093624223243e-05,
+      "loss": 0.9473,
+      "step": 10596
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7023386887409302,
+      "learning_rate": 1.757459547042583e-05,
+      "loss": 0.812,
+      "step": 10597
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7728832433063537,
+      "learning_rate": 1.757409727252673e-05,
+      "loss": 0.8353,
+      "step": 10598
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.600880167260463,
+      "learning_rate": 1.7573599030528834e-05,
+      "loss": 0.9292,
+      "step": 10599
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.506397433130876,
+      "learning_rate": 1.7573100744435044e-05,
+      "loss": 0.7981,
+      "step": 10600
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7994112032823921,
+      "learning_rate": 1.757260241424827e-05,
+      "loss": 0.9167,
+      "step": 10601
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.725726979364499,
+      "learning_rate": 1.7572104039971405e-05,
+      "loss": 0.8856,
+      "step": 10602
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5700986629885507,
+      "learning_rate": 1.757160562160735e-05,
+      "loss": 0.8332,
+      "step": 10603
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.745214674327516,
+      "learning_rate": 1.7571107159159013e-05,
+      "loss": 0.9583,
+      "step": 10604
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5934054456988316,
+      "learning_rate": 1.757060865262929e-05,
+      "loss": 0.9361,
+      "step": 10605
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5539410133317149,
+      "learning_rate": 1.757011010202109e-05,
+      "loss": 0.9117,
+      "step": 10606
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7860185271320592,
+      "learning_rate": 1.756961150733731e-05,
+      "loss": 0.7816,
+      "step": 10607
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4421099080030793,
+      "learning_rate": 1.756911286858086e-05,
+      "loss": 0.8973,
+      "step": 10608
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.969464967175546,
+      "learning_rate": 1.7568614185754635e-05,
+      "loss": 0.8398,
+      "step": 10609
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6655479905591404,
+      "learning_rate": 1.7568115458861542e-05,
+      "loss": 0.9786,
+      "step": 10610
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.19328233098124,
+      "learning_rate": 1.7567616687904484e-05,
+      "loss": 0.8814,
+      "step": 10611
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4739253112179562,
+      "learning_rate": 1.7567117872886368e-05,
+      "loss": 0.8079,
+      "step": 10612
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.557396998358071,
+      "learning_rate": 1.7566619013810094e-05,
+      "loss": 0.9301,
+      "step": 10613
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5660824630938555,
+      "learning_rate": 1.7566120110678572e-05,
+      "loss": 0.9043,
+      "step": 10614
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.807289405726736,
+      "learning_rate": 1.7565621163494697e-05,
+      "loss": 0.7711,
+      "step": 10615
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8469964394002252,
+      "learning_rate": 1.7565122172261384e-05,
+      "loss": 0.8873,
+      "step": 10616
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7292922349539417,
+      "learning_rate": 1.7564623136981532e-05,
+      "loss": 0.8471,
+      "step": 10617
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0022144441026783,
+      "learning_rate": 1.7564124057658057e-05,
+      "loss": 0.7754,
+      "step": 10618
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7447947822112704,
+      "learning_rate": 1.756362493429385e-05,
+      "loss": 0.8043,
+      "step": 10619
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.661928011599705,
+      "learning_rate": 1.7563125766891826e-05,
+      "loss": 0.9151,
+      "step": 10620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4776703952812564,
+      "learning_rate": 1.7562626555454887e-05,
+      "loss": 0.8837,
+      "step": 10621
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5173371679789749,
+      "learning_rate": 1.7562127299985944e-05,
+      "loss": 0.9041,
+      "step": 10622
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8567348173406912,
+      "learning_rate": 1.7561628000487902e-05,
+      "loss": 0.9459,
+      "step": 10623
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6429573627978378,
+      "learning_rate": 1.7561128656963667e-05,
+      "loss": 0.9804,
+      "step": 10624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8898719460582418,
+      "learning_rate": 1.756062926941615e-05,
+      "loss": 0.9111,
+      "step": 10625
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0026824324190793,
+      "learning_rate": 1.7560129837848255e-05,
+      "loss": 0.9007,
+      "step": 10626
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6961478806387058,
+      "learning_rate": 1.755963036226289e-05,
+      "loss": 0.8721,
+      "step": 10627
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7357688925859023,
+      "learning_rate": 1.755913084266296e-05,
+      "loss": 0.9726,
+      "step": 10628
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5569428737509232,
+      "learning_rate": 1.755863127905138e-05,
+      "loss": 0.8922,
+      "step": 10629
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8422108776055632,
+      "learning_rate": 1.7558131671431056e-05,
+      "loss": 0.7693,
+      "step": 10630
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.083097090184993,
+      "learning_rate": 1.7557632019804894e-05,
+      "loss": 0.8477,
+      "step": 10631
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8951387254874354,
+      "learning_rate": 1.755713232417581e-05,
+      "loss": 0.9176,
+      "step": 10632
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6540257892138264,
+      "learning_rate": 1.7556632584546705e-05,
+      "loss": 0.8804,
+      "step": 10633
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8198375128210039,
+      "learning_rate": 1.7556132800920494e-05,
+      "loss": 0.7653,
+      "step": 10634
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7743550402652041,
+      "learning_rate": 1.7555632973300087e-05,
+      "loss": 0.7148,
+      "step": 10635
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5424637178481748,
+      "learning_rate": 1.755513310168839e-05,
+      "loss": 0.9716,
+      "step": 10636
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6343801907414544,
+      "learning_rate": 1.755463318608832e-05,
+      "loss": 0.7634,
+      "step": 10637
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5192170719678322,
+      "learning_rate": 1.7554133226502782e-05,
+      "loss": 0.844,
+      "step": 10638
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6186627536902107,
+      "learning_rate": 1.755363322293469e-05,
+      "loss": 0.9608,
+      "step": 10639
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7105965047592202,
+      "learning_rate": 1.755313317538695e-05,
+      "loss": 0.8622,
+      "step": 10640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.589187621777294,
+      "learning_rate": 1.755263308386248e-05,
+      "loss": 0.8626,
+      "step": 10641
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8031833654471183,
+      "learning_rate": 1.755213294836419e-05,
+      "loss": 0.7669,
+      "step": 10642
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.530919890087968,
+      "learning_rate": 1.755163276889499e-05,
+      "loss": 0.8418,
+      "step": 10643
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7889981156338075,
+      "learning_rate": 1.7551132545457793e-05,
+      "loss": 0.6684,
+      "step": 10644
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8200082202930792,
+      "learning_rate": 1.7550632278055514e-05,
+      "loss": 0.8285,
+      "step": 10645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.541149358931869,
+      "learning_rate": 1.7550131966691057e-05,
+      "loss": 0.9703,
+      "step": 10646
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8764844947385781,
+      "learning_rate": 1.7549631611367345e-05,
+      "loss": 0.9186,
+      "step": 10647
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.331534335909335,
+      "learning_rate": 1.7549131212087288e-05,
+      "loss": 0.7879,
+      "step": 10648
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.781464552901554,
+      "learning_rate": 1.75486307688538e-05,
+      "loss": 1.0926,
+      "step": 10649
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8816836552973955,
+      "learning_rate": 1.7548130281669793e-05,
+      "loss": 0.9166,
+      "step": 10650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6728118939033287,
+      "learning_rate": 1.7547629750538182e-05,
+      "loss": 0.9797,
+      "step": 10651
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5611379203360252,
+      "learning_rate": 1.7547129175461882e-05,
+      "loss": 0.9209,
+      "step": 10652
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5907538564335186,
+      "learning_rate": 1.7546628556443806e-05,
+      "loss": 0.828,
+      "step": 10653
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5966905010218853,
+      "learning_rate": 1.754612789348687e-05,
+      "loss": 0.7979,
+      "step": 10654
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8009755516890303,
+      "learning_rate": 1.7545627186593986e-05,
+      "loss": 0.7835,
+      "step": 10655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6252159230021934,
+      "learning_rate": 1.754512643576807e-05,
+      "loss": 0.8485,
+      "step": 10656
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6730627279936947,
+      "learning_rate": 1.7544625641012044e-05,
+      "loss": 0.9533,
+      "step": 10657
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4279505178392577,
+      "learning_rate": 1.7544124802328814e-05,
+      "loss": 0.9356,
+      "step": 10658
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5464736744568168,
+      "learning_rate": 1.7543623919721305e-05,
+      "loss": 0.8701,
+      "step": 10659
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9437403444751513,
+      "learning_rate": 1.754312299319243e-05,
+      "loss": 0.9076,
+      "step": 10660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6875612172404644,
+      "learning_rate": 1.7542622022745102e-05,
+      "loss": 0.88,
+      "step": 10661
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1722363334197987,
+      "learning_rate": 1.754212100838224e-05,
+      "loss": 1.0186,
+      "step": 10662
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6609185364686438,
+      "learning_rate": 1.7541619950106764e-05,
+      "loss": 1.004,
+      "step": 10663
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8654509470288818,
+      "learning_rate": 1.7541118847921586e-05,
+      "loss": 0.9298,
+      "step": 10664
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.786020744658914,
+      "learning_rate": 1.754061770182963e-05,
+      "loss": 0.8677,
+      "step": 10665
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7111836339480688,
+      "learning_rate": 1.754011651183381e-05,
+      "loss": 0.7805,
+      "step": 10666
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0058981304129837,
+      "learning_rate": 1.7539615277937045e-05,
+      "loss": 0.9418,
+      "step": 10667
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.480372343228599,
+      "learning_rate": 1.753911400014225e-05,
+      "loss": 0.9055,
+      "step": 10668
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.806389925113204,
+      "learning_rate": 1.753861267845235e-05,
+      "loss": 0.8533,
+      "step": 10669
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4528812468902745,
+      "learning_rate": 1.7538111312870255e-05,
+      "loss": 0.9081,
+      "step": 10670
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7164731944190534,
+      "learning_rate": 1.7537609903398895e-05,
+      "loss": 1.0166,
+      "step": 10671
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4365821044869085,
+      "learning_rate": 1.7537108450041182e-05,
+      "loss": 0.8741,
+      "step": 10672
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7490054025065658,
+      "learning_rate": 1.7536606952800034e-05,
+      "loss": 0.7027,
+      "step": 10673
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.549899059809289,
+      "learning_rate": 1.753610541167838e-05,
+      "loss": 0.9212,
+      "step": 10674
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5462351650931985,
+      "learning_rate": 1.7535603826679132e-05,
+      "loss": 0.8886,
+      "step": 10675
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.724075157433689,
+      "learning_rate": 1.7535102197805212e-05,
+      "loss": 0.7271,
+      "step": 10676
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6196225993043984,
+      "learning_rate": 1.753460052505954e-05,
+      "loss": 0.8091,
+      "step": 10677
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5473456509269803,
+      "learning_rate": 1.753409880844504e-05,
+      "loss": 0.9775,
+      "step": 10678
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.795137150937844,
+      "learning_rate": 1.7533597047964633e-05,
+      "loss": 0.8921,
+      "step": 10679
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.806655809412758,
+      "learning_rate": 1.7533095243621237e-05,
+      "loss": 0.9446,
+      "step": 10680
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5439595875390781,
+      "learning_rate": 1.7532593395417775e-05,
+      "loss": 1.0297,
+      "step": 10681
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.258461754397175,
+      "learning_rate": 1.7532091503357172e-05,
+      "loss": 0.906,
+      "step": 10682
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5513769221198537,
+      "learning_rate": 1.7531589567442347e-05,
+      "loss": 0.9179,
+      "step": 10683
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8452386723154792,
+      "learning_rate": 1.7531087587676223e-05,
+      "loss": 0.8819,
+      "step": 10684
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.567394087207208,
+      "learning_rate": 1.753058556406172e-05,
+      "loss": 0.8658,
+      "step": 10685
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.692338397548806,
+      "learning_rate": 1.7530083496601763e-05,
+      "loss": 0.9373,
+      "step": 10686
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.682964579629334,
+      "learning_rate": 1.7529581385299283e-05,
+      "loss": 1.0222,
+      "step": 10687
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8907840246925918,
+      "learning_rate": 1.7529079230157192e-05,
+      "loss": 0.944,
+      "step": 10688
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.956605502033506,
+      "learning_rate": 1.7528577031178416e-05,
+      "loss": 0.798,
+      "step": 10689
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6148455560773993,
+      "learning_rate": 1.7528074788365885e-05,
+      "loss": 1.0884,
+      "step": 10690
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.655890097432084,
+      "learning_rate": 1.7527572501722516e-05,
+      "loss": 1.0117,
+      "step": 10691
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.652845011171343,
+      "learning_rate": 1.7527070171251236e-05,
+      "loss": 0.8381,
+      "step": 10692
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4523691978871567,
+      "learning_rate": 1.7526567796954973e-05,
+      "loss": 0.856,
+      "step": 10693
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.942060361976154,
+      "learning_rate": 1.752606537883665e-05,
+      "loss": 0.8851,
+      "step": 10694
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.558245784280649,
+      "learning_rate": 1.752556291689919e-05,
+      "loss": 0.9631,
+      "step": 10695
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3385211258792369,
+      "learning_rate": 1.752506041114552e-05,
+      "loss": 0.8713,
+      "step": 10696
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6753472577647512,
+      "learning_rate": 1.7524557861578567e-05,
+      "loss": 0.9624,
+      "step": 10697
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9151981111634486,
+      "learning_rate": 1.7524055268201254e-05,
+      "loss": 0.9044,
+      "step": 10698
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5757887023242227,
+      "learning_rate": 1.752355263101651e-05,
+      "loss": 0.8917,
+      "step": 10699
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4996881381787786,
+      "learning_rate": 1.7523049950027263e-05,
+      "loss": 1.0007,
+      "step": 10700
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4302200224948616,
+      "learning_rate": 1.7522547225236434e-05,
+      "loss": 1.0026,
+      "step": 10701
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5548119867672485,
+      "learning_rate": 1.7522044456646957e-05,
+      "loss": 1.0305,
+      "step": 10702
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8663578485115349,
+      "learning_rate": 1.7521541644261754e-05,
+      "loss": 1.0073,
+      "step": 10703
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5772188208954618,
+      "learning_rate": 1.7521038788083755e-05,
+      "loss": 0.89,
+      "step": 10704
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5890209912154776,
+      "learning_rate": 1.7520535888115887e-05,
+      "loss": 0.923,
+      "step": 10705
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8967522447232057,
+      "learning_rate": 1.7520032944361074e-05,
+      "loss": 0.8811,
+      "step": 10706
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6986243687919647,
+      "learning_rate": 1.751952995682225e-05,
+      "loss": 0.8311,
+      "step": 10707
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.725562391514474,
+      "learning_rate": 1.7519026925502344e-05,
+      "loss": 0.9156,
+      "step": 10708
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6819955297648368,
+      "learning_rate": 1.7518523850404285e-05,
+      "loss": 0.9619,
+      "step": 10709
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5782518992356407,
+      "learning_rate": 1.7518020731530998e-05,
+      "loss": 1.0322,
+      "step": 10710
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8116965149991744,
+      "learning_rate": 1.7517517568885413e-05,
+      "loss": 0.812,
+      "step": 10711
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.797077862804859,
+      "learning_rate": 1.751701436247046e-05,
+      "loss": 0.9303,
+      "step": 10712
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.67699170366327,
+      "learning_rate": 1.7516511112289068e-05,
+      "loss": 0.9727,
+      "step": 10713
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.597097524364554,
+      "learning_rate": 1.751600781834417e-05,
+      "loss": 0.8608,
+      "step": 10714
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7465914203477018,
+      "learning_rate": 1.75155044806387e-05,
+      "loss": 0.9645,
+      "step": 10715
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7675210159762549,
+      "learning_rate": 1.7515001099175578e-05,
+      "loss": 0.7343,
+      "step": 10716
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7649469226745038,
+      "learning_rate": 1.751449767395774e-05,
+      "loss": 0.8622,
+      "step": 10717
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8563283001914908,
+      "learning_rate": 1.751399420498812e-05,
+      "loss": 1.093,
+      "step": 10718
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.01812038857361,
+      "learning_rate": 1.751349069226965e-05,
+      "loss": 0.8685,
+      "step": 10719
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8542259397061542,
+      "learning_rate": 1.7512987135805254e-05,
+      "loss": 0.9114,
+      "step": 10720
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1356024560010005,
+      "learning_rate": 1.7512483535597868e-05,
+      "loss": 0.8332,
+      "step": 10721
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5906746229112156,
+      "learning_rate": 1.751197989165043e-05,
+      "loss": 1.1611,
+      "step": 10722
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6648072958197007,
+      "learning_rate": 1.7511476203965857e-05,
+      "loss": 0.8743,
+      "step": 10723
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8464973636512438,
+      "learning_rate": 1.75109724725471e-05,
+      "loss": 1.0983,
+      "step": 10724
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4731238068833599,
+      "learning_rate": 1.751046869739708e-05,
+      "loss": 0.9929,
+      "step": 10725
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.839806954599682,
+      "learning_rate": 1.7509964878518736e-05,
+      "loss": 0.8854,
+      "step": 10726
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7664130358980052,
+      "learning_rate": 1.7509461015914992e-05,
+      "loss": 0.8492,
+      "step": 10727
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.615811777488586,
+      "learning_rate": 1.7508957109588794e-05,
+      "loss": 0.9462,
+      "step": 10728
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7417928348951779,
+      "learning_rate": 1.7508453159543072e-05,
+      "loss": 0.8998,
+      "step": 10729
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7660545355698092,
+      "learning_rate": 1.7507949165780753e-05,
+      "loss": 1.008,
+      "step": 10730
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5912462710542552,
+      "learning_rate": 1.750744512830478e-05,
+      "loss": 0.8973,
+      "step": 10731
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.985771071359004,
+      "learning_rate": 1.7506941047118084e-05,
+      "loss": 0.8276,
+      "step": 10732
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.788625962003585,
+      "learning_rate": 1.7506436922223602e-05,
+      "loss": 0.7529,
+      "step": 10733
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.755135604358449,
+      "learning_rate": 1.750593275362427e-05,
+      "loss": 0.8154,
+      "step": 10734
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6321299809563774,
+      "learning_rate": 1.750542854132302e-05,
+      "loss": 0.8639,
+      "step": 10735
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8660519678079723,
+      "learning_rate": 1.7504924285322782e-05,
+      "loss": 0.7804,
+      "step": 10736
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1245577003795892,
+      "learning_rate": 1.7504419985626504e-05,
+      "loss": 0.8029,
+      "step": 10737
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.527185668014591,
+      "learning_rate": 1.7503915642237116e-05,
+      "loss": 0.9926,
+      "step": 10738
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8540460548099599,
+      "learning_rate": 1.7503411255157558e-05,
+      "loss": 0.7725,
+      "step": 10739
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8935263020122581,
+      "learning_rate": 1.750290682439076e-05,
+      "loss": 0.9757,
+      "step": 10740
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6086255864507943,
+      "learning_rate": 1.7502402349939667e-05,
+      "loss": 0.8556,
+      "step": 10741
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5915557856105529,
+      "learning_rate": 1.750189783180721e-05,
+      "loss": 0.8626,
+      "step": 10742
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0000040501868086,
+      "learning_rate": 1.7501393269996332e-05,
+      "loss": 0.7828,
+      "step": 10743
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5245724100136642,
+      "learning_rate": 1.7500888664509962e-05,
+      "loss": 0.9156,
+      "step": 10744
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7760862450867105,
+      "learning_rate": 1.7500384015351048e-05,
+      "loss": 0.9049,
+      "step": 10745
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6209512607195609,
+      "learning_rate": 1.7499879322522524e-05,
+      "loss": 0.9194,
+      "step": 10746
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7912468195042696,
+      "learning_rate": 1.7499374586027323e-05,
+      "loss": 0.7303,
+      "step": 10747
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.661605827604434,
+      "learning_rate": 1.7498869805868394e-05,
+      "loss": 0.8336,
+      "step": 10748
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5809673428559996,
+      "learning_rate": 1.749836498204867e-05,
+      "loss": 0.8887,
+      "step": 10749
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8848415952094302,
+      "learning_rate": 1.749786011457109e-05,
+      "loss": 0.8768,
+      "step": 10750
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7102856696871627,
+      "learning_rate": 1.7497355203438594e-05,
+      "loss": 0.9132,
+      "step": 10751
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7417831474869205,
+      "learning_rate": 1.7496850248654123e-05,
+      "loss": 0.9125,
+      "step": 10752
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7450379393741458,
+      "learning_rate": 1.7496345250220617e-05,
+      "loss": 0.9561,
+      "step": 10753
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7085881021398766,
+      "learning_rate": 1.7495840208141015e-05,
+      "loss": 0.8456,
+      "step": 10754
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.277810400449216,
+      "learning_rate": 1.749533512241826e-05,
+      "loss": 0.9867,
+      "step": 10755
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0703920224977757,
+      "learning_rate": 1.7494829993055288e-05,
+      "loss": 0.9517,
+      "step": 10756
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7961546168321725,
+      "learning_rate": 1.7494324820055045e-05,
+      "loss": 0.8826,
+      "step": 10757
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5893927439040971,
+      "learning_rate": 1.7493819603420468e-05,
+      "loss": 1.0328,
+      "step": 10758
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3989432821920453,
+      "learning_rate": 1.7493314343154503e-05,
+      "loss": 0.9253,
+      "step": 10759
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.035120584887918,
+      "learning_rate": 1.7492809039260086e-05,
+      "loss": 0.8294,
+      "step": 10760
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.615364325249829,
+      "learning_rate": 1.7492303691740166e-05,
+      "loss": 0.8507,
+      "step": 10761
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7426969314181817,
+      "learning_rate": 1.7491798300597684e-05,
+      "loss": 0.9068,
+      "step": 10762
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.531753449255459,
+      "learning_rate": 1.7491292865835576e-05,
+      "loss": 0.9153,
+      "step": 10763
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7773587676379285,
+      "learning_rate": 1.749078738745679e-05,
+      "loss": 0.875,
+      "step": 10764
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7606923258919677,
+      "learning_rate": 1.7490281865464268e-05,
+      "loss": 0.7153,
+      "step": 10765
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5926152751542237,
+      "learning_rate": 1.748977629986096e-05,
+      "loss": 0.845,
+      "step": 10766
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6411319318825,
+      "learning_rate": 1.7489270690649794e-05,
+      "loss": 0.8682,
+      "step": 10767
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5323532704485467,
+      "learning_rate": 1.748876503783373e-05,
+      "loss": 1.0566,
+      "step": 10768
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6122294131091919,
+      "learning_rate": 1.7488259341415702e-05,
+      "loss": 0.9292,
+      "step": 10769
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6120773083397348,
+      "learning_rate": 1.7487753601398655e-05,
+      "loss": 0.8242,
+      "step": 10770
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5183834854900229,
+      "learning_rate": 1.748724781778554e-05,
+      "loss": 0.9542,
+      "step": 10771
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.408672732294936,
+      "learning_rate": 1.7486741990579298e-05,
+      "loss": 1.1188,
+      "step": 10772
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7694618725122095,
+      "learning_rate": 1.748623611978287e-05,
+      "loss": 0.7014,
+      "step": 10773
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7388066149028136,
+      "learning_rate": 1.7485730205399208e-05,
+      "loss": 0.8866,
+      "step": 10774
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.816378416933217,
+      "learning_rate": 1.7485224247431255e-05,
+      "loss": 0.7923,
+      "step": 10775
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6769987542381104,
+      "learning_rate": 1.7484718245881957e-05,
+      "loss": 0.9368,
+      "step": 10776
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9199557176793736,
+      "learning_rate": 1.7484212200754256e-05,
+      "loss": 0.9006,
+      "step": 10777
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7979098096695384,
+      "learning_rate": 1.7483706112051105e-05,
+      "loss": 0.7338,
+      "step": 10778
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4813196634485926,
+      "learning_rate": 1.7483199979775447e-05,
+      "loss": 0.9475,
+      "step": 10779
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5920329193906186,
+      "learning_rate": 1.7482693803930228e-05,
+      "loss": 0.9986,
+      "step": 10780
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7588253029462328,
+      "learning_rate": 1.7482187584518397e-05,
+      "loss": 0.7434,
+      "step": 10781
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.95646387523873,
+      "learning_rate": 1.7481681321542906e-05,
+      "loss": 0.7857,
+      "step": 10782
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8954536963296473,
+      "learning_rate": 1.7481175015006693e-05,
+      "loss": 0.8628,
+      "step": 10783
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.556515662582499,
+      "learning_rate": 1.748066866491271e-05,
+      "loss": 0.8411,
+      "step": 10784
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.221464978679329,
+      "learning_rate": 1.748016227126391e-05,
+      "loss": 0.9243,
+      "step": 10785
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6135854271590815,
+      "learning_rate": 1.7479655834063235e-05,
+      "loss": 0.8245,
+      "step": 10786
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7470000236164063,
+      "learning_rate": 1.7479149353313635e-05,
+      "loss": 0.922,
+      "step": 10787
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7561949157635233,
+      "learning_rate": 1.747864282901806e-05,
+      "loss": 0.7713,
+      "step": 10788
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8380204523159365,
+      "learning_rate": 1.7478136261179458e-05,
+      "loss": 0.8874,
+      "step": 10789
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0559318854477926,
+      "learning_rate": 1.747762964980078e-05,
+      "loss": 0.7398,
+      "step": 10790
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.341567843392128,
+      "learning_rate": 1.7477122994884973e-05,
+      "loss": 0.6961,
+      "step": 10791
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5909037084817284,
+      "learning_rate": 1.747661629643499e-05,
+      "loss": 0.8051,
+      "step": 10792
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9314118465653596,
+      "learning_rate": 1.7476109554453783e-05,
+      "loss": 0.8357,
+      "step": 10793
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6906778133083709,
+      "learning_rate": 1.7475602768944297e-05,
+      "loss": 0.832,
+      "step": 10794
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7549908953424054,
+      "learning_rate": 1.747509593990948e-05,
+      "loss": 0.9279,
+      "step": 10795
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.552806565624126,
+      "learning_rate": 1.7474589067352296e-05,
+      "loss": 0.7769,
+      "step": 10796
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9010154571329523,
+      "learning_rate": 1.7474082151275683e-05,
+      "loss": 0.9107,
+      "step": 10797
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8640323469822808,
+      "learning_rate": 1.74735751916826e-05,
+      "loss": 1.0351,
+      "step": 10798
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9236999296546773,
+      "learning_rate": 1.747306818857599e-05,
+      "loss": 0.8556,
+      "step": 10799
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8860409049176943,
+      "learning_rate": 1.7472561141958818e-05,
+      "loss": 0.7899,
+      "step": 10800
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2017310400880477,
+      "learning_rate": 1.7472054051834028e-05,
+      "loss": 0.9254,
+      "step": 10801
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0733688011097198,
+      "learning_rate": 1.7471546918204575e-05,
+      "loss": 0.9469,
+      "step": 10802
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.002844387441039,
+      "learning_rate": 1.7471039741073407e-05,
+      "loss": 0.8968,
+      "step": 10803
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6542503950039356,
+      "learning_rate": 1.7470532520443483e-05,
+      "loss": 1.0255,
+      "step": 10804
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8976430543722307,
+      "learning_rate": 1.7470025256317752e-05,
+      "loss": 0.8431,
+      "step": 10805
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5925009027162123,
+      "learning_rate": 1.7469517948699167e-05,
+      "loss": 0.973,
+      "step": 10806
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7746487665752047,
+      "learning_rate": 1.7469010597590688e-05,
+      "loss": 0.8957,
+      "step": 10807
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5908465442549613,
+      "learning_rate": 1.7468503202995263e-05,
+      "loss": 0.9168,
+      "step": 10808
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7815542785970059,
+      "learning_rate": 1.7467995764915847e-05,
+      "loss": 0.7493,
+      "step": 10809
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7420328061657269,
+      "learning_rate": 1.7467488283355396e-05,
+      "loss": 0.8681,
+      "step": 10810
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6966269080287444,
+      "learning_rate": 1.7466980758316864e-05,
+      "loss": 1.0143,
+      "step": 10811
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6429979863455766,
+      "learning_rate": 1.7466473189803203e-05,
+      "loss": 0.8802,
+      "step": 10812
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2317618886180983,
+      "learning_rate": 1.7465965577817375e-05,
+      "loss": 0.8698,
+      "step": 10813
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7735247075913033,
+      "learning_rate": 1.746545792236233e-05,
+      "loss": 0.7596,
+      "step": 10814
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6426176103523418,
+      "learning_rate": 1.7464950223441026e-05,
+      "loss": 0.8737,
+      "step": 10815
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7702650816794188,
+      "learning_rate": 1.7464442481056418e-05,
+      "loss": 0.6797,
+      "step": 10816
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5545216533980863,
+      "learning_rate": 1.7463934695211463e-05,
+      "loss": 0.8715,
+      "step": 10817
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3785610258179863,
+      "learning_rate": 1.7463426865909118e-05,
+      "loss": 0.8957,
+      "step": 10818
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.732215717079353,
+      "learning_rate": 1.746291899315234e-05,
+      "loss": 0.9156,
+      "step": 10819
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7109970425071381,
+      "learning_rate": 1.7462411076944084e-05,
+      "loss": 0.8225,
+      "step": 10820
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7462743251505242,
+      "learning_rate": 1.7461903117287305e-05,
+      "loss": 0.7312,
+      "step": 10821
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5805474096917855,
+      "learning_rate": 1.746139511418497e-05,
+      "loss": 0.8598,
+      "step": 10822
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9319848592444266,
+      "learning_rate": 1.7460887067640025e-05,
+      "loss": 0.9124,
+      "step": 10823
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6115954513309345,
+      "learning_rate": 1.7460378977655432e-05,
+      "loss": 0.8764,
+      "step": 10824
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8034772779344261,
+      "learning_rate": 1.7459870844234154e-05,
+      "loss": 0.8656,
+      "step": 10825
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6781666232546302,
+      "learning_rate": 1.745936266737915e-05,
+      "loss": 0.9976,
+      "step": 10826
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7455128033424352,
+      "learning_rate": 1.745885444709337e-05,
+      "loss": 0.8905,
+      "step": 10827
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8257382971775256,
+      "learning_rate": 1.7458346183379778e-05,
+      "loss": 0.8571,
+      "step": 10828
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.119319688300801,
+      "learning_rate": 1.7457837876241332e-05,
+      "loss": 0.8643,
+      "step": 10829
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9205486556584417,
+      "learning_rate": 1.7457329525680996e-05,
+      "loss": 0.9772,
+      "step": 10830
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7027291100919826,
+      "learning_rate": 1.7456821131701724e-05,
+      "loss": 0.8661,
+      "step": 10831
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6778864467571002,
+      "learning_rate": 1.745631269430648e-05,
+      "loss": 0.8991,
+      "step": 10832
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.034779704022088,
+      "learning_rate": 1.7455804213498223e-05,
+      "loss": 1.0274,
+      "step": 10833
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6279821798779563,
+      "learning_rate": 1.7455295689279913e-05,
+      "loss": 0.8612,
+      "step": 10834
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.509036849138934,
+      "learning_rate": 1.745478712165451e-05,
+      "loss": 0.9898,
+      "step": 10835
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7209137000779928,
+      "learning_rate": 1.745427851062498e-05,
+      "loss": 0.9146,
+      "step": 10836
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5993077907522657,
+      "learning_rate": 1.7453769856194277e-05,
+      "loss": 0.886,
+      "step": 10837
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8039577298319764,
+      "learning_rate": 1.7453261158365367e-05,
+      "loss": 0.77,
+      "step": 10838
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6911866726598284,
+      "learning_rate": 1.745275241714121e-05,
+      "loss": 0.9782,
+      "step": 10839
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5907328169700057,
+      "learning_rate": 1.745224363252477e-05,
+      "loss": 0.847,
+      "step": 10840
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.735083014678998,
+      "learning_rate": 1.745173480451901e-05,
+      "loss": 0.911,
+      "step": 10841
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7946888282967752,
+      "learning_rate": 1.745122593312689e-05,
+      "loss": 0.8867,
+      "step": 10842
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7218218322355807,
+      "learning_rate": 1.745071701835137e-05,
+      "loss": 0.8912,
+      "step": 10843
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.591153317171932,
+      "learning_rate": 1.7450208060195423e-05,
+      "loss": 0.8907,
+      "step": 10844
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8300225269052504,
+      "learning_rate": 1.7449699058662e-05,
+      "loss": 0.9924,
+      "step": 10845
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8487744945532463,
+      "learning_rate": 1.7449190013754075e-05,
+      "loss": 0.9841,
+      "step": 10846
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7031989005347419,
+      "learning_rate": 1.7448680925474605e-05,
+      "loss": 0.866,
+      "step": 10847
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7281541731312762,
+      "learning_rate": 1.7448171793826553e-05,
+      "loss": 0.8289,
+      "step": 10848
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.494083222248021,
+      "learning_rate": 1.744766261881289e-05,
+      "loss": 0.8368,
+      "step": 10849
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8456241060458949,
+      "learning_rate": 1.7447153400436577e-05,
+      "loss": 0.8371,
+      "step": 10850
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.878695931077512,
+      "learning_rate": 1.744664413870058e-05,
+      "loss": 0.8086,
+      "step": 10851
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.451335119901995,
+      "learning_rate": 1.7446134833607863e-05,
+      "loss": 0.947,
+      "step": 10852
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.731945169464345,
+      "learning_rate": 1.7445625485161393e-05,
+      "loss": 0.7785,
+      "step": 10853
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8019182612211888,
+      "learning_rate": 1.7445116093364133e-05,
+      "loss": 0.9742,
+      "step": 10854
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7677817243371077,
+      "learning_rate": 1.744460665821905e-05,
+      "loss": 0.7646,
+      "step": 10855
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.02102802614974,
+      "learning_rate": 1.7444097179729112e-05,
+      "loss": 0.8976,
+      "step": 10856
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7160332782812824,
+      "learning_rate": 1.7443587657897278e-05,
+      "loss": 0.8596,
+      "step": 10857
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8012610982198907,
+      "learning_rate": 1.7443078092726525e-05,
+      "loss": 0.7525,
+      "step": 10858
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5976127817377705,
+      "learning_rate": 1.7442568484219812e-05,
+      "loss": 0.9883,
+      "step": 10859
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7209890200789055,
+      "learning_rate": 1.7442058832380113e-05,
+      "loss": 0.9114,
+      "step": 10860
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8291193036426001,
+      "learning_rate": 1.7441549137210386e-05,
+      "loss": 0.9511,
+      "step": 10861
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6913500090197942,
+      "learning_rate": 1.744103939871361e-05,
+      "loss": 0.8525,
+      "step": 10862
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.178593614486722,
+      "learning_rate": 1.744052961689274e-05,
+      "loss": 1.001,
+      "step": 10863
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.770775276955077,
+      "learning_rate": 1.7440019791750757e-05,
+      "loss": 0.6702,
+      "step": 10864
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.809421953272047,
+      "learning_rate": 1.743950992329062e-05,
+      "loss": 0.9969,
+      "step": 10865
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4625703542987667,
+      "learning_rate": 1.74390000115153e-05,
+      "loss": 0.891,
+      "step": 10866
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5676078991280775,
+      "learning_rate": 1.743849005642777e-05,
+      "loss": 0.8269,
+      "step": 10867
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4395157536224286,
+      "learning_rate": 1.7437980058030993e-05,
+      "loss": 0.9735,
+      "step": 10868
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.854808755369104,
+      "learning_rate": 1.7437470016327944e-05,
+      "loss": 0.857,
+      "step": 10869
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.93484953961313,
+      "learning_rate": 1.743695993132159e-05,
+      "loss": 0.9803,
+      "step": 10870
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.711567153547034,
+      "learning_rate": 1.74364498030149e-05,
+      "loss": 0.8915,
+      "step": 10871
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9862614941240497,
+      "learning_rate": 1.7435939631410845e-05,
+      "loss": 0.938,
+      "step": 10872
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9976975532048573,
+      "learning_rate": 1.7435429416512394e-05,
+      "loss": 0.9136,
+      "step": 10873
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7977538994920342,
+      "learning_rate": 1.7434919158322516e-05,
+      "loss": 0.7421,
+      "step": 10874
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7400244736657988,
+      "learning_rate": 1.743440885684419e-05,
+      "loss": 0.8593,
+      "step": 10875
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.85799226281806,
+      "learning_rate": 1.743389851208038e-05,
+      "loss": 0.9283,
+      "step": 10876
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6730239400115763,
+      "learning_rate": 1.7433388124034062e-05,
+      "loss": 0.9582,
+      "step": 10877
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7394709403636421,
+      "learning_rate": 1.74328776927082e-05,
+      "loss": 0.8866,
+      "step": 10878
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0996231229187847,
+      "learning_rate": 1.7432367218105772e-05,
+      "loss": 1.0102,
+      "step": 10879
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6512689957232123,
+      "learning_rate": 1.7431856700229752e-05,
+      "loss": 0.9209,
+      "step": 10880
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8481394887374158,
+      "learning_rate": 1.7431346139083108e-05,
+      "loss": 0.9119,
+      "step": 10881
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7702004675770768,
+      "learning_rate": 1.7430835534668814e-05,
+      "loss": 0.7853,
+      "step": 10882
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4742817787198057,
+      "learning_rate": 1.7430324886989843e-05,
+      "loss": 0.9668,
+      "step": 10883
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0314117338596014,
+      "learning_rate": 1.7429814196049166e-05,
+      "loss": 0.9111,
+      "step": 10884
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5280489686405574,
+      "learning_rate": 1.7429303461849764e-05,
+      "loss": 0.9527,
+      "step": 10885
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6605691502210436,
+      "learning_rate": 1.74287926843946e-05,
+      "loss": 0.8302,
+      "step": 10886
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6757253914063432,
+      "learning_rate": 1.7428281863686654e-05,
+      "loss": 0.7967,
+      "step": 10887
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7946556243699298,
+      "learning_rate": 1.74277709997289e-05,
+      "loss": 0.7377,
+      "step": 10888
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7485294218942709,
+      "learning_rate": 1.742726009252431e-05,
+      "loss": 0.7816,
+      "step": 10889
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8002535112116782,
+      "learning_rate": 1.7426749142075863e-05,
+      "loss": 0.7383,
+      "step": 10890
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7713462987615258,
+      "learning_rate": 1.742623814838653e-05,
+      "loss": 0.843,
+      "step": 10891
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6511897440860392,
+      "learning_rate": 1.742572711145929e-05,
+      "loss": 0.8862,
+      "step": 10892
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5946746552989022,
+      "learning_rate": 1.7425216031297113e-05,
+      "loss": 0.9098,
+      "step": 10893
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6455882714088976,
+      "learning_rate": 1.7424704907902975e-05,
+      "loss": 0.8805,
+      "step": 10894
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6761970192856963,
+      "learning_rate": 1.742419374127986e-05,
+      "loss": 0.8414,
+      "step": 10895
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7522524127816768,
+      "learning_rate": 1.7423682531430736e-05,
+      "loss": 0.8686,
+      "step": 10896
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.109629883146708,
+      "learning_rate": 1.742317127835858e-05,
+      "loss": 0.902,
+      "step": 10897
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5750594449290796,
+      "learning_rate": 1.7422659982066373e-05,
+      "loss": 0.9426,
+      "step": 10898
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5962653714389516,
+      "learning_rate": 1.7422148642557088e-05,
+      "loss": 0.9145,
+      "step": 10899
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7783733351579304,
+      "learning_rate": 1.7421637259833707e-05,
+      "loss": 0.7429,
+      "step": 10900
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5960645444189034,
+      "learning_rate": 1.7421125833899203e-05,
+      "loss": 0.796,
+      "step": 10901
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9326827928786317,
+      "learning_rate": 1.7420614364756553e-05,
+      "loss": 0.8306,
+      "step": 10902
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6247535617598603,
+      "learning_rate": 1.7420102852408735e-05,
+      "loss": 1.0651,
+      "step": 10903
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8089605953151857,
+      "learning_rate": 1.7419591296858733e-05,
+      "loss": 0.9678,
+      "step": 10904
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7740073419413355,
+      "learning_rate": 1.741907969810952e-05,
+      "loss": 0.7208,
+      "step": 10905
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5835614503370343,
+      "learning_rate": 1.7418568056164077e-05,
+      "loss": 0.9259,
+      "step": 10906
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7279735139888786,
+      "learning_rate": 1.7418056371025383e-05,
+      "loss": 0.7552,
+      "step": 10907
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7828566461020231,
+      "learning_rate": 1.7417544642696414e-05,
+      "loss": 0.7549,
+      "step": 10908
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8412745753373774,
+      "learning_rate": 1.741703287118015e-05,
+      "loss": 1.0531,
+      "step": 10909
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7250599935145656,
+      "learning_rate": 1.7416521056479577e-05,
+      "loss": 1.0259,
+      "step": 10910
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8898315142319304,
+      "learning_rate": 1.7416009198597668e-05,
+      "loss": 0.7804,
+      "step": 10911
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7183659478181579,
+      "learning_rate": 1.7415497297537407e-05,
+      "loss": 0.9535,
+      "step": 10912
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.51243827189316,
+      "learning_rate": 1.7414985353301773e-05,
+      "loss": 0.9144,
+      "step": 10913
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8115613697338574,
+      "learning_rate": 1.7414473365893744e-05,
+      "loss": 0.7791,
+      "step": 10914
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.72025229108135,
+      "learning_rate": 1.7413961335316305e-05,
+      "loss": 1.0226,
+      "step": 10915
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6437869255309772,
+      "learning_rate": 1.7413449261572437e-05,
+      "loss": 1.0565,
+      "step": 10916
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1353230817240827,
+      "learning_rate": 1.741293714466512e-05,
+      "loss": 0.8641,
+      "step": 10917
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2095712373933987,
+      "learning_rate": 1.7412424984597334e-05,
+      "loss": 0.8881,
+      "step": 10918
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.796496238499999,
+      "learning_rate": 1.7411912781372066e-05,
+      "loss": 0.7417,
+      "step": 10919
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.48971207057701,
+      "learning_rate": 1.7411400534992292e-05,
+      "loss": 0.8187,
+      "step": 10920
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6971508931210737,
+      "learning_rate": 1.7410888245461e-05,
+      "loss": 0.7572,
+      "step": 10921
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7416939885390057,
+      "learning_rate": 1.7410375912781173e-05,
+      "loss": 0.9427,
+      "step": 10922
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.787205408807641,
+      "learning_rate": 1.7409863536955788e-05,
+      "loss": 1.0087,
+      "step": 10923
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.721324510043915,
+      "learning_rate": 1.7409351117987833e-05,
+      "loss": 0.8241,
+      "step": 10924
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.723565118648729,
+      "learning_rate": 1.7408838655880287e-05,
+      "loss": 0.9529,
+      "step": 10925
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7810695142985866,
+      "learning_rate": 1.7408326150636138e-05,
+      "loss": 1.063,
+      "step": 10926
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.389727900146725,
+      "learning_rate": 1.740781360225837e-05,
+      "loss": 0.8554,
+      "step": 10927
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6563723629096336,
+      "learning_rate": 1.7407301010749967e-05,
+      "loss": 0.7891,
+      "step": 10928
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9122078551667663,
+      "learning_rate": 1.740678837611391e-05,
+      "loss": 0.8906,
+      "step": 10929
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9675822870203823,
+      "learning_rate": 1.7406275698353184e-05,
+      "loss": 0.9462,
+      "step": 10930
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5801998886927326,
+      "learning_rate": 1.740576297747078e-05,
+      "loss": 0.9159,
+      "step": 10931
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5828404641105365,
+      "learning_rate": 1.7405250213469678e-05,
+      "loss": 0.9214,
+      "step": 10932
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5603070796604022,
+      "learning_rate": 1.7404737406352866e-05,
+      "loss": 0.8301,
+      "step": 10933
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.43867695190203,
+      "learning_rate": 1.7404224556123324e-05,
+      "loss": 0.8433,
+      "step": 10934
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8457641273117076,
+      "learning_rate": 1.740371166278405e-05,
+      "loss": 0.9102,
+      "step": 10935
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7311210866887723,
+      "learning_rate": 1.7403198726338017e-05,
+      "loss": 0.7307,
+      "step": 10936
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6138750975962157,
+      "learning_rate": 1.740268574678822e-05,
+      "loss": 0.9511,
+      "step": 10937
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6025104759518358,
+      "learning_rate": 1.7402172724137643e-05,
+      "loss": 0.9956,
+      "step": 10938
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.439941065296839,
+      "learning_rate": 1.7401659658389268e-05,
+      "loss": 0.8788,
+      "step": 10939
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.592753579454749,
+      "learning_rate": 1.7401146549546093e-05,
+      "loss": 0.8991,
+      "step": 10940
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7380545471333144,
+      "learning_rate": 1.7400633397611097e-05,
+      "loss": 0.9418,
+      "step": 10941
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6553973288716284,
+      "learning_rate": 1.7400120202587273e-05,
+      "loss": 0.9193,
+      "step": 10942
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8050470657430115,
+      "learning_rate": 1.7399606964477604e-05,
+      "loss": 0.9367,
+      "step": 10943
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8398826642872952,
+      "learning_rate": 1.7399093683285082e-05,
+      "loss": 0.8333,
+      "step": 10944
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.714237865974463,
+      "learning_rate": 1.739858035901269e-05,
+      "loss": 0.9074,
+      "step": 10945
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6137595377293772,
+      "learning_rate": 1.7398066991663426e-05,
+      "loss": 0.9528,
+      "step": 10946
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7731493723577858,
+      "learning_rate": 1.7397553581240272e-05,
+      "loss": 0.7814,
+      "step": 10947
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.404349349671131,
+      "learning_rate": 1.7397040127746215e-05,
+      "loss": 0.9551,
+      "step": 10948
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.141528539300322,
+      "learning_rate": 1.7396526631184256e-05,
+      "loss": 0.8704,
+      "step": 10949
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7435010503827648,
+      "learning_rate": 1.739601309155737e-05,
+      "loss": 0.8635,
+      "step": 10950
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7377362365283504,
+      "learning_rate": 1.739549950886856e-05,
+      "loss": 0.8691,
+      "step": 10951
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8831094761807301,
+      "learning_rate": 1.7394985883120807e-05,
+      "loss": 0.7343,
+      "step": 10952
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8192538500221833,
+      "learning_rate": 1.7394472214317107e-05,
+      "loss": 0.8438,
+      "step": 10953
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7469009907212232,
+      "learning_rate": 1.7393958502460447e-05,
+      "loss": 0.8424,
+      "step": 10954
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6557701798986115,
+      "learning_rate": 1.739344474755382e-05,
+      "loss": 0.8632,
+      "step": 10955
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.888308068175342,
+      "learning_rate": 1.7392930949600217e-05,
+      "loss": 0.8043,
+      "step": 10956
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7579638509197202,
+      "learning_rate": 1.739241710860263e-05,
+      "loss": 0.8998,
+      "step": 10957
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5396690510374988,
+      "learning_rate": 1.739190322456405e-05,
+      "loss": 0.9241,
+      "step": 10958
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6896072119319732,
+      "learning_rate": 1.7391389297487472e-05,
+      "loss": 0.8419,
+      "step": 10959
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4615938408330051,
+      "learning_rate": 1.739087532737588e-05,
+      "loss": 0.9285,
+      "step": 10960
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6826168902378074,
+      "learning_rate": 1.7390361314232275e-05,
+      "loss": 0.9837,
+      "step": 10961
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5627861100933205,
+      "learning_rate": 1.7389847258059646e-05,
+      "loss": 0.9307,
+      "step": 10962
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5671209940720765,
+      "learning_rate": 1.738933315886099e-05,
+      "loss": 0.9122,
+      "step": 10963
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.484467158762391,
+      "learning_rate": 1.7388819016639292e-05,
+      "loss": 0.912,
+      "step": 10964
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8160538877043897,
+      "learning_rate": 1.738830483139755e-05,
+      "loss": 0.8353,
+      "step": 10965
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8042791811188081,
+      "learning_rate": 1.7387790603138764e-05,
+      "loss": 0.7007,
+      "step": 10966
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8936577659476632,
+      "learning_rate": 1.7387276331865916e-05,
+      "loss": 0.8102,
+      "step": 10967
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1236537377298728,
+      "learning_rate": 1.738676201758201e-05,
+      "loss": 0.9152,
+      "step": 10968
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9017463638674386,
+      "learning_rate": 1.7386247660290038e-05,
+      "loss": 0.7823,
+      "step": 10969
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6715990468537727,
+      "learning_rate": 1.738573325999299e-05,
+      "loss": 0.8896,
+      "step": 10970
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6453967920263857,
+      "learning_rate": 1.7385218816693864e-05,
+      "loss": 0.9314,
+      "step": 10971
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.771634132696132,
+      "learning_rate": 1.738470433039566e-05,
+      "loss": 0.7907,
+      "step": 10972
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7808134332246832,
+      "learning_rate": 1.738418980110137e-05,
+      "loss": 0.7631,
+      "step": 10973
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.63093989720974,
+      "learning_rate": 1.7383675228813986e-05,
+      "loss": 0.8512,
+      "step": 10974
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6761318395425786,
+      "learning_rate": 1.7383160613536508e-05,
+      "loss": 0.8435,
+      "step": 10975
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8560358191432824,
+      "learning_rate": 1.738264595527193e-05,
+      "loss": 1.0572,
+      "step": 10976
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.025180626346289,
+      "learning_rate": 1.7382131254023252e-05,
+      "loss": 1.0333,
+      "step": 10977
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.970764066322988,
+      "learning_rate": 1.7381616509793468e-05,
+      "loss": 0.9241,
+      "step": 10978
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8110425047373104,
+      "learning_rate": 1.7381101722585577e-05,
+      "loss": 0.8814,
+      "step": 10979
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7764962650981995,
+      "learning_rate": 1.7380586892402572e-05,
+      "loss": 0.7067,
+      "step": 10980
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5859241709837586,
+      "learning_rate": 1.7380072019247455e-05,
+      "loss": 0.8944,
+      "step": 10981
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6050744204372245,
+      "learning_rate": 1.7379557103123223e-05,
+      "loss": 0.8577,
+      "step": 10982
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0274963879593515,
+      "learning_rate": 1.7379042144032873e-05,
+      "loss": 1.0545,
+      "step": 10983
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7083933500954752,
+      "learning_rate": 1.7378527141979402e-05,
+      "loss": 0.6791,
+      "step": 10984
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4637128545889375,
+      "learning_rate": 1.7378012096965813e-05,
+      "loss": 0.9959,
+      "step": 10985
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.511863567468387,
+      "learning_rate": 1.73774970089951e-05,
+      "loss": 0.9856,
+      "step": 10986
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.776965958706754,
+      "learning_rate": 1.7376981878070265e-05,
+      "loss": 0.7313,
+      "step": 10987
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.34682174168128,
+      "learning_rate": 1.73764667041943e-05,
+      "loss": 0.9156,
+      "step": 10988
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8721218903594674,
+      "learning_rate": 1.737595148737022e-05,
+      "loss": 0.9718,
+      "step": 10989
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5267150445000957,
+      "learning_rate": 1.7375436227601007e-05,
+      "loss": 0.8254,
+      "step": 10990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5430951447151293,
+      "learning_rate": 1.7374920924889675e-05,
+      "loss": 0.8938,
+      "step": 10991
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8255341479971532,
+      "learning_rate": 1.7374405579239217e-05,
+      "loss": 0.7377,
+      "step": 10992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6518775307782965,
+      "learning_rate": 1.737389019065263e-05,
+      "loss": 0.9237,
+      "step": 10993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.061275253539868,
+      "learning_rate": 1.7373374759132927e-05,
+      "loss": 0.8637,
+      "step": 10994
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6662282465304665,
+      "learning_rate": 1.73728592846831e-05,
+      "loss": 0.8004,
+      "step": 10995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7587474711836388,
+      "learning_rate": 1.7372343767306148e-05,
+      "loss": 0.7462,
+      "step": 10996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5083209317111064,
+      "learning_rate": 1.737182820700508e-05,
+      "loss": 0.895,
+      "step": 10997
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2945625343682603,
+      "learning_rate": 1.7371312603782893e-05,
+      "loss": 0.8761,
+      "step": 10998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5469268556233315,
+      "learning_rate": 1.737079695764259e-05,
+      "loss": 1.0196,
+      "step": 10999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2328016196148326,
+      "learning_rate": 1.7370281268587175e-05,
+      "loss": 0.9986,
+      "step": 11000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.344978772084474,
+      "learning_rate": 1.736976553661965e-05,
+      "loss": 0.9758,
+      "step": 11001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5890605291744218,
+      "learning_rate": 1.7369249761743015e-05,
+      "loss": 0.9289,
+      "step": 11002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7012657118843606,
+      "learning_rate": 1.7368733943960278e-05,
+      "loss": 0.7578,
+      "step": 11003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3431965840028486,
+      "learning_rate": 1.736821808327443e-05,
+      "loss": 0.9379,
+      "step": 11004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7835860650063605,
+      "learning_rate": 1.7367702179688494e-05,
+      "loss": 0.7101,
+      "step": 11005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7883219371942352,
+      "learning_rate": 1.7367186233205457e-05,
+      "loss": 0.8006,
+      "step": 11006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5401031968773402,
+      "learning_rate": 1.7366670243828333e-05,
+      "loss": 0.983,
+      "step": 11007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2646245870345947,
+      "learning_rate": 1.736615421156012e-05,
+      "loss": 0.9432,
+      "step": 11008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.50963523553919,
+      "learning_rate": 1.7365638136403827e-05,
+      "loss": 0.9232,
+      "step": 11009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6331083722531572,
+      "learning_rate": 1.7365122018362456e-05,
+      "loss": 0.8517,
+      "step": 11010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7751333882134164,
+      "learning_rate": 1.736460585743901e-05,
+      "loss": 0.755,
+      "step": 11011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8248427049432316,
+      "learning_rate": 1.7364089653636503e-05,
+      "loss": 0.9367,
+      "step": 11012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5328998402381273,
+      "learning_rate": 1.736357340695793e-05,
+      "loss": 0.8022,
+      "step": 11013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.025092771477765,
+      "learning_rate": 1.7363057117406304e-05,
+      "loss": 0.9578,
+      "step": 11014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0878337546895374,
+      "learning_rate": 1.7362540784984627e-05,
+      "loss": 0.9682,
+      "step": 11015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8761840316623637,
+      "learning_rate": 1.7362024409695908e-05,
+      "loss": 0.8793,
+      "step": 11016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.058639528996746,
+      "learning_rate": 1.736150799154315e-05,
+      "loss": 0.8977,
+      "step": 11017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9067003708690007,
+      "learning_rate": 1.7360991530529363e-05,
+      "loss": 0.9318,
+      "step": 11018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4912860912812,
+      "learning_rate": 1.7360475026657552e-05,
+      "loss": 0.892,
+      "step": 11019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5994063496054316,
+      "learning_rate": 1.7359958479930726e-05,
+      "loss": 0.8791,
+      "step": 11020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8038079233607622,
+      "learning_rate": 1.7359441890351893e-05,
+      "loss": 0.7192,
+      "step": 11021
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8533252649653646,
+      "learning_rate": 1.735892525792406e-05,
+      "loss": 0.8322,
+      "step": 11022
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5209447896840436,
+      "learning_rate": 1.735840858265023e-05,
+      "loss": 0.9889,
+      "step": 11023
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6072920291121824,
+      "learning_rate": 1.7357891864533418e-05,
+      "loss": 0.9444,
+      "step": 11024
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.393350820901506,
+      "learning_rate": 1.7357375103576628e-05,
+      "loss": 0.8979,
+      "step": 11025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5461434473191586,
+      "learning_rate": 1.735685829978287e-05,
+      "loss": 0.8688,
+      "step": 11026
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9615195848473566,
+      "learning_rate": 1.735634145315516e-05,
+      "loss": 0.8975,
+      "step": 11027
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5813287418644886,
+      "learning_rate": 1.7355824563696496e-05,
+      "loss": 0.8347,
+      "step": 11028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4785269433629868,
+      "learning_rate": 1.7355307631409894e-05,
+      "loss": 0.7743,
+      "step": 11029
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8423125699646543,
+      "learning_rate": 1.7354790656298364e-05,
+      "loss": 0.8237,
+      "step": 11030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6156904086085033,
+      "learning_rate": 1.7354273638364914e-05,
+      "loss": 0.7944,
+      "step": 11031
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8133971894911678,
+      "learning_rate": 1.7353756577612552e-05,
+      "loss": 0.8778,
+      "step": 11032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.581481297240255,
+      "learning_rate": 1.735323947404429e-05,
+      "loss": 0.8647,
+      "step": 11033
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5149530568142398,
+      "learning_rate": 1.7352722327663146e-05,
+      "loss": 0.9003,
+      "step": 11034
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6160814079229506,
+      "learning_rate": 1.735220513847212e-05,
+      "loss": 0.8097,
+      "step": 11035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6544011059019281,
+      "learning_rate": 1.735168790647423e-05,
+      "loss": 0.9942,
+      "step": 11036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5957735140707432,
+      "learning_rate": 1.735117063167248e-05,
+      "loss": 0.9849,
+      "step": 11037
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7383998100595424,
+      "learning_rate": 1.7350653314069898e-05,
+      "loss": 0.8043,
+      "step": 11038
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8614296530987804,
+      "learning_rate": 1.7350135953669477e-05,
+      "loss": 0.9573,
+      "step": 11039
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6965768859996384,
+      "learning_rate": 1.7349618550474237e-05,
+      "loss": 0.8901,
+      "step": 11040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7340150683637423,
+      "learning_rate": 1.7349101104487195e-05,
+      "loss": 0.9514,
+      "step": 11041
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.850132911471564,
+      "learning_rate": 1.734858361571136e-05,
+      "loss": 0.9176,
+      "step": 11042
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9027442977824849,
+      "learning_rate": 1.7348066084149743e-05,
+      "loss": 0.9339,
+      "step": 11043
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9335784732089885,
+      "learning_rate": 1.7347548509805357e-05,
+      "loss": 0.896,
+      "step": 11044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.341761011744059,
+      "learning_rate": 1.7347030892681217e-05,
+      "loss": 0.884,
+      "step": 11045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8162623590567037,
+      "learning_rate": 1.734651323278034e-05,
+      "loss": 0.886,
+      "step": 11046
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.514571158202679,
+      "learning_rate": 1.7345995530105737e-05,
+      "loss": 0.8533,
+      "step": 11047
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8468403253582004,
+      "learning_rate": 1.734547778466042e-05,
+      "loss": 0.7887,
+      "step": 11048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7549318504168858,
+      "learning_rate": 1.734495999644741e-05,
+      "loss": 0.9118,
+      "step": 11049
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.772586386141931,
+      "learning_rate": 1.7344442165469714e-05,
+      "loss": 1.0789,
+      "step": 11050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6248722409468381,
+      "learning_rate": 1.7343924291730352e-05,
+      "loss": 0.8918,
+      "step": 11051
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7717104136649842,
+      "learning_rate": 1.7343406375232333e-05,
+      "loss": 0.7874,
+      "step": 11052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5505402355608555,
+      "learning_rate": 1.7342888415978683e-05,
+      "loss": 0.8754,
+      "step": 11053
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7718547050797444,
+      "learning_rate": 1.734237041397241e-05,
+      "loss": 0.8858,
+      "step": 11054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6480321067168477,
+      "learning_rate": 1.734185236921653e-05,
+      "loss": 1.0212,
+      "step": 11055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7341729396765209,
+      "learning_rate": 1.7341334281714067e-05,
+      "loss": 0.8583,
+      "step": 11056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.765950835730778,
+      "learning_rate": 1.7340816151468026e-05,
+      "loss": 0.986,
+      "step": 11057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.550451922809546,
+      "learning_rate": 1.734029797848143e-05,
+      "loss": 0.8833,
+      "step": 11058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.668038000435332,
+      "learning_rate": 1.73397797627573e-05,
+      "loss": 0.9512,
+      "step": 11059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.41700365247482,
+      "learning_rate": 1.7339261504298645e-05,
+      "loss": 0.8841,
+      "step": 11060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.569020153188788,
+      "learning_rate": 1.7338743203108484e-05,
+      "loss": 0.8068,
+      "step": 11061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7757361759703162,
+      "learning_rate": 1.7338224859189837e-05,
+      "loss": 0.6586,
+      "step": 11062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7858608278098179,
+      "learning_rate": 1.7337706472545727e-05,
+      "loss": 0.7733,
+      "step": 11063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5813996703181785,
+      "learning_rate": 1.733718804317916e-05,
+      "loss": 0.8878,
+      "step": 11064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.730599336461254,
+      "learning_rate": 1.7336669571093165e-05,
+      "loss": 0.9296,
+      "step": 11065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7996241656345182,
+      "learning_rate": 1.7336151056290758e-05,
+      "loss": 0.7871,
+      "step": 11066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.750955230190145,
+      "learning_rate": 1.7335632498774956e-05,
+      "loss": 0.8438,
+      "step": 11067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7765053344083886,
+      "learning_rate": 1.7335113898548775e-05,
+      "loss": 0.6721,
+      "step": 11068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7745494770520032,
+      "learning_rate": 1.7334595255615243e-05,
+      "loss": 0.8607,
+      "step": 11069
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6144030110337602,
+      "learning_rate": 1.7334076569977372e-05,
+      "loss": 0.8539,
+      "step": 11070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6188995092427276,
+      "learning_rate": 1.733355784163819e-05,
+      "loss": 0.8919,
+      "step": 11071
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7656189532995332,
+      "learning_rate": 1.733303907060071e-05,
+      "loss": 1.0173,
+      "step": 11072
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5270537759076135,
+      "learning_rate": 1.7332520256867954e-05,
+      "loss": 0.9661,
+      "step": 11073
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6502700704498068,
+      "learning_rate": 1.7332001400442946e-05,
+      "loss": 0.8635,
+      "step": 11074
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6848265133963511,
+      "learning_rate": 1.7331482501328704e-05,
+      "loss": 0.9798,
+      "step": 11075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7380408708686008,
+      "learning_rate": 1.733096355952825e-05,
+      "loss": 0.9403,
+      "step": 11076
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6503304656762268,
+      "learning_rate": 1.7330444575044603e-05,
+      "loss": 0.9804,
+      "step": 11077
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6119959494193306,
+      "learning_rate": 1.7329925547880785e-05,
+      "loss": 0.8733,
+      "step": 11078
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4673826251409805,
+      "learning_rate": 1.7329406478039825e-05,
+      "loss": 0.813,
+      "step": 11079
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7005070006105263,
+      "learning_rate": 1.732888736552474e-05,
+      "loss": 0.8857,
+      "step": 11080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5344388341966544,
+      "learning_rate": 1.732836821033855e-05,
+      "loss": 0.8901,
+      "step": 11081
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7096086979533125,
+      "learning_rate": 1.732784901248428e-05,
+      "loss": 0.8329,
+      "step": 11082
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8399039367987657,
+      "learning_rate": 1.7327329771964952e-05,
+      "loss": 0.9878,
+      "step": 11083
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6103589547963846,
+      "learning_rate": 1.7326810488783593e-05,
+      "loss": 0.9453,
+      "step": 11084
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6328956447048841,
+      "learning_rate": 1.732629116294322e-05,
+      "loss": 0.849,
+      "step": 11085
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2394477917378386,
+      "learning_rate": 1.7325771794446865e-05,
+      "loss": 0.9431,
+      "step": 11086
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6096780942798472,
+      "learning_rate": 1.7325252383297547e-05,
+      "loss": 0.9078,
+      "step": 11087
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.649342544851126,
+      "learning_rate": 1.7324732929498287e-05,
+      "loss": 0.9207,
+      "step": 11088
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7049797325076972,
+      "learning_rate": 1.7324213433052115e-05,
+      "loss": 0.8694,
+      "step": 11089
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.541801268930513,
+      "learning_rate": 1.7323693893962055e-05,
+      "loss": 0.9547,
+      "step": 11090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7044106560596324,
+      "learning_rate": 1.732317431223113e-05,
+      "loss": 0.8245,
+      "step": 11091
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4571653985592439,
+      "learning_rate": 1.732265468786236e-05,
+      "loss": 0.9409,
+      "step": 11092
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7832538795259716,
+      "learning_rate": 1.732213502085878e-05,
+      "loss": 0.7876,
+      "step": 11093
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.612701519639388,
+      "learning_rate": 1.7321615311223413e-05,
+      "loss": 0.9744,
+      "step": 11094
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9956335505946288,
+      "learning_rate": 1.7321095558959284e-05,
+      "loss": 0.8609,
+      "step": 11095
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5030355914239522,
+      "learning_rate": 1.732057576406942e-05,
+      "loss": 0.8059,
+      "step": 11096
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8942535336788675,
+      "learning_rate": 1.7320055926556843e-05,
+      "loss": 0.9176,
+      "step": 11097
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0171903020996647,
+      "learning_rate": 1.7319536046424583e-05,
+      "loss": 0.8372,
+      "step": 11098
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.797080668730871,
+      "learning_rate": 1.731901612367567e-05,
+      "loss": 0.9247,
+      "step": 11099
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7197303252379363,
+      "learning_rate": 1.7318496158313124e-05,
+      "loss": 0.7561,
+      "step": 11100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.087746892556377,
+      "learning_rate": 1.731797615033998e-05,
+      "loss": 0.9483,
+      "step": 11101
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7583595543334505,
+      "learning_rate": 1.731745609975926e-05,
+      "loss": 0.7257,
+      "step": 11102
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.071962390768,
+      "learning_rate": 1.7316936006573994e-05,
+      "loss": 0.8802,
+      "step": 11103
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.749292062077607,
+      "learning_rate": 1.731641587078721e-05,
+      "loss": 0.9269,
+      "step": 11104
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5152440842520607,
+      "learning_rate": 1.7315895692401937e-05,
+      "loss": 0.9221,
+      "step": 11105
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6793802878291049,
+      "learning_rate": 1.7315375471421203e-05,
+      "loss": 0.8643,
+      "step": 11106
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.773927730023857,
+      "learning_rate": 1.7314855207848037e-05,
+      "loss": 0.8939,
+      "step": 11107
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7659702054221653,
+      "learning_rate": 1.731433490168547e-05,
+      "loss": 0.8518,
+      "step": 11108
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5613612391530234,
+      "learning_rate": 1.7313814552936527e-05,
+      "loss": 0.9152,
+      "step": 11109
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6801131212400764,
+      "learning_rate": 1.7313294161604236e-05,
+      "loss": 0.9744,
+      "step": 11110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4989914271147398,
+      "learning_rate": 1.7312773727691637e-05,
+      "loss": 0.9089,
+      "step": 11111
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7297031497162547,
+      "learning_rate": 1.7312253251201754e-05,
+      "loss": 0.9754,
+      "step": 11112
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7344307240778922,
+      "learning_rate": 1.7311732732137617e-05,
+      "loss": 0.9182,
+      "step": 11113
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6564119699340605,
+      "learning_rate": 1.7311212170502254e-05,
+      "loss": 0.8188,
+      "step": 11114
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.852872356055363,
+      "learning_rate": 1.7310691566298702e-05,
+      "loss": 0.8632,
+      "step": 11115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5668167123983752,
+      "learning_rate": 1.7310170919529993e-05,
+      "loss": 0.8721,
+      "step": 11116
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8014779628218687,
+      "learning_rate": 1.7309650230199148e-05,
+      "loss": 0.9133,
+      "step": 11117
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6041231727241758,
+      "learning_rate": 1.7309129498309208e-05,
+      "loss": 0.8785,
+      "step": 11118
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6352225953224364,
+      "learning_rate": 1.73086087238632e-05,
+      "loss": 0.8373,
+      "step": 11119
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7211965320092818,
+      "learning_rate": 1.730808790686416e-05,
+      "loss": 0.707,
+      "step": 11120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.776764382819821,
+      "learning_rate": 1.7307567047315118e-05,
+      "loss": 0.7736,
+      "step": 11121
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.49064546883329,
+      "learning_rate": 1.7307046145219108e-05,
+      "loss": 1.0965,
+      "step": 11122
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6276378003956693,
+      "learning_rate": 1.730652520057916e-05,
+      "loss": 0.871,
+      "step": 11123
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3834832498856735,
+      "learning_rate": 1.730600421339831e-05,
+      "loss": 0.8831,
+      "step": 11124
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6933826429603775,
+      "learning_rate": 1.7305483183679592e-05,
+      "loss": 0.8649,
+      "step": 11125
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9695083949652252,
+      "learning_rate": 1.7304962111426036e-05,
+      "loss": 0.8647,
+      "step": 11126
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9209743290465693,
+      "learning_rate": 1.730444099664068e-05,
+      "loss": 0.9024,
+      "step": 11127
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4908662493403195,
+      "learning_rate": 1.7303919839326554e-05,
+      "loss": 0.92,
+      "step": 11128
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.648038318715491,
+      "learning_rate": 1.7303398639486696e-05,
+      "loss": 0.8003,
+      "step": 11129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7098500079218715,
+      "learning_rate": 1.7302877397124137e-05,
+      "loss": 0.7985,
+      "step": 11130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.716227303637787,
+      "learning_rate": 1.7302356112241913e-05,
+      "loss": 0.9417,
+      "step": 11131
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.805102240115237,
+      "learning_rate": 1.7301834784843064e-05,
+      "loss": 0.9897,
+      "step": 11132
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6613119077947678,
+      "learning_rate": 1.7301313414930618e-05,
+      "loss": 0.8695,
+      "step": 11133
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3966418538712713,
+      "learning_rate": 1.7300792002507614e-05,
+      "loss": 0.9493,
+      "step": 11134
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9947890957121628,
+      "learning_rate": 1.730027054757709e-05,
+      "loss": 0.7727,
+      "step": 11135
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7954092329629056,
+      "learning_rate": 1.7299749050142074e-05,
+      "loss": 0.7355,
+      "step": 11136
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.914821753246417,
+      "learning_rate": 1.7299227510205613e-05,
+      "loss": 0.8854,
+      "step": 11137
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7542555207580284,
+      "learning_rate": 1.7298705927770737e-05,
+      "loss": 0.6929,
+      "step": 11138
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8252305598080414,
+      "learning_rate": 1.7298184302840485e-05,
+      "loss": 0.8576,
+      "step": 11139
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7886415294822853,
+      "learning_rate": 1.729766263541789e-05,
+      "loss": 0.8417,
+      "step": 11140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8239382034105351,
+      "learning_rate": 1.7297140925505996e-05,
+      "loss": 0.9392,
+      "step": 11141
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6032840676131195,
+      "learning_rate": 1.7296619173107837e-05,
+      "loss": 0.8555,
+      "step": 11142
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4384992581349119,
+      "learning_rate": 1.7296097378226452e-05,
+      "loss": 0.8652,
+      "step": 11143
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8092784773967228,
+      "learning_rate": 1.7295575540864878e-05,
+      "loss": 0.8493,
+      "step": 11144
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6711467272649156,
+      "learning_rate": 1.729505366102615e-05,
+      "loss": 0.7799,
+      "step": 11145
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0933319473598604,
+      "learning_rate": 1.7294531738713313e-05,
+      "loss": 0.9529,
+      "step": 11146
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4959349622698936,
+      "learning_rate": 1.7294009773929405e-05,
+      "loss": 1.0353,
+      "step": 11147
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9938556852752167,
+      "learning_rate": 1.7293487766677457e-05,
+      "loss": 0.8198,
+      "step": 11148
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9354648011825775,
+      "learning_rate": 1.7292965716960518e-05,
+      "loss": 0.9586,
+      "step": 11149
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4960764139353135,
+      "learning_rate": 1.7292443624781624e-05,
+      "loss": 0.9563,
+      "step": 11150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6174851840171653,
+      "learning_rate": 1.729192149014381e-05,
+      "loss": 0.7742,
+      "step": 11151
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5222747871225006,
+      "learning_rate": 1.729139931305013e-05,
+      "loss": 0.7799,
+      "step": 11152
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5420166056196283,
+      "learning_rate": 1.7290877093503606e-05,
+      "loss": 0.8862,
+      "step": 11153
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.479627935252292,
+      "learning_rate": 1.729035483150729e-05,
+      "loss": 0.8873,
+      "step": 11154
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8695238697149783,
+      "learning_rate": 1.728983252706422e-05,
+      "loss": 0.8068,
+      "step": 11155
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.614913036955143,
+      "learning_rate": 1.7289310180177438e-05,
+      "loss": 0.8642,
+      "step": 11156
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.009611277450526,
+      "learning_rate": 1.7288787790849984e-05,
+      "loss": 0.9319,
+      "step": 11157
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8416233786667148,
+      "learning_rate": 1.72882653590849e-05,
+      "loss": 0.7388,
+      "step": 11158
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7085294589425801,
+      "learning_rate": 1.7287742884885225e-05,
+      "loss": 0.9712,
+      "step": 11159
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4068735985034175,
+      "learning_rate": 1.728722036825401e-05,
+      "loss": 0.8076,
+      "step": 11160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.932791866061625,
+      "learning_rate": 1.7286697809194285e-05,
+      "loss": 0.8493,
+      "step": 11161
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5714122950493958,
+      "learning_rate": 1.72861752077091e-05,
+      "loss": 0.8993,
+      "step": 11162
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5359820400775934,
+      "learning_rate": 1.7285652563801493e-05,
+      "loss": 0.8383,
+      "step": 11163
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7355060104981028,
+      "learning_rate": 1.7285129877474516e-05,
+      "loss": 0.8934,
+      "step": 11164
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.771519332664238,
+      "learning_rate": 1.7284607148731206e-05,
+      "loss": 0.9183,
+      "step": 11165
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.634553645051188,
+      "learning_rate": 1.7284084377574605e-05,
+      "loss": 0.9664,
+      "step": 11166
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6969537742228813,
+      "learning_rate": 1.728356156400776e-05,
+      "loss": 0.9759,
+      "step": 11167
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9461280841685105,
+      "learning_rate": 1.7283038708033707e-05,
+      "loss": 0.9075,
+      "step": 11168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6532937936422014,
+      "learning_rate": 1.72825158096555e-05,
+      "loss": 0.9311,
+      "step": 11169
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4219955431556408,
+      "learning_rate": 1.728199286887619e-05,
+      "loss": 1.0248,
+      "step": 11170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9242028705262693,
+      "learning_rate": 1.72814698856988e-05,
+      "loss": 1.0222,
+      "step": 11171
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0657735396044576,
+      "learning_rate": 1.7280946860126394e-05,
+      "loss": 0.8964,
+      "step": 11172
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6894766149410716,
+      "learning_rate": 1.7280423792162008e-05,
+      "loss": 1.0169,
+      "step": 11173
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5689639776647337,
+      "learning_rate": 1.727990068180869e-05,
+      "loss": 0.8444,
+      "step": 11174
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6069957029440554,
+      "learning_rate": 1.7279377529069484e-05,
+      "loss": 0.8473,
+      "step": 11175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.383935679256483,
+      "learning_rate": 1.7278854333947438e-05,
+      "loss": 0.8715,
+      "step": 11176
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5008485689042446,
+      "learning_rate": 1.72783310964456e-05,
+      "loss": 0.9452,
+      "step": 11177
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.05397910479886,
+      "learning_rate": 1.727780781656701e-05,
+      "loss": 0.9375,
+      "step": 11178
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0243601029294513,
+      "learning_rate": 1.727728449431472e-05,
+      "loss": 0.919,
+      "step": 11179
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.421698214201191,
+      "learning_rate": 1.7276761129691776e-05,
+      "loss": 0.9554,
+      "step": 11180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8132901298896635,
+      "learning_rate": 1.7276237722701227e-05,
+      "loss": 0.7962,
+      "step": 11181
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6975826382442891,
+      "learning_rate": 1.7275714273346117e-05,
+      "loss": 0.845,
+      "step": 11182
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6026478796338375,
+      "learning_rate": 1.7275190781629494e-05,
+      "loss": 0.9576,
+      "step": 11183
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.840572173074179,
+      "learning_rate": 1.7274667247554407e-05,
+      "loss": 0.725,
+      "step": 11184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7163425452203318,
+      "learning_rate": 1.7274143671123905e-05,
+      "loss": 1.0316,
+      "step": 11185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.281914192788524,
+      "learning_rate": 1.7273620052341037e-05,
+      "loss": 0.9947,
+      "step": 11186
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5498154487300855,
+      "learning_rate": 1.727309639120885e-05,
+      "loss": 0.9138,
+      "step": 11187
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9796284859564623,
+      "learning_rate": 1.727257268773039e-05,
+      "loss": 0.9761,
+      "step": 11188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6868024098605099,
+      "learning_rate": 1.7272048941908713e-05,
+      "loss": 0.9236,
+      "step": 11189
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6962301522712757,
+      "learning_rate": 1.7271525153746865e-05,
+      "loss": 0.8411,
+      "step": 11190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8007340917649617,
+      "learning_rate": 1.7271001323247892e-05,
+      "loss": 0.9959,
+      "step": 11191
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4970747622931118,
+      "learning_rate": 1.727047745041485e-05,
+      "loss": 0.9614,
+      "step": 11192
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.763353092843813,
+      "learning_rate": 1.7269953535250782e-05,
+      "loss": 0.963,
+      "step": 11193
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5231664716820978,
+      "learning_rate": 1.726942957775875e-05,
+      "loss": 0.8935,
+      "step": 11194
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6203353483742018,
+      "learning_rate": 1.726890557794179e-05,
+      "loss": 1.0255,
+      "step": 11195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6496795789745948,
+      "learning_rate": 1.7268381535802966e-05,
+      "loss": 0.9137,
+      "step": 11196
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8976700815558631,
+      "learning_rate": 1.726785745134532e-05,
+      "loss": 0.8797,
+      "step": 11197
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7405906044353567,
+      "learning_rate": 1.7267333324571912e-05,
+      "loss": 0.7142,
+      "step": 11198
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.702014632500241,
+      "learning_rate": 1.7266809155485783e-05,
+      "loss": 0.8766,
+      "step": 11199
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6347806961109452,
+      "learning_rate": 1.7266284944089994e-05,
+      "loss": 0.849,
+      "step": 11200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9658001462620485,
+      "learning_rate": 1.7265760690387594e-05,
+      "loss": 0.9011,
+      "step": 11201
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5471741606124274,
+      "learning_rate": 1.7265236394381634e-05,
+      "loss": 0.9706,
+      "step": 11202
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1711984514823817,
+      "learning_rate": 1.7264712056075167e-05,
+      "loss": 0.9577,
+      "step": 11203
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6120378233083499,
+      "learning_rate": 1.7264187675471246e-05,
+      "loss": 0.9009,
+      "step": 11204
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.760606759048485,
+      "learning_rate": 1.7263663252572923e-05,
+      "loss": 0.6806,
+      "step": 11205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5964975670605368,
+      "learning_rate": 1.7263138787383255e-05,
+      "loss": 0.9519,
+      "step": 11206
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6536731376593343,
+      "learning_rate": 1.7262614279905295e-05,
+      "loss": 0.796,
+      "step": 11207
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6489325164375532,
+      "learning_rate": 1.7262089730142094e-05,
+      "loss": 0.868,
+      "step": 11208
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5759956109597344,
+      "learning_rate": 1.7261565138096706e-05,
+      "loss": 0.969,
+      "step": 11209
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6427849751754307,
+      "learning_rate": 1.7261040503772187e-05,
+      "loss": 0.9161,
+      "step": 11210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6256063920046095,
+      "learning_rate": 1.7260515827171596e-05,
+      "loss": 0.8763,
+      "step": 11211
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.590195780907781,
+      "learning_rate": 1.7259991108297976e-05,
+      "loss": 0.9881,
+      "step": 11212
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8067963606655013,
+      "learning_rate": 1.7259466347154393e-05,
+      "loss": 0.894,
+      "step": 11213
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7142600513663444,
+      "learning_rate": 1.72589415437439e-05,
+      "loss": 0.8185,
+      "step": 11214
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0049478725125973,
+      "learning_rate": 1.725841669806955e-05,
+      "loss": 0.8576,
+      "step": 11215
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6321420586635578,
+      "learning_rate": 1.72578918101344e-05,
+      "loss": 0.8931,
+      "step": 11216
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6473977852567798,
+      "learning_rate": 1.7257366879941508e-05,
+      "loss": 0.9012,
+      "step": 11217
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6462135666891191,
+      "learning_rate": 1.7256841907493928e-05,
+      "loss": 0.9154,
+      "step": 11218
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9817238975692895,
+      "learning_rate": 1.7256316892794712e-05,
+      "loss": 0.9474,
+      "step": 11219
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9029542497004348,
+      "learning_rate": 1.7255791835846926e-05,
+      "loss": 0.9391,
+      "step": 11220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.778790276884935,
+      "learning_rate": 1.725526673665362e-05,
+      "loss": 0.8835,
+      "step": 11221
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.658187976717464,
+      "learning_rate": 1.7254741595217856e-05,
+      "loss": 0.9043,
+      "step": 11222
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.748089205506392,
+      "learning_rate": 1.725421641154269e-05,
+      "loss": 0.9297,
+      "step": 11223
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5680751909381236,
+      "learning_rate": 1.725369118563118e-05,
+      "loss": 0.8362,
+      "step": 11224
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.685320052703521,
+      "learning_rate": 1.725316591748638e-05,
+      "loss": 0.9279,
+      "step": 11225
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.783819423099934,
+      "learning_rate": 1.7252640607111354e-05,
+      "loss": 0.8824,
+      "step": 11226
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8669735480611802,
+      "learning_rate": 1.725211525450916e-05,
+      "loss": 0.8638,
+      "step": 11227
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.618834268910168,
+      "learning_rate": 1.725158985968285e-05,
+      "loss": 0.8042,
+      "step": 11228
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2493963446812195,
+      "learning_rate": 1.725106442263549e-05,
+      "loss": 0.8884,
+      "step": 11229
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7837913180995975,
+      "learning_rate": 1.7250538943370142e-05,
+      "loss": 0.7312,
+      "step": 11230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1489913356580663,
+      "learning_rate": 1.7250013421889857e-05,
+      "loss": 0.9777,
+      "step": 11231
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.600569784020343,
+      "learning_rate": 1.7249487858197698e-05,
+      "loss": 1.0069,
+      "step": 11232
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7965153091030077,
+      "learning_rate": 1.7248962252296725e-05,
+      "loss": 0.9312,
+      "step": 11233
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6389373016701194,
+      "learning_rate": 1.724843660419e-05,
+      "loss": 0.8734,
+      "step": 11234
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.629638905864513,
+      "learning_rate": 1.7247910913880583e-05,
+      "loss": 1.0621,
+      "step": 11235
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5852213489804008,
+      "learning_rate": 1.7247385181371535e-05,
+      "loss": 0.9623,
+      "step": 11236
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.628334961063071,
+      "learning_rate": 1.7246859406665916e-05,
+      "loss": 1.047,
+      "step": 11237
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1095211724430367,
+      "learning_rate": 1.7246333589766786e-05,
+      "loss": 1.079,
+      "step": 11238
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5330557611512563,
+      "learning_rate": 1.7245807730677206e-05,
+      "loss": 0.8657,
+      "step": 11239
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8679764342163703,
+      "learning_rate": 1.724528182940024e-05,
+      "loss": 0.9518,
+      "step": 11240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0677278583915553,
+      "learning_rate": 1.724475588593895e-05,
+      "loss": 0.7062,
+      "step": 11241
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9131004156380462,
+      "learning_rate": 1.7244229900296398e-05,
+      "loss": 0.9029,
+      "step": 11242
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7856604035028063,
+      "learning_rate": 1.724370387247565e-05,
+      "loss": 0.7858,
+      "step": 11243
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7592552661701568,
+      "learning_rate": 1.7243177802479758e-05,
+      "loss": 0.9792,
+      "step": 11244
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7108514817522364,
+      "learning_rate": 1.7242651690311794e-05,
+      "loss": 1.0143,
+      "step": 11245
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8016769486982587,
+      "learning_rate": 1.724212553597482e-05,
+      "loss": 0.8838,
+      "step": 11246
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9991776030892554,
+      "learning_rate": 1.7241599339471894e-05,
+      "loss": 0.9935,
+      "step": 11247
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9337421838003417,
+      "learning_rate": 1.724107310080609e-05,
+      "loss": 0.9255,
+      "step": 11248
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8301308033930332,
+      "learning_rate": 1.724054681998046e-05,
+      "loss": 0.9516,
+      "step": 11249
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7223368078083208,
+      "learning_rate": 1.724002049699808e-05,
+      "loss": 0.9417,
+      "step": 11250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.545190578156371,
+      "learning_rate": 1.7239494131862003e-05,
+      "loss": 0.9508,
+      "step": 11251
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5614395294255807,
+      "learning_rate": 1.7238967724575305e-05,
+      "loss": 0.8154,
+      "step": 11252
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8034621148825207,
+      "learning_rate": 1.723844127514104e-05,
+      "loss": 1.1328,
+      "step": 11253
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.506475606125901,
+      "learning_rate": 1.7237914783562283e-05,
+      "loss": 0.9779,
+      "step": 11254
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5140434348622334,
+      "learning_rate": 1.723738824984209e-05,
+      "loss": 0.776,
+      "step": 11255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7338355506554168,
+      "learning_rate": 1.7236861673983534e-05,
+      "loss": 0.9246,
+      "step": 11256
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5391724292749487,
+      "learning_rate": 1.7236335055989676e-05,
+      "loss": 0.9414,
+      "step": 11257
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.693048685319219,
+      "learning_rate": 1.7235808395863586e-05,
+      "loss": 0.894,
+      "step": 11258
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.695580694245642,
+      "learning_rate": 1.723528169360833e-05,
+      "loss": 0.9547,
+      "step": 11259
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6672072684559263,
+      "learning_rate": 1.723475494922697e-05,
+      "loss": 0.9078,
+      "step": 11260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7958809578254501,
+      "learning_rate": 1.723422816272258e-05,
+      "loss": 0.784,
+      "step": 11261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6527448117699308,
+      "learning_rate": 1.723370133409822e-05,
+      "loss": 0.8931,
+      "step": 11262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8374889333703227,
+      "learning_rate": 1.723317446335696e-05,
+      "loss": 0.781,
+      "step": 11263
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7971647211569689,
+      "learning_rate": 1.7232647550501875e-05,
+      "loss": 0.7361,
+      "step": 11264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6286364663453734,
+      "learning_rate": 1.723212059553602e-05,
+      "loss": 0.8631,
+      "step": 11265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.548139742726575,
+      "learning_rate": 1.723159359846247e-05,
+      "loss": 0.8915,
+      "step": 11266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8014487041370806,
+      "learning_rate": 1.7231066559284293e-05,
+      "loss": 0.744,
+      "step": 11267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8262741078738154,
+      "learning_rate": 1.723053947800456e-05,
+      "loss": 0.9331,
+      "step": 11268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3882315449433156,
+      "learning_rate": 1.7230012354626334e-05,
+      "loss": 0.8708,
+      "step": 11269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.788724913625474,
+      "learning_rate": 1.722948518915269e-05,
+      "loss": 0.8726,
+      "step": 11270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6190992058568603,
+      "learning_rate": 1.722895798158669e-05,
+      "loss": 1.005,
+      "step": 11271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7661449920205736,
+      "learning_rate": 1.7228430731931407e-05,
+      "loss": 0.9172,
+      "step": 11272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0758116615324296,
+      "learning_rate": 1.7227903440189914e-05,
+      "loss": 0.8356,
+      "step": 11273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6977128885019233,
+      "learning_rate": 1.722737610636528e-05,
+      "loss": 0.7718,
+      "step": 11274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9390813064073857,
+      "learning_rate": 1.7226848730460574e-05,
+      "loss": 0.9238,
+      "step": 11275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7721746040683954,
+      "learning_rate": 1.722632131247887e-05,
+      "loss": 0.9429,
+      "step": 11276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.720952739519491,
+      "learning_rate": 1.722579385242323e-05,
+      "loss": 0.8529,
+      "step": 11277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7456530466104202,
+      "learning_rate": 1.7225266350296733e-05,
+      "loss": 0.6949,
+      "step": 11278
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7600531449688919,
+      "learning_rate": 1.7224738806102447e-05,
+      "loss": 0.7523,
+      "step": 11279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.816555241721806,
+      "learning_rate": 1.7224211219843444e-05,
+      "loss": 0.8977,
+      "step": 11280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7630649686098394,
+      "learning_rate": 1.7223683591522798e-05,
+      "loss": 1.0031,
+      "step": 11281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6002009211206454,
+      "learning_rate": 1.7223155921143578e-05,
+      "loss": 0.8052,
+      "step": 11282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6907965053711214,
+      "learning_rate": 1.7222628208708857e-05,
+      "loss": 0.9456,
+      "step": 11283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5926429296838565,
+      "learning_rate": 1.722210045422171e-05,
+      "loss": 0.9393,
+      "step": 11284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9639486358693254,
+      "learning_rate": 1.7221572657685205e-05,
+      "loss": 0.8866,
+      "step": 11285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9344488619974722,
+      "learning_rate": 1.722104481910242e-05,
+      "loss": 0.9125,
+      "step": 11286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5100937172381352,
+      "learning_rate": 1.722051693847642e-05,
+      "loss": 0.9067,
+      "step": 11287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7558435097558447,
+      "learning_rate": 1.721998901581029e-05,
+      "loss": 0.6873,
+      "step": 11288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.713635623441992,
+      "learning_rate": 1.7219461051107102e-05,
+      "loss": 0.9856,
+      "step": 11289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6176241991293467,
+      "learning_rate": 1.721893304436992e-05,
+      "loss": 0.989,
+      "step": 11290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9627725113435537,
+      "learning_rate": 1.7218404995601823e-05,
+      "loss": 1.034,
+      "step": 11291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7182544023193254,
+      "learning_rate": 1.721787690480589e-05,
+      "loss": 0.915,
+      "step": 11292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8484783078916316,
+      "learning_rate": 1.7217348771985194e-05,
+      "loss": 0.8485,
+      "step": 11293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1014217793442813,
+      "learning_rate": 1.7216820597142804e-05,
+      "loss": 0.8476,
+      "step": 11294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5046318062127844,
+      "learning_rate": 1.7216292380281802e-05,
+      "loss": 0.9158,
+      "step": 11295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6298006010863746,
+      "learning_rate": 1.721576412140526e-05,
+      "loss": 0.8946,
+      "step": 11296
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7017292014854601,
+      "learning_rate": 1.721523582051626e-05,
+      "loss": 0.8639,
+      "step": 11297
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5891646959219692,
+      "learning_rate": 1.7214707477617866e-05,
+      "loss": 0.8646,
+      "step": 11298
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9661452062936862,
+      "learning_rate": 1.7214179092713165e-05,
+      "loss": 0.7629,
+      "step": 11299
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4940710045490595,
+      "learning_rate": 1.7213650665805224e-05,
+      "loss": 0.8895,
+      "step": 11300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.064246695152596,
+      "learning_rate": 1.7213122196897135e-05,
+      "loss": 0.8659,
+      "step": 11301
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5544900127197787,
+      "learning_rate": 1.7212593685991956e-05,
+      "loss": 1.0081,
+      "step": 11302
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7493057734741979,
+      "learning_rate": 1.7212065133092778e-05,
+      "loss": 0.7272,
+      "step": 11303
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7605431456266555,
+      "learning_rate": 1.7211536538202673e-05,
+      "loss": 0.754,
+      "step": 11304
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.670786203758318,
+      "learning_rate": 1.7211007901324714e-05,
+      "loss": 0.8923,
+      "step": 11305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7466931716968341,
+      "learning_rate": 1.721047922246199e-05,
+      "loss": 0.8597,
+      "step": 11306
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6790624608354603,
+      "learning_rate": 1.720995050161757e-05,
+      "loss": 0.9384,
+      "step": 11307
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.482129598292086,
+      "learning_rate": 1.7209421738794536e-05,
+      "loss": 0.9313,
+      "step": 11308
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.725101387651939,
+      "learning_rate": 1.720889293399597e-05,
+      "loss": 0.8388,
+      "step": 11309
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4813723406694637,
+      "learning_rate": 1.720836408722494e-05,
+      "loss": 0.9802,
+      "step": 11310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7826534448022034,
+      "learning_rate": 1.7207835198484537e-05,
+      "loss": 0.72,
+      "step": 11311
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5932778404918355,
+      "learning_rate": 1.720730626777783e-05,
+      "loss": 0.8632,
+      "step": 11312
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5023018850568495,
+      "learning_rate": 1.720677729510791e-05,
+      "loss": 0.869,
+      "step": 11313
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8237916628147275,
+      "learning_rate": 1.720624828047785e-05,
+      "loss": 0.8761,
+      "step": 11314
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6036324480625281,
+      "learning_rate": 1.7205719223890732e-05,
+      "loss": 0.8242,
+      "step": 11315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.84786959235784,
+      "learning_rate": 1.7205190125349633e-05,
+      "loss": 0.7946,
+      "step": 11316
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6518769312512103,
+      "learning_rate": 1.7204660984857638e-05,
+      "loss": 0.8105,
+      "step": 11317
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9374742429914145,
+      "learning_rate": 1.7204131802417826e-05,
+      "loss": 0.9409,
+      "step": 11318
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.866314372506022,
+      "learning_rate": 1.7203602578033275e-05,
+      "loss": 0.8457,
+      "step": 11319
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6184687666049486,
+      "learning_rate": 1.7203073311707072e-05,
+      "loss": 0.833,
+      "step": 11320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.223499025143294,
+      "learning_rate": 1.720254400344229e-05,
+      "loss": 0.8152,
+      "step": 11321
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6617084987464124,
+      "learning_rate": 1.720201465324202e-05,
+      "loss": 0.9054,
+      "step": 11322
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7327258088105648,
+      "learning_rate": 1.7201485261109345e-05,
+      "loss": 0.963,
+      "step": 11323
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.707849537863288,
+      "learning_rate": 1.720095582704734e-05,
+      "loss": 0.9366,
+      "step": 11324
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6123715090955575,
+      "learning_rate": 1.7200426351059093e-05,
+      "loss": 0.9022,
+      "step": 11325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6752608992853193,
+      "learning_rate": 1.719989683314768e-05,
+      "loss": 0.9343,
+      "step": 11326
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4754632632607083,
+      "learning_rate": 1.7199367273316187e-05,
+      "loss": 0.9213,
+      "step": 11327
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4320990954885167,
+      "learning_rate": 1.71988376715677e-05,
+      "loss": 0.9752,
+      "step": 11328
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4188208542048584,
+      "learning_rate": 1.71983080279053e-05,
+      "loss": 1.0033,
+      "step": 11329
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5547874586805186,
+      "learning_rate": 1.7197778342332075e-05,
+      "loss": 0.7959,
+      "step": 11330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7286259069122283,
+      "learning_rate": 1.7197248614851103e-05,
+      "loss": 0.9021,
+      "step": 11331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1832495161058953,
+      "learning_rate": 1.7196718845465472e-05,
+      "loss": 0.8731,
+      "step": 11332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6884186171553843,
+      "learning_rate": 1.7196189034178267e-05,
+      "loss": 0.9219,
+      "step": 11333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6835899185302465,
+      "learning_rate": 1.719565918099257e-05,
+      "loss": 0.9077,
+      "step": 11334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9358579402080727,
+      "learning_rate": 1.7195129285911465e-05,
+      "loss": 0.9847,
+      "step": 11335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7102982723872873,
+      "learning_rate": 1.7194599348938042e-05,
+      "loss": 0.8825,
+      "step": 11336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6228966846267947,
+      "learning_rate": 1.719406937007538e-05,
+      "loss": 0.9178,
+      "step": 11337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.864234890595187,
+      "learning_rate": 1.7193539349326573e-05,
+      "loss": 0.8576,
+      "step": 11338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6129149892964778,
+      "learning_rate": 1.71930092866947e-05,
+      "loss": 0.8575,
+      "step": 11339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5567319481540596,
+      "learning_rate": 1.7192479182182853e-05,
+      "loss": 0.9521,
+      "step": 11340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3537756846570717,
+      "learning_rate": 1.7191949035794113e-05,
+      "loss": 0.8831,
+      "step": 11341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7276896351285411,
+      "learning_rate": 1.7191418847531567e-05,
+      "loss": 0.9031,
+      "step": 11342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7134366496436604,
+      "learning_rate": 1.7190888617398306e-05,
+      "loss": 1.0522,
+      "step": 11343
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6000227715472919,
+      "learning_rate": 1.7190358345397412e-05,
+      "loss": 0.9542,
+      "step": 11344
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7221256991250737,
+      "learning_rate": 1.718982803153198e-05,
+      "loss": 0.9275,
+      "step": 11345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.029983013036052,
+      "learning_rate": 1.7189297675805088e-05,
+      "loss": 0.9952,
+      "step": 11346
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.812962330575727,
+      "learning_rate": 1.718876727821983e-05,
+      "loss": 0.9231,
+      "step": 11347
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7276170995590296,
+      "learning_rate": 1.7188236838779297e-05,
+      "loss": 0.919,
+      "step": 11348
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6652889958370245,
+      "learning_rate": 1.718770635748657e-05,
+      "loss": 0.9245,
+      "step": 11349
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.884620367886352,
+      "learning_rate": 1.718717583434474e-05,
+      "loss": 1.0199,
+      "step": 11350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4294895380478747,
+      "learning_rate": 1.71866452693569e-05,
+      "loss": 0.887,
+      "step": 11351
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.664550344976814,
+      "learning_rate": 1.718611466252613e-05,
+      "loss": 0.8352,
+      "step": 11352
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7777236928647173,
+      "learning_rate": 1.718558401385553e-05,
+      "loss": 0.9145,
+      "step": 11353
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6071855779939133,
+      "learning_rate": 1.7185053323348187e-05,
+      "loss": 0.8236,
+      "step": 11354
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6634168075772098,
+      "learning_rate": 1.7184522591007184e-05,
+      "loss": 0.9458,
+      "step": 11355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6831713643633082,
+      "learning_rate": 1.718399181683562e-05,
+      "loss": 0.8388,
+      "step": 11356
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.060290120244752,
+      "learning_rate": 1.718346100083658e-05,
+      "loss": 0.8938,
+      "step": 11357
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7815693027317372,
+      "learning_rate": 1.7182930143013154e-05,
+      "loss": 0.7371,
+      "step": 11358
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.99737099914396,
+      "learning_rate": 1.7182399243368436e-05,
+      "loss": 0.8011,
+      "step": 11359
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5891593819835346,
+      "learning_rate": 1.7181868301905515e-05,
+      "loss": 0.8847,
+      "step": 11360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7486090891279441,
+      "learning_rate": 1.7181337318627485e-05,
+      "loss": 0.697,
+      "step": 11361
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7838175708539987,
+      "learning_rate": 1.7180806293537435e-05,
+      "loss": 0.9215,
+      "step": 11362
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6268777642006638,
+      "learning_rate": 1.7180275226638455e-05,
+      "loss": 0.9211,
+      "step": 11363
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6892436525409875,
+      "learning_rate": 1.7179744117933642e-05,
+      "loss": 0.9193,
+      "step": 11364
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7408899828720742,
+      "learning_rate": 1.7179212967426083e-05,
+      "loss": 0.7483,
+      "step": 11365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9553531016948666,
+      "learning_rate": 1.7178681775118874e-05,
+      "loss": 0.9037,
+      "step": 11366
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6628089132335357,
+      "learning_rate": 1.717815054101511e-05,
+      "loss": 0.921,
+      "step": 11367
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6483661514666113,
+      "learning_rate": 1.717761926511788e-05,
+      "loss": 0.9677,
+      "step": 11368
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0159751244982163,
+      "learning_rate": 1.7177087947430275e-05,
+      "loss": 0.9093,
+      "step": 11369
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7087919671363105,
+      "learning_rate": 1.717655658795539e-05,
+      "loss": 0.9811,
+      "step": 11370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7630925400306491,
+      "learning_rate": 1.717602518669633e-05,
+      "loss": 0.7712,
+      "step": 11371
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6510479556184774,
+      "learning_rate": 1.717549374365617e-05,
+      "loss": 0.902,
+      "step": 11372
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8017233180382011,
+      "learning_rate": 1.717496225883802e-05,
+      "loss": 0.7689,
+      "step": 11373
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6004742404078445,
+      "learning_rate": 1.717443073224496e-05,
+      "loss": 0.8413,
+      "step": 11374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5563305530574527,
+      "learning_rate": 1.7173899163880105e-05,
+      "loss": 0.845,
+      "step": 11375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5430737316543943,
+      "learning_rate": 1.7173367553746526e-05,
+      "loss": 0.931,
+      "step": 11376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5939679953430137,
+      "learning_rate": 1.7172835901847338e-05,
+      "loss": 0.9471,
+      "step": 11377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7706775136941897,
+      "learning_rate": 1.7172304208185627e-05,
+      "loss": 0.7899,
+      "step": 11378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5502524809168576,
+      "learning_rate": 1.7171772472764488e-05,
+      "loss": 0.9827,
+      "step": 11379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7205845951194103,
+      "learning_rate": 1.717124069558702e-05,
+      "loss": 0.9159,
+      "step": 11380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8632694651435446,
+      "learning_rate": 1.717070887665632e-05,
+      "loss": 0.9788,
+      "step": 11381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9918350832606233,
+      "learning_rate": 1.717017701597548e-05,
+      "loss": 0.8796,
+      "step": 11382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.724311361061355,
+      "learning_rate": 1.7169645113547598e-05,
+      "loss": 0.9351,
+      "step": 11383
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7722197835915205,
+      "learning_rate": 1.7169113169375778e-05,
+      "loss": 0.9821,
+      "step": 11384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8019631594753835,
+      "learning_rate": 1.7168581183463106e-05,
+      "loss": 0.9072,
+      "step": 11385
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.569312208602369,
+      "learning_rate": 1.7168049155812686e-05,
+      "loss": 0.798,
+      "step": 11386
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9836636805170319,
+      "learning_rate": 1.7167517086427618e-05,
+      "loss": 0.7554,
+      "step": 11387
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6585107150428935,
+      "learning_rate": 1.716698497531099e-05,
+      "loss": 0.8491,
+      "step": 11388
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.792064240082765,
+      "learning_rate": 1.7166452822465913e-05,
+      "loss": 0.7361,
+      "step": 11389
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7266500953834336,
+      "learning_rate": 1.7165920627895477e-05,
+      "loss": 0.9502,
+      "step": 11390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.885580828291886,
+      "learning_rate": 1.7165388391602783e-05,
+      "loss": 0.8997,
+      "step": 11391
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7242805637590843,
+      "learning_rate": 1.7164856113590925e-05,
+      "loss": 1.0064,
+      "step": 11392
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.117948987909777,
+      "learning_rate": 1.716432379386301e-05,
+      "loss": 0.9639,
+      "step": 11393
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6581071939200476,
+      "learning_rate": 1.7163791432422133e-05,
+      "loss": 0.9431,
+      "step": 11394
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8907740719281043,
+      "learning_rate": 1.7163259029271392e-05,
+      "loss": 0.9358,
+      "step": 11395
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7846874319820816,
+      "learning_rate": 1.716272658441389e-05,
+      "loss": 0.8349,
+      "step": 11396
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2276299348390243,
+      "learning_rate": 1.716219409785273e-05,
+      "loss": 0.8088,
+      "step": 11397
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9505351426663295,
+      "learning_rate": 1.7161661569591004e-05,
+      "loss": 0.8613,
+      "step": 11398
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7486554313071256,
+      "learning_rate": 1.716112899963182e-05,
+      "loss": 0.9406,
+      "step": 11399
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8580160586235932,
+      "learning_rate": 1.7160596387978273e-05,
+      "loss": 0.9647,
+      "step": 11400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5055787627915374,
+      "learning_rate": 1.7160063734633468e-05,
+      "loss": 0.9109,
+      "step": 11401
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.651107062159186,
+      "learning_rate": 1.7159531039600507e-05,
+      "loss": 0.9321,
+      "step": 11402
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0525341508371495,
+      "learning_rate": 1.715899830288249e-05,
+      "loss": 0.9714,
+      "step": 11403
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0678787652405957,
+      "learning_rate": 1.7158465524482512e-05,
+      "loss": 0.9351,
+      "step": 11404
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.700316851372971,
+      "learning_rate": 1.715793270440369e-05,
+      "loss": 0.9165,
+      "step": 11405
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5146110781468416,
+      "learning_rate": 1.715739984264911e-05,
+      "loss": 0.9486,
+      "step": 11406
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4764986583791377,
+      "learning_rate": 1.7156866939221888e-05,
+      "loss": 0.8541,
+      "step": 11407
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.509073235931666,
+      "learning_rate": 1.715633399412512e-05,
+      "loss": 0.8933,
+      "step": 11408
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8634205818938623,
+      "learning_rate": 1.7155801007361908e-05,
+      "loss": 0.8122,
+      "step": 11409
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2657771771343445,
+      "learning_rate": 1.7155267978935355e-05,
+      "loss": 0.8844,
+      "step": 11410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6901788608444128,
+      "learning_rate": 1.7154734908848568e-05,
+      "loss": 0.9455,
+      "step": 11411
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.69787243008886,
+      "learning_rate": 1.715420179710465e-05,
+      "loss": 0.9864,
+      "step": 11412
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9184971946337576,
+      "learning_rate": 1.7153668643706704e-05,
+      "loss": 0.728,
+      "step": 11413
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.654107066068019,
+      "learning_rate": 1.7153135448657834e-05,
+      "loss": 1.0503,
+      "step": 11414
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5938385212132609,
+      "learning_rate": 1.7152602211961146e-05,
+      "loss": 1.0343,
+      "step": 11415
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.185665832539685,
+      "learning_rate": 1.7152068933619742e-05,
+      "loss": 0.893,
+      "step": 11416
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.698769008641849,
+      "learning_rate": 1.715153561363673e-05,
+      "loss": 0.8279,
+      "step": 11417
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8266726841853216,
+      "learning_rate": 1.7151002252015213e-05,
+      "loss": 0.7609,
+      "step": 11418
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5348022080707162,
+      "learning_rate": 1.7150468848758296e-05,
+      "loss": 0.9296,
+      "step": 11419
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4762152985484702,
+      "learning_rate": 1.7149935403869086e-05,
+      "loss": 0.9137,
+      "step": 11420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8296819678047944,
+      "learning_rate": 1.714940191735069e-05,
+      "loss": 0.7192,
+      "step": 11421
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5295132273161378,
+      "learning_rate": 1.714886838920621e-05,
+      "loss": 0.7568,
+      "step": 11422
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7158015953181838,
+      "learning_rate": 1.7148334819438758e-05,
+      "loss": 0.9473,
+      "step": 11423
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5229523340395725,
+      "learning_rate": 1.7147801208051432e-05,
+      "loss": 0.9053,
+      "step": 11424
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9131144032677798,
+      "learning_rate": 1.7147267555047352e-05,
+      "loss": 0.9814,
+      "step": 11425
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.709766913356567,
+      "learning_rate": 1.7146733860429614e-05,
+      "loss": 0.7762,
+      "step": 11426
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8774543830431787,
+      "learning_rate": 1.7146200124201327e-05,
+      "loss": 0.8153,
+      "step": 11427
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.784513431606304,
+      "learning_rate": 1.71456663463656e-05,
+      "loss": 0.7083,
+      "step": 11428
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.632739147671576,
+      "learning_rate": 1.7145132526925543e-05,
+      "loss": 0.7625,
+      "step": 11429
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.847207579952492,
+      "learning_rate": 1.714459866588426e-05,
+      "loss": 0.8976,
+      "step": 11430
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.496365639621765,
+      "learning_rate": 1.7144064763244864e-05,
+      "loss": 0.8635,
+      "step": 11431
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.793260699680695,
+      "learning_rate": 1.7143530819010463e-05,
+      "loss": 0.8857,
+      "step": 11432
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9733504224154714,
+      "learning_rate": 1.7142996833184163e-05,
+      "loss": 0.9617,
+      "step": 11433
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5266522377724228,
+      "learning_rate": 1.7142462805769073e-05,
+      "loss": 0.8865,
+      "step": 11434
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5958925890568825,
+      "learning_rate": 1.71419287367683e-05,
+      "loss": 0.9828,
+      "step": 11435
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7219423879462592,
+      "learning_rate": 1.7141394626184965e-05,
+      "loss": 0.8242,
+      "step": 11436
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9463892333484476,
+      "learning_rate": 1.7140860474022163e-05,
+      "loss": 0.9328,
+      "step": 11437
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6974985745910258,
+      "learning_rate": 1.7140326280283013e-05,
+      "loss": 0.9066,
+      "step": 11438
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5998846748798268,
+      "learning_rate": 1.7139792044970622e-05,
+      "loss": 0.7738,
+      "step": 11439
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0056420018115815,
+      "learning_rate": 1.71392577680881e-05,
+      "loss": 0.8951,
+      "step": 11440
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5902853173771543,
+      "learning_rate": 1.7138723449638562e-05,
+      "loss": 0.9801,
+      "step": 11441
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7008720708192695,
+      "learning_rate": 1.7138189089625113e-05,
+      "loss": 0.9327,
+      "step": 11442
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9086400890715933,
+      "learning_rate": 1.7137654688050868e-05,
+      "loss": 0.948,
+      "step": 11443
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.82634833085467,
+      "learning_rate": 1.713712024491894e-05,
+      "loss": 1.0446,
+      "step": 11444
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4573236803517646,
+      "learning_rate": 1.7136585760232436e-05,
+      "loss": 0.7637,
+      "step": 11445
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.856085892677114,
+      "learning_rate": 1.713605123399447e-05,
+      "loss": 0.8405,
+      "step": 11446
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.884965514558646,
+      "learning_rate": 1.7135516666208156e-05,
+      "loss": 1.0073,
+      "step": 11447
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6370582298383074,
+      "learning_rate": 1.7134982056876604e-05,
+      "loss": 0.8969,
+      "step": 11448
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6713977829061006,
+      "learning_rate": 1.713444740600293e-05,
+      "loss": 0.9881,
+      "step": 11449
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.563106137645751,
+      "learning_rate": 1.7133912713590243e-05,
+      "loss": 0.9424,
+      "step": 11450
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5649329350401413,
+      "learning_rate": 1.7133377979641655e-05,
+      "loss": 0.8067,
+      "step": 11451
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4140733172548932,
+      "learning_rate": 1.7132843204160284e-05,
+      "loss": 0.926,
+      "step": 11452
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5671471331674454,
+      "learning_rate": 1.713230838714924e-05,
+      "loss": 0.8972,
+      "step": 11453
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7355772918241645,
+      "learning_rate": 1.7131773528611643e-05,
+      "loss": 0.8605,
+      "step": 11454
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.73405012860579,
+      "learning_rate": 1.71312386285506e-05,
+      "loss": 0.8694,
+      "step": 11455
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.684067853908445,
+      "learning_rate": 1.713070368696923e-05,
+      "loss": 1.0996,
+      "step": 11456
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6548761586434868,
+      "learning_rate": 1.713016870387064e-05,
+      "loss": 0.9217,
+      "step": 11457
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6632430275021293,
+      "learning_rate": 1.7129633679257956e-05,
+      "loss": 0.8935,
+      "step": 11458
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.655553297701889,
+      "learning_rate": 1.7129098613134285e-05,
+      "loss": 0.9158,
+      "step": 11459
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7651227451940005,
+      "learning_rate": 1.7128563505502744e-05,
+      "loss": 0.733,
+      "step": 11460
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.030158161845256,
+      "learning_rate": 1.712802835636645e-05,
+      "loss": 0.8896,
+      "step": 11461
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6694314389683522,
+      "learning_rate": 1.712749316572852e-05,
+      "loss": 0.8657,
+      "step": 11462
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.642280822214301,
+      "learning_rate": 1.7126957933592067e-05,
+      "loss": 0.744,
+      "step": 11463
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7250163474573859,
+      "learning_rate": 1.7126422659960208e-05,
+      "loss": 0.8718,
+      "step": 11464
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5979261685289068,
+      "learning_rate": 1.712588734483606e-05,
+      "loss": 0.9046,
+      "step": 11465
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.172789264694019,
+      "learning_rate": 1.712535198822274e-05,
+      "loss": 0.8133,
+      "step": 11466
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0580157185168506,
+      "learning_rate": 1.7124816590123368e-05,
+      "loss": 0.9026,
+      "step": 11467
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6521612910419123,
+      "learning_rate": 1.7124281150541054e-05,
+      "loss": 0.835,
+      "step": 11468
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8126573241775293,
+      "learning_rate": 1.712374566947892e-05,
+      "loss": 0.7476,
+      "step": 11469
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7350145441143874,
+      "learning_rate": 1.7123210146940087e-05,
+      "loss": 0.9359,
+      "step": 11470
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8508188811294215,
+      "learning_rate": 1.7122674582927668e-05,
+      "loss": 0.9486,
+      "step": 11471
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.793439353709271,
+      "learning_rate": 1.712213897744478e-05,
+      "loss": 0.7518,
+      "step": 11472
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5642368373114717,
+      "learning_rate": 1.7121603330494547e-05,
+      "loss": 0.8294,
+      "step": 11473
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7060712271116822,
+      "learning_rate": 1.712106764208008e-05,
+      "loss": 0.9247,
+      "step": 11474
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8524823687547372,
+      "learning_rate": 1.7120531912204507e-05,
+      "loss": 0.8532,
+      "step": 11475
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8723249740730281,
+      "learning_rate": 1.711999614087094e-05,
+      "loss": 0.9933,
+      "step": 11476
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.548342008074934,
+      "learning_rate": 1.7119460328082506e-05,
+      "loss": 0.9152,
+      "step": 11477
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4665655124557557,
+      "learning_rate": 1.7118924473842318e-05,
+      "loss": 0.9202,
+      "step": 11478
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8481006026244005,
+      "learning_rate": 1.7118388578153498e-05,
+      "loss": 0.9294,
+      "step": 11479
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4693167043695068,
+      "learning_rate": 1.7117852641019162e-05,
+      "loss": 0.7709,
+      "step": 11480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4861664603048645,
+      "learning_rate": 1.711731666244244e-05,
+      "loss": 0.8891,
+      "step": 11481
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6093034464149893,
+      "learning_rate": 1.7116780642426444e-05,
+      "loss": 0.8917,
+      "step": 11482
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.443912487010528,
+      "learning_rate": 1.71162445809743e-05,
+      "loss": 0.8138,
+      "step": 11483
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6825692847265623,
+      "learning_rate": 1.7115708478089124e-05,
+      "loss": 0.9325,
+      "step": 11484
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5913940021888524,
+      "learning_rate": 1.7115172333774043e-05,
+      "loss": 0.9109,
+      "step": 11485
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.235862488883453,
+      "learning_rate": 1.7114636148032176e-05,
+      "loss": 0.803,
+      "step": 11486
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.314730365046471,
+      "learning_rate": 1.7114099920866644e-05,
+      "loss": 0.8849,
+      "step": 11487
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6739731373929418,
+      "learning_rate": 1.7113563652280568e-05,
+      "loss": 0.8517,
+      "step": 11488
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8688437043902857,
+      "learning_rate": 1.7113027342277077e-05,
+      "loss": 0.8822,
+      "step": 11489
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8106988360336522,
+      "learning_rate": 1.7112490990859286e-05,
+      "loss": 0.9453,
+      "step": 11490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7447717594394727,
+      "learning_rate": 1.711195459803032e-05,
+      "loss": 0.7551,
+      "step": 11491
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5311540437028734,
+      "learning_rate": 1.71114181637933e-05,
+      "loss": 0.9338,
+      "step": 11492
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7601577874406893,
+      "learning_rate": 1.711088168815136e-05,
+      "loss": 0.9278,
+      "step": 11493
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8367757688779272,
+      "learning_rate": 1.711034517110761e-05,
+      "loss": 0.8374,
+      "step": 11494
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8752309042343396,
+      "learning_rate": 1.710980861266518e-05,
+      "loss": 0.8239,
+      "step": 11495
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3819686404904938,
+      "learning_rate": 1.7109272012827195e-05,
+      "loss": 0.7747,
+      "step": 11496
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7432221228182592,
+      "learning_rate": 1.7108735371596776e-05,
+      "loss": 0.9635,
+      "step": 11497
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.099846757773804,
+      "learning_rate": 1.7108198688977048e-05,
+      "loss": 0.9248,
+      "step": 11498
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6399114414978462,
+      "learning_rate": 1.7107661964971137e-05,
+      "loss": 0.9105,
+      "step": 11499
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2414903692373236,
+      "learning_rate": 1.710712519958217e-05,
+      "loss": 0.9317,
+      "step": 11500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.689169544970982,
+      "learning_rate": 1.710658839281327e-05,
+      "loss": 0.8635,
+      "step": 11501
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.491453427146651,
+      "learning_rate": 1.710605154466756e-05,
+      "loss": 1.0583,
+      "step": 11502
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6965908288672695,
+      "learning_rate": 1.7105514655148173e-05,
+      "loss": 0.8874,
+      "step": 11503
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.041858424067533,
+      "learning_rate": 1.710497772425823e-05,
+      "loss": 0.8696,
+      "step": 11504
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.64036717384663,
+      "learning_rate": 1.710444075200085e-05,
+      "loss": 0.9459,
+      "step": 11505
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.383384338976247,
+      "learning_rate": 1.710390373837917e-05,
+      "loss": 0.8068,
+      "step": 11506
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7997763773035107,
+      "learning_rate": 1.7103366683396316e-05,
+      "loss": 0.7179,
+      "step": 11507
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.563479109991507,
+      "learning_rate": 1.7102829587055413e-05,
+      "loss": 0.8612,
+      "step": 11508
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8060742427372333,
+      "learning_rate": 1.710229244935958e-05,
+      "loss": 0.7331,
+      "step": 11509
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8631151978370708,
+      "learning_rate": 1.7101755270311957e-05,
+      "loss": 0.9383,
+      "step": 11510
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4804608817728107,
+      "learning_rate": 1.7101218049915666e-05,
+      "loss": 0.8898,
+      "step": 11511
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7468019727880773,
+      "learning_rate": 1.7100680788173836e-05,
+      "loss": 0.881,
+      "step": 11512
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.063880866753923,
+      "learning_rate": 1.7100143485089596e-05,
+      "loss": 0.7968,
+      "step": 11513
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5272847821847806,
+      "learning_rate": 1.7099606140666068e-05,
+      "loss": 0.9091,
+      "step": 11514
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2064354526994747,
+      "learning_rate": 1.709906875490639e-05,
+      "loss": 0.9595,
+      "step": 11515
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4953020483256032,
+      "learning_rate": 1.7098531327813684e-05,
+      "loss": 0.8659,
+      "step": 11516
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6709141598493518,
+      "learning_rate": 1.7097993859391082e-05,
+      "loss": 0.8964,
+      "step": 11517
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5405667044164995,
+      "learning_rate": 1.709745634964171e-05,
+      "loss": 0.8231,
+      "step": 11518
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0025721329476065,
+      "learning_rate": 1.7096918798568705e-05,
+      "loss": 0.7671,
+      "step": 11519
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5528758412296766,
+      "learning_rate": 1.7096381206175188e-05,
+      "loss": 0.7917,
+      "step": 11520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.619629657174909,
+      "learning_rate": 1.7095843572464294e-05,
+      "loss": 1.0247,
+      "step": 11521
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6291577219172864,
+      "learning_rate": 1.7095305897439152e-05,
+      "loss": 0.8581,
+      "step": 11522
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9690246949778445,
+      "learning_rate": 1.7094768181102895e-05,
+      "loss": 0.8534,
+      "step": 11523
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4771765131676844,
+      "learning_rate": 1.709423042345865e-05,
+      "loss": 0.994,
+      "step": 11524
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5230142291173887,
+      "learning_rate": 1.7093692624509545e-05,
+      "loss": 0.7815,
+      "step": 11525
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6558259196702714,
+      "learning_rate": 1.709315478425872e-05,
+      "loss": 0.8829,
+      "step": 11526
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.740447632436287,
+      "learning_rate": 1.7092616902709304e-05,
+      "loss": 0.812,
+      "step": 11527
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.104244157237844,
+      "learning_rate": 1.7092078979864425e-05,
+      "loss": 0.9141,
+      "step": 11528
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2204561893708736,
+      "learning_rate": 1.7091541015727216e-05,
+      "loss": 0.781,
+      "step": 11529
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8044155389607317,
+      "learning_rate": 1.709100301030081e-05,
+      "loss": 0.92,
+      "step": 11530
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5012236375434176,
+      "learning_rate": 1.709046496358834e-05,
+      "loss": 1.0298,
+      "step": 11531
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7751973742578129,
+      "learning_rate": 1.7089926875592938e-05,
+      "loss": 0.7283,
+      "step": 11532
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8313101817376523,
+      "learning_rate": 1.7089388746317738e-05,
+      "loss": 0.9262,
+      "step": 11533
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8226443422572279,
+      "learning_rate": 1.708885057576587e-05,
+      "loss": 0.7717,
+      "step": 11534
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.947389257459457,
+      "learning_rate": 1.7088312363940472e-05,
+      "loss": 0.8683,
+      "step": 11535
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7176079547516587,
+      "learning_rate": 1.7087774110844674e-05,
+      "loss": 1.0006,
+      "step": 11536
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6235719281528203,
+      "learning_rate": 1.7087235816481614e-05,
+      "loss": 0.9139,
+      "step": 11537
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8619647355586075,
+      "learning_rate": 1.7086697480854423e-05,
+      "loss": 0.9479,
+      "step": 11538
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7731913947923459,
+      "learning_rate": 1.708615910396623e-05,
+      "loss": 0.7378,
+      "step": 11539
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.766301409229372,
+      "learning_rate": 1.708562068582018e-05,
+      "loss": 0.8791,
+      "step": 11540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6592940282476747,
+      "learning_rate": 1.7085082226419402e-05,
+      "loss": 0.9762,
+      "step": 11541
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8042910529548501,
+      "learning_rate": 1.7084543725767035e-05,
+      "loss": 0.7978,
+      "step": 11542
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7088939749786394,
+      "learning_rate": 1.7084005183866206e-05,
+      "loss": 0.9735,
+      "step": 11543
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4625315212710355,
+      "learning_rate": 1.708346660072006e-05,
+      "loss": 0.7912,
+      "step": 11544
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.523003844538643,
+      "learning_rate": 1.7082927976331726e-05,
+      "loss": 0.946,
+      "step": 11545
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4970947760992555,
+      "learning_rate": 1.7082389310704347e-05,
+      "loss": 0.8077,
+      "step": 11546
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5979042564619261,
+      "learning_rate": 1.7081850603841053e-05,
+      "loss": 0.8997,
+      "step": 11547
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9367478723520104,
+      "learning_rate": 1.708131185574498e-05,
+      "loss": 0.9274,
+      "step": 11548
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.699647751539209,
+      "learning_rate": 1.7080773066419272e-05,
+      "loss": 0.8691,
+      "step": 11549
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5322891681695765,
+      "learning_rate": 1.708023423586706e-05,
+      "loss": 0.8798,
+      "step": 11550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6707734581891414,
+      "learning_rate": 1.7079695364091483e-05,
+      "loss": 0.9398,
+      "step": 11551
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5298035437633744,
+      "learning_rate": 1.7079156451095677e-05,
+      "loss": 0.9224,
+      "step": 11552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6849300160469252,
+      "learning_rate": 1.707861749688278e-05,
+      "loss": 0.8955,
+      "step": 11553
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.128649544153777,
+      "learning_rate": 1.707807850145593e-05,
+      "loss": 0.888,
+      "step": 11554
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.146374697997401,
+      "learning_rate": 1.707753946481827e-05,
+      "loss": 0.8764,
+      "step": 11555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.953048645494002,
+      "learning_rate": 1.7077000386972934e-05,
+      "loss": 0.9809,
+      "step": 11556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5838911213394093,
+      "learning_rate": 1.707646126792306e-05,
+      "loss": 0.8842,
+      "step": 11557
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8594342395930947,
+      "learning_rate": 1.7075922107671786e-05,
+      "loss": 1.0105,
+      "step": 11558
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6713811330210209,
+      "learning_rate": 1.7075382906222254e-05,
+      "loss": 0.9447,
+      "step": 11559
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7603390427176494,
+      "learning_rate": 1.7074843663577607e-05,
+      "loss": 0.7517,
+      "step": 11560
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.573155191949161,
+      "learning_rate": 1.7074304379740975e-05,
+      "loss": 0.8753,
+      "step": 11561
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4847781396206463,
+      "learning_rate": 1.7073765054715506e-05,
+      "loss": 0.7831,
+      "step": 11562
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.616623884212629,
+      "learning_rate": 1.7073225688504337e-05,
+      "loss": 0.889,
+      "step": 11563
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5942949287535555,
+      "learning_rate": 1.707268628111061e-05,
+      "loss": 0.952,
+      "step": 11564
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9026382411602571,
+      "learning_rate": 1.7072146832537464e-05,
+      "loss": 0.8855,
+      "step": 11565
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.974565155979667,
+      "learning_rate": 1.707160734278804e-05,
+      "loss": 1.0009,
+      "step": 11566
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.601178808688326,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 1.0636,
+      "step": 11567
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5055221186601484,
+      "learning_rate": 1.707052823977292e-05,
+      "loss": 0.8135,
+      "step": 11568
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.421369984919532,
+      "learning_rate": 1.7069988626513508e-05,
+      "loss": 0.9863,
+      "step": 11569
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.012314352473784,
+      "learning_rate": 1.7069448972090387e-05,
+      "loss": 0.8518,
+      "step": 11570
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3992473291418934,
+      "learning_rate": 1.7068909276506692e-05,
+      "loss": 0.9066,
+      "step": 11571
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2054382848717315,
+      "learning_rate": 1.7068369539765572e-05,
+      "loss": 0.8454,
+      "step": 11572
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5091226622910294,
+      "learning_rate": 1.7067829761870166e-05,
+      "loss": 0.9735,
+      "step": 11573
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8124596527140593,
+      "learning_rate": 1.7067289942823617e-05,
+      "loss": 0.9416,
+      "step": 11574
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7978354874505188,
+      "learning_rate": 1.7066750082629067e-05,
+      "loss": 0.766,
+      "step": 11575
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8509937077114436,
+      "learning_rate": 1.7066210181289666e-05,
+      "loss": 0.9423,
+      "step": 11576
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6216038623765752,
+      "learning_rate": 1.7065670238808545e-05,
+      "loss": 1.0593,
+      "step": 11577
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8508823908752063,
+      "learning_rate": 1.706513025518886e-05,
+      "loss": 0.8515,
+      "step": 11578
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7483662607870998,
+      "learning_rate": 1.706459023043375e-05,
+      "loss": 0.9007,
+      "step": 11579
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.548558060168807,
+      "learning_rate": 1.7064050164546356e-05,
+      "loss": 0.827,
+      "step": 11580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6598767811796902,
+      "learning_rate": 1.7063510057529825e-05,
+      "loss": 0.9289,
+      "step": 11581
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.484989509074972,
+      "learning_rate": 1.706296990938731e-05,
+      "loss": 0.9321,
+      "step": 11582
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9358991189692525,
+      "learning_rate": 1.7062429720121937e-05,
+      "loss": 0.9187,
+      "step": 11583
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6527288021845372,
+      "learning_rate": 1.706188948973687e-05,
+      "loss": 0.9765,
+      "step": 11584
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7614483214765372,
+      "learning_rate": 1.7061349218235243e-05,
+      "loss": 0.7976,
+      "step": 11585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7777884425968749,
+      "learning_rate": 1.7060808905620207e-05,
+      "loss": 0.93,
+      "step": 11586
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0426921694586864,
+      "learning_rate": 1.7060268551894904e-05,
+      "loss": 0.9511,
+      "step": 11587
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5752192731632482,
+      "learning_rate": 1.7059728157062485e-05,
+      "loss": 0.8978,
+      "step": 11588
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.691087471147994,
+      "learning_rate": 1.7059187721126093e-05,
+      "loss": 0.9421,
+      "step": 11589
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5364401199064779,
+      "learning_rate": 1.7058647244088873e-05,
+      "loss": 0.8177,
+      "step": 11590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.62598650744192,
+      "learning_rate": 1.7058106725953977e-05,
+      "loss": 0.9865,
+      "step": 11591
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6388637538511839,
+      "learning_rate": 1.7057566166724545e-05,
+      "loss": 0.8934,
+      "step": 11592
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4790186925555402,
+      "learning_rate": 1.7057025566403733e-05,
+      "loss": 0.8886,
+      "step": 11593
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7864371867429285,
+      "learning_rate": 1.705648492499468e-05,
+      "loss": 0.7522,
+      "step": 11594
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6756395468350593,
+      "learning_rate": 1.7055944242500543e-05,
+      "loss": 0.9222,
+      "step": 11595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6058661399672196,
+      "learning_rate": 1.705540351892446e-05,
+      "loss": 1.0004,
+      "step": 11596
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5057697924297244,
+      "learning_rate": 1.7054862754269586e-05,
+      "loss": 0.8708,
+      "step": 11597
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7497575966759404,
+      "learning_rate": 1.7054321948539066e-05,
+      "loss": 1.0279,
+      "step": 11598
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7435701999380724,
+      "learning_rate": 1.705378110173605e-05,
+      "loss": 0.8617,
+      "step": 11599
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5958955084779445,
+      "learning_rate": 1.705324021386369e-05,
+      "loss": 0.7847,
+      "step": 11600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6314445686684005,
+      "learning_rate": 1.705269928492513e-05,
+      "loss": 1.05,
+      "step": 11601
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.454813979542336,
+      "learning_rate": 1.7052158314923524e-05,
+      "loss": 0.8645,
+      "step": 11602
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.118807077224749,
+      "learning_rate": 1.705161730386202e-05,
+      "loss": 0.8619,
+      "step": 11603
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8112111372932981,
+      "learning_rate": 1.7051076251743763e-05,
+      "loss": 0.7783,
+      "step": 11604
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.008347335416011,
+      "learning_rate": 1.705053515857191e-05,
+      "loss": 0.992,
+      "step": 11605
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.612084755242021,
+      "learning_rate": 1.7049994024349613e-05,
+      "loss": 1.029,
+      "step": 11606
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.850902896790132,
+      "learning_rate": 1.7049452849080014e-05,
+      "loss": 0.9849,
+      "step": 11607
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.03705725534324,
+      "learning_rate": 1.704891163276627e-05,
+      "loss": 0.8796,
+      "step": 11608
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6926023873390157,
+      "learning_rate": 1.7048370375411533e-05,
+      "loss": 0.9729,
+      "step": 11609
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8088223847589475,
+      "learning_rate": 1.704782907701895e-05,
+      "loss": 0.8285,
+      "step": 11610
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8984387696950202,
+      "learning_rate": 1.7047287737591672e-05,
+      "loss": 0.8345,
+      "step": 11611
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5371919271475185,
+      "learning_rate": 1.7046746357132858e-05,
+      "loss": 0.912,
+      "step": 11612
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6970608140325327,
+      "learning_rate": 1.7046204935645655e-05,
+      "loss": 0.9539,
+      "step": 11613
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7810545532308008,
+      "learning_rate": 1.7045663473133215e-05,
+      "loss": 0.7696,
+      "step": 11614
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4386063793825112,
+      "learning_rate": 1.7045121969598694e-05,
+      "loss": 0.8095,
+      "step": 11615
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5024166576436453,
+      "learning_rate": 1.7044580425045244e-05,
+      "loss": 0.9481,
+      "step": 11616
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8723801597469543,
+      "learning_rate": 1.7044038839476013e-05,
+      "loss": 0.8191,
+      "step": 11617
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.436289927272891,
+      "learning_rate": 1.7043497212894154e-05,
+      "loss": 0.873,
+      "step": 11618
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5066327440816907,
+      "learning_rate": 1.704295554530283e-05,
+      "loss": 0.9547,
+      "step": 11619
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5755446524003984,
+      "learning_rate": 1.7042413836705187e-05,
+      "loss": 0.9543,
+      "step": 11620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.167476553195292,
+      "learning_rate": 1.7041872087104383e-05,
+      "loss": 0.9843,
+      "step": 11621
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7626272042422932,
+      "learning_rate": 1.7041330296503566e-05,
+      "loss": 0.8155,
+      "step": 11622
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5453148478757988,
+      "learning_rate": 1.70407884649059e-05,
+      "loss": 0.8031,
+      "step": 11623
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6544119421279364,
+      "learning_rate": 1.704024659231453e-05,
+      "loss": 0.8627,
+      "step": 11624
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.569068614965317,
+      "learning_rate": 1.703970467873262e-05,
+      "loss": 0.9086,
+      "step": 11625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.79597187021612,
+      "learning_rate": 1.7039162724163317e-05,
+      "loss": 0.7799,
+      "step": 11626
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5951947031276372,
+      "learning_rate": 1.703862072860978e-05,
+      "loss": 0.9288,
+      "step": 11627
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.545825079649154,
+      "learning_rate": 1.7038078692075166e-05,
+      "loss": 0.8081,
+      "step": 11628
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6224621538511783,
+      "learning_rate": 1.703753661456263e-05,
+      "loss": 1.0044,
+      "step": 11629
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8634548446274124,
+      "learning_rate": 1.703699449607533e-05,
+      "loss": 0.8689,
+      "step": 11630
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4885075930459806,
+      "learning_rate": 1.7036452336616413e-05,
+      "loss": 0.9165,
+      "step": 11631
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8295654274127038,
+      "learning_rate": 1.703591013618905e-05,
+      "loss": 0.7595,
+      "step": 11632
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7062926184849219,
+      "learning_rate": 1.703536789479639e-05,
+      "loss": 1.0101,
+      "step": 11633
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6344655578890566,
+      "learning_rate": 1.7034825612441585e-05,
+      "loss": 0.9645,
+      "step": 11634
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7290127501528179,
+      "learning_rate": 1.7034283289127806e-05,
+      "loss": 0.9603,
+      "step": 11635
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.126329626307387,
+      "learning_rate": 1.7033740924858194e-05,
+      "loss": 0.9481,
+      "step": 11636
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8505908038432672,
+      "learning_rate": 1.7033198519635922e-05,
+      "loss": 0.868,
+      "step": 11637
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8245951084774321,
+      "learning_rate": 1.7032656073464137e-05,
+      "loss": 0.8579,
+      "step": 11638
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7419980310716987,
+      "learning_rate": 1.7032113586346003e-05,
+      "loss": 0.8585,
+      "step": 11639
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.733716841892521,
+      "learning_rate": 1.7031571058284678e-05,
+      "loss": 0.6445,
+      "step": 11640
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8017692179015206,
+      "learning_rate": 1.7031028489283324e-05,
+      "loss": 0.7345,
+      "step": 11641
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.533057798875491,
+      "learning_rate": 1.703048587934509e-05,
+      "loss": 0.855,
+      "step": 11642
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.612249838370602,
+      "learning_rate": 1.7029943228473147e-05,
+      "loss": 0.957,
+      "step": 11643
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8104479545960976,
+      "learning_rate": 1.7029400536670646e-05,
+      "loss": 0.8913,
+      "step": 11644
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5632728075727855,
+      "learning_rate": 1.702885780394075e-05,
+      "loss": 0.9213,
+      "step": 11645
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5641747713068592,
+      "learning_rate": 1.7028315030286617e-05,
+      "loss": 0.8896,
+      "step": 11646
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6485341945066552,
+      "learning_rate": 1.7027772215711414e-05,
+      "loss": 0.7348,
+      "step": 11647
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.450582131721184,
+      "learning_rate": 1.702722936021829e-05,
+      "loss": 0.8141,
+      "step": 11648
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9122798987507357,
+      "learning_rate": 1.7026686463810413e-05,
+      "loss": 0.9515,
+      "step": 11649
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.754413946032629,
+      "learning_rate": 1.7026143526490947e-05,
+      "loss": 0.8256,
+      "step": 11650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7991336139684752,
+      "learning_rate": 1.7025600548263043e-05,
+      "loss": 0.6946,
+      "step": 11651
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7916969164218977,
+      "learning_rate": 1.702505752912987e-05,
+      "loss": 0.8766,
+      "step": 11652
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7513610077689972,
+      "learning_rate": 1.7024514469094592e-05,
+      "loss": 0.7513,
+      "step": 11653
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8375046562895636,
+      "learning_rate": 1.7023971368160362e-05,
+      "loss": 0.8765,
+      "step": 11654
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7034972679033236,
+      "learning_rate": 1.7023428226330348e-05,
+      "loss": 0.9341,
+      "step": 11655
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6906531525459696,
+      "learning_rate": 1.7022885043607714e-05,
+      "loss": 0.902,
+      "step": 11656
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3890252709688073,
+      "learning_rate": 1.7022341819995622e-05,
+      "loss": 0.839,
+      "step": 11657
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6624740166176695,
+      "learning_rate": 1.7021798555497225e-05,
+      "loss": 0.8398,
+      "step": 11658
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.855462958772305,
+      "learning_rate": 1.70212552501157e-05,
+      "loss": 0.9177,
+      "step": 11659
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.884215360217214,
+      "learning_rate": 1.7020711903854202e-05,
+      "loss": 0.9526,
+      "step": 11660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.763553426430557,
+      "learning_rate": 1.7020168516715894e-05,
+      "loss": 0.7892,
+      "step": 11661
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8993698921408677,
+      "learning_rate": 1.7019625088703944e-05,
+      "loss": 0.7793,
+      "step": 11662
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8718450311080719,
+      "learning_rate": 1.701908161982152e-05,
+      "loss": 0.7619,
+      "step": 11663
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7609129654678044,
+      "learning_rate": 1.7018538110071773e-05,
+      "loss": 0.9008,
+      "step": 11664
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6224190829646066,
+      "learning_rate": 1.7017994559457878e-05,
+      "loss": 0.8726,
+      "step": 11665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7876525201744174,
+      "learning_rate": 1.7017450967982995e-05,
+      "loss": 0.736,
+      "step": 11666
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.049807599464536,
+      "learning_rate": 1.7016907335650292e-05,
+      "loss": 0.8136,
+      "step": 11667
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7304853624941666,
+      "learning_rate": 1.701636366246293e-05,
+      "loss": 0.7776,
+      "step": 11668
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7446483086575281,
+      "learning_rate": 1.7015819948424085e-05,
+      "loss": 0.781,
+      "step": 11669
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4324661856997896,
+      "learning_rate": 1.7015276193536912e-05,
+      "loss": 0.8586,
+      "step": 11670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6342608289593683,
+      "learning_rate": 1.7014732397804578e-05,
+      "loss": 0.957,
+      "step": 11671
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8038382791835952,
+      "learning_rate": 1.7014188561230252e-05,
+      "loss": 0.8228,
+      "step": 11672
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.631376285460188,
+      "learning_rate": 1.7013644683817098e-05,
+      "loss": 0.8491,
+      "step": 11673
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2892777627329988,
+      "learning_rate": 1.7013100765568284e-05,
+      "loss": 0.8822,
+      "step": 11674
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7797179963283083,
+      "learning_rate": 1.701255680648698e-05,
+      "loss": 0.9279,
+      "step": 11675
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5899804710255008,
+      "learning_rate": 1.701201280657635e-05,
+      "loss": 0.8014,
+      "step": 11676
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.515181330104914,
+      "learning_rate": 1.7011468765839558e-05,
+      "loss": 0.8338,
+      "step": 11677
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8769210734605402,
+      "learning_rate": 1.701092468427978e-05,
+      "loss": 0.8731,
+      "step": 11678
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5210546233900208,
+      "learning_rate": 1.7010380561900174e-05,
+      "loss": 0.8598,
+      "step": 11679
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8414562980301477,
+      "learning_rate": 1.7009836398703913e-05,
+      "loss": 0.9638,
+      "step": 11680
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8092985620626014,
+      "learning_rate": 1.700929219469417e-05,
+      "loss": 0.9442,
+      "step": 11681
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4559006462583273,
+      "learning_rate": 1.7008747949874105e-05,
+      "loss": 0.8112,
+      "step": 11682
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.782665200258611,
+      "learning_rate": 1.7008203664246892e-05,
+      "loss": 0.8602,
+      "step": 11683
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6595352775792258,
+      "learning_rate": 1.70076593378157e-05,
+      "loss": 0.9313,
+      "step": 11684
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7640471853832336,
+      "learning_rate": 1.7007114970583694e-05,
+      "loss": 0.9202,
+      "step": 11685
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7646554565236945,
+      "learning_rate": 1.7006570562554047e-05,
+      "loss": 0.7232,
+      "step": 11686
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.418957675015573,
+      "learning_rate": 1.700602611372993e-05,
+      "loss": 0.7738,
+      "step": 11687
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.696889147488339,
+      "learning_rate": 1.7005481624114506e-05,
+      "loss": 0.9439,
+      "step": 11688
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1784936955777208,
+      "learning_rate": 1.7004937093710953e-05,
+      "loss": 0.913,
+      "step": 11689
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.509921711491408,
+      "learning_rate": 1.700439252252244e-05,
+      "loss": 1.0142,
+      "step": 11690
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.497683348596163,
+      "learning_rate": 1.7003847910552134e-05,
+      "loss": 1.0168,
+      "step": 11691
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9003713408852407,
+      "learning_rate": 1.7003303257803207e-05,
+      "loss": 1.0131,
+      "step": 11692
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9309718406947407,
+      "learning_rate": 1.7002758564278832e-05,
+      "loss": 0.9427,
+      "step": 11693
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6618067608657803,
+      "learning_rate": 1.700221382998218e-05,
+      "loss": 0.9894,
+      "step": 11694
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7333815703423898,
+      "learning_rate": 1.7001669054916424e-05,
+      "loss": 0.72,
+      "step": 11695
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7921731973844167,
+      "learning_rate": 1.700112423908473e-05,
+      "loss": 0.7292,
+      "step": 11696
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7949189711171787,
+      "learning_rate": 1.700057938249028e-05,
+      "loss": 0.7218,
+      "step": 11697
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.637594030094039,
+      "learning_rate": 1.700003448513624e-05,
+      "loss": 0.9529,
+      "step": 11698
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6146381344679348,
+      "learning_rate": 1.6999489547025782e-05,
+      "loss": 0.8221,
+      "step": 11699
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1265070175117717,
+      "learning_rate": 1.6998944568162075e-05,
+      "loss": 0.7473,
+      "step": 11700
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4770120007052623,
+      "learning_rate": 1.69983995485483e-05,
+      "loss": 0.9365,
+      "step": 11701
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5958501398323597,
+      "learning_rate": 1.6997854488187632e-05,
+      "loss": 0.9591,
+      "step": 11702
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7792652323236597,
+      "learning_rate": 1.699730938708323e-05,
+      "loss": 0.8799,
+      "step": 11703
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7738219337217808,
+      "learning_rate": 1.6996764245238282e-05,
+      "loss": 0.7454,
+      "step": 11704
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8379439901132917,
+      "learning_rate": 1.6996219062655963e-05,
+      "loss": 0.995,
+      "step": 11705
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6327014602930239,
+      "learning_rate": 1.6995673839339435e-05,
+      "loss": 1.007,
+      "step": 11706
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4756574415585098,
+      "learning_rate": 1.6995128575291882e-05,
+      "loss": 0.8365,
+      "step": 11707
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4862523880654746,
+      "learning_rate": 1.699458327051647e-05,
+      "loss": 0.9485,
+      "step": 11708
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6226519881145096,
+      "learning_rate": 1.6994037925016386e-05,
+      "loss": 0.918,
+      "step": 11709
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6780875514822837,
+      "learning_rate": 1.69934925387948e-05,
+      "loss": 0.8894,
+      "step": 11710
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1753244508259044,
+      "learning_rate": 1.699294711185488e-05,
+      "loss": 0.9655,
+      "step": 11711
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.825145436906505,
+      "learning_rate": 1.6992401644199815e-05,
+      "loss": 0.8742,
+      "step": 11712
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9445042381057092,
+      "learning_rate": 1.6991856135832767e-05,
+      "loss": 0.9015,
+      "step": 11713
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7924251854418753,
+      "learning_rate": 1.6991310586756923e-05,
+      "loss": 0.9875,
+      "step": 11714
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6227628063580077,
+      "learning_rate": 1.6990764996975456e-05,
+      "loss": 0.8277,
+      "step": 11715
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7806791769394419,
+      "learning_rate": 1.699021936649154e-05,
+      "loss": 0.9141,
+      "step": 11716
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5968452456002689,
+      "learning_rate": 1.6989673695308354e-05,
+      "loss": 0.7863,
+      "step": 11717
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5840141115406237,
+      "learning_rate": 1.6989127983429077e-05,
+      "loss": 0.9857,
+      "step": 11718
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5678172599005444,
+      "learning_rate": 1.698858223085688e-05,
+      "loss": 0.8871,
+      "step": 11719
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8526576316873573,
+      "learning_rate": 1.698803643759495e-05,
+      "loss": 0.7137,
+      "step": 11720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.637294748647991,
+      "learning_rate": 1.698749060364646e-05,
+      "loss": 0.8451,
+      "step": 11721
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.731866886591587,
+      "learning_rate": 1.698694472901458e-05,
+      "loss": 0.7473,
+      "step": 11722
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3983868999648796,
+      "learning_rate": 1.69863988137025e-05,
+      "loss": 0.8635,
+      "step": 11723
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4558722189833153,
+      "learning_rate": 1.6985852857713397e-05,
+      "loss": 0.8253,
+      "step": 11724
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7576222157801167,
+      "learning_rate": 1.6985306861050443e-05,
+      "loss": 1.006,
+      "step": 11725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9660779598774714,
+      "learning_rate": 1.6984760823716818e-05,
+      "loss": 0.9509,
+      "step": 11726
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7970637156327673,
+      "learning_rate": 1.6984214745715712e-05,
+      "loss": 0.8198,
+      "step": 11727
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7881548760555834,
+      "learning_rate": 1.6983668627050292e-05,
+      "loss": 0.9761,
+      "step": 11728
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6867551054122765,
+      "learning_rate": 1.6983122467723743e-05,
+      "loss": 0.9488,
+      "step": 11729
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7191606018790266,
+      "learning_rate": 1.6982576267739247e-05,
+      "loss": 0.927,
+      "step": 11730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7875958019763798,
+      "learning_rate": 1.6982030027099975e-05,
+      "loss": 0.795,
+      "step": 11731
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7296718736665442,
+      "learning_rate": 1.698148374580912e-05,
+      "loss": 0.8928,
+      "step": 11732
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5375130016389997,
+      "learning_rate": 1.6980937423869852e-05,
+      "loss": 0.8215,
+      "step": 11733
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1393124914817405,
+      "learning_rate": 1.698039106128536e-05,
+      "loss": 0.8169,
+      "step": 11734
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9380797319599765,
+      "learning_rate": 1.697984465805882e-05,
+      "loss": 0.8854,
+      "step": 11735
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8388505206932955,
+      "learning_rate": 1.6979298214193412e-05,
+      "loss": 0.8836,
+      "step": 11736
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6811760639807418,
+      "learning_rate": 1.697875172969232e-05,
+      "loss": 0.8724,
+      "step": 11737
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9660451152787946,
+      "learning_rate": 1.6978205204558728e-05,
+      "loss": 0.9012,
+      "step": 11738
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8920793218781566,
+      "learning_rate": 1.6977658638795816e-05,
+      "loss": 0.9624,
+      "step": 11739
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0842437103484395,
+      "learning_rate": 1.697711203240677e-05,
+      "loss": 0.9942,
+      "step": 11740
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3570114695930786,
+      "learning_rate": 1.6976565385394763e-05,
+      "loss": 0.85,
+      "step": 11741
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.920789693201905,
+      "learning_rate": 1.6976018697762985e-05,
+      "loss": 0.9388,
+      "step": 11742
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5316308313593003,
+      "learning_rate": 1.6975471969514617e-05,
+      "loss": 0.7974,
+      "step": 11743
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.718827789379138,
+      "learning_rate": 1.6974925200652846e-05,
+      "loss": 0.7341,
+      "step": 11744
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4977306719569545,
+      "learning_rate": 1.697437839118085e-05,
+      "loss": 0.9405,
+      "step": 11745
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8159744827683976,
+      "learning_rate": 1.6973831541101816e-05,
+      "loss": 0.7433,
+      "step": 11746
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0088752446501315,
+      "learning_rate": 1.6973284650418923e-05,
+      "loss": 0.9953,
+      "step": 11747
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5119425563632165,
+      "learning_rate": 1.6972737719135365e-05,
+      "loss": 0.85,
+      "step": 11748
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5715135571186816,
+      "learning_rate": 1.697219074725432e-05,
+      "loss": 0.8665,
+      "step": 11749
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4815236269118013,
+      "learning_rate": 1.6971643734778967e-05,
+      "loss": 0.9377,
+      "step": 11750
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5718257145913428,
+      "learning_rate": 1.6971096681712502e-05,
+      "loss": 0.9077,
+      "step": 11751
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.663492417437709,
+      "learning_rate": 1.6970549588058106e-05,
+      "loss": 0.8596,
+      "step": 11752
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.894969689316515,
+      "learning_rate": 1.6970002453818963e-05,
+      "loss": 0.8019,
+      "step": 11753
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.015609256832839,
+      "learning_rate": 1.6969455278998254e-05,
+      "loss": 0.952,
+      "step": 11754
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.859769992665121,
+      "learning_rate": 1.6968908063599175e-05,
+      "loss": 0.8647,
+      "step": 11755
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1158886055662505,
+      "learning_rate": 1.6968360807624904e-05,
+      "loss": 0.9043,
+      "step": 11756
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5691938144525777,
+      "learning_rate": 1.6967813511078633e-05,
+      "loss": 0.8319,
+      "step": 11757
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5998307713386943,
+      "learning_rate": 1.6967266173963545e-05,
+      "loss": 0.878,
+      "step": 11758
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8762496109300948,
+      "learning_rate": 1.6966718796282827e-05,
+      "loss": 0.9348,
+      "step": 11759
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.657285825579996,
+      "learning_rate": 1.6966171378039665e-05,
+      "loss": 0.9445,
+      "step": 11760
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6256698285964193,
+      "learning_rate": 1.696562391923725e-05,
+      "loss": 0.7911,
+      "step": 11761
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6594069705795516,
+      "learning_rate": 1.6965076419878764e-05,
+      "loss": 0.7245,
+      "step": 11762
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6452219910103254,
+      "learning_rate": 1.69645288799674e-05,
+      "loss": 0.7426,
+      "step": 11763
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4482176665122757,
+      "learning_rate": 1.6963981299506348e-05,
+      "loss": 1.0772,
+      "step": 11764
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6794013118831608,
+      "learning_rate": 1.6963433678498786e-05,
+      "loss": 0.9452,
+      "step": 11765
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5959217811670694,
+      "learning_rate": 1.696288601694791e-05,
+      "loss": 0.8746,
+      "step": 11766
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7520254331043675,
+      "learning_rate": 1.696233831485691e-05,
+      "loss": 0.7663,
+      "step": 11767
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7649138751787847,
+      "learning_rate": 1.6961790572228967e-05,
+      "loss": 0.7404,
+      "step": 11768
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.361475186352525,
+      "learning_rate": 1.6961242789067282e-05,
+      "loss": 0.8849,
+      "step": 11769
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7978488312784521,
+      "learning_rate": 1.6960694965375033e-05,
+      "loss": 0.7454,
+      "step": 11770
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5230815292751594,
+      "learning_rate": 1.6960147101155415e-05,
+      "loss": 0.8911,
+      "step": 11771
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5663193372721818,
+      "learning_rate": 1.695959919641162e-05,
+      "loss": 0.8711,
+      "step": 11772
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7516059369679647,
+      "learning_rate": 1.695905125114683e-05,
+      "loss": 0.7401,
+      "step": 11773
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.203737670691406,
+      "learning_rate": 1.6958503265364243e-05,
+      "loss": 0.9517,
+      "step": 11774
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8725568362902933,
+      "learning_rate": 1.6957955239067046e-05,
+      "loss": 0.8256,
+      "step": 11775
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.331811401320834,
+      "learning_rate": 1.695740717225843e-05,
+      "loss": 0.9631,
+      "step": 11776
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8352988827339414,
+      "learning_rate": 1.695685906494159e-05,
+      "loss": 0.8665,
+      "step": 11777
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.92067499120524,
+      "learning_rate": 1.6956310917119712e-05,
+      "loss": 0.8564,
+      "step": 11778
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5772629306927997,
+      "learning_rate": 1.695576272879599e-05,
+      "loss": 0.8836,
+      "step": 11779
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8863951795883753,
+      "learning_rate": 1.6955214499973616e-05,
+      "loss": 0.9443,
+      "step": 11780
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7923073591826852,
+      "learning_rate": 1.695466623065578e-05,
+      "loss": 0.864,
+      "step": 11781
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5276146631926337,
+      "learning_rate": 1.6954117920845672e-05,
+      "loss": 1.0153,
+      "step": 11782
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9553100855501913,
+      "learning_rate": 1.6953569570546494e-05,
+      "loss": 0.9385,
+      "step": 11783
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.992373326694736,
+      "learning_rate": 1.695302117976143e-05,
+      "loss": 0.738,
+      "step": 11784
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.662204892113177,
+      "learning_rate": 1.6952472748493675e-05,
+      "loss": 0.8323,
+      "step": 11785
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7729955686266823,
+      "learning_rate": 1.6951924276746425e-05,
+      "loss": 0.9231,
+      "step": 11786
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6278784895588203,
+      "learning_rate": 1.6951375764522868e-05,
+      "loss": 0.7964,
+      "step": 11787
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8100300158386906,
+      "learning_rate": 1.69508272118262e-05,
+      "loss": 0.8086,
+      "step": 11788
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7209612095208804,
+      "learning_rate": 1.695027861865962e-05,
+      "loss": 0.9272,
+      "step": 11789
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6716960909854597,
+      "learning_rate": 1.6949729985026314e-05,
+      "loss": 0.9514,
+      "step": 11790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8073449650469792,
+      "learning_rate": 1.6949181310929476e-05,
+      "loss": 0.7331,
+      "step": 11791
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5737813289987297,
+      "learning_rate": 1.694863259637231e-05,
+      "loss": 0.7299,
+      "step": 11792
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7227959523345764,
+      "learning_rate": 1.6948083841358004e-05,
+      "loss": 0.8751,
+      "step": 11793
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.778925222319704,
+      "learning_rate": 1.6947535045889754e-05,
+      "loss": 0.7808,
+      "step": 11794
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5387906553065946,
+      "learning_rate": 1.6946986209970754e-05,
+      "loss": 0.9634,
+      "step": 11795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.42739676335661,
+      "learning_rate": 1.6946437333604203e-05,
+      "loss": 0.8505,
+      "step": 11796
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5377896982589299,
+      "learning_rate": 1.6945888416793296e-05,
+      "loss": 0.7545,
+      "step": 11797
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.937208008100082,
+      "learning_rate": 1.694533945954122e-05,
+      "loss": 0.8662,
+      "step": 11798
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5969193434911093,
+      "learning_rate": 1.6944790461851186e-05,
+      "loss": 0.8208,
+      "step": 11799
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.557672956045301,
+      "learning_rate": 1.694424142372638e-05,
+      "loss": 0.8947,
+      "step": 11800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8033866422701393,
+      "learning_rate": 1.6943692345170005e-05,
+      "loss": 0.8249,
+      "step": 11801
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8263358348071398,
+      "learning_rate": 1.6943143226185252e-05,
+      "loss": 0.786,
+      "step": 11802
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8953041607863408,
+      "learning_rate": 1.694259406677532e-05,
+      "loss": 0.9065,
+      "step": 11803
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3051411437092924,
+      "learning_rate": 1.6942044866943414e-05,
+      "loss": 0.8131,
+      "step": 11804
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.620608626553952,
+      "learning_rate": 1.6941495626692718e-05,
+      "loss": 0.9349,
+      "step": 11805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9348846706049638,
+      "learning_rate": 1.694094634602644e-05,
+      "loss": 0.8505,
+      "step": 11806
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5408631131524857,
+      "learning_rate": 1.6940397024947773e-05,
+      "loss": 0.9978,
+      "step": 11807
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.784639088443313,
+      "learning_rate": 1.693984766345992e-05,
+      "loss": 0.9993,
+      "step": 11808
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7036370473130855,
+      "learning_rate": 1.6939298261566075e-05,
+      "loss": 0.8568,
+      "step": 11809
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.126256863324569,
+      "learning_rate": 1.6938748819269436e-05,
+      "loss": 0.8982,
+      "step": 11810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4191214542834882,
+      "learning_rate": 1.6938199336573208e-05,
+      "loss": 0.864,
+      "step": 11811
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4879360113779163,
+      "learning_rate": 1.6937649813480586e-05,
+      "loss": 1.0191,
+      "step": 11812
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9166313017276815,
+      "learning_rate": 1.6937100249994772e-05,
+      "loss": 0.9131,
+      "step": 11813
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.913633239974259,
+      "learning_rate": 1.693655064611896e-05,
+      "loss": 0.8629,
+      "step": 11814
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7879326747756474,
+      "learning_rate": 1.693600100185636e-05,
+      "loss": 0.8888,
+      "step": 11815
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7474508681704937,
+      "learning_rate": 1.693545131721016e-05,
+      "loss": 0.7115,
+      "step": 11816
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.215623540345661,
+      "learning_rate": 1.693490159218357e-05,
+      "loss": 0.9121,
+      "step": 11817
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5870393781149805,
+      "learning_rate": 1.6934351826779787e-05,
+      "loss": 0.8422,
+      "step": 11818
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1352435911458505,
+      "learning_rate": 1.6933802021002012e-05,
+      "loss": 1.031,
+      "step": 11819
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7746737343655556,
+      "learning_rate": 1.6933252174853446e-05,
+      "loss": 0.7531,
+      "step": 11820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.82900029980677,
+      "learning_rate": 1.6932702288337293e-05,
+      "loss": 0.8398,
+      "step": 11821
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.285302844348197,
+      "learning_rate": 1.693215236145675e-05,
+      "loss": 0.965,
+      "step": 11822
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6793435402124128,
+      "learning_rate": 1.693160239421502e-05,
+      "loss": 0.7836,
+      "step": 11823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.962294659532668,
+      "learning_rate": 1.693105238661531e-05,
+      "loss": 0.8887,
+      "step": 11824
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7390143972909352,
+      "learning_rate": 1.6930502338660817e-05,
+      "loss": 0.7286,
+      "step": 11825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4767632944742919,
+      "learning_rate": 1.6929952250354747e-05,
+      "loss": 0.9042,
+      "step": 11826
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7995932215828256,
+      "learning_rate": 1.6929402121700298e-05,
+      "loss": 0.7712,
+      "step": 11827
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9225512221766123,
+      "learning_rate": 1.692885195270068e-05,
+      "loss": 0.9867,
+      "step": 11828
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8461414705026633,
+      "learning_rate": 1.6928301743359087e-05,
+      "loss": 0.8562,
+      "step": 11829
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8984031588902992,
+      "learning_rate": 1.692775149367873e-05,
+      "loss": 0.9673,
+      "step": 11830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8151307361654732,
+      "learning_rate": 1.692720120366281e-05,
+      "loss": 0.8548,
+      "step": 11831
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0379092175747178,
+      "learning_rate": 1.6926650873314535e-05,
+      "loss": 0.8701,
+      "step": 11832
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9928794227250706,
+      "learning_rate": 1.69261005026371e-05,
+      "loss": 0.8055,
+      "step": 11833
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3158127865575935,
+      "learning_rate": 1.692555009163372e-05,
+      "loss": 0.835,
+      "step": 11834
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6250793223359496,
+      "learning_rate": 1.6924999640307596e-05,
+      "loss": 0.8565,
+      "step": 11835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6887492303546836,
+      "learning_rate": 1.6924449148661927e-05,
+      "loss": 0.9517,
+      "step": 11836
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.026678408944157,
+      "learning_rate": 1.6923898616699925e-05,
+      "loss": 0.9666,
+      "step": 11837
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.418059722940006,
+      "learning_rate": 1.6923348044424793e-05,
+      "loss": 0.8993,
+      "step": 11838
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8557674629915655,
+      "learning_rate": 1.6922797431839736e-05,
+      "loss": 0.8766,
+      "step": 11839
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0984299082168905,
+      "learning_rate": 1.692224677894796e-05,
+      "loss": 1.0187,
+      "step": 11840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6726636547756737,
+      "learning_rate": 1.6921696085752674e-05,
+      "loss": 0.8536,
+      "step": 11841
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.623338146697342,
+      "learning_rate": 1.692114535225708e-05,
+      "loss": 0.7335,
+      "step": 11842
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9625473682230714,
+      "learning_rate": 1.6920594578464386e-05,
+      "loss": 0.9151,
+      "step": 11843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6515993773020263,
+      "learning_rate": 1.6920043764377804e-05,
+      "loss": 1.042,
+      "step": 11844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6004110040001764,
+      "learning_rate": 1.6919492910000534e-05,
+      "loss": 0.8001,
+      "step": 11845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.628485278464238,
+      "learning_rate": 1.6918942015335785e-05,
+      "loss": 0.9143,
+      "step": 11846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1998805640891113,
+      "learning_rate": 1.6918391080386763e-05,
+      "loss": 0.9322,
+      "step": 11847
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6515090717243937,
+      "learning_rate": 1.691784010515668e-05,
+      "loss": 0.8631,
+      "step": 11848
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7799180562099924,
+      "learning_rate": 1.6917289089648742e-05,
+      "loss": 0.7206,
+      "step": 11849
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9186087638840819,
+      "learning_rate": 1.6916738033866153e-05,
+      "loss": 0.8532,
+      "step": 11850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7643630838437012,
+      "learning_rate": 1.6916186937812132e-05,
+      "loss": 0.9157,
+      "step": 11851
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.604320214113011,
+      "learning_rate": 1.6915635801489876e-05,
+      "loss": 0.8937,
+      "step": 11852
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9297402876067704,
+      "learning_rate": 1.6915084624902602e-05,
+      "loss": 0.8402,
+      "step": 11853
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7487951657013134,
+      "learning_rate": 1.6914533408053513e-05,
+      "loss": 0.8851,
+      "step": 11854
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.12662363746002,
+      "learning_rate": 1.6913982150945823e-05,
+      "loss": 0.7553,
+      "step": 11855
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5946411879673514,
+      "learning_rate": 1.691343085358274e-05,
+      "loss": 0.9844,
+      "step": 11856
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5953049442899454,
+      "learning_rate": 1.691287951596747e-05,
+      "loss": 0.9184,
+      "step": 11857
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8332270345945123,
+      "learning_rate": 1.6912328138103237e-05,
+      "loss": 0.977,
+      "step": 11858
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.404762524844973,
+      "learning_rate": 1.6911776719993232e-05,
+      "loss": 0.9572,
+      "step": 11859
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6507398223578482,
+      "learning_rate": 1.6911225261640678e-05,
+      "loss": 1.0288,
+      "step": 11860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8749000078257825,
+      "learning_rate": 1.691067376304878e-05,
+      "loss": 0.9495,
+      "step": 11861
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.788358974089642,
+      "learning_rate": 1.6910122224220757e-05,
+      "loss": 0.7143,
+      "step": 11862
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6734895753108283,
+      "learning_rate": 1.6909570645159808e-05,
+      "loss": 0.8636,
+      "step": 11863
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8415755274199193,
+      "learning_rate": 1.6909019025869158e-05,
+      "loss": 0.9172,
+      "step": 11864
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7010902137357211,
+      "learning_rate": 1.6908467366352004e-05,
+      "loss": 0.992,
+      "step": 11865
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.134548792767442,
+      "learning_rate": 1.690791566661157e-05,
+      "loss": 0.9045,
+      "step": 11866
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.874180628995273,
+      "learning_rate": 1.690736392665107e-05,
+      "loss": 0.8631,
+      "step": 11867
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0231819772221793,
+      "learning_rate": 1.6906812146473703e-05,
+      "loss": 0.8171,
+      "step": 11868
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6660341292395662,
+      "learning_rate": 1.6906260326082692e-05,
+      "loss": 0.8527,
+      "step": 11869
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4255569638378933,
+      "learning_rate": 1.6905708465481242e-05,
+      "loss": 0.9183,
+      "step": 11870
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.527830793648295,
+      "learning_rate": 1.6905156564672575e-05,
+      "loss": 0.7614,
+      "step": 11871
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7411145104939842,
+      "learning_rate": 1.6904604623659898e-05,
+      "loss": 0.9314,
+      "step": 11872
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6760440794812015,
+      "learning_rate": 1.690405264244643e-05,
+      "loss": 0.9693,
+      "step": 11873
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.698532523195556,
+      "learning_rate": 1.690350062103538e-05,
+      "loss": 0.9242,
+      "step": 11874
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4881781953789393,
+      "learning_rate": 1.6902948559429962e-05,
+      "loss": 0.7851,
+      "step": 11875
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6988388438411828,
+      "learning_rate": 1.6902396457633397e-05,
+      "loss": 0.9068,
+      "step": 11876
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.655128401071479,
+      "learning_rate": 1.6901844315648893e-05,
+      "loss": 1.05,
+      "step": 11877
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8847777338917546,
+      "learning_rate": 1.690129213347966e-05,
+      "loss": 0.9092,
+      "step": 11878
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.755126190692796,
+      "learning_rate": 1.690073991112893e-05,
+      "loss": 0.8385,
+      "step": 11879
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6102744413155885,
+      "learning_rate": 1.69001876485999e-05,
+      "loss": 0.9025,
+      "step": 11880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8720092430474187,
+      "learning_rate": 1.6899635345895796e-05,
+      "loss": 0.8161,
+      "step": 11881
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7459203412934068,
+      "learning_rate": 1.689908300301983e-05,
+      "loss": 0.9023,
+      "step": 11882
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6030273731256668,
+      "learning_rate": 1.6898530619975218e-05,
+      "loss": 0.7709,
+      "step": 11883
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9247385348240937,
+      "learning_rate": 1.6897978196765174e-05,
+      "loss": 0.7821,
+      "step": 11884
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6024038216091834,
+      "learning_rate": 1.689742573339292e-05,
+      "loss": 0.8203,
+      "step": 11885
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8617433637097933,
+      "learning_rate": 1.689687322986167e-05,
+      "loss": 0.972,
+      "step": 11886
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5104518365473514,
+      "learning_rate": 1.689632068617464e-05,
+      "loss": 0.9517,
+      "step": 11887
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7456215036095901,
+      "learning_rate": 1.6895768102335045e-05,
+      "loss": 0.9851,
+      "step": 11888
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8262608480644833,
+      "learning_rate": 1.689521547834611e-05,
+      "loss": 1.1219,
+      "step": 11889
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6277196512995065,
+      "learning_rate": 1.6894662814211046e-05,
+      "loss": 0.9363,
+      "step": 11890
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8132194822854347,
+      "learning_rate": 1.689411010993307e-05,
+      "loss": 0.7297,
+      "step": 11891
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6040367795349533,
+      "learning_rate": 1.6893557365515404e-05,
+      "loss": 1.0483,
+      "step": 11892
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8824627182278642,
+      "learning_rate": 1.6893004580961264e-05,
+      "loss": 0.7831,
+      "step": 11893
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6361125951718978,
+      "learning_rate": 1.689245175627387e-05,
+      "loss": 0.7656,
+      "step": 11894
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.551665810209733,
+      "learning_rate": 1.689189889145644e-05,
+      "loss": 0.8508,
+      "step": 11895
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7778888714625977,
+      "learning_rate": 1.689134598651219e-05,
+      "loss": 0.7351,
+      "step": 11896
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7923144933205744,
+      "learning_rate": 1.6890793041444345e-05,
+      "loss": 0.7273,
+      "step": 11897
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7208415879197188,
+      "learning_rate": 1.689024005625612e-05,
+      "loss": 0.7496,
+      "step": 11898
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5518025733672036,
+      "learning_rate": 1.6889687030950736e-05,
+      "loss": 0.9027,
+      "step": 11899
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.871029523461936,
+      "learning_rate": 1.688913396553141e-05,
+      "loss": 0.8209,
+      "step": 11900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9199037423689178,
+      "learning_rate": 1.688858086000137e-05,
+      "loss": 0.9341,
+      "step": 11901
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8274445056385737,
+      "learning_rate": 1.688802771436383e-05,
+      "loss": 0.9114,
+      "step": 11902
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5831373592735207,
+      "learning_rate": 1.688747452862201e-05,
+      "loss": 0.7782,
+      "step": 11903
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6737770546030644,
+      "learning_rate": 1.6886921302779134e-05,
+      "loss": 0.9152,
+      "step": 11904
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6156489442117086,
+      "learning_rate": 1.6886368036838422e-05,
+      "loss": 0.9294,
+      "step": 11905
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6866337738228052,
+      "learning_rate": 1.6885814730803095e-05,
+      "loss": 0.8749,
+      "step": 11906
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.791577152866485,
+      "learning_rate": 1.6885261384676373e-05,
+      "loss": 0.8781,
+      "step": 11907
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8601860409643,
+      "learning_rate": 1.688470799846148e-05,
+      "loss": 0.9045,
+      "step": 11908
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5081979585173522,
+      "learning_rate": 1.6884154572161636e-05,
+      "loss": 0.7655,
+      "step": 11909
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.515456510828473,
+      "learning_rate": 1.6883601105780068e-05,
+      "loss": 0.8049,
+      "step": 11910
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7779449998902882,
+      "learning_rate": 1.6883047599319994e-05,
+      "loss": 0.7024,
+      "step": 11911
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7694748496863049,
+      "learning_rate": 1.6882494052784637e-05,
+      "loss": 0.8589,
+      "step": 11912
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5646180848542084,
+      "learning_rate": 1.688194046617722e-05,
+      "loss": 0.942,
+      "step": 11913
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5456179215144736,
+      "learning_rate": 1.688138683950097e-05,
+      "loss": 0.8851,
+      "step": 11914
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8629385273547945,
+      "learning_rate": 1.6880833172759105e-05,
+      "loss": 1.0411,
+      "step": 11915
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8696924277041789,
+      "learning_rate": 1.688027946595485e-05,
+      "loss": 0.9068,
+      "step": 11916
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8093850193856522,
+      "learning_rate": 1.687972571909143e-05,
+      "loss": 0.9642,
+      "step": 11917
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7119229560475238,
+      "learning_rate": 1.6879171932172074e-05,
+      "loss": 0.9055,
+      "step": 11918
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5237004523498159,
+      "learning_rate": 1.6878618105199995e-05,
+      "loss": 0.8481,
+      "step": 11919
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.178830112405341,
+      "learning_rate": 1.6878064238178423e-05,
+      "loss": 0.9361,
+      "step": 11920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6751924325889995,
+      "learning_rate": 1.6877510331110586e-05,
+      "loss": 0.8742,
+      "step": 11921
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5174598840521414,
+      "learning_rate": 1.6876956383999707e-05,
+      "loss": 0.8193,
+      "step": 11922
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5927471932600201,
+      "learning_rate": 1.687640239684901e-05,
+      "loss": 0.9493,
+      "step": 11923
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6320087710091213,
+      "learning_rate": 1.6875848369661722e-05,
+      "loss": 0.846,
+      "step": 11924
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0236681971838073,
+      "learning_rate": 1.6875294302441067e-05,
+      "loss": 0.824,
+      "step": 11925
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8059948510508657,
+      "learning_rate": 1.687474019519027e-05,
+      "loss": 0.7554,
+      "step": 11926
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.612880332176102,
+      "learning_rate": 1.6874186047912565e-05,
+      "loss": 1.01,
+      "step": 11927
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.427346443889244,
+      "learning_rate": 1.6873631860611165e-05,
+      "loss": 0.9605,
+      "step": 11928
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5882196352768796,
+      "learning_rate": 1.687307763328931e-05,
+      "loss": 0.9055,
+      "step": 11929
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6243369311505065,
+      "learning_rate": 1.6872523365950218e-05,
+      "loss": 0.894,
+      "step": 11930
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6417551179628413,
+      "learning_rate": 1.6871969058597118e-05,
+      "loss": 0.8638,
+      "step": 11931
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4852237339179353,
+      "learning_rate": 1.687141471123324e-05,
+      "loss": 1.021,
+      "step": 11932
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5135975500094294,
+      "learning_rate": 1.6870860323861813e-05,
+      "loss": 0.7815,
+      "step": 11933
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7869269301425696,
+      "learning_rate": 1.6870305896486058e-05,
+      "loss": 0.6953,
+      "step": 11934
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7118126221475751,
+      "learning_rate": 1.686975142910921e-05,
+      "loss": 0.9083,
+      "step": 11935
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7131249495944432,
+      "learning_rate": 1.686919692173449e-05,
+      "loss": 0.7661,
+      "step": 11936
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.890865598370462,
+      "learning_rate": 1.6868642374365135e-05,
+      "loss": 0.9425,
+      "step": 11937
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5990574060745164,
+      "learning_rate": 1.686808778700437e-05,
+      "loss": 0.8791,
+      "step": 11938
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.52940346357347,
+      "learning_rate": 1.686753315965542e-05,
+      "loss": 0.8707,
+      "step": 11939
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4877697772568135,
+      "learning_rate": 1.686697849232152e-05,
+      "loss": 0.8304,
+      "step": 11940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8246345157903339,
+      "learning_rate": 1.6866423785005895e-05,
+      "loss": 0.9389,
+      "step": 11941
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.626327601834971,
+      "learning_rate": 1.6865869037711776e-05,
+      "loss": 0.8803,
+      "step": 11942
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.680680569286317,
+      "learning_rate": 1.68653142504424e-05,
+      "loss": 0.938,
+      "step": 11943
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8900692370960732,
+      "learning_rate": 1.6864759423200984e-05,
+      "loss": 0.8942,
+      "step": 11944
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1387147140621563,
+      "learning_rate": 1.6864204555990768e-05,
+      "loss": 0.9377,
+      "step": 11945
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.505174174538712,
+      "learning_rate": 1.6863649648814982e-05,
+      "loss": 0.996,
+      "step": 11946
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9097670416881334,
+      "learning_rate": 1.686309470167685e-05,
+      "loss": 0.8916,
+      "step": 11947
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7915491469037939,
+      "learning_rate": 1.686253971457961e-05,
+      "loss": 0.8375,
+      "step": 11948
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6473067173076699,
+      "learning_rate": 1.6861984687526494e-05,
+      "loss": 0.9209,
+      "step": 11949
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6920297350709872,
+      "learning_rate": 1.6861429620520725e-05,
+      "loss": 1.0455,
+      "step": 11950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.76326928415404,
+      "learning_rate": 1.6860874513565544e-05,
+      "loss": 0.6994,
+      "step": 11951
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2167199354116653,
+      "learning_rate": 1.6860319366664182e-05,
+      "loss": 0.9548,
+      "step": 11952
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7533962146293378,
+      "learning_rate": 1.6859764179819866e-05,
+      "loss": 0.9289,
+      "step": 11953
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5844255153097195,
+      "learning_rate": 1.6859208953035832e-05,
+      "loss": 0.9392,
+      "step": 11954
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0442228668185543,
+      "learning_rate": 1.685865368631531e-05,
+      "loss": 0.8882,
+      "step": 11955
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.070634698509685,
+      "learning_rate": 1.6858098379661538e-05,
+      "loss": 0.9656,
+      "step": 11956
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5455420772049355,
+      "learning_rate": 1.685754303307774e-05,
+      "loss": 0.8801,
+      "step": 11957
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7402729737217782,
+      "learning_rate": 1.6856987646567163e-05,
+      "loss": 0.7426,
+      "step": 11958
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7954610396272822,
+      "learning_rate": 1.685643222013303e-05,
+      "loss": 0.7029,
+      "step": 11959
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7757253189407964,
+      "learning_rate": 1.6855876753778574e-05,
+      "loss": 0.9252,
+      "step": 11960
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.893315054471856,
+      "learning_rate": 1.6855321247507034e-05,
+      "loss": 0.9018,
+      "step": 11961
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7831425662719714,
+      "learning_rate": 1.6854765701321648e-05,
+      "loss": 0.6974,
+      "step": 11962
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.536807499597227,
+      "learning_rate": 1.6854210115225642e-05,
+      "loss": 0.9197,
+      "step": 11963
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6305531725772264,
+      "learning_rate": 1.6853654489222258e-05,
+      "loss": 0.8365,
+      "step": 11964
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7062035776875484,
+      "learning_rate": 1.6853098823314728e-05,
+      "loss": 0.9447,
+      "step": 11965
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.535571008686289,
+      "learning_rate": 1.685254311750628e-05,
+      "loss": 1.0688,
+      "step": 11966
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5788159770519228,
+      "learning_rate": 1.6851987371800164e-05,
+      "loss": 0.9078,
+      "step": 11967
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1274157156174507,
+      "learning_rate": 1.6851431586199604e-05,
+      "loss": 0.9108,
+      "step": 11968
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.866340581707766,
+      "learning_rate": 1.6850875760707842e-05,
+      "loss": 0.836,
+      "step": 11969
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.105344429618733,
+      "learning_rate": 1.6850319895328112e-05,
+      "loss": 0.8866,
+      "step": 11970
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9657941433995334,
+      "learning_rate": 1.6849763990063654e-05,
+      "loss": 0.823,
+      "step": 11971
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5267912136862072,
+      "learning_rate": 1.6849208044917696e-05,
+      "loss": 0.7875,
+      "step": 11972
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0609907217721593,
+      "learning_rate": 1.6848652059893485e-05,
+      "loss": 0.9281,
+      "step": 11973
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9292368859833333,
+      "learning_rate": 1.6848096034994253e-05,
+      "loss": 0.7822,
+      "step": 11974
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6565847050571256,
+      "learning_rate": 1.684753997022323e-05,
+      "loss": 0.8968,
+      "step": 11975
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4260049889555932,
+      "learning_rate": 1.6846983865583674e-05,
+      "loss": 0.7829,
+      "step": 11976
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.818365802229671,
+      "learning_rate": 1.6846427721078798e-05,
+      "loss": 0.7833,
+      "step": 11977
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9245493009586625,
+      "learning_rate": 1.684587153671186e-05,
+      "loss": 0.9453,
+      "step": 11978
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8368783545027416,
+      "learning_rate": 1.6845315312486088e-05,
+      "loss": 0.9554,
+      "step": 11979
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.609424948604408,
+      "learning_rate": 1.6844759048404724e-05,
+      "loss": 0.9869,
+      "step": 11980
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7866375996969315,
+      "learning_rate": 1.6844202744471006e-05,
+      "loss": 0.9116,
+      "step": 11981
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6722354443620058,
+      "learning_rate": 1.684364640068817e-05,
+      "loss": 0.8513,
+      "step": 11982
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4986188382909524,
+      "learning_rate": 1.684309001705946e-05,
+      "loss": 0.8861,
+      "step": 11983
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5004775006260447,
+      "learning_rate": 1.6842533593588116e-05,
+      "loss": 0.9398,
+      "step": 11984
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.558206720956472,
+      "learning_rate": 1.684197713027737e-05,
+      "loss": 0.9898,
+      "step": 11985
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.096053578522094,
+      "learning_rate": 1.684142062713047e-05,
+      "loss": 0.9228,
+      "step": 11986
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9905456899423644,
+      "learning_rate": 1.6840864084150658e-05,
+      "loss": 0.9454,
+      "step": 11987
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0406175796783166,
+      "learning_rate": 1.684030750134116e-05,
+      "loss": 0.9105,
+      "step": 11988
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8094813194994734,
+      "learning_rate": 1.683975087870523e-05,
+      "loss": 0.9252,
+      "step": 11989
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.49765538640789,
+      "learning_rate": 1.683919421624611e-05,
+      "loss": 0.8326,
+      "step": 11990
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.855900640479037,
+      "learning_rate": 1.683863751396703e-05,
+      "loss": 0.9225,
+      "step": 11991
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6281656358575407,
+      "learning_rate": 1.683808077187124e-05,
+      "loss": 0.9282,
+      "step": 11992
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7998804997126636,
+      "learning_rate": 1.6837523989961973e-05,
+      "loss": 0.8072,
+      "step": 11993
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.67364191416572,
+      "learning_rate": 1.6836967168242483e-05,
+      "loss": 0.928,
+      "step": 11994
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7854462943324119,
+      "learning_rate": 1.6836410306716005e-05,
+      "loss": 0.9554,
+      "step": 11995
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6829020417295673,
+      "learning_rate": 1.6835853405385778e-05,
+      "loss": 0.9988,
+      "step": 11996
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.83268277602567,
+      "learning_rate": 1.6835296464255052e-05,
+      "loss": 0.9173,
+      "step": 11997
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5678082296557043,
+      "learning_rate": 1.6834739483327066e-05,
+      "loss": 1.0384,
+      "step": 11998
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.104277962191381,
+      "learning_rate": 1.6834182462605062e-05,
+      "loss": 0.8246,
+      "step": 11999
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7602686888898487,
+      "learning_rate": 1.6833625402092286e-05,
+      "loss": 0.8645,
+      "step": 12000
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6041213267131018,
+      "learning_rate": 1.6833068301791977e-05,
+      "loss": 0.9113,
+      "step": 12001
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5987947473327595,
+      "learning_rate": 1.683251116170738e-05,
+      "loss": 0.9146,
+      "step": 12002
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7538991054384148,
+      "learning_rate": 1.683195398184174e-05,
+      "loss": 0.7696,
+      "step": 12003
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.598615945971454,
+      "learning_rate": 1.6831396762198303e-05,
+      "loss": 0.8994,
+      "step": 12004
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6312025312528575,
+      "learning_rate": 1.683083950278031e-05,
+      "loss": 0.9173,
+      "step": 12005
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1469722069445285,
+      "learning_rate": 1.6830282203591007e-05,
+      "loss": 0.8982,
+      "step": 12006
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6194089528708333,
+      "learning_rate": 1.682972486463364e-05,
+      "loss": 0.8182,
+      "step": 12007
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6205936556553933,
+      "learning_rate": 1.682916748591145e-05,
+      "loss": 0.9215,
+      "step": 12008
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9300994080498424,
+      "learning_rate": 1.6828610067427686e-05,
+      "loss": 0.9709,
+      "step": 12009
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6038221956885672,
+      "learning_rate": 1.6828052609185592e-05,
+      "loss": 0.884,
+      "step": 12010
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8997622918148926,
+      "learning_rate": 1.6827495111188413e-05,
+      "loss": 0.8515,
+      "step": 12011
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1720136531035363,
+      "learning_rate": 1.6826937573439395e-05,
+      "loss": 0.8107,
+      "step": 12012
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6670433151228083,
+      "learning_rate": 1.682637999594179e-05,
+      "loss": 0.9115,
+      "step": 12013
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.781347145843065,
+      "learning_rate": 1.6825822378698836e-05,
+      "loss": 0.754,
+      "step": 12014
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7219968902578384,
+      "learning_rate": 1.6825264721713787e-05,
+      "loss": 0.9366,
+      "step": 12015
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6875884437487558,
+      "learning_rate": 1.6824707024989885e-05,
+      "loss": 0.9451,
+      "step": 12016
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5995693316035462,
+      "learning_rate": 1.6824149288530374e-05,
+      "loss": 0.8895,
+      "step": 12017
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.149500730875499,
+      "learning_rate": 1.6823591512338508e-05,
+      "loss": 0.9348,
+      "step": 12018
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.760873645347549,
+      "learning_rate": 1.682303369641753e-05,
+      "loss": 0.8645,
+      "step": 12019
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8905186844145445,
+      "learning_rate": 1.6822475840770692e-05,
+      "loss": 0.907,
+      "step": 12020
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5571656158862432,
+      "learning_rate": 1.6821917945401237e-05,
+      "loss": 0.8709,
+      "step": 12021
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.260418202020794,
+      "learning_rate": 1.6821360010312418e-05,
+      "loss": 0.9313,
+      "step": 12022
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4934270165979413,
+      "learning_rate": 1.682080203550748e-05,
+      "loss": 0.8573,
+      "step": 12023
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.034536272609702,
+      "learning_rate": 1.6820244020989672e-05,
+      "loss": 0.9021,
+      "step": 12024
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0570860638442308,
+      "learning_rate": 1.681968596676225e-05,
+      "loss": 0.9167,
+      "step": 12025
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.83266942293986,
+      "learning_rate": 1.681912787282845e-05,
+      "loss": 0.9359,
+      "step": 12026
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.607199162803655,
+      "learning_rate": 1.6818569739191535e-05,
+      "loss": 0.8248,
+      "step": 12027
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.940844247380595,
+      "learning_rate": 1.6818011565854744e-05,
+      "loss": 0.9108,
+      "step": 12028
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6376121189011217,
+      "learning_rate": 1.6817453352821333e-05,
+      "loss": 0.9435,
+      "step": 12029
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.691841416330259,
+      "learning_rate": 1.681689510009455e-05,
+      "loss": 0.9331,
+      "step": 12030
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6628542023659327,
+      "learning_rate": 1.6816336807677644e-05,
+      "loss": 0.8653,
+      "step": 12031
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.294024461941376,
+      "learning_rate": 1.681577847557387e-05,
+      "loss": 0.7179,
+      "step": 12032
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9792058539684019,
+      "learning_rate": 1.6815220103786472e-05,
+      "loss": 0.9435,
+      "step": 12033
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4942126219558383,
+      "learning_rate": 1.6814661692318707e-05,
+      "loss": 0.9001,
+      "step": 12034
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8648791980329587,
+      "learning_rate": 1.6814103241173825e-05,
+      "loss": 0.8529,
+      "step": 12035
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8701655761344942,
+      "learning_rate": 1.6813544750355074e-05,
+      "loss": 0.9545,
+      "step": 12036
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7721434012550688,
+      "learning_rate": 1.6812986219865712e-05,
+      "loss": 0.8937,
+      "step": 12037
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6869979922361382,
+      "learning_rate": 1.6812427649708984e-05,
+      "loss": 1.0893,
+      "step": 12038
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.23489195066584,
+      "learning_rate": 1.681186903988815e-05,
+      "loss": 0.9954,
+      "step": 12039
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.861121565290976,
+      "learning_rate": 1.6811310390406454e-05,
+      "loss": 0.7943,
+      "step": 12040
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5769377507551614,
+      "learning_rate": 1.681075170126715e-05,
+      "loss": 0.8206,
+      "step": 12041
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3856169391168405,
+      "learning_rate": 1.6810192972473497e-05,
+      "loss": 0.8743,
+      "step": 12042
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9042744798290965,
+      "learning_rate": 1.6809634204028745e-05,
+      "loss": 0.8607,
+      "step": 12043
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.898374813530906,
+      "learning_rate": 1.6809075395936142e-05,
+      "loss": 0.7559,
+      "step": 12044
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6287797757903792,
+      "learning_rate": 1.6808516548198955e-05,
+      "loss": 1.0029,
+      "step": 12045
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5626154270229073,
+      "learning_rate": 1.6807957660820423e-05,
+      "loss": 0.9255,
+      "step": 12046
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.553058150223732,
+      "learning_rate": 1.6807398733803804e-05,
+      "loss": 0.9342,
+      "step": 12047
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.246229687656165,
+      "learning_rate": 1.680683976715236e-05,
+      "loss": 0.8822,
+      "step": 12048
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.567625828461718,
+      "learning_rate": 1.6806280760869338e-05,
+      "loss": 0.9194,
+      "step": 12049
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8442038817741306,
+      "learning_rate": 1.6805721714957995e-05,
+      "loss": 0.8615,
+      "step": 12050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4476134548520896,
+      "learning_rate": 1.6805162629421583e-05,
+      "loss": 1.0393,
+      "step": 12051
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5180204249317317,
+      "learning_rate": 1.6804603504263362e-05,
+      "loss": 0.9184,
+      "step": 12052
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4237982661121347,
+      "learning_rate": 1.6804044339486584e-05,
+      "loss": 0.9437,
+      "step": 12053
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5683298241787367,
+      "learning_rate": 1.6803485135094506e-05,
+      "loss": 0.9303,
+      "step": 12054
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.611735242885587,
+      "learning_rate": 1.6802925891090383e-05,
+      "loss": 0.8213,
+      "step": 12055
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7501577104244833,
+      "learning_rate": 1.680236660747747e-05,
+      "loss": 0.7556,
+      "step": 12056
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.974586014174337,
+      "learning_rate": 1.680180728425903e-05,
+      "loss": 0.8201,
+      "step": 12057
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.993298985399936,
+      "learning_rate": 1.6801247921438313e-05,
+      "loss": 1.0068,
+      "step": 12058
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0756698308222354,
+      "learning_rate": 1.6800688519018575e-05,
+      "loss": 0.8988,
+      "step": 12059
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8244927472270251,
+      "learning_rate": 1.680012907700308e-05,
+      "loss": 0.8861,
+      "step": 12060
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9336884554123603,
+      "learning_rate": 1.6799569595395074e-05,
+      "loss": 0.8323,
+      "step": 12061
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.088100338386899,
+      "learning_rate": 1.6799010074197827e-05,
+      "loss": 0.9371,
+      "step": 12062
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5040603419417602,
+      "learning_rate": 1.6798450513414586e-05,
+      "loss": 0.9496,
+      "step": 12063
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.915075576516552,
+      "learning_rate": 1.6797890913048615e-05,
+      "loss": 0.8665,
+      "step": 12064
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6633815755659207,
+      "learning_rate": 1.6797331273103172e-05,
+      "loss": 0.9223,
+      "step": 12065
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6068577900312653,
+      "learning_rate": 1.6796771593581515e-05,
+      "loss": 0.883,
+      "step": 12066
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2480114650244136,
+      "learning_rate": 1.67962118744869e-05,
+      "loss": 0.9311,
+      "step": 12067
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.652746162691458,
+      "learning_rate": 1.6795652115822592e-05,
+      "loss": 0.8531,
+      "step": 12068
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5109277988128746,
+      "learning_rate": 1.6795092317591845e-05,
+      "loss": 0.9331,
+      "step": 12069
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7087959573625484,
+      "learning_rate": 1.6794532479797917e-05,
+      "loss": 0.904,
+      "step": 12070
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7058387289330337,
+      "learning_rate": 1.6793972602444068e-05,
+      "loss": 0.8387,
+      "step": 12071
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0870066098098943,
+      "learning_rate": 1.679341268553356e-05,
+      "loss": 0.809,
+      "step": 12072
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6100550936958016,
+      "learning_rate": 1.6792852729069655e-05,
+      "loss": 0.8466,
+      "step": 12073
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5805900355126292,
+      "learning_rate": 1.679229273305561e-05,
+      "loss": 0.7826,
+      "step": 12074
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7415190818920951,
+      "learning_rate": 1.6791732697494686e-05,
+      "loss": 0.9487,
+      "step": 12075
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8296216456389471,
+      "learning_rate": 1.6791172622390143e-05,
+      "loss": 0.7222,
+      "step": 12076
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7052077722070471,
+      "learning_rate": 1.6790612507745248e-05,
+      "loss": 1.0562,
+      "step": 12077
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.580337238134025,
+      "learning_rate": 1.6790052353563254e-05,
+      "loss": 0.8409,
+      "step": 12078
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.822350772669777,
+      "learning_rate": 1.6789492159847423e-05,
+      "loss": 0.9346,
+      "step": 12079
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.461389939320792,
+      "learning_rate": 1.6788931926601023e-05,
+      "loss": 0.9343,
+      "step": 12080
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7284629918577966,
+      "learning_rate": 1.6788371653827308e-05,
+      "loss": 0.7117,
+      "step": 12081
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7375188360014948,
+      "learning_rate": 1.6787811341529545e-05,
+      "loss": 0.9151,
+      "step": 12082
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9043100554976007,
+      "learning_rate": 1.6787250989711e-05,
+      "loss": 0.9057,
+      "step": 12083
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7688161928596076,
+      "learning_rate": 1.6786690598374925e-05,
+      "loss": 0.9178,
+      "step": 12084
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7536739895370543,
+      "learning_rate": 1.678613016752459e-05,
+      "loss": 1.035,
+      "step": 12085
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.564625828669157,
+      "learning_rate": 1.678556969716326e-05,
+      "loss": 1.0008,
+      "step": 12086
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7050321826502366,
+      "learning_rate": 1.678500918729419e-05,
+      "loss": 0.9831,
+      "step": 12087
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5072557766213537,
+      "learning_rate": 1.6784448637920654e-05,
+      "loss": 0.909,
+      "step": 12088
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.422028918444073,
+      "learning_rate": 1.6783888049045905e-05,
+      "loss": 0.8803,
+      "step": 12089
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7171492684636207,
+      "learning_rate": 1.6783327420673217e-05,
+      "loss": 0.7521,
+      "step": 12090
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8454929868911767,
+      "learning_rate": 1.6782766752805843e-05,
+      "loss": 0.9211,
+      "step": 12091
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7361819570618178,
+      "learning_rate": 1.678220604544706e-05,
+      "loss": 0.8216,
+      "step": 12092
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9304340204213981,
+      "learning_rate": 1.6781645298600122e-05,
+      "loss": 0.8727,
+      "step": 12093
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.776117311973096,
+      "learning_rate": 1.67810845122683e-05,
+      "loss": 0.9127,
+      "step": 12094
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5399637669154793,
+      "learning_rate": 1.6780523686454856e-05,
+      "loss": 0.9363,
+      "step": 12095
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6170712388149475,
+      "learning_rate": 1.6779962821163055e-05,
+      "loss": 1.0037,
+      "step": 12096
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5360259193886876,
+      "learning_rate": 1.6779401916396167e-05,
+      "loss": 0.9539,
+      "step": 12097
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5607749567476685,
+      "learning_rate": 1.6778840972157453e-05,
+      "loss": 0.8498,
+      "step": 12098
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6577397219977938,
+      "learning_rate": 1.677827998845018e-05,
+      "loss": 0.9981,
+      "step": 12099
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5415657281108988,
+      "learning_rate": 1.6777718965277618e-05,
+      "loss": 0.8641,
+      "step": 12100
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.43334072632315,
+      "learning_rate": 1.677715790264303e-05,
+      "loss": 0.9486,
+      "step": 12101
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.471890966943418,
+      "learning_rate": 1.677659680054968e-05,
+      "loss": 0.875,
+      "step": 12102
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.179018444133653,
+      "learning_rate": 1.677603565900084e-05,
+      "loss": 0.8822,
+      "step": 12103
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7533661773748173,
+      "learning_rate": 1.6775474477999775e-05,
+      "loss": 0.9695,
+      "step": 12104
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.008509169304168,
+      "learning_rate": 1.6774913257549753e-05,
+      "loss": 0.8203,
+      "step": 12105
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9657269970522908,
+      "learning_rate": 1.677435199765404e-05,
+      "loss": 0.8282,
+      "step": 12106
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9046154481391864,
+      "learning_rate": 1.6773790698315904e-05,
+      "loss": 0.9145,
+      "step": 12107
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4726416515920369,
+      "learning_rate": 1.6773229359538615e-05,
+      "loss": 1.0641,
+      "step": 12108
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5989849417906798,
+      "learning_rate": 1.677266798132544e-05,
+      "loss": 0.9467,
+      "step": 12109
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6305797335370829,
+      "learning_rate": 1.6772106563679652e-05,
+      "loss": 0.8047,
+      "step": 12110
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8789901008354555,
+      "learning_rate": 1.677154510660451e-05,
+      "loss": 0.9507,
+      "step": 12111
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7087077680067277,
+      "learning_rate": 1.677098361010329e-05,
+      "loss": 0.9321,
+      "step": 12112
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6245033208333588,
+      "learning_rate": 1.677042207417926e-05,
+      "loss": 0.8337,
+      "step": 12113
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0861704392464477,
+      "learning_rate": 1.676986049883569e-05,
+      "loss": 0.8713,
+      "step": 12114
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5559999931229478,
+      "learning_rate": 1.676929888407585e-05,
+      "loss": 0.9811,
+      "step": 12115
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7788298332957857,
+      "learning_rate": 1.6768737229903006e-05,
+      "loss": 0.8652,
+      "step": 12116
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7310375239800375,
+      "learning_rate": 1.6768175536320433e-05,
+      "loss": 0.9742,
+      "step": 12117
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.687911955457781,
+      "learning_rate": 1.67676138033314e-05,
+      "loss": 0.8578,
+      "step": 12118
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9241766580632238,
+      "learning_rate": 1.6767052030939176e-05,
+      "loss": 0.9828,
+      "step": 12119
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8210837379813571,
+      "learning_rate": 1.676649021914703e-05,
+      "loss": 0.8899,
+      "step": 12120
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8266179346458793,
+      "learning_rate": 1.676592836795824e-05,
+      "loss": 0.761,
+      "step": 12121
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7521943065678918,
+      "learning_rate": 1.6765366477376067e-05,
+      "loss": 0.931,
+      "step": 12122
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6810450572288365,
+      "learning_rate": 1.6764804547403794e-05,
+      "loss": 1.0164,
+      "step": 12123
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0881655090465707,
+      "learning_rate": 1.6764242578044685e-05,
+      "loss": 0.9771,
+      "step": 12124
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.840120166898804,
+      "learning_rate": 1.6763680569302013e-05,
+      "loss": 0.7018,
+      "step": 12125
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.084788705612784,
+      "learning_rate": 1.6763118521179055e-05,
+      "loss": 0.8606,
+      "step": 12126
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7089786146952417,
+      "learning_rate": 1.6762556433679077e-05,
+      "loss": 0.8665,
+      "step": 12127
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6642982879796724,
+      "learning_rate": 1.6761994306805352e-05,
+      "loss": 0.9995,
+      "step": 12128
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7484918603713964,
+      "learning_rate": 1.676143214056116e-05,
+      "loss": 0.8172,
+      "step": 12129
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.076751291994439,
+      "learning_rate": 1.676086993494977e-05,
+      "loss": 0.8749,
+      "step": 12130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.529647058915622,
+      "learning_rate": 1.676030768997445e-05,
+      "loss": 0.8599,
+      "step": 12131
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.616544986292,
+      "learning_rate": 1.6759745405638483e-05,
+      "loss": 0.7398,
+      "step": 12132
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6861570760859006,
+      "learning_rate": 1.6759183081945132e-05,
+      "loss": 1.0307,
+      "step": 12133
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5774028665487227,
+      "learning_rate": 1.6758620718897686e-05,
+      "loss": 0.9331,
+      "step": 12134
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.210057736562824,
+      "learning_rate": 1.6758058316499404e-05,
+      "loss": 0.8175,
+      "step": 12135
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.784458488434124,
+      "learning_rate": 1.675749587475357e-05,
+      "loss": 0.8138,
+      "step": 12136
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7834373456057644,
+      "learning_rate": 1.6756933393663454e-05,
+      "loss": 0.9499,
+      "step": 12137
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.822015964080425,
+      "learning_rate": 1.6756370873232333e-05,
+      "loss": 0.8173,
+      "step": 12138
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7345039250218726,
+      "learning_rate": 1.6755808313463483e-05,
+      "loss": 0.9035,
+      "step": 12139
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5994606425611946,
+      "learning_rate": 1.6755245714360176e-05,
+      "loss": 0.8773,
+      "step": 12140
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0041589416405596,
+      "learning_rate": 1.675468307592569e-05,
+      "loss": 0.761,
+      "step": 12141
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.321829587968344,
+      "learning_rate": 1.67541203981633e-05,
+      "loss": 0.9192,
+      "step": 12142
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4119262645429393,
+      "learning_rate": 1.6753557681076283e-05,
+      "loss": 0.8508,
+      "step": 12143
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6164565770717425,
+      "learning_rate": 1.675299492466792e-05,
+      "loss": 0.8657,
+      "step": 12144
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5609053779149802,
+      "learning_rate": 1.6752432128941478e-05,
+      "loss": 0.8392,
+      "step": 12145
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5584484179504128,
+      "learning_rate": 1.675186929390024e-05,
+      "loss": 0.9377,
+      "step": 12146
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6403155738420643,
+      "learning_rate": 1.6751306419547477e-05,
+      "loss": 0.8465,
+      "step": 12147
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9983518320603546,
+      "learning_rate": 1.6750743505886477e-05,
+      "loss": 0.9305,
+      "step": 12148
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.64933580327383,
+      "learning_rate": 1.6750180552920505e-05,
+      "loss": 0.9025,
+      "step": 12149
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5881623972916037,
+      "learning_rate": 1.674961756065285e-05,
+      "loss": 0.8205,
+      "step": 12150
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8367942528765582,
+      "learning_rate": 1.674905452908678e-05,
+      "loss": 0.9311,
+      "step": 12151
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.666641151845576,
+      "learning_rate": 1.6748491458225583e-05,
+      "loss": 0.9506,
+      "step": 12152
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6777519100362814,
+      "learning_rate": 1.674792834807253e-05,
+      "loss": 0.8885,
+      "step": 12153
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6113348533749987,
+      "learning_rate": 1.67473651986309e-05,
+      "loss": 1.0026,
+      "step": 12154
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2823353391240704,
+      "learning_rate": 1.6746802009903973e-05,
+      "loss": 0.8203,
+      "step": 12155
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5753867272467543,
+      "learning_rate": 1.674623878189503e-05,
+      "loss": 0.9068,
+      "step": 12156
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6497852301373186,
+      "learning_rate": 1.6745675514607352e-05,
+      "loss": 0.8995,
+      "step": 12157
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8378951667923098,
+      "learning_rate": 1.674511220804421e-05,
+      "loss": 0.9481,
+      "step": 12158
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8090166756785467,
+      "learning_rate": 1.674454886220889e-05,
+      "loss": 0.7426,
+      "step": 12159
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.527107917163093,
+      "learning_rate": 1.6743985477104674e-05,
+      "loss": 0.838,
+      "step": 12160
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8460183229425031,
+      "learning_rate": 1.674342205273484e-05,
+      "loss": 0.9113,
+      "step": 12161
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8351275469964452,
+      "learning_rate": 1.6742858589102667e-05,
+      "loss": 0.7911,
+      "step": 12162
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6413491391224464,
+      "learning_rate": 1.6742295086211437e-05,
+      "loss": 0.9077,
+      "step": 12163
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8271721310368887,
+      "learning_rate": 1.674173154406443e-05,
+      "loss": 0.9921,
+      "step": 12164
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4418028660718507,
+      "learning_rate": 1.6741167962664926e-05,
+      "loss": 0.9529,
+      "step": 12165
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7032142658180354,
+      "learning_rate": 1.674060434201621e-05,
+      "loss": 1.0048,
+      "step": 12166
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5846264727040846,
+      "learning_rate": 1.674004068212156e-05,
+      "loss": 0.921,
+      "step": 12167
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.694832108765549,
+      "learning_rate": 1.6739476982984262e-05,
+      "loss": 0.8191,
+      "step": 12168
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9690399925512339,
+      "learning_rate": 1.6738913244607596e-05,
+      "loss": 0.8237,
+      "step": 12169
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7428082487825272,
+      "learning_rate": 1.6738349466994837e-05,
+      "loss": 0.9667,
+      "step": 12170
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5451265020904041,
+      "learning_rate": 1.673778565014928e-05,
+      "loss": 0.9163,
+      "step": 12171
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9022869335023378,
+      "learning_rate": 1.6737221794074198e-05,
+      "loss": 0.8417,
+      "step": 12172
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7762526395965342,
+      "learning_rate": 1.673665789877288e-05,
+      "loss": 0.7331,
+      "step": 12173
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8382672379496612,
+      "learning_rate": 1.6736093964248607e-05,
+      "loss": 1.0269,
+      "step": 12174
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.67148565099998,
+      "learning_rate": 1.673552999050466e-05,
+      "loss": 0.881,
+      "step": 12175
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.397806401015637,
+      "learning_rate": 1.673496597754433e-05,
+      "loss": 0.8278,
+      "step": 12176
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6899164368514699,
+      "learning_rate": 1.673440192537089e-05,
+      "loss": 0.8748,
+      "step": 12177
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6126227141547906,
+      "learning_rate": 1.6733837833987634e-05,
+      "loss": 0.8675,
+      "step": 12178
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.590075726023872,
+      "learning_rate": 1.6733273703397837e-05,
+      "loss": 0.8853,
+      "step": 12179
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6388584561499524,
+      "learning_rate": 1.6732709533604795e-05,
+      "loss": 0.8153,
+      "step": 12180
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7386396255141179,
+      "learning_rate": 1.6732145324611782e-05,
+      "loss": 1.015,
+      "step": 12181
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5836815003004827,
+      "learning_rate": 1.6731581076422093e-05,
+      "loss": 0.8938,
+      "step": 12182
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8083329867565539,
+      "learning_rate": 1.6731016789039004e-05,
+      "loss": 0.8738,
+      "step": 12183
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7253086142697782,
+      "learning_rate": 1.6730452462465804e-05,
+      "loss": 0.9699,
+      "step": 12184
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.539157268562032,
+      "learning_rate": 1.672988809670578e-05,
+      "loss": 0.8994,
+      "step": 12185
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.911845262125024,
+      "learning_rate": 1.6729323691762215e-05,
+      "loss": 0.9144,
+      "step": 12186
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.307513678875182,
+      "learning_rate": 1.6728759247638403e-05,
+      "loss": 1.0184,
+      "step": 12187
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3732074325557118,
+      "learning_rate": 1.6728194764337616e-05,
+      "loss": 0.7939,
+      "step": 12188
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7672123626003609,
+      "learning_rate": 1.6727630241863154e-05,
+      "loss": 0.7142,
+      "step": 12189
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6831646914848786,
+      "learning_rate": 1.6727065680218298e-05,
+      "loss": 0.8365,
+      "step": 12190
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8590114813478344,
+      "learning_rate": 1.6726501079406333e-05,
+      "loss": 0.9574,
+      "step": 12191
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8890531818483094,
+      "learning_rate": 1.6725936439430556e-05,
+      "loss": 0.9694,
+      "step": 12192
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.709951148728141,
+      "learning_rate": 1.672537176029424e-05,
+      "loss": 1.0271,
+      "step": 12193
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7717054653670089,
+      "learning_rate": 1.6724807042000683e-05,
+      "loss": 0.7312,
+      "step": 12194
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7692635953143021,
+      "learning_rate": 1.6724242284553173e-05,
+      "loss": 0.7686,
+      "step": 12195
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.707284662808486,
+      "learning_rate": 1.6723677487954994e-05,
+      "loss": 0.8104,
+      "step": 12196
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8201775811496401,
+      "learning_rate": 1.6723112652209435e-05,
+      "loss": 0.8345,
+      "step": 12197
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8422120693794468,
+      "learning_rate": 1.6722547777319788e-05,
+      "loss": 1.0803,
+      "step": 12198
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9902740230123044,
+      "learning_rate": 1.6721982863289337e-05,
+      "loss": 0.8451,
+      "step": 12199
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9275162040683491,
+      "learning_rate": 1.6721417910121373e-05,
+      "loss": 0.793,
+      "step": 12200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8698808682269792,
+      "learning_rate": 1.672085291781919e-05,
+      "loss": 0.9471,
+      "step": 12201
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8060056000803537,
+      "learning_rate": 1.6720287886386065e-05,
+      "loss": 0.8978,
+      "step": 12202
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5594995453639005,
+      "learning_rate": 1.6719722815825304e-05,
+      "loss": 0.9585,
+      "step": 12203
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.26632461852345,
+      "learning_rate": 1.6719157706140187e-05,
+      "loss": 0.9207,
+      "step": 12204
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1293797245901667,
+      "learning_rate": 1.671859255733401e-05,
+      "loss": 0.8914,
+      "step": 12205
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0022078535797516,
+      "learning_rate": 1.6718027369410057e-05,
+      "loss": 0.7809,
+      "step": 12206
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5206307658815879,
+      "learning_rate": 1.6717462142371622e-05,
+      "loss": 0.8409,
+      "step": 12207
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5406809911566528,
+      "learning_rate": 1.6716896876221996e-05,
+      "loss": 0.9265,
+      "step": 12208
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.568422139717541,
+      "learning_rate": 1.6716331570964465e-05,
+      "loss": 0.8889,
+      "step": 12209
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5251274211123274,
+      "learning_rate": 1.6715766226602334e-05,
+      "loss": 1.0001,
+      "step": 12210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7925874700520106,
+      "learning_rate": 1.671520084313888e-05,
+      "loss": 0.8464,
+      "step": 12211
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9092642788549528,
+      "learning_rate": 1.6714635420577405e-05,
+      "loss": 0.9288,
+      "step": 12212
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7281647398614608,
+      "learning_rate": 1.6714069958921193e-05,
+      "loss": 0.9314,
+      "step": 12213
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6334876442546635,
+      "learning_rate": 1.6713504458173538e-05,
+      "loss": 0.8306,
+      "step": 12214
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7079482274338498,
+      "learning_rate": 1.671293891833774e-05,
+      "loss": 0.8822,
+      "step": 12215
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7325062751202407,
+      "learning_rate": 1.6712373339417082e-05,
+      "loss": 0.9231,
+      "step": 12216
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9959321324322807,
+      "learning_rate": 1.6711807721414865e-05,
+      "loss": 0.8938,
+      "step": 12217
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7868623516545794,
+      "learning_rate": 1.6711242064334377e-05,
+      "loss": 0.9871,
+      "step": 12218
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5337025477427622,
+      "learning_rate": 1.6710676368178913e-05,
+      "loss": 0.8855,
+      "step": 12219
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.732656145498076,
+      "learning_rate": 1.6710110632951765e-05,
+      "loss": 0.7739,
+      "step": 12220
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8750496210794094,
+      "learning_rate": 1.6709544858656232e-05,
+      "loss": 0.9248,
+      "step": 12221
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6113848564608044,
+      "learning_rate": 1.67089790452956e-05,
+      "loss": 0.8885,
+      "step": 12222
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.379260451135479,
+      "learning_rate": 1.670841319287317e-05,
+      "loss": 0.9252,
+      "step": 12223
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8943207397854485,
+      "learning_rate": 1.6707847301392237e-05,
+      "loss": 1.0494,
+      "step": 12224
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7446486497389646,
+      "learning_rate": 1.670728137085609e-05,
+      "loss": 0.8869,
+      "step": 12225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.669502953584017,
+      "learning_rate": 1.670671540126803e-05,
+      "loss": 0.9199,
+      "step": 12226
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.05287157908762,
+      "learning_rate": 1.670614939263135e-05,
+      "loss": 0.8999,
+      "step": 12227
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6909647957171001,
+      "learning_rate": 1.670558334494934e-05,
+      "loss": 0.8032,
+      "step": 12228
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7775839311822306,
+      "learning_rate": 1.6705017258225306e-05,
+      "loss": 0.7236,
+      "step": 12229
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7493131864286027,
+      "learning_rate": 1.6704451132462534e-05,
+      "loss": 0.7561,
+      "step": 12230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6701024053753322,
+      "learning_rate": 1.670388496766433e-05,
+      "loss": 0.801,
+      "step": 12231
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8037260554270789,
+      "learning_rate": 1.6703318763833985e-05,
+      "loss": 0.7923,
+      "step": 12232
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.669883109902486,
+      "learning_rate": 1.670275252097479e-05,
+      "loss": 0.8292,
+      "step": 12233
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3779884142065795,
+      "learning_rate": 1.6702186239090058e-05,
+      "loss": 0.8572,
+      "step": 12234
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5118176745204672,
+      "learning_rate": 1.6701619918183066e-05,
+      "loss": 0.807,
+      "step": 12235
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6364504410292038,
+      "learning_rate": 1.6701053558257125e-05,
+      "loss": 0.7506,
+      "step": 12236
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.778540625965692,
+      "learning_rate": 1.670048715931553e-05,
+      "loss": 0.9124,
+      "step": 12237
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7695970815628368,
+      "learning_rate": 1.6699920721361576e-05,
+      "loss": 0.873,
+      "step": 12238
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7118783256912948,
+      "learning_rate": 1.669935424439856e-05,
+      "loss": 0.8509,
+      "step": 12239
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.568225321523873,
+      "learning_rate": 1.6698787728429786e-05,
+      "loss": 0.9628,
+      "step": 12240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8214935628966862,
+      "learning_rate": 1.6698221173458547e-05,
+      "loss": 0.9508,
+      "step": 12241
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.357107152748715,
+      "learning_rate": 1.6697654579488147e-05,
+      "loss": 0.9354,
+      "step": 12242
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.099712591711463,
+      "learning_rate": 1.669708794652188e-05,
+      "loss": 0.9586,
+      "step": 12243
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4923506075047415,
+      "learning_rate": 1.6696521274563046e-05,
+      "loss": 0.935,
+      "step": 12244
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6558292014337364,
+      "learning_rate": 1.6695954563614945e-05,
+      "loss": 0.9342,
+      "step": 12245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4161736339345727,
+      "learning_rate": 1.6695387813680873e-05,
+      "loss": 0.9898,
+      "step": 12246
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.578176725029422,
+      "learning_rate": 1.6694821024764142e-05,
+      "loss": 0.9507,
+      "step": 12247
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5798493606958028,
+      "learning_rate": 1.6694254196868036e-05,
+      "loss": 0.8018,
+      "step": 12248
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5992636709679737,
+      "learning_rate": 1.6693687329995864e-05,
+      "loss": 1.0479,
+      "step": 12249
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8119488289879347,
+      "learning_rate": 1.669312042415093e-05,
+      "loss": 0.7321,
+      "step": 12250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4239673768595857,
+      "learning_rate": 1.6692553479336522e-05,
+      "loss": 0.8838,
+      "step": 12251
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.734874791350191,
+      "learning_rate": 1.6691986495555955e-05,
+      "loss": 0.9081,
+      "step": 12252
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8654154396301834,
+      "learning_rate": 1.6691419472812523e-05,
+      "loss": 0.9531,
+      "step": 12253
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8046171747766437,
+      "learning_rate": 1.669085241110953e-05,
+      "loss": 0.7887,
+      "step": 12254
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5715258928765696,
+      "learning_rate": 1.669028531045027e-05,
+      "loss": 0.8028,
+      "step": 12255
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5744562749853817,
+      "learning_rate": 1.6689718170838058e-05,
+      "loss": 1.0136,
+      "step": 12256
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5472922858812632,
+      "learning_rate": 1.6689150992276183e-05,
+      "loss": 0.9597,
+      "step": 12257
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5469401139946641,
+      "learning_rate": 1.6688583774767954e-05,
+      "loss": 0.8696,
+      "step": 12258
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.666617530135532,
+      "learning_rate": 1.6688016518316676e-05,
+      "loss": 0.9394,
+      "step": 12259
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7011088428271393,
+      "learning_rate": 1.6687449222925644e-05,
+      "loss": 0.8714,
+      "step": 12260
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8769095071972701,
+      "learning_rate": 1.668688188859817e-05,
+      "loss": 0.8834,
+      "step": 12261
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7839522581055267,
+      "learning_rate": 1.668631451533755e-05,
+      "loss": 0.7527,
+      "step": 12262
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.660715825154198,
+      "learning_rate": 1.668574710314709e-05,
+      "loss": 0.8473,
+      "step": 12263
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7819216528854769,
+      "learning_rate": 1.6685179652030096e-05,
+      "loss": 0.7558,
+      "step": 12264
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7519906666289344,
+      "learning_rate": 1.6684612161989866e-05,
+      "loss": 0.7519,
+      "step": 12265
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6014410970921857,
+      "learning_rate": 1.668404463302971e-05,
+      "loss": 0.8583,
+      "step": 12266
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7582936920829766,
+      "learning_rate": 1.6683477065152928e-05,
+      "loss": 0.7893,
+      "step": 12267
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9225030505899916,
+      "learning_rate": 1.668290945836283e-05,
+      "loss": 1.0459,
+      "step": 12268
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2207496979664394,
+      "learning_rate": 1.6682341812662715e-05,
+      "loss": 0.9435,
+      "step": 12269
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8646621386945001,
+      "learning_rate": 1.6681774128055893e-05,
+      "loss": 0.8616,
+      "step": 12270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.684157281910385,
+      "learning_rate": 1.6681206404545665e-05,
+      "loss": 0.9558,
+      "step": 12271
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.627757204331449,
+      "learning_rate": 1.6680638642135337e-05,
+      "loss": 0.9955,
+      "step": 12272
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.723455592431574,
+      "learning_rate": 1.6680070840828215e-05,
+      "loss": 0.8891,
+      "step": 12273
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4599873012136697,
+      "learning_rate": 1.6679503000627607e-05,
+      "loss": 0.8658,
+      "step": 12274
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4682776481882587,
+      "learning_rate": 1.6678935121536817e-05,
+      "loss": 0.8566,
+      "step": 12275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9400572814199228,
+      "learning_rate": 1.6678367203559154e-05,
+      "loss": 0.863,
+      "step": 12276
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9335789225351774,
+      "learning_rate": 1.6677799246697924e-05,
+      "loss": 0.9213,
+      "step": 12277
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8849549836134487,
+      "learning_rate": 1.667723125095643e-05,
+      "loss": 0.9934,
+      "step": 12278
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.766935439761052,
+      "learning_rate": 1.667666321633798e-05,
+      "loss": 0.8688,
+      "step": 12279
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4438720996519214,
+      "learning_rate": 1.6676095142845883e-05,
+      "loss": 0.8699,
+      "step": 12280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.728291103630508,
+      "learning_rate": 1.6675527030483452e-05,
+      "loss": 1.0034,
+      "step": 12281
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8809251011715278,
+      "learning_rate": 1.6674958879253986e-05,
+      "loss": 0.941,
+      "step": 12282
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6794413028878576,
+      "learning_rate": 1.6674390689160795e-05,
+      "loss": 1.0009,
+      "step": 12283
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6543983039364842,
+      "learning_rate": 1.667382246020719e-05,
+      "loss": 0.9173,
+      "step": 12284
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7860959328125816,
+      "learning_rate": 1.667325419239648e-05,
+      "loss": 0.9201,
+      "step": 12285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.206627340749698,
+      "learning_rate": 1.6672685885731964e-05,
+      "loss": 0.9139,
+      "step": 12286
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5783713629501168,
+      "learning_rate": 1.6672117540216962e-05,
+      "loss": 0.8396,
+      "step": 12287
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.716399418345919,
+      "learning_rate": 1.6671549155854783e-05,
+      "loss": 0.9156,
+      "step": 12288
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6903218991008275,
+      "learning_rate": 1.6670980732648726e-05,
+      "loss": 0.7826,
+      "step": 12289
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1336144370961314,
+      "learning_rate": 1.6670412270602115e-05,
+      "loss": 1.0281,
+      "step": 12290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7279165073206113,
+      "learning_rate": 1.6669843769718245e-05,
+      "loss": 1.0206,
+      "step": 12291
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8822676876190743,
+      "learning_rate": 1.6669275230000438e-05,
+      "loss": 0.8305,
+      "step": 12292
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6117625969653189,
+      "learning_rate": 1.6668706651451994e-05,
+      "loss": 0.8035,
+      "step": 12293
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7937566841552137,
+      "learning_rate": 1.6668138034076235e-05,
+      "loss": 0.9185,
+      "step": 12294
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4710250547425299,
+      "learning_rate": 1.666756937787646e-05,
+      "loss": 0.8532,
+      "step": 12295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7022814553019228,
+      "learning_rate": 1.6667000682855988e-05,
+      "loss": 0.9504,
+      "step": 12296
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8402156453337422,
+      "learning_rate": 1.666643194901812e-05,
+      "loss": 0.9303,
+      "step": 12297
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5704857606435816,
+      "learning_rate": 1.6665863176366186e-05,
+      "loss": 0.9642,
+      "step": 12298
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6253123861433203,
+      "learning_rate": 1.6665294364903482e-05,
+      "loss": 0.8874,
+      "step": 12299
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9581402793866538,
+      "learning_rate": 1.666472551463332e-05,
+      "loss": 0.8924,
+      "step": 12300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6892515049267898,
+      "learning_rate": 1.666415662555902e-05,
+      "loss": 0.9902,
+      "step": 12301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5766855568536504,
+      "learning_rate": 1.666358769768389e-05,
+      "loss": 0.8887,
+      "step": 12302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7577192453145378,
+      "learning_rate": 1.666301873101124e-05,
+      "loss": 0.8587,
+      "step": 12303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6552417157799173,
+      "learning_rate": 1.6662449725544386e-05,
+      "loss": 0.7826,
+      "step": 12304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8785490603078208,
+      "learning_rate": 1.6661880681286647e-05,
+      "loss": 0.794,
+      "step": 12305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8908194295900622,
+      "learning_rate": 1.6661311598241325e-05,
+      "loss": 0.8863,
+      "step": 12306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3980117405735806,
+      "learning_rate": 1.6660742476411736e-05,
+      "loss": 0.9378,
+      "step": 12307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.58416023227016,
+      "learning_rate": 1.6660173315801194e-05,
+      "loss": 0.822,
+      "step": 12308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8644591582627947,
+      "learning_rate": 1.6659604116413016e-05,
+      "loss": 0.9413,
+      "step": 12309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5331017093920953,
+      "learning_rate": 1.665903487825052e-05,
+      "loss": 0.9194,
+      "step": 12310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4799170126505596,
+      "learning_rate": 1.6658465601317008e-05,
+      "loss": 0.844,
+      "step": 12311
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9319909566254545,
+      "learning_rate": 1.6657896285615804e-05,
+      "loss": 0.9935,
+      "step": 12312
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.070700103114267,
+      "learning_rate": 1.6657326931150222e-05,
+      "loss": 0.8953,
+      "step": 12313
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6172336293419107,
+      "learning_rate": 1.6656757537923572e-05,
+      "loss": 0.9716,
+      "step": 12314
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6310794910024244,
+      "learning_rate": 1.6656188105939172e-05,
+      "loss": 0.9302,
+      "step": 12315
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8971320299818641,
+      "learning_rate": 1.665561863520034e-05,
+      "loss": 0.9151,
+      "step": 12316
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.736470191149903,
+      "learning_rate": 1.665504912571039e-05,
+      "loss": 0.9484,
+      "step": 12317
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6279318956950544,
+      "learning_rate": 1.665447957747263e-05,
+      "loss": 0.9467,
+      "step": 12318
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9334966897235022,
+      "learning_rate": 1.665390999049039e-05,
+      "loss": 0.958,
+      "step": 12319
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7719169144365599,
+      "learning_rate": 1.6653340364766978e-05,
+      "loss": 0.9212,
+      "step": 12320
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5125219834409933,
+      "learning_rate": 1.6652770700305708e-05,
+      "loss": 0.9738,
+      "step": 12321
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8087532685532643,
+      "learning_rate": 1.6652200997109906e-05,
+      "loss": 0.8755,
+      "step": 12322
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5968160585435691,
+      "learning_rate": 1.665163125518288e-05,
+      "loss": 1.012,
+      "step": 12323
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9473321589679378,
+      "learning_rate": 1.6651061474527953e-05,
+      "loss": 0.928,
+      "step": 12324
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5942480191405095,
+      "learning_rate": 1.665049165514844e-05,
+      "loss": 0.9122,
+      "step": 12325
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7380775131349784,
+      "learning_rate": 1.6649921797047658e-05,
+      "loss": 0.9767,
+      "step": 12326
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5727733898717682,
+      "learning_rate": 1.664935190022893e-05,
+      "loss": 0.9269,
+      "step": 12327
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6581584125942213,
+      "learning_rate": 1.6648781964695565e-05,
+      "loss": 0.8798,
+      "step": 12328
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.085039497775422,
+      "learning_rate": 1.664821199045089e-05,
+      "loss": 0.8907,
+      "step": 12329
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7011905498675808,
+      "learning_rate": 1.6647641977498215e-05,
+      "loss": 0.8789,
+      "step": 12330
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7651592109264136,
+      "learning_rate": 1.6647071925840868e-05,
+      "loss": 0.8595,
+      "step": 12331
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7933733261228861,
+      "learning_rate": 1.6646501835482162e-05,
+      "loss": 0.7963,
+      "step": 12332
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.507486813729768,
+      "learning_rate": 1.664593170642542e-05,
+      "loss": 0.8573,
+      "step": 12333
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6484609787838709,
+      "learning_rate": 1.6645361538673962e-05,
+      "loss": 0.854,
+      "step": 12334
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5813308269095445,
+      "learning_rate": 1.66447913322311e-05,
+      "loss": 1.0647,
+      "step": 12335
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1801038788545153,
+      "learning_rate": 1.6644221087100164e-05,
+      "loss": 0.937,
+      "step": 12336
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.787563594581469,
+      "learning_rate": 1.6643650803284462e-05,
+      "loss": 0.7713,
+      "step": 12337
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7275017154735235,
+      "learning_rate": 1.6643080480787328e-05,
+      "loss": 0.8655,
+      "step": 12338
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.065116485984859,
+      "learning_rate": 1.6642510119612073e-05,
+      "loss": 0.7833,
+      "step": 12339
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7060828246009705,
+      "learning_rate": 1.6641939719762023e-05,
+      "loss": 0.8046,
+      "step": 12340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8453269440793627,
+      "learning_rate": 1.6641369281240497e-05,
+      "loss": 0.9786,
+      "step": 12341
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8012910752244902,
+      "learning_rate": 1.6640798804050816e-05,
+      "loss": 0.7556,
+      "step": 12342
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5308843744721088,
+      "learning_rate": 1.6640228288196303e-05,
+      "loss": 0.9296,
+      "step": 12343
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.737642490652528,
+      "learning_rate": 1.6639657733680276e-05,
+      "loss": 0.8437,
+      "step": 12344
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.698978431382576,
+      "learning_rate": 1.663908714050606e-05,
+      "loss": 0.9014,
+      "step": 12345
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.610952920686975,
+      "learning_rate": 1.6638516508676977e-05,
+      "loss": 1.0385,
+      "step": 12346
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5209681413639942,
+      "learning_rate": 1.6637945838196356e-05,
+      "loss": 0.8812,
+      "step": 12347
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8091922010826214,
+      "learning_rate": 1.6637375129067502e-05,
+      "loss": 0.7622,
+      "step": 12348
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9585759600570634,
+      "learning_rate": 1.6636804381293755e-05,
+      "loss": 0.8998,
+      "step": 12349
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.848051859640521,
+      "learning_rate": 1.6636233594878434e-05,
+      "loss": 0.9756,
+      "step": 12350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9402208040483353,
+      "learning_rate": 1.6635662769824854e-05,
+      "loss": 0.9358,
+      "step": 12351
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0561839140256595,
+      "learning_rate": 1.6635091906136346e-05,
+      "loss": 0.9857,
+      "step": 12352
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7065791136982313,
+      "learning_rate": 1.6634521003816235e-05,
+      "loss": 0.9536,
+      "step": 12353
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8281716021758135,
+      "learning_rate": 1.663395006286784e-05,
+      "loss": 0.7957,
+      "step": 12354
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3297558036049972,
+      "learning_rate": 1.6633379083294487e-05,
+      "loss": 0.9298,
+      "step": 12355
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8159158647140275,
+      "learning_rate": 1.6632808065099503e-05,
+      "loss": 0.7684,
+      "step": 12356
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5707137784780996,
+      "learning_rate": 1.663223700828621e-05,
+      "loss": 0.8785,
+      "step": 12357
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6814034805454174,
+      "learning_rate": 1.663166591285793e-05,
+      "loss": 0.9133,
+      "step": 12358
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4503937763600603,
+      "learning_rate": 1.6631094778817997e-05,
+      "loss": 0.781,
+      "step": 12359
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.194924135819242,
+      "learning_rate": 1.6630523606169726e-05,
+      "loss": 0.9546,
+      "step": 12360
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5253818454051704,
+      "learning_rate": 1.6629952394916447e-05,
+      "loss": 0.8996,
+      "step": 12361
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4668536798250065,
+      "learning_rate": 1.6629381145061492e-05,
+      "loss": 0.9968,
+      "step": 12362
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5603568359883468,
+      "learning_rate": 1.6628809856608177e-05,
+      "loss": 0.9087,
+      "step": 12363
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7071356459116276,
+      "learning_rate": 1.6628238529559834e-05,
+      "loss": 0.9637,
+      "step": 12364
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.173739605641109,
+      "learning_rate": 1.6627667163919788e-05,
+      "loss": 0.8942,
+      "step": 12365
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5016821938104739,
+      "learning_rate": 1.6627095759691364e-05,
+      "loss": 0.959,
+      "step": 12366
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6556425408708844,
+      "learning_rate": 1.662652431687789e-05,
+      "loss": 0.9674,
+      "step": 12367
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3935317380399856,
+      "learning_rate": 1.6625952835482696e-05,
+      "loss": 1.0093,
+      "step": 12368
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5679124163019151,
+      "learning_rate": 1.6625381315509103e-05,
+      "loss": 0.9,
+      "step": 12369
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7268788148723457,
+      "learning_rate": 1.6624809756960445e-05,
+      "loss": 0.9095,
+      "step": 12370
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8919170102582898,
+      "learning_rate": 1.6624238159840043e-05,
+      "loss": 0.936,
+      "step": 12371
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.52140468160026,
+      "learning_rate": 1.6623666524151233e-05,
+      "loss": 1.0062,
+      "step": 12372
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.673129431360089,
+      "learning_rate": 1.6623094849897342e-05,
+      "loss": 0.8764,
+      "step": 12373
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.721181444416572,
+      "learning_rate": 1.6622523137081693e-05,
+      "loss": 0.8416,
+      "step": 12374
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.70022299736537,
+      "learning_rate": 1.6621951385707617e-05,
+      "loss": 0.9386,
+      "step": 12375
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7811153066649196,
+      "learning_rate": 1.6621379595778444e-05,
+      "loss": 0.7529,
+      "step": 12376
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1186814244354397,
+      "learning_rate": 1.6620807767297505e-05,
+      "loss": 0.9592,
+      "step": 12377
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5846133397793458,
+      "learning_rate": 1.6620235900268127e-05,
+      "loss": 0.9892,
+      "step": 12378
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.013109660817868,
+      "learning_rate": 1.6619663994693634e-05,
+      "loss": 0.8492,
+      "step": 12379
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0948575221486005,
+      "learning_rate": 1.6619092050577367e-05,
+      "loss": 0.8426,
+      "step": 12380
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5192902375285138,
+      "learning_rate": 1.661852006792265e-05,
+      "loss": 0.9254,
+      "step": 12381
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6364137240406902,
+      "learning_rate": 1.6617948046732814e-05,
+      "loss": 0.9067,
+      "step": 12382
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9348586695639756,
+      "learning_rate": 1.6617375987011186e-05,
+      "loss": 0.906,
+      "step": 12383
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4590646135318748,
+      "learning_rate": 1.6616803888761105e-05,
+      "loss": 0.8899,
+      "step": 12384
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.721278023140308,
+      "learning_rate": 1.6616231751985893e-05,
+      "loss": 1.0528,
+      "step": 12385
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8881592477347118,
+      "learning_rate": 1.6615659576688886e-05,
+      "loss": 0.8614,
+      "step": 12386
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9370035403504267,
+      "learning_rate": 1.6615087362873415e-05,
+      "loss": 0.9785,
+      "step": 12387
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8024223662404868,
+      "learning_rate": 1.6614515110542807e-05,
+      "loss": 0.7696,
+      "step": 12388
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5241390675902078,
+      "learning_rate": 1.6613942819700402e-05,
+      "loss": 0.8594,
+      "step": 12389
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5655149512026225,
+      "learning_rate": 1.661337049034953e-05,
+      "loss": 0.85,
+      "step": 12390
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8562662092731468,
+      "learning_rate": 1.661279812249352e-05,
+      "loss": 0.8614,
+      "step": 12391
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0499085936522152,
+      "learning_rate": 1.6612225716135703e-05,
+      "loss": 0.7483,
+      "step": 12392
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.126306062148004,
+      "learning_rate": 1.6611653271279415e-05,
+      "loss": 0.9214,
+      "step": 12393
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9970361760015345,
+      "learning_rate": 1.661108078792799e-05,
+      "loss": 0.9011,
+      "step": 12394
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.618935057963343,
+      "learning_rate": 1.661050826608476e-05,
+      "loss": 0.8472,
+      "step": 12395
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.669865161402569,
+      "learning_rate": 1.6609935705753057e-05,
+      "loss": 0.9448,
+      "step": 12396
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4149877215664934,
+      "learning_rate": 1.6609363106936215e-05,
+      "loss": 0.8628,
+      "step": 12397
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5878739224700038,
+      "learning_rate": 1.660879046963757e-05,
+      "loss": 0.8633,
+      "step": 12398
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5146007339333536,
+      "learning_rate": 1.6608217793860455e-05,
+      "loss": 0.9836,
+      "step": 12399
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8789795195636763,
+      "learning_rate": 1.66076450796082e-05,
+      "loss": 0.7568,
+      "step": 12400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6979757028449252,
+      "learning_rate": 1.660707232688415e-05,
+      "loss": 0.956,
+      "step": 12401
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.500427909394399,
+      "learning_rate": 1.6606499535691623e-05,
+      "loss": 0.8488,
+      "step": 12402
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6089630645552249,
+      "learning_rate": 1.6605926706033973e-05,
+      "loss": 0.8867,
+      "step": 12403
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8508437747345459,
+      "learning_rate": 1.6605353837914522e-05,
+      "loss": 0.7926,
+      "step": 12404
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5592909957229333,
+      "learning_rate": 1.660478093133661e-05,
+      "loss": 0.9414,
+      "step": 12405
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5263947396985849,
+      "learning_rate": 1.6604207986303577e-05,
+      "loss": 0.8553,
+      "step": 12406
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.890055453499283,
+      "learning_rate": 1.660363500281875e-05,
+      "loss": 0.9412,
+      "step": 12407
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.905197623267496,
+      "learning_rate": 1.660306198088547e-05,
+      "loss": 0.9616,
+      "step": 12408
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6006842372509393,
+      "learning_rate": 1.6602488920507075e-05,
+      "loss": 0.9508,
+      "step": 12409
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8765328850260976,
+      "learning_rate": 1.6601915821686895e-05,
+      "loss": 0.8056,
+      "step": 12410
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6615510442923849,
+      "learning_rate": 1.6601342684428274e-05,
+      "loss": 0.9828,
+      "step": 12411
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6464495197898061,
+      "learning_rate": 1.6600769508734545e-05,
+      "loss": 0.9056,
+      "step": 12412
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7108594666909338,
+      "learning_rate": 1.6600196294609047e-05,
+      "loss": 0.9969,
+      "step": 12413
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4569919870085877,
+      "learning_rate": 1.6599623042055114e-05,
+      "loss": 0.9222,
+      "step": 12414
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7369460122421327,
+      "learning_rate": 1.659904975107609e-05,
+      "loss": 0.9434,
+      "step": 12415
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8487943434289995,
+      "learning_rate": 1.6598476421675303e-05,
+      "loss": 0.8945,
+      "step": 12416
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6363474385866088,
+      "learning_rate": 1.6597903053856104e-05,
+      "loss": 0.9559,
+      "step": 12417
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.762205820807905,
+      "learning_rate": 1.6597329647621822e-05,
+      "loss": 0.9041,
+      "step": 12418
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6424749923494606,
+      "learning_rate": 1.65967562029758e-05,
+      "loss": 1.1081,
+      "step": 12419
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.769802577369969,
+      "learning_rate": 1.659618271992137e-05,
+      "loss": 0.9792,
+      "step": 12420
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7767831809317305,
+      "learning_rate": 1.6595609198461878e-05,
+      "loss": 1.0391,
+      "step": 12421
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7901524439858039,
+      "learning_rate": 1.6595035638600662e-05,
+      "loss": 0.7316,
+      "step": 12422
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5719017359433596,
+      "learning_rate": 1.659446204034106e-05,
+      "loss": 0.8507,
+      "step": 12423
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.060623685904592,
+      "learning_rate": 1.6593888403686417e-05,
+      "loss": 0.7878,
+      "step": 12424
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4872159402676133,
+      "learning_rate": 1.6593314728640063e-05,
+      "loss": 0.9692,
+      "step": 12425
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7780443697177473,
+      "learning_rate": 1.6592741015205343e-05,
+      "loss": 0.8847,
+      "step": 12426
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8555299143999606,
+      "learning_rate": 1.65921672633856e-05,
+      "loss": 0.922,
+      "step": 12427
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9722574689888284,
+      "learning_rate": 1.659159347318417e-05,
+      "loss": 0.8291,
+      "step": 12428
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.708378600199253,
+      "learning_rate": 1.65910196446044e-05,
+      "loss": 0.9437,
+      "step": 12429
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.50651213327216,
+      "learning_rate": 1.659044577764962e-05,
+      "loss": 0.841,
+      "step": 12430
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5880006129482342,
+      "learning_rate": 1.6589871872323185e-05,
+      "loss": 0.8139,
+      "step": 12431
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8287486261025165,
+      "learning_rate": 1.6589297928628423e-05,
+      "loss": 0.9988,
+      "step": 12432
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6971909721930618,
+      "learning_rate": 1.658872394656869e-05,
+      "loss": 1.0148,
+      "step": 12433
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7816083022840867,
+      "learning_rate": 1.658814992614731e-05,
+      "loss": 0.7952,
+      "step": 12434
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6521421839300543,
+      "learning_rate": 1.6587575867367645e-05,
+      "loss": 0.8259,
+      "step": 12435
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9100101776562906,
+      "learning_rate": 1.6587001770233024e-05,
+      "loss": 0.9603,
+      "step": 12436
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9651148575540143,
+      "learning_rate": 1.6586427634746794e-05,
+      "loss": 0.8795,
+      "step": 12437
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7377481739674325,
+      "learning_rate": 1.6585853460912295e-05,
+      "loss": 0.8117,
+      "step": 12438
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8828528123259494,
+      "learning_rate": 1.6585279248732873e-05,
+      "loss": 0.8819,
+      "step": 12439
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.608773014865421,
+      "learning_rate": 1.658470499821187e-05,
+      "loss": 0.8734,
+      "step": 12440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7605960099335882,
+      "learning_rate": 1.658413070935263e-05,
+      "loss": 0.8459,
+      "step": 12441
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7970479089607763,
+      "learning_rate": 1.65835563821585e-05,
+      "loss": 0.7204,
+      "step": 12442
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7186280940691347,
+      "learning_rate": 1.6582982016632818e-05,
+      "loss": 0.9729,
+      "step": 12443
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6324783193831447,
+      "learning_rate": 1.658240761277893e-05,
+      "loss": 0.976,
+      "step": 12444
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1272252434631294,
+      "learning_rate": 1.6581833170600182e-05,
+      "loss": 1.0526,
+      "step": 12445
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8009911426679113,
+      "learning_rate": 1.6581258690099916e-05,
+      "loss": 0.93,
+      "step": 12446
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1238207067006103,
+      "learning_rate": 1.6580684171281478e-05,
+      "loss": 0.7885,
+      "step": 12447
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.690476768315683,
+      "learning_rate": 1.6580109614148213e-05,
+      "loss": 0.9279,
+      "step": 12448
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5214922340127963,
+      "learning_rate": 1.6579535018703472e-05,
+      "loss": 0.9095,
+      "step": 12449
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5365930495181914,
+      "learning_rate": 1.6578960384950588e-05,
+      "loss": 0.971,
+      "step": 12450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6791019795625792,
+      "learning_rate": 1.657838571289292e-05,
+      "loss": 0.8603,
+      "step": 12451
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9813175161085175,
+      "learning_rate": 1.6577811002533803e-05,
+      "loss": 0.9601,
+      "step": 12452
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7313133043420725,
+      "learning_rate": 1.657723625387659e-05,
+      "loss": 0.8794,
+      "step": 12453
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8447617294100722,
+      "learning_rate": 1.6576661466924625e-05,
+      "loss": 0.9528,
+      "step": 12454
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5184662174667591,
+      "learning_rate": 1.657608664168125e-05,
+      "loss": 0.9888,
+      "step": 12455
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5859284832486973,
+      "learning_rate": 1.6575511778149823e-05,
+      "loss": 0.8883,
+      "step": 12456
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7638601858629971,
+      "learning_rate": 1.657493687633368e-05,
+      "loss": 0.747,
+      "step": 12457
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4253055720035062,
+      "learning_rate": 1.6574361936236177e-05,
+      "loss": 1.0465,
+      "step": 12458
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.663783605052354,
+      "learning_rate": 1.6573786957860658e-05,
+      "loss": 0.8601,
+      "step": 12459
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6779937229358202,
+      "learning_rate": 1.6573211941210467e-05,
+      "loss": 1.0118,
+      "step": 12460
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4610106787659176,
+      "learning_rate": 1.6572636886288957e-05,
+      "loss": 0.8747,
+      "step": 12461
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8187017357598494,
+      "learning_rate": 1.6572061793099473e-05,
+      "loss": 0.9164,
+      "step": 12462
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.815818450899879,
+      "learning_rate": 1.6571486661645367e-05,
+      "loss": 0.9278,
+      "step": 12463
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7746348979030931,
+      "learning_rate": 1.6570911491929984e-05,
+      "loss": 0.7661,
+      "step": 12464
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8881838375171054,
+      "learning_rate": 1.657033628395667e-05,
+      "loss": 1.0494,
+      "step": 12465
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7231725770254809,
+      "learning_rate": 1.656976103772878e-05,
+      "loss": 0.9585,
+      "step": 12466
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9829245804628635,
+      "learning_rate": 1.6569185753249662e-05,
+      "loss": 0.844,
+      "step": 12467
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9534500390282343,
+      "learning_rate": 1.656861043052267e-05,
+      "loss": 0.7989,
+      "step": 12468
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.343567576200772,
+      "learning_rate": 1.656803506955114e-05,
+      "loss": 0.8192,
+      "step": 12469
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7880945839124653,
+      "learning_rate": 1.6567459670338434e-05,
+      "loss": 0.8782,
+      "step": 12470
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5942889604392418,
+      "learning_rate": 1.65668842328879e-05,
+      "loss": 0.8052,
+      "step": 12471
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.715804923712623,
+      "learning_rate": 1.6566308757202883e-05,
+      "loss": 0.828,
+      "step": 12472
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6820408642681866,
+      "learning_rate": 1.656573324328674e-05,
+      "loss": 1.0088,
+      "step": 12473
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.687165150034548,
+      "learning_rate": 1.656515769114282e-05,
+      "loss": 0.836,
+      "step": 12474
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7561081354808734,
+      "learning_rate": 1.6564582100774472e-05,
+      "loss": 0.7485,
+      "step": 12475
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6584603705401548,
+      "learning_rate": 1.656400647218505e-05,
+      "loss": 0.9157,
+      "step": 12476
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.203524444989961,
+      "learning_rate": 1.6563430805377903e-05,
+      "loss": 0.8137,
+      "step": 12477
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2972536105705044,
+      "learning_rate": 1.6562855100356386e-05,
+      "loss": 0.8908,
+      "step": 12478
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8735159531102061,
+      "learning_rate": 1.656227935712385e-05,
+      "loss": 0.8963,
+      "step": 12479
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5116461779280899,
+      "learning_rate": 1.6561703575683645e-05,
+      "loss": 0.9679,
+      "step": 12480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6706837135594916,
+      "learning_rate": 1.6561127756039122e-05,
+      "loss": 0.9187,
+      "step": 12481
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.901686399783727,
+      "learning_rate": 1.6560551898193637e-05,
+      "loss": 0.8407,
+      "step": 12482
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.70722075195508,
+      "learning_rate": 1.655997600215054e-05,
+      "loss": 0.8294,
+      "step": 12483
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7120751778507237,
+      "learning_rate": 1.6559400067913187e-05,
+      "loss": 0.9512,
+      "step": 12484
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7606617152252335,
+      "learning_rate": 1.655882409548493e-05,
+      "loss": 0.8824,
+      "step": 12485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6686411806532575,
+      "learning_rate": 1.6558248084869125e-05,
+      "loss": 1.0202,
+      "step": 12486
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.604446851958499,
+      "learning_rate": 1.6557672036069122e-05,
+      "loss": 0.9886,
+      "step": 12487
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.67382234260919,
+      "learning_rate": 1.6557095949088275e-05,
+      "loss": 0.8858,
+      "step": 12488
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8690076416061525,
+      "learning_rate": 1.655651982392994e-05,
+      "loss": 0.9147,
+      "step": 12489
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5672737697709371,
+      "learning_rate": 1.655594366059747e-05,
+      "loss": 0.8841,
+      "step": 12490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6404206310967329,
+      "learning_rate": 1.6555367459094223e-05,
+      "loss": 1.0062,
+      "step": 12491
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5812644336347015,
+      "learning_rate": 1.6554791219423548e-05,
+      "loss": 0.8956,
+      "step": 12492
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4176601396910753,
+      "learning_rate": 1.6554214941588807e-05,
+      "loss": 0.8884,
+      "step": 12493
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4873184701590279,
+      "learning_rate": 1.655363862559335e-05,
+      "loss": 0.8585,
+      "step": 12494
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2956422485533174,
+      "learning_rate": 1.6553062271440536e-05,
+      "loss": 0.7741,
+      "step": 12495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7087286449240466,
+      "learning_rate": 1.6552485879133714e-05,
+      "loss": 0.8118,
+      "step": 12496
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.707273886589387,
+      "learning_rate": 1.655190944867625e-05,
+      "loss": 0.9284,
+      "step": 12497
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9130632527322302,
+      "learning_rate": 1.6551332980071492e-05,
+      "loss": 0.9253,
+      "step": 12498
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7790652735219696,
+      "learning_rate": 1.6550756473322797e-05,
+      "loss": 0.9805,
+      "step": 12499
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6296442837207765,
+      "learning_rate": 1.6550179928433527e-05,
+      "loss": 0.9048,
+      "step": 12500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5263378089916553,
+      "learning_rate": 1.6549603345407035e-05,
+      "loss": 0.9409,
+      "step": 12501
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9080976329814432,
+      "learning_rate": 1.654902672424668e-05,
+      "loss": 0.7782,
+      "step": 12502
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7529917977600955,
+      "learning_rate": 1.6548450064955815e-05,
+      "loss": 1.0012,
+      "step": 12503
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6870131725506514,
+      "learning_rate": 1.6547873367537805e-05,
+      "loss": 0.9409,
+      "step": 12504
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.247440382270785,
+      "learning_rate": 1.6547296631996e-05,
+      "loss": 0.9357,
+      "step": 12505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8788022657819778,
+      "learning_rate": 1.654671985833376e-05,
+      "loss": 0.8547,
+      "step": 12506
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0360996940175484,
+      "learning_rate": 1.6546143046554447e-05,
+      "loss": 0.9315,
+      "step": 12507
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3235985727029784,
+      "learning_rate": 1.6545566196661416e-05,
+      "loss": 0.9278,
+      "step": 12508
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6786042111067023,
+      "learning_rate": 1.6544989308658025e-05,
+      "loss": 0.9233,
+      "step": 12509
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9301446014812815,
+      "learning_rate": 1.6544412382547638e-05,
+      "loss": 0.8459,
+      "step": 12510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6618030521261227,
+      "learning_rate": 1.6543835418333606e-05,
+      "loss": 0.8999,
+      "step": 12511
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6294495630373382,
+      "learning_rate": 1.6543258416019293e-05,
+      "loss": 0.9619,
+      "step": 12512
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6588089270757083,
+      "learning_rate": 1.654268137560806e-05,
+      "loss": 0.8613,
+      "step": 12513
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.811094355325948,
+      "learning_rate": 1.6542104297103264e-05,
+      "loss": 0.952,
+      "step": 12514
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7365096354576954,
+      "learning_rate": 1.6541527180508265e-05,
+      "loss": 0.842,
+      "step": 12515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.51651726708065,
+      "learning_rate": 1.6540950025826424e-05,
+      "loss": 0.8799,
+      "step": 12516
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5221692945682659,
+      "learning_rate": 1.65403728330611e-05,
+      "loss": 0.9864,
+      "step": 12517
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.748648599932897,
+      "learning_rate": 1.6539795602215655e-05,
+      "loss": 0.9995,
+      "step": 12518
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7739705612884613,
+      "learning_rate": 1.6539218333293453e-05,
+      "loss": 0.8739,
+      "step": 12519
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9873705064280107,
+      "learning_rate": 1.653864102629785e-05,
+      "loss": 1.0299,
+      "step": 12520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5695576945624037,
+      "learning_rate": 1.6538063681232203e-05,
+      "loss": 0.7908,
+      "step": 12521
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4765494360587106,
+      "learning_rate": 1.6537486298099884e-05,
+      "loss": 0.8338,
+      "step": 12522
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4663326703355097,
+      "learning_rate": 1.653690887690425e-05,
+      "loss": 0.9562,
+      "step": 12523
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6747036745128252,
+      "learning_rate": 1.6536331417648664e-05,
+      "loss": 0.8389,
+      "step": 12524
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.708344545004428,
+      "learning_rate": 1.653575392033649e-05,
+      "loss": 0.8454,
+      "step": 12525
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1320030750275034,
+      "learning_rate": 1.6535176384971082e-05,
+      "loss": 0.8267,
+      "step": 12526
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9035249248895303,
+      "learning_rate": 1.653459881155581e-05,
+      "loss": 0.7963,
+      "step": 12527
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4259848806676192,
+      "learning_rate": 1.6534021200094033e-05,
+      "loss": 0.9228,
+      "step": 12528
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5979963473241527,
+      "learning_rate": 1.6533443550589116e-05,
+      "loss": 0.9369,
+      "step": 12529
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7676112751624937,
+      "learning_rate": 1.6532865863044424e-05,
+      "loss": 0.9961,
+      "step": 12530
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7672487293133794,
+      "learning_rate": 1.653228813746332e-05,
+      "loss": 0.7104,
+      "step": 12531
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7910427687950412,
+      "learning_rate": 1.6531710373849166e-05,
+      "loss": 0.8218,
+      "step": 12532
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8487445167938519,
+      "learning_rate": 1.6531132572205324e-05,
+      "loss": 0.9736,
+      "step": 12533
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.842049873541206,
+      "learning_rate": 1.653055473253516e-05,
+      "loss": 0.7822,
+      "step": 12534
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.608236335151212,
+      "learning_rate": 1.6529976854842042e-05,
+      "loss": 0.8638,
+      "step": 12535
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8254009282332346,
+      "learning_rate": 1.6529398939129328e-05,
+      "loss": 0.747,
+      "step": 12536
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7132120282191745,
+      "learning_rate": 1.652882098540039e-05,
+      "loss": 0.8475,
+      "step": 12537
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0149758045679036,
+      "learning_rate": 1.6528242993658586e-05,
+      "loss": 1.0113,
+      "step": 12538
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5540198448210965,
+      "learning_rate": 1.6527664963907288e-05,
+      "loss": 0.8629,
+      "step": 12539
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8219331935239541,
+      "learning_rate": 1.652708689614985e-05,
+      "loss": 0.7456,
+      "step": 12540
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6228748050575457,
+      "learning_rate": 1.6526508790389655e-05,
+      "loss": 1.0844,
+      "step": 12541
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9434915986888157,
+      "learning_rate": 1.6525930646630057e-05,
+      "loss": 0.8276,
+      "step": 12542
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7995491715460996,
+      "learning_rate": 1.652535246487442e-05,
+      "loss": 0.8765,
+      "step": 12543
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9029525868721007,
+      "learning_rate": 1.652477424512612e-05,
+      "loss": 0.8164,
+      "step": 12544
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7212729142551657,
+      "learning_rate": 1.652419598738852e-05,
+      "loss": 0.7936,
+      "step": 12545
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8778267420170716,
+      "learning_rate": 1.6523617691664982e-05,
+      "loss": 0.9619,
+      "step": 12546
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9594262256867638,
+      "learning_rate": 1.6523039357958876e-05,
+      "loss": 0.8952,
+      "step": 12547
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.560099345519682,
+      "learning_rate": 1.652246098627357e-05,
+      "loss": 0.9795,
+      "step": 12548
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6529841107607404,
+      "learning_rate": 1.6521882576612433e-05,
+      "loss": 0.9619,
+      "step": 12549
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5778485149975539,
+      "learning_rate": 1.6521304128978826e-05,
+      "loss": 0.8486,
+      "step": 12550
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4746979874708153,
+      "learning_rate": 1.652072564337613e-05,
+      "loss": 0.8356,
+      "step": 12551
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7506077196866365,
+      "learning_rate": 1.6520147119807698e-05,
+      "loss": 0.7613,
+      "step": 12552
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7330598429960706,
+      "learning_rate": 1.6519568558276908e-05,
+      "loss": 0.9692,
+      "step": 12553
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2049713994956677,
+      "learning_rate": 1.6518989958787126e-05,
+      "loss": 0.9109,
+      "step": 12554
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.477308116399067,
+      "learning_rate": 1.651841132134172e-05,
+      "loss": 0.9728,
+      "step": 12555
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4901937655990958,
+      "learning_rate": 1.651783264594406e-05,
+      "loss": 0.94,
+      "step": 12556
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6089614842290154,
+      "learning_rate": 1.6517253932597515e-05,
+      "loss": 0.9986,
+      "step": 12557
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7785967250205361,
+      "learning_rate": 1.6516675181305456e-05,
+      "loss": 1.0527,
+      "step": 12558
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6480676072299012,
+      "learning_rate": 1.6516096392071248e-05,
+      "loss": 0.8221,
+      "step": 12559
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.696522689275679,
+      "learning_rate": 1.651551756489826e-05,
+      "loss": 0.9797,
+      "step": 12560
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5312379556709517,
+      "learning_rate": 1.6514938699789877e-05,
+      "loss": 0.8476,
+      "step": 12561
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5457454591095265,
+      "learning_rate": 1.651435979674945e-05,
+      "loss": 0.9108,
+      "step": 12562
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8078600808782608,
+      "learning_rate": 1.6513780855780358e-05,
+      "loss": 0.8888,
+      "step": 12563
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.537778832093164,
+      "learning_rate": 1.6513201876885972e-05,
+      "loss": 0.9319,
+      "step": 12564
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7699667131008336,
+      "learning_rate": 1.6512622860069667e-05,
+      "loss": 0.9455,
+      "step": 12565
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5631278613755497,
+      "learning_rate": 1.6512043805334808e-05,
+      "loss": 0.8483,
+      "step": 12566
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9314566129753745,
+      "learning_rate": 1.6511464712684764e-05,
+      "loss": 0.859,
+      "step": 12567
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6103633345369717,
+      "learning_rate": 1.6510885582122912e-05,
+      "loss": 0.938,
+      "step": 12568
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7473359523056573,
+      "learning_rate": 1.651030641365263e-05,
+      "loss": 0.9826,
+      "step": 12569
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.532606489897966,
+      "learning_rate": 1.6509727207277273e-05,
+      "loss": 0.9357,
+      "step": 12570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9894306807555242,
+      "learning_rate": 1.650914796300023e-05,
+      "loss": 0.9361,
+      "step": 12571
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7764814994435214,
+      "learning_rate": 1.6508568680824865e-05,
+      "loss": 0.7405,
+      "step": 12572
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6989534701594384,
+      "learning_rate": 1.6507989360754553e-05,
+      "loss": 0.8781,
+      "step": 12573
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8262812468301517,
+      "learning_rate": 1.6507410002792665e-05,
+      "loss": 0.8704,
+      "step": 12574
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8367153875102848,
+      "learning_rate": 1.6506830606942578e-05,
+      "loss": 0.8972,
+      "step": 12575
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5521889232743238,
+      "learning_rate": 1.650625117320766e-05,
+      "loss": 0.9879,
+      "step": 12576
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.002412500637353,
+      "learning_rate": 1.650567170159129e-05,
+      "loss": 0.908,
+      "step": 12577
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4456708113955319,
+      "learning_rate": 1.6505092192096837e-05,
+      "loss": 0.9974,
+      "step": 12578
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.48290098848779,
+      "learning_rate": 1.650451264472768e-05,
+      "loss": 0.96,
+      "step": 12579
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.620594576004967,
+      "learning_rate": 1.6503933059487195e-05,
+      "loss": 0.8476,
+      "step": 12580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.678172496173363,
+      "learning_rate": 1.6503353436378747e-05,
+      "loss": 0.844,
+      "step": 12581
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5390241777830653,
+      "learning_rate": 1.6502773775405717e-05,
+      "loss": 0.9521,
+      "step": 12582
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.482926168930758,
+      "learning_rate": 1.650219407657148e-05,
+      "loss": 0.9215,
+      "step": 12583
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8229668409780964,
+      "learning_rate": 1.6501614339879413e-05,
+      "loss": 0.9805,
+      "step": 12584
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.375797747057036,
+      "learning_rate": 1.650103456533289e-05,
+      "loss": 0.968,
+      "step": 12585
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.773534233286462,
+      "learning_rate": 1.650045475293528e-05,
+      "loss": 0.7618,
+      "step": 12586
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6628740936326065,
+      "learning_rate": 1.6499874902689968e-05,
+      "loss": 0.9221,
+      "step": 12587
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5456194442312237,
+      "learning_rate": 1.6499295014600327e-05,
+      "loss": 0.973,
+      "step": 12588
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.578514060623624,
+      "learning_rate": 1.649871508866973e-05,
+      "loss": 0.8677,
+      "step": 12589
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6732856260728168,
+      "learning_rate": 1.6498135124901557e-05,
+      "loss": 0.955,
+      "step": 12590
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6734210713084436,
+      "learning_rate": 1.6497555123299184e-05,
+      "loss": 0.8454,
+      "step": 12591
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.284670486462056,
+      "learning_rate": 1.649697508386599e-05,
+      "loss": 0.8612,
+      "step": 12592
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4748837739514487,
+      "learning_rate": 1.6496395006605346e-05,
+      "loss": 0.9153,
+      "step": 12593
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6935193640113249,
+      "learning_rate": 1.6495814891520636e-05,
+      "loss": 0.8072,
+      "step": 12594
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6453769648793497,
+      "learning_rate": 1.6495234738615238e-05,
+      "loss": 0.9928,
+      "step": 12595
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6655270804655515,
+      "learning_rate": 1.6494654547892522e-05,
+      "loss": 0.8069,
+      "step": 12596
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7332860652671882,
+      "learning_rate": 1.6494074319355874e-05,
+      "loss": 0.9597,
+      "step": 12597
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5584716293407859,
+      "learning_rate": 1.649349405300867e-05,
+      "loss": 0.8893,
+      "step": 12598
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3465580251644056,
+      "learning_rate": 1.6492913748854285e-05,
+      "loss": 1.1,
+      "step": 12599
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6119867374297783,
+      "learning_rate": 1.6492333406896103e-05,
+      "loss": 0.8688,
+      "step": 12600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.489032656506618,
+      "learning_rate": 1.6491753027137497e-05,
+      "loss": 0.9054,
+      "step": 12601
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9589069784146453,
+      "learning_rate": 1.6491172609581856e-05,
+      "loss": 1.0995,
+      "step": 12602
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0324835674921222,
+      "learning_rate": 1.649059215423255e-05,
+      "loss": 0.8397,
+      "step": 12603
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7335179065112396,
+      "learning_rate": 1.649001166109296e-05,
+      "loss": 1.022,
+      "step": 12604
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.019547009516114,
+      "learning_rate": 1.648943113016647e-05,
+      "loss": 1.0115,
+      "step": 12605
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7692996423771428,
+      "learning_rate": 1.648885056145645e-05,
+      "loss": 0.8341,
+      "step": 12606
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5638282863570336,
+      "learning_rate": 1.6488269954966295e-05,
+      "loss": 0.9358,
+      "step": 12607
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6987614941833937,
+      "learning_rate": 1.6487689310699377e-05,
+      "loss": 0.8142,
+      "step": 12608
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6718518661005357,
+      "learning_rate": 1.648710862865908e-05,
+      "loss": 0.9252,
+      "step": 12609
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2752145251849663,
+      "learning_rate": 1.648652790884878e-05,
+      "loss": 0.9144,
+      "step": 12610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.676734374484602,
+      "learning_rate": 1.648594715127186e-05,
+      "loss": 0.8925,
+      "step": 12611
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.805887951984885,
+      "learning_rate": 1.64853663559317e-05,
+      "loss": 0.7908,
+      "step": 12612
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6000145221081203,
+      "learning_rate": 1.648478552283169e-05,
+      "loss": 0.8359,
+      "step": 12613
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4498515764896736,
+      "learning_rate": 1.6484204651975203e-05,
+      "loss": 0.8294,
+      "step": 12614
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7757579340878209,
+      "learning_rate": 1.6483623743365624e-05,
+      "loss": 0.7278,
+      "step": 12615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6964654103518686,
+      "learning_rate": 1.6483042797006334e-05,
+      "loss": 0.904,
+      "step": 12616
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5871429823123733,
+      "learning_rate": 1.6482461812900718e-05,
+      "loss": 0.9828,
+      "step": 12617
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.698641223785912,
+      "learning_rate": 1.6481880791052158e-05,
+      "loss": 0.9113,
+      "step": 12618
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.608829494089598,
+      "learning_rate": 1.648129973146403e-05,
+      "loss": 0.9651,
+      "step": 12619
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4962423522506487,
+      "learning_rate": 1.6480718634139726e-05,
+      "loss": 0.9535,
+      "step": 12620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7809019125030477,
+      "learning_rate": 1.648013749908263e-05,
+      "loss": 0.7884,
+      "step": 12621
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7282543494065922,
+      "learning_rate": 1.6479556326296116e-05,
+      "loss": 0.7103,
+      "step": 12622
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8772168566989658,
+      "learning_rate": 1.6478975115783577e-05,
+      "loss": 0.7358,
+      "step": 12623
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8237221390975689,
+      "learning_rate": 1.6478393867548396e-05,
+      "loss": 0.8515,
+      "step": 12624
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7560606639060015,
+      "learning_rate": 1.647781258159395e-05,
+      "loss": 0.7589,
+      "step": 12625
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7991030014289275,
+      "learning_rate": 1.647723125792363e-05,
+      "loss": 0.9104,
+      "step": 12626
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.531969615875157,
+      "learning_rate": 1.647664989654082e-05,
+      "loss": 0.8871,
+      "step": 12627
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.920777077458059,
+      "learning_rate": 1.6476068497448903e-05,
+      "loss": 0.908,
+      "step": 12628
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8083738446204989,
+      "learning_rate": 1.6475487060651265e-05,
+      "loss": 1.0061,
+      "step": 12629
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8031364758810078,
+      "learning_rate": 1.6474905586151293e-05,
+      "loss": 0.8411,
+      "step": 12630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8308589104893846,
+      "learning_rate": 1.647432407395237e-05,
+      "loss": 0.7637,
+      "step": 12631
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5386864504486362,
+      "learning_rate": 1.647374252405788e-05,
+      "loss": 0.9511,
+      "step": 12632
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0085028278759407,
+      "learning_rate": 1.6473160936471214e-05,
+      "loss": 0.9686,
+      "step": 12633
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5815727365089918,
+      "learning_rate": 1.6472579311195752e-05,
+      "loss": 0.8693,
+      "step": 12634
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7678188926959285,
+      "learning_rate": 1.6471997648234886e-05,
+      "loss": 0.7166,
+      "step": 12635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4337333690859966,
+      "learning_rate": 1.6471415947592e-05,
+      "loss": 0.9194,
+      "step": 12636
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8603606016710408,
+      "learning_rate": 1.647083420927049e-05,
+      "loss": 1.0016,
+      "step": 12637
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1093450308752546,
+      "learning_rate": 1.6470252433273723e-05,
+      "loss": 0.8929,
+      "step": 12638
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6928876882928887,
+      "learning_rate": 1.6469670619605103e-05,
+      "loss": 0.9385,
+      "step": 12639
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6208816322578374,
+      "learning_rate": 1.6469088768268013e-05,
+      "loss": 0.8765,
+      "step": 12640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6696356577939484,
+      "learning_rate": 1.646850687926584e-05,
+      "loss": 1.0485,
+      "step": 12641
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.693680336904277,
+      "learning_rate": 1.646792495260197e-05,
+      "loss": 0.9364,
+      "step": 12642
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6251287475949632,
+      "learning_rate": 1.6467342988279793e-05,
+      "loss": 0.9739,
+      "step": 12643
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5136148912332594,
+      "learning_rate": 1.64667609863027e-05,
+      "loss": 0.7858,
+      "step": 12644
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4852397715649488,
+      "learning_rate": 1.646617894667408e-05,
+      "loss": 0.9132,
+      "step": 12645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7064558056713621,
+      "learning_rate": 1.646559686939731e-05,
+      "loss": 0.9024,
+      "step": 12646
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8008319928299041,
+      "learning_rate": 1.6465014754475794e-05,
+      "loss": 0.8574,
+      "step": 12647
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9148908250567094,
+      "learning_rate": 1.6464432601912914e-05,
+      "loss": 0.7542,
+      "step": 12648
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7877093853688848,
+      "learning_rate": 1.6463850411712063e-05,
+      "loss": 0.7816,
+      "step": 12649
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5941163250987163,
+      "learning_rate": 1.6463268183876627e-05,
+      "loss": 0.8872,
+      "step": 12650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6415394739224567,
+      "learning_rate": 1.6462685918409998e-05,
+      "loss": 0.8351,
+      "step": 12651
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5747664879508252,
+      "learning_rate": 1.6462103615315563e-05,
+      "loss": 0.9182,
+      "step": 12652
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6277903142359176,
+      "learning_rate": 1.6461521274596717e-05,
+      "loss": 0.9527,
+      "step": 12653
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5102131536788581,
+      "learning_rate": 1.6460938896256848e-05,
+      "loss": 0.9492,
+      "step": 12654
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5191634664186373,
+      "learning_rate": 1.6460356480299345e-05,
+      "loss": 0.995,
+      "step": 12655
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7696793383542471,
+      "learning_rate": 1.6459774026727606e-05,
+      "loss": 0.6877,
+      "step": 12656
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.795785709521723,
+      "learning_rate": 1.6459191535545013e-05,
+      "loss": 0.9261,
+      "step": 12657
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7429201727839025,
+      "learning_rate": 1.6458609006754966e-05,
+      "loss": 0.9294,
+      "step": 12658
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5246587693719436,
+      "learning_rate": 1.6458026440360847e-05,
+      "loss": 0.8276,
+      "step": 12659
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8831545517337924,
+      "learning_rate": 1.6457443836366062e-05,
+      "loss": 0.9342,
+      "step": 12660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.028977901533004,
+      "learning_rate": 1.6456861194773988e-05,
+      "loss": 0.9191,
+      "step": 12661
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7123814049708626,
+      "learning_rate": 1.6456278515588023e-05,
+      "loss": 1.1107,
+      "step": 12662
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.688230486666659,
+      "learning_rate": 1.6455695798811567e-05,
+      "loss": 0.9187,
+      "step": 12663
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5357223650594698,
+      "learning_rate": 1.6455113044448e-05,
+      "loss": 0.8262,
+      "step": 12664
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7430090746345017,
+      "learning_rate": 1.6454530252500723e-05,
+      "loss": 0.9917,
+      "step": 12665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8130964902728026,
+      "learning_rate": 1.645394742297313e-05,
+      "loss": 0.7649,
+      "step": 12666
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7867050960757254,
+      "learning_rate": 1.645336455586861e-05,
+      "loss": 0.8309,
+      "step": 12667
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0556840779537464,
+      "learning_rate": 1.645278165119056e-05,
+      "loss": 0.9671,
+      "step": 12668
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6056737861487071,
+      "learning_rate": 1.6452198708942368e-05,
+      "loss": 0.789,
+      "step": 12669
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.488456605138759,
+      "learning_rate": 1.6451615729127436e-05,
+      "loss": 0.8822,
+      "step": 12670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7678260573807504,
+      "learning_rate": 1.6451032711749154e-05,
+      "loss": 0.8002,
+      "step": 12671
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6751920412351775,
+      "learning_rate": 1.6450449656810914e-05,
+      "loss": 0.8484,
+      "step": 12672
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8097411481695852,
+      "learning_rate": 1.6449866564316118e-05,
+      "loss": 0.7807,
+      "step": 12673
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7064943694151342,
+      "learning_rate": 1.6449283434268153e-05,
+      "loss": 0.8583,
+      "step": 12674
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.760467256484119,
+      "learning_rate": 1.644870026667042e-05,
+      "loss": 0.852,
+      "step": 12675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7658560232917008,
+      "learning_rate": 1.6448117061526314e-05,
+      "loss": 0.7527,
+      "step": 12676
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.782717565970901,
+      "learning_rate": 1.644753381883923e-05,
+      "loss": 0.7611,
+      "step": 12677
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5667981584601531,
+      "learning_rate": 1.644695053861256e-05,
+      "loss": 0.9934,
+      "step": 12678
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.97830122258075,
+      "learning_rate": 1.6446367220849704e-05,
+      "loss": 0.9668,
+      "step": 12679
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.780568455867297,
+      "learning_rate": 1.6445783865554056e-05,
+      "loss": 0.9676,
+      "step": 12680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7771271967027271,
+      "learning_rate": 1.6445200472729016e-05,
+      "loss": 0.9311,
+      "step": 12681
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5602813517826317,
+      "learning_rate": 1.6444617042377978e-05,
+      "loss": 0.946,
+      "step": 12682
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.093228506700597,
+      "learning_rate": 1.6444033574504336e-05,
+      "loss": 0.8941,
+      "step": 12683
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6915626066110467,
+      "learning_rate": 1.644345006911149e-05,
+      "loss": 0.9392,
+      "step": 12684
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6571753988322875,
+      "learning_rate": 1.6442866526202842e-05,
+      "loss": 0.9555,
+      "step": 12685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.685433137903053,
+      "learning_rate": 1.6442282945781784e-05,
+      "loss": 0.8113,
+      "step": 12686
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8447062326761756,
+      "learning_rate": 1.6441699327851715e-05,
+      "loss": 0.7916,
+      "step": 12687
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6628357340450541,
+      "learning_rate": 1.6441115672416033e-05,
+      "loss": 1.0741,
+      "step": 12688
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5766627428384168,
+      "learning_rate": 1.6440531979478137e-05,
+      "loss": 0.7886,
+      "step": 12689
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6210766683678173,
+      "learning_rate": 1.643994824904142e-05,
+      "loss": 0.9164,
+      "step": 12690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9744139926589637,
+      "learning_rate": 1.6439364481109292e-05,
+      "loss": 0.9011,
+      "step": 12691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.996925061132911,
+      "learning_rate": 1.643878067568514e-05,
+      "loss": 1.043,
+      "step": 12692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5895585130935703,
+      "learning_rate": 1.6438196832772372e-05,
+      "loss": 0.9251,
+      "step": 12693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.424481169725021,
+      "learning_rate": 1.6437612952374382e-05,
+      "loss": 0.8621,
+      "step": 12694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0126402385441184,
+      "learning_rate": 1.6437029034494574e-05,
+      "loss": 0.8743,
+      "step": 12695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9096843275412196,
+      "learning_rate": 1.6436445079136344e-05,
+      "loss": 0.8556,
+      "step": 12696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.759608672405273,
+      "learning_rate": 1.6435861086303092e-05,
+      "loss": 0.917,
+      "step": 12697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6968162855646993,
+      "learning_rate": 1.643527705599822e-05,
+      "loss": 0.8719,
+      "step": 12698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6792118218179983,
+      "learning_rate": 1.6434692988225125e-05,
+      "loss": 0.8493,
+      "step": 12699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5440364109690443,
+      "learning_rate": 1.6434108882987213e-05,
+      "loss": 1.0517,
+      "step": 12700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4017470779538326,
+      "learning_rate": 1.643352474028788e-05,
+      "loss": 0.9034,
+      "step": 12701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5400722094314672,
+      "learning_rate": 1.643294056013053e-05,
+      "loss": 1.0233,
+      "step": 12702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5687194154462332,
+      "learning_rate": 1.6432356342518564e-05,
+      "loss": 0.9057,
+      "step": 12703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.569719786307193,
+      "learning_rate": 1.6431772087455384e-05,
+      "loss": 0.9729,
+      "step": 12704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.76935453261447,
+      "learning_rate": 1.643118779494439e-05,
+      "loss": 0.9593,
+      "step": 12705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8839346862827882,
+      "learning_rate": 1.6430603464988982e-05,
+      "loss": 0.8295,
+      "step": 12706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.21114189786239,
+      "learning_rate": 1.6430019097592567e-05,
+      "loss": 0.844,
+      "step": 12707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5444527620174695,
+      "learning_rate": 1.6429434692758542e-05,
+      "loss": 0.9894,
+      "step": 12708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4039737823260006,
+      "learning_rate": 1.6428850250490317e-05,
+      "loss": 0.8089,
+      "step": 12709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8634149792178707,
+      "learning_rate": 1.642826577079129e-05,
+      "loss": 0.9,
+      "step": 12710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6145665561649727,
+      "learning_rate": 1.6427681253664863e-05,
+      "loss": 0.8977,
+      "step": 12711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5499514590911352,
+      "learning_rate": 1.642709669911444e-05,
+      "loss": 0.7554,
+      "step": 12712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6144036989312172,
+      "learning_rate": 1.642651210714343e-05,
+      "loss": 0.8531,
+      "step": 12713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5993323669206696,
+      "learning_rate": 1.6425927477755227e-05,
+      "loss": 0.8204,
+      "step": 12714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8693888430079513,
+      "learning_rate": 1.642534281095324e-05,
+      "loss": 0.891,
+      "step": 12715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6334175257912775,
+      "learning_rate": 1.6424758106740874e-05,
+      "loss": 0.9687,
+      "step": 12716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6792998689480663,
+      "learning_rate": 1.642417336512153e-05,
+      "loss": 0.914,
+      "step": 12717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8357341921735465,
+      "learning_rate": 1.642358858609862e-05,
+      "loss": 0.6853,
+      "step": 12718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6567062936070427,
+      "learning_rate": 1.642300376967554e-05,
+      "loss": 0.9678,
+      "step": 12719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.88484072829485,
+      "learning_rate": 1.64224189158557e-05,
+      "loss": 0.8866,
+      "step": 12720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8379271309728884,
+      "learning_rate": 1.64218340246425e-05,
+      "loss": 0.6944,
+      "step": 12721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.737613117894055,
+      "learning_rate": 1.6421249096039352e-05,
+      "loss": 0.8656,
+      "step": 12722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.042342393293249,
+      "learning_rate": 1.6420664130049663e-05,
+      "loss": 0.927,
+      "step": 12723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8502158450404118,
+      "learning_rate": 1.6420079126676828e-05,
+      "loss": 0.7239,
+      "step": 12724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.682705686916955,
+      "learning_rate": 1.6419494085924267e-05,
+      "loss": 0.8467,
+      "step": 12725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8330676801048547,
+      "learning_rate": 1.6418909007795374e-05,
+      "loss": 0.8018,
+      "step": 12726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7127942084602337,
+      "learning_rate": 1.641832389229356e-05,
+      "loss": 0.9653,
+      "step": 12727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8644343927482594,
+      "learning_rate": 1.6417738739422234e-05,
+      "loss": 0.9624,
+      "step": 12728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.529511881282163,
+      "learning_rate": 1.64171535491848e-05,
+      "loss": 0.8881,
+      "step": 12729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5762863326828658,
+      "learning_rate": 1.6416568321584668e-05,
+      "loss": 0.8619,
+      "step": 12730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7382667898188722,
+      "learning_rate": 1.641598305662524e-05,
+      "loss": 0.8002,
+      "step": 12731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6658611930736191,
+      "learning_rate": 1.641539775430993e-05,
+      "loss": 0.8743,
+      "step": 12732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6293978253812245,
+      "learning_rate": 1.6414812414642144e-05,
+      "loss": 0.8947,
+      "step": 12733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7789797781612189,
+      "learning_rate": 1.6414227037625288e-05,
+      "loss": 0.747,
+      "step": 12734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8322975975529627,
+      "learning_rate": 1.6413641623262774e-05,
+      "loss": 0.9139,
+      "step": 12735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.57625825619699,
+      "learning_rate": 1.6413056171558006e-05,
+      "loss": 0.9247,
+      "step": 12736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7992217973061835,
+      "learning_rate": 1.6412470682514396e-05,
+      "loss": 0.8824,
+      "step": 12737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0764007816031356,
+      "learning_rate": 1.641188515613535e-05,
+      "loss": 0.9008,
+      "step": 12738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6586252317375287,
+      "learning_rate": 1.6411299592424276e-05,
+      "loss": 0.9018,
+      "step": 12739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6683414202594598,
+      "learning_rate": 1.641071399138459e-05,
+      "loss": 0.8913,
+      "step": 12740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8504437700231267,
+      "learning_rate": 1.6410128353019697e-05,
+      "loss": 0.8791,
+      "step": 12741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5386448314767698,
+      "learning_rate": 1.6409542677333007e-05,
+      "loss": 0.8776,
+      "step": 12742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.508588370016059,
+      "learning_rate": 1.640895696432793e-05,
+      "loss": 0.8646,
+      "step": 12743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9756674666041365,
+      "learning_rate": 1.6408371214007876e-05,
+      "loss": 0.9563,
+      "step": 12744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7034440610675692,
+      "learning_rate": 1.6407785426376258e-05,
+      "loss": 0.862,
+      "step": 12745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.888878643316328,
+      "learning_rate": 1.640719960143648e-05,
+      "loss": 0.9109,
+      "step": 12746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5005490916323536,
+      "learning_rate": 1.6406613739191964e-05,
+      "loss": 0.9555,
+      "step": 12747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7228457151440122,
+      "learning_rate": 1.640602783964611e-05,
+      "loss": 0.9245,
+      "step": 12748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6648532718504356,
+      "learning_rate": 1.6405441902802334e-05,
+      "loss": 0.9558,
+      "step": 12749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.43347695523253,
+      "learning_rate": 1.640485592866405e-05,
+      "loss": 0.8842,
+      "step": 12750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8202044253312699,
+      "learning_rate": 1.6404269917234662e-05,
+      "loss": 0.7318,
+      "step": 12751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7964682224875849,
+      "learning_rate": 1.6403683868517593e-05,
+      "loss": 0.7877,
+      "step": 12752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5232583460630635,
+      "learning_rate": 1.6403097782516244e-05,
+      "loss": 0.9892,
+      "step": 12753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2212407953884,
+      "learning_rate": 1.6402511659234038e-05,
+      "loss": 0.8195,
+      "step": 12754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4911702468797055,
+      "learning_rate": 1.640192549867438e-05,
+      "loss": 0.8245,
+      "step": 12755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7084085511414875,
+      "learning_rate": 1.6401339300840683e-05,
+      "loss": 0.8746,
+      "step": 12756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.077659962862526,
+      "learning_rate": 1.640075306573636e-05,
+      "loss": 0.8987,
+      "step": 12757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8404539065833108,
+      "learning_rate": 1.6400166793364828e-05,
+      "loss": 0.9309,
+      "step": 12758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6439193221689645,
+      "learning_rate": 1.63995804837295e-05,
+      "loss": 1.0063,
+      "step": 12759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6877373694616915,
+      "learning_rate": 1.639899413683379e-05,
+      "loss": 0.8167,
+      "step": 12760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2790556758990355,
+      "learning_rate": 1.6398407752681108e-05,
+      "loss": 0.9474,
+      "step": 12761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5857153891904505,
+      "learning_rate": 1.6397821331274872e-05,
+      "loss": 0.9777,
+      "step": 12762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.026584276325081,
+      "learning_rate": 1.639723487261849e-05,
+      "loss": 0.8778,
+      "step": 12763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6155990207507454,
+      "learning_rate": 1.639664837671539e-05,
+      "loss": 0.9319,
+      "step": 12764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6258179487191011,
+      "learning_rate": 1.6396061843568973e-05,
+      "loss": 0.9787,
+      "step": 12765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8819588703379448,
+      "learning_rate": 1.6395475273182655e-05,
+      "loss": 1.0489,
+      "step": 12766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9441999878145146,
+      "learning_rate": 1.6394888665559857e-05,
+      "loss": 0.8956,
+      "step": 12767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5818106572754553,
+      "learning_rate": 1.6394302020703995e-05,
+      "loss": 0.8288,
+      "step": 12768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.664173362905175,
+      "learning_rate": 1.6393715338618483e-05,
+      "loss": 0.9504,
+      "step": 12769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.259188030034506,
+      "learning_rate": 1.6393128619306734e-05,
+      "loss": 0.98,
+      "step": 12770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0846159745380124,
+      "learning_rate": 1.6392541862772166e-05,
+      "loss": 0.8207,
+      "step": 12771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7475679500050082,
+      "learning_rate": 1.6391955069018198e-05,
+      "loss": 0.7697,
+      "step": 12772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6405390293474778,
+      "learning_rate": 1.639136823804824e-05,
+      "loss": 0.8485,
+      "step": 12773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7381125066523966,
+      "learning_rate": 1.6390781369865715e-05,
+      "loss": 0.9102,
+      "step": 12774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6903563830096886,
+      "learning_rate": 1.6390194464474038e-05,
+      "loss": 0.8877,
+      "step": 12775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7621563662267818,
+      "learning_rate": 1.6389607521876625e-05,
+      "loss": 0.8411,
+      "step": 12776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.336151651316932,
+      "learning_rate": 1.6389020542076898e-05,
+      "loss": 0.8571,
+      "step": 12777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.733982169421019,
+      "learning_rate": 1.6388433525078266e-05,
+      "loss": 0.7526,
+      "step": 12778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6084483973228527,
+      "learning_rate": 1.6387846470884154e-05,
+      "loss": 0.9251,
+      "step": 12779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.267468156230357,
+      "learning_rate": 1.6387259379497973e-05,
+      "loss": 0.8746,
+      "step": 12780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5346877582178646,
+      "learning_rate": 1.638667225092315e-05,
+      "loss": 0.9921,
+      "step": 12781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9820419447782867,
+      "learning_rate": 1.6386085085163096e-05,
+      "loss": 0.9032,
+      "step": 12782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1260578233788805,
+      "learning_rate": 1.6385497882221236e-05,
+      "loss": 0.858,
+      "step": 12783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6207722688557575,
+      "learning_rate": 1.6384910642100983e-05,
+      "loss": 0.785,
+      "step": 12784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7783645155862117,
+      "learning_rate": 1.6384323364805762e-05,
+      "loss": 0.7437,
+      "step": 12785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9101246522834352,
+      "learning_rate": 1.6383736050338984e-05,
+      "loss": 0.9141,
+      "step": 12786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7193441073686582,
+      "learning_rate": 1.638314869870408e-05,
+      "loss": 0.9042,
+      "step": 12787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6523179256983478,
+      "learning_rate": 1.6382561309904457e-05,
+      "loss": 0.9753,
+      "step": 12788
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0728348658474554,
+      "learning_rate": 1.6381973883943546e-05,
+      "loss": 0.9171,
+      "step": 12789
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6935259740927417,
+      "learning_rate": 1.6381386420824763e-05,
+      "loss": 0.8849,
+      "step": 12790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7515782944511504,
+      "learning_rate": 1.6380798920551526e-05,
+      "loss": 0.8309,
+      "step": 12791
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8296490302208404,
+      "learning_rate": 1.6380211383127255e-05,
+      "loss": 0.9221,
+      "step": 12792
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8689823324019477,
+      "learning_rate": 1.637962380855538e-05,
+      "loss": 0.8066,
+      "step": 12793
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7376657165815874,
+      "learning_rate": 1.637903619683931e-05,
+      "loss": 0.806,
+      "step": 12794
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4810757242158656,
+      "learning_rate": 1.6378448547982473e-05,
+      "loss": 0.8234,
+      "step": 12795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.778269412678106,
+      "learning_rate": 1.6377860861988288e-05,
+      "loss": 1.0414,
+      "step": 12796
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5736207290555737,
+      "learning_rate": 1.6377273138860182e-05,
+      "loss": 1.0684,
+      "step": 12797
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9108990429484334,
+      "learning_rate": 1.637668537860157e-05,
+      "loss": 0.9255,
+      "step": 12798
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.910671810530706,
+      "learning_rate": 1.6376097581215874e-05,
+      "loss": 0.9441,
+      "step": 12799
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9657321679272581,
+      "learning_rate": 1.6375509746706523e-05,
+      "loss": 0.9048,
+      "step": 12800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.610216791138144,
+      "learning_rate": 1.6374921875076933e-05,
+      "loss": 0.8677,
+      "step": 12801
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6751099849359652,
+      "learning_rate": 1.6374333966330535e-05,
+      "loss": 0.8704,
+      "step": 12802
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5602707083802219,
+      "learning_rate": 1.637374602047074e-05,
+      "loss": 0.9089,
+      "step": 12803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8335633389215515,
+      "learning_rate": 1.6373158037500984e-05,
+      "loss": 0.9157,
+      "step": 12804
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5000245657124602,
+      "learning_rate": 1.6372570017424682e-05,
+      "loss": 0.9369,
+      "step": 12805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6499295580643996,
+      "learning_rate": 1.6371981960245257e-05,
+      "loss": 1.0529,
+      "step": 12806
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8624058394247289,
+      "learning_rate": 1.6371393865966137e-05,
+      "loss": 0.7886,
+      "step": 12807
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5372443571020245,
+      "learning_rate": 1.6370805734590746e-05,
+      "loss": 0.9216,
+      "step": 12808
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1620474209799796,
+      "learning_rate": 1.6370217566122505e-05,
+      "loss": 0.9007,
+      "step": 12809
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6673107881211873,
+      "learning_rate": 1.6369629360564845e-05,
+      "loss": 0.8178,
+      "step": 12810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7554407649081292,
+      "learning_rate": 1.6369041117921183e-05,
+      "loss": 0.8546,
+      "step": 12811
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5235032446824275,
+      "learning_rate": 1.6368452838194947e-05,
+      "loss": 0.8848,
+      "step": 12812
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6472053406934561,
+      "learning_rate": 1.6367864521389563e-05,
+      "loss": 1.0285,
+      "step": 12813
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.949421187141664,
+      "learning_rate": 1.6367276167508456e-05,
+      "loss": 0.9809,
+      "step": 12814
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5509372470459675,
+      "learning_rate": 1.636668777655505e-05,
+      "loss": 0.7733,
+      "step": 12815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.64881510969619,
+      "learning_rate": 1.6366099348532777e-05,
+      "loss": 0.8832,
+      "step": 12816
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.841983135880127,
+      "learning_rate": 1.636551088344505e-05,
+      "loss": 0.8983,
+      "step": 12817
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.740249231009645,
+      "learning_rate": 1.636492238129531e-05,
+      "loss": 0.8038,
+      "step": 12818
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6934644227625244,
+      "learning_rate": 1.636433384208697e-05,
+      "loss": 0.8878,
+      "step": 12819
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.698168868468206,
+      "learning_rate": 1.636374526582347e-05,
+      "loss": 0.8546,
+      "step": 12820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6613468267812106,
+      "learning_rate": 1.636315665250823e-05,
+      "loss": 0.9891,
+      "step": 12821
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6448672729404379,
+      "learning_rate": 1.636256800214467e-05,
+      "loss": 0.8712,
+      "step": 12822
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0135220894896446,
+      "learning_rate": 1.6361979314736233e-05,
+      "loss": 0.947,
+      "step": 12823
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6953336395000993,
+      "learning_rate": 1.6361390590286336e-05,
+      "loss": 0.8827,
+      "step": 12824
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5677549064415652,
+      "learning_rate": 1.636080182879841e-05,
+      "loss": 0.9679,
+      "step": 12825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.042751117946678,
+      "learning_rate": 1.636021303027588e-05,
+      "loss": 0.8739,
+      "step": 12826
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7536117271842493,
+      "learning_rate": 1.6359624194722176e-05,
+      "loss": 0.9807,
+      "step": 12827
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8229931298643394,
+      "learning_rate": 1.6359035322140727e-05,
+      "loss": 0.8609,
+      "step": 12828
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.919245380907971,
+      "learning_rate": 1.635844641253496e-05,
+      "loss": 0.7683,
+      "step": 12829
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7950669289109467,
+      "learning_rate": 1.6357857465908308e-05,
+      "loss": 0.8458,
+      "step": 12830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8178132066961037,
+      "learning_rate": 1.6357268482264195e-05,
+      "loss": 0.9011,
+      "step": 12831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.079059637964745,
+      "learning_rate": 1.6356679461606052e-05,
+      "loss": 0.8731,
+      "step": 12832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.595108493855602,
+      "learning_rate": 1.6356090403937307e-05,
+      "loss": 0.9242,
+      "step": 12833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8360354646901407,
+      "learning_rate": 1.6355501309261398e-05,
+      "loss": 0.8362,
+      "step": 12834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5869720938415608,
+      "learning_rate": 1.6354912177581743e-05,
+      "loss": 0.9642,
+      "step": 12835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5456547843883832,
+      "learning_rate": 1.6354323008901774e-05,
+      "loss": 0.9813,
+      "step": 12836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6128558624783782,
+      "learning_rate": 1.635373380322493e-05,
+      "loss": 0.9588,
+      "step": 12837
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7981551726050766,
+      "learning_rate": 1.6353144560554636e-05,
+      "loss": 0.9416,
+      "step": 12838
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6269641581901961,
+      "learning_rate": 1.6352555280894323e-05,
+      "loss": 0.8761,
+      "step": 12839
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7433884730829896,
+      "learning_rate": 1.635196596424742e-05,
+      "loss": 0.9443,
+      "step": 12840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3584789643948354,
+      "learning_rate": 1.635137661061736e-05,
+      "loss": 0.8927,
+      "step": 12841
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.724981428692839,
+      "learning_rate": 1.6350787220007576e-05,
+      "loss": 1.0974,
+      "step": 12842
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5657386205000088,
+      "learning_rate": 1.6350197792421497e-05,
+      "loss": 1.0133,
+      "step": 12843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.317910519922149,
+      "learning_rate": 1.634960832786256e-05,
+      "loss": 0.848,
+      "step": 12844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7909962644017927,
+      "learning_rate": 1.6349018826334188e-05,
+      "loss": 0.7373,
+      "step": 12845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9651299643742872,
+      "learning_rate": 1.634842928783982e-05,
+      "loss": 0.9403,
+      "step": 12846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6947510021954937,
+      "learning_rate": 1.6347839712382887e-05,
+      "loss": 0.8739,
+      "step": 12847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7187272701433451,
+      "learning_rate": 1.6347250099966823e-05,
+      "loss": 0.9269,
+      "step": 12848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3871849257123934,
+      "learning_rate": 1.6346660450595058e-05,
+      "loss": 0.9584,
+      "step": 12849
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.911027888206191,
+      "learning_rate": 1.6346070764271025e-05,
+      "loss": 0.9906,
+      "step": 12850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.186985607794059,
+      "learning_rate": 1.6345481040998163e-05,
+      "loss": 0.8016,
+      "step": 12851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.534522710267964,
+      "learning_rate": 1.6344891280779894e-05,
+      "loss": 0.9244,
+      "step": 12852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6225453606756697,
+      "learning_rate": 1.6344301483619666e-05,
+      "loss": 0.8659,
+      "step": 12853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.840491320614252,
+      "learning_rate": 1.6343711649520904e-05,
+      "loss": 0.7579,
+      "step": 12854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6690989123560658,
+      "learning_rate": 1.6343121778487047e-05,
+      "loss": 0.8981,
+      "step": 12855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5985077893083122,
+      "learning_rate": 1.6342531870521523e-05,
+      "loss": 0.9555,
+      "step": 12856
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5082371645212937,
+      "learning_rate": 1.634194192562777e-05,
+      "loss": 0.9175,
+      "step": 12857
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8475202780756008,
+      "learning_rate": 1.6341351943809228e-05,
+      "loss": 0.7134,
+      "step": 12858
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.704501516017322,
+      "learning_rate": 1.634076192506932e-05,
+      "loss": 0.9489,
+      "step": 12859
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7112380193121852,
+      "learning_rate": 1.6340171869411497e-05,
+      "loss": 0.967,
+      "step": 12860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7713694423283044,
+      "learning_rate": 1.633958177683918e-05,
+      "loss": 0.9708,
+      "step": 12861
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4648417418585364,
+      "learning_rate": 1.6338991647355812e-05,
+      "loss": 0.9664,
+      "step": 12862
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9736975364056035,
+      "learning_rate": 1.6338401480964832e-05,
+      "loss": 0.9451,
+      "step": 12863
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.822081420136393,
+      "learning_rate": 1.633781127766967e-05,
+      "loss": 1.0876,
+      "step": 12864
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9151057348792364,
+      "learning_rate": 1.633722103747376e-05,
+      "loss": 0.916,
+      "step": 12865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2177770725536035,
+      "learning_rate": 1.6336630760380546e-05,
+      "loss": 0.8675,
+      "step": 12866
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6706136364659676,
+      "learning_rate": 1.633604044639346e-05,
+      "loss": 1.0051,
+      "step": 12867
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7964623265337485,
+      "learning_rate": 1.6335450095515944e-05,
+      "loss": 0.9459,
+      "step": 12868
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5890439067080004,
+      "learning_rate": 1.633485970775143e-05,
+      "loss": 1.0121,
+      "step": 12869
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.905585896576535,
+      "learning_rate": 1.6334269283103353e-05,
+      "loss": 0.8586,
+      "step": 12870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5232194219063269,
+      "learning_rate": 1.633367882157516e-05,
+      "loss": 0.9131,
+      "step": 12871
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.943550389961239,
+      "learning_rate": 1.6333088323170285e-05,
+      "loss": 0.9754,
+      "step": 12872
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.195589364449735,
+      "learning_rate": 1.633249778789216e-05,
+      "loss": 0.7872,
+      "step": 12873
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6140476063151559,
+      "learning_rate": 1.633190721574423e-05,
+      "loss": 0.8585,
+      "step": 12874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5817685964448631,
+      "learning_rate": 1.633131660672993e-05,
+      "loss": 0.9745,
+      "step": 12875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.585737317772975,
+      "learning_rate": 1.6330725960852706e-05,
+      "loss": 0.8864,
+      "step": 12876
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.236048368432106,
+      "learning_rate": 1.6330135278115988e-05,
+      "loss": 0.8332,
+      "step": 12877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6544071629807708,
+      "learning_rate": 1.6329544558523212e-05,
+      "loss": 0.951,
+      "step": 12878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.864343940595966,
+      "learning_rate": 1.6328953802077833e-05,
+      "loss": 0.9693,
+      "step": 12879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6808166415771415,
+      "learning_rate": 1.6328363008783274e-05,
+      "loss": 0.9067,
+      "step": 12880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7329253551158437,
+      "learning_rate": 1.6327772178642986e-05,
+      "loss": 0.8106,
+      "step": 12881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.935016382597704,
+      "learning_rate": 1.6327181311660406e-05,
+      "loss": 0.8988,
+      "step": 12882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8352431978142267,
+      "learning_rate": 1.632659040783897e-05,
+      "loss": 0.7576,
+      "step": 12883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6009814665826183,
+      "learning_rate": 1.6325999467182123e-05,
+      "loss": 1.0477,
+      "step": 12884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.805369348346411,
+      "learning_rate": 1.6325408489693308e-05,
+      "loss": 0.9047,
+      "step": 12885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7736287736922853,
+      "learning_rate": 1.6324817475375957e-05,
+      "loss": 0.9151,
+      "step": 12886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7662698596812607,
+      "learning_rate": 1.6324226424233517e-05,
+      "loss": 0.7571,
+      "step": 12887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7380520789094815,
+      "learning_rate": 1.632363533626943e-05,
+      "loss": 0.9624,
+      "step": 12888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6393649729424336,
+      "learning_rate": 1.6323044211487135e-05,
+      "loss": 0.8879,
+      "step": 12889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7329264906504114,
+      "learning_rate": 1.6322453049890078e-05,
+      "loss": 1.0315,
+      "step": 12890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6374715821839914,
+      "learning_rate": 1.6321861851481693e-05,
+      "loss": 0.8524,
+      "step": 12891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7949833676269152,
+      "learning_rate": 1.632127061626543e-05,
+      "loss": 0.6936,
+      "step": 12892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4744797932494838,
+      "learning_rate": 1.6320679344244727e-05,
+      "loss": 0.983,
+      "step": 12893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6075107841488068,
+      "learning_rate": 1.6320088035423027e-05,
+      "loss": 0.9346,
+      "step": 12894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6266602800045102,
+      "learning_rate": 1.6319496689803775e-05,
+      "loss": 0.842,
+      "step": 12895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5470811870406405,
+      "learning_rate": 1.6318905307390412e-05,
+      "loss": 0.8658,
+      "step": 12896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9434230544359605,
+      "learning_rate": 1.6318313888186383e-05,
+      "loss": 0.9379,
+      "step": 12897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1970009902963046,
+      "learning_rate": 1.6317722432195127e-05,
+      "loss": 0.9698,
+      "step": 12898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5096636187297459,
+      "learning_rate": 1.6317130939420093e-05,
+      "loss": 0.8739,
+      "step": 12899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7885053867933757,
+      "learning_rate": 1.6316539409864724e-05,
+      "loss": 0.7454,
+      "step": 12900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9533688360239148,
+      "learning_rate": 1.631594784353246e-05,
+      "loss": 0.912,
+      "step": 12901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6615916370766306,
+      "learning_rate": 1.631535624042675e-05,
+      "loss": 0.9849,
+      "step": 12902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4330562415148913,
+      "learning_rate": 1.6314764600551035e-05,
+      "loss": 0.8695,
+      "step": 12903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5404316511978406,
+      "learning_rate": 1.6314172923908764e-05,
+      "loss": 0.9865,
+      "step": 12904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6986545650535563,
+      "learning_rate": 1.6313581210503375e-05,
+      "loss": 0.9354,
+      "step": 12905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5423475559300348,
+      "learning_rate": 1.631298946033832e-05,
+      "loss": 0.7542,
+      "step": 12906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.300435934885311,
+      "learning_rate": 1.6312397673417043e-05,
+      "loss": 0.9208,
+      "step": 12907
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.771428258063162,
+      "learning_rate": 1.6311805849742984e-05,
+      "loss": 0.7882,
+      "step": 12908
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3027635155087536,
+      "learning_rate": 1.6311213989319598e-05,
+      "loss": 0.8546,
+      "step": 12909
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.499922811861658,
+      "learning_rate": 1.6310622092150323e-05,
+      "loss": 0.9151,
+      "step": 12910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7431585858429965,
+      "learning_rate": 1.631003015823861e-05,
+      "loss": 0.9763,
+      "step": 12911
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5051530446508214,
+      "learning_rate": 1.63094381875879e-05,
+      "loss": 0.7997,
+      "step": 12912
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5496967790549254,
+      "learning_rate": 1.630884618020165e-05,
+      "loss": 0.9104,
+      "step": 12913
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9545024709324035,
+      "learning_rate": 1.6308254136083295e-05,
+      "loss": 1.0091,
+      "step": 12914
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5804964845882243,
+      "learning_rate": 1.6307662055236288e-05,
+      "loss": 1.0018,
+      "step": 12915
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.646123796088457,
+      "learning_rate": 1.6307069937664077e-05,
+      "loss": 0.8515,
+      "step": 12916
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8062588663681306,
+      "learning_rate": 1.6306477783370108e-05,
+      "loss": 0.7604,
+      "step": 12917
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5749459114121385,
+      "learning_rate": 1.6305885592357826e-05,
+      "loss": 0.89,
+      "step": 12918
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.79327095929695,
+      "learning_rate": 1.6305293364630684e-05,
+      "loss": 0.985,
+      "step": 12919
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6881298698265423,
+      "learning_rate": 1.630470110019213e-05,
+      "loss": 0.9028,
+      "step": 12920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5646264029979757,
+      "learning_rate": 1.6304108799045607e-05,
+      "loss": 1.054,
+      "step": 12921
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.567025145294003,
+      "learning_rate": 1.6303516461194564e-05,
+      "loss": 0.9815,
+      "step": 12922
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.553475071716242,
+      "learning_rate": 1.6302924086642456e-05,
+      "loss": 0.8337,
+      "step": 12923
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5431308240682855,
+      "learning_rate": 1.6302331675392728e-05,
+      "loss": 0.9554,
+      "step": 12924
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.06446086761192,
+      "learning_rate": 1.630173922744883e-05,
+      "loss": 0.9961,
+      "step": 12925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8683615616819882,
+      "learning_rate": 1.6301146742814215e-05,
+      "loss": 0.7525,
+      "step": 12926
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.556286526629191,
+      "learning_rate": 1.630055422149232e-05,
+      "loss": 0.8785,
+      "step": 12927
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.737954735562127,
+      "learning_rate": 1.6299961663486613e-05,
+      "loss": 0.946,
+      "step": 12928
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6471552641732812,
+      "learning_rate": 1.629936906880053e-05,
+      "loss": 0.8685,
+      "step": 12929
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.737161048320795,
+      "learning_rate": 1.6298776437437526e-05,
+      "loss": 0.7214,
+      "step": 12930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6254642298548523,
+      "learning_rate": 1.629818376940105e-05,
+      "loss": 0.952,
+      "step": 12931
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4893985516264734,
+      "learning_rate": 1.6297591064694557e-05,
+      "loss": 0.9163,
+      "step": 12932
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.640520391897133,
+      "learning_rate": 1.629699832332149e-05,
+      "loss": 0.9452,
+      "step": 12933
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6199469707225362,
+      "learning_rate": 1.629640554528531e-05,
+      "loss": 0.9004,
+      "step": 12934
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5366640389120043,
+      "learning_rate": 1.629581273058946e-05,
+      "loss": 0.8905,
+      "step": 12935
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7432851786871626,
+      "learning_rate": 1.62952198792374e-05,
+      "loss": 1.0031,
+      "step": 12936
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.789918263089038,
+      "learning_rate": 1.6294626991232574e-05,
+      "loss": 0.7061,
+      "step": 12937
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5295345629162471,
+      "learning_rate": 1.6294034066578436e-05,
+      "loss": 0.9169,
+      "step": 12938
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0089032508152833,
+      "learning_rate": 1.629344110527844e-05,
+      "loss": 0.9576,
+      "step": 12939
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0612686230201156,
+      "learning_rate": 1.6292848107336038e-05,
+      "loss": 0.917,
+      "step": 12940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.04519732435264,
+      "learning_rate": 1.6292255072754682e-05,
+      "loss": 0.8358,
+      "step": 12941
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9495632141039239,
+      "learning_rate": 1.6291662001537827e-05,
+      "loss": 0.8949,
+      "step": 12942
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.21379624910625,
+      "learning_rate": 1.629106889368892e-05,
+      "loss": 0.9182,
+      "step": 12943
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9118804038986161,
+      "learning_rate": 1.6290475749211417e-05,
+      "loss": 0.8342,
+      "step": 12944
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8180509073152917,
+      "learning_rate": 1.6289882568108778e-05,
+      "loss": 1.0084,
+      "step": 12945
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7992584738611076,
+      "learning_rate": 1.628928935038445e-05,
+      "loss": 0.7273,
+      "step": 12946
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7874234352929368,
+      "learning_rate": 1.6288696096041886e-05,
+      "loss": 0.7046,
+      "step": 12947
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6528692314900448,
+      "learning_rate": 1.628810280508454e-05,
+      "loss": 0.7826,
+      "step": 12948
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9910980592395855,
+      "learning_rate": 1.6287509477515872e-05,
+      "loss": 1.0737,
+      "step": 12949
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9530035690020688,
+      "learning_rate": 1.6286916113339333e-05,
+      "loss": 0.899,
+      "step": 12950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7277354129071291,
+      "learning_rate": 1.628632271255838e-05,
+      "loss": 0.7277,
+      "step": 12951
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.602798477784272,
+      "learning_rate": 1.6285729275176463e-05,
+      "loss": 0.8482,
+      "step": 12952
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.666896418242102,
+      "learning_rate": 1.628513580119704e-05,
+      "loss": 1.0034,
+      "step": 12953
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.04944686315921,
+      "learning_rate": 1.6284542290623568e-05,
+      "loss": 0.9475,
+      "step": 12954
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6953325988005519,
+      "learning_rate": 1.62839487434595e-05,
+      "loss": 0.9425,
+      "step": 12955
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8357234286137356,
+      "learning_rate": 1.6283355159708297e-05,
+      "loss": 0.8209,
+      "step": 12956
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6114962280613052,
+      "learning_rate": 1.6282761539373402e-05,
+      "loss": 1.0107,
+      "step": 12957
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8968809488983154,
+      "learning_rate": 1.628216788245829e-05,
+      "loss": 0.9189,
+      "step": 12958
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6214740768785416,
+      "learning_rate": 1.6281574188966402e-05,
+      "loss": 0.8709,
+      "step": 12959
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1503584486724328,
+      "learning_rate": 1.6280980458901205e-05,
+      "loss": 0.9854,
+      "step": 12960
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8715295783964765,
+      "learning_rate": 1.6280386692266142e-05,
+      "loss": 0.9351,
+      "step": 12961
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.485364568090339,
+      "learning_rate": 1.6279792889064688e-05,
+      "loss": 0.8794,
+      "step": 12962
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7710836056350723,
+      "learning_rate": 1.6279199049300288e-05,
+      "loss": 0.9002,
+      "step": 12963
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0681274929827143,
+      "learning_rate": 1.6278605172976404e-05,
+      "loss": 0.8322,
+      "step": 12964
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.049202377986188,
+      "learning_rate": 1.6278011260096495e-05,
+      "loss": 0.9773,
+      "step": 12965
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.361081133877135,
+      "learning_rate": 1.6277417310664015e-05,
+      "loss": 0.9947,
+      "step": 12966
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.534218247226385,
+      "learning_rate": 1.6276823324682426e-05,
+      "loss": 0.8288,
+      "step": 12967
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0912535802355188,
+      "learning_rate": 1.6276229302155183e-05,
+      "loss": 0.8705,
+      "step": 12968
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5507150344696041,
+      "learning_rate": 1.6275635243085743e-05,
+      "loss": 0.7954,
+      "step": 12969
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6619579723766296,
+      "learning_rate": 1.6275041147477572e-05,
+      "loss": 0.8892,
+      "step": 12970
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9183878257778708,
+      "learning_rate": 1.6274447015334122e-05,
+      "loss": 0.8941,
+      "step": 12971
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.689191695803817,
+      "learning_rate": 1.6273852846658856e-05,
+      "loss": 0.9305,
+      "step": 12972
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9605897606334066,
+      "learning_rate": 1.6273258641455233e-05,
+      "loss": 0.9401,
+      "step": 12973
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0268336738089143,
+      "learning_rate": 1.6272664399726712e-05,
+      "loss": 0.965,
+      "step": 12974
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.856972290804312,
+      "learning_rate": 1.6272070121476753e-05,
+      "loss": 0.8301,
+      "step": 12975
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9661967400030633,
+      "learning_rate": 1.6271475806708815e-05,
+      "loss": 0.7852,
+      "step": 12976
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9162780420275256,
+      "learning_rate": 1.6270881455426358e-05,
+      "loss": 0.8535,
+      "step": 12977
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4713984055425893,
+      "learning_rate": 1.6270287067632847e-05,
+      "loss": 0.8608,
+      "step": 12978
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8458806255621147,
+      "learning_rate": 1.626969264333174e-05,
+      "loss": 0.8761,
+      "step": 12979
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.94539852827629,
+      "learning_rate": 1.6269098182526494e-05,
+      "loss": 0.9886,
+      "step": 12980
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7875444771424736,
+      "learning_rate": 1.6268503685220577e-05,
+      "loss": 0.7682,
+      "step": 12981
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6091199064800799,
+      "learning_rate": 1.6267909151417444e-05,
+      "loss": 0.7872,
+      "step": 12982
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8003084331062433,
+      "learning_rate": 1.6267314581120564e-05,
+      "loss": 0.9974,
+      "step": 12983
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7079074330111603,
+      "learning_rate": 1.6266719974333387e-05,
+      "loss": 0.936,
+      "step": 12984
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7811021217074543,
+      "learning_rate": 1.6266125331059387e-05,
+      "loss": 0.742,
+      "step": 12985
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5771505733009026,
+      "learning_rate": 1.626553065130202e-05,
+      "loss": 0.9085,
+      "step": 12986
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6937771898948488,
+      "learning_rate": 1.6264935935064753e-05,
+      "loss": 0.9102,
+      "step": 12987
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.524665984904311,
+      "learning_rate": 1.6264341182351043e-05,
+      "loss": 0.8843,
+      "step": 12988
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9199858856832956,
+      "learning_rate": 1.6263746393164356e-05,
+      "loss": 0.8839,
+      "step": 12989
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.80693524245953,
+      "learning_rate": 1.6263151567508154e-05,
+      "loss": 0.8458,
+      "step": 12990
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9476308663036164,
+      "learning_rate": 1.62625567053859e-05,
+      "loss": 0.8683,
+      "step": 12991
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.018864450582992,
+      "learning_rate": 1.6261961806801062e-05,
+      "loss": 1.0216,
+      "step": 12992
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1438036087094914,
+      "learning_rate": 1.6261366871757097e-05,
+      "loss": 0.8606,
+      "step": 12993
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5424352322179495,
+      "learning_rate": 1.626077190025747e-05,
+      "loss": 0.816,
+      "step": 12994
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4295268347280867,
+      "learning_rate": 1.6260176892305654e-05,
+      "loss": 0.8194,
+      "step": 12995
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1284535484054135,
+      "learning_rate": 1.6259581847905097e-05,
+      "loss": 0.9311,
+      "step": 12996
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6694729708440066,
+      "learning_rate": 1.6258986767059278e-05,
+      "loss": 0.8916,
+      "step": 12997
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7200950733631837,
+      "learning_rate": 1.6258391649771654e-05,
+      "loss": 1.0826,
+      "step": 12998
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6919398417941736,
+      "learning_rate": 1.6257796496045694e-05,
+      "loss": 0.8999,
+      "step": 12999
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7141314329313841,
+      "learning_rate": 1.625720130588486e-05,
+      "loss": 0.8446,
+      "step": 13000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.205580509108831,
+      "learning_rate": 1.625660607929262e-05,
+      "loss": 0.7244,
+      "step": 13001
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7954695006868964,
+      "learning_rate": 1.625601081627244e-05,
+      "loss": 0.7371,
+      "step": 13002
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.517102370994666,
+      "learning_rate": 1.6255415516827783e-05,
+      "loss": 0.92,
+      "step": 13003
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4511302506334087,
+      "learning_rate": 1.6254820180962115e-05,
+      "loss": 0.9019,
+      "step": 13004
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5489539524720946,
+      "learning_rate": 1.6254224808678906e-05,
+      "loss": 0.8554,
+      "step": 13005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.99639828539157,
+      "learning_rate": 1.625362939998162e-05,
+      "loss": 0.9834,
+      "step": 13006
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5489047880737588,
+      "learning_rate": 1.6253033954873725e-05,
+      "loss": 0.927,
+      "step": 13007
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5348532664128285,
+      "learning_rate": 1.6252438473358682e-05,
+      "loss": 0.9476,
+      "step": 13008
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5430002154804223,
+      "learning_rate": 1.625184295543997e-05,
+      "loss": 1.034,
+      "step": 13009
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2236963159706975,
+      "learning_rate": 1.625124740112104e-05,
+      "loss": 0.8356,
+      "step": 13010
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8078235129974094,
+      "learning_rate": 1.6250651810405375e-05,
+      "loss": 0.9687,
+      "step": 13011
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.818757739308693,
+      "learning_rate": 1.6250056183296434e-05,
+      "loss": 0.9119,
+      "step": 13012
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6367025446046848,
+      "learning_rate": 1.6249460519797687e-05,
+      "loss": 0.8598,
+      "step": 13013
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6851095399765603,
+      "learning_rate": 1.6248864819912603e-05,
+      "loss": 0.9768,
+      "step": 13014
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8044905862586806,
+      "learning_rate": 1.624826908364465e-05,
+      "loss": 0.9788,
+      "step": 13015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7599083102872957,
+      "learning_rate": 1.6247673310997294e-05,
+      "loss": 0.8722,
+      "step": 13016
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7980654708285073,
+      "learning_rate": 1.6247077501974008e-05,
+      "loss": 0.9577,
+      "step": 13017
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5875995096838904,
+      "learning_rate": 1.624648165657826e-05,
+      "loss": 0.8772,
+      "step": 13018
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.623081259056278,
+      "learning_rate": 1.6245885774813514e-05,
+      "loss": 0.9822,
+      "step": 13019
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7008068346454719,
+      "learning_rate": 1.624528985668325e-05,
+      "loss": 1.0615,
+      "step": 13020
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6674235946512312,
+      "learning_rate": 1.6244693902190928e-05,
+      "loss": 0.9116,
+      "step": 13021
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7186039838721976,
+      "learning_rate": 1.6244097911340018e-05,
+      "loss": 0.9155,
+      "step": 13022
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8955435864495105,
+      "learning_rate": 1.6243501884133997e-05,
+      "loss": 0.8436,
+      "step": 13023
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6717171952438452,
+      "learning_rate": 1.6242905820576327e-05,
+      "loss": 0.9294,
+      "step": 13024
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6210473140491615,
+      "learning_rate": 1.6242309720670492e-05,
+      "loss": 0.891,
+      "step": 13025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.580316823738266,
+      "learning_rate": 1.6241713584419946e-05,
+      "loss": 0.9908,
+      "step": 13026
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7007972107313536,
+      "learning_rate": 1.6241117411828167e-05,
+      "loss": 0.8796,
+      "step": 13027
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.658386991695198,
+      "learning_rate": 1.624052120289863e-05,
+      "loss": 0.9369,
+      "step": 13028
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7047974014290417,
+      "learning_rate": 1.6239924957634802e-05,
+      "loss": 0.8552,
+      "step": 13029
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6238283245801974,
+      "learning_rate": 1.6239328676040155e-05,
+      "loss": 0.9384,
+      "step": 13030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8324624733906638,
+      "learning_rate": 1.623873235811816e-05,
+      "loss": 0.9318,
+      "step": 13031
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8347605415149555,
+      "learning_rate": 1.623813600387229e-05,
+      "loss": 0.9242,
+      "step": 13032
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.680717070159626,
+      "learning_rate": 1.6237539613306018e-05,
+      "loss": 0.8468,
+      "step": 13033
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5858547720823704,
+      "learning_rate": 1.6236943186422815e-05,
+      "loss": 0.8824,
+      "step": 13034
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7605874837144847,
+      "learning_rate": 1.6236346723226154e-05,
+      "loss": 0.7307,
+      "step": 13035
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7737958602201247,
+      "learning_rate": 1.623575022371951e-05,
+      "loss": 0.8263,
+      "step": 13036
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7145268623275505,
+      "learning_rate": 1.6235153687906354e-05,
+      "loss": 0.9278,
+      "step": 13037
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.540477211063941,
+      "learning_rate": 1.6234557115790157e-05,
+      "loss": 0.8215,
+      "step": 13038
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.598096237117986,
+      "learning_rate": 1.6233960507374397e-05,
+      "loss": 0.8544,
+      "step": 13039
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1733647840864787,
+      "learning_rate": 1.6233363862662543e-05,
+      "loss": 0.8591,
+      "step": 13040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7394218847184555,
+      "learning_rate": 1.6232767181658072e-05,
+      "loss": 0.8515,
+      "step": 13041
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5992756612711045,
+      "learning_rate": 1.623217046436446e-05,
+      "loss": 0.8602,
+      "step": 13042
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3983333339235926,
+      "learning_rate": 1.6231573710785172e-05,
+      "loss": 0.7833,
+      "step": 13043
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5654656242635097,
+      "learning_rate": 1.6230976920923696e-05,
+      "loss": 0.7053,
+      "step": 13044
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6180526088753378,
+      "learning_rate": 1.6230380094783497e-05,
+      "loss": 0.8533,
+      "step": 13045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7505961563809498,
+      "learning_rate": 1.622978323236805e-05,
+      "loss": 0.8965,
+      "step": 13046
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6016080092078844,
+      "learning_rate": 1.6229186333680832e-05,
+      "loss": 0.8262,
+      "step": 13047
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.84924015209165,
+      "learning_rate": 1.6228589398725326e-05,
+      "loss": 0.9509,
+      "step": 13048
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7978325907206365,
+      "learning_rate": 1.6227992427504995e-05,
+      "loss": 0.8457,
+      "step": 13049
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8277170636168213,
+      "learning_rate": 1.6227395420023323e-05,
+      "loss": 0.7937,
+      "step": 13050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6078953568173948,
+      "learning_rate": 1.6226798376283782e-05,
+      "loss": 0.8005,
+      "step": 13051
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6888276123259711,
+      "learning_rate": 1.6226201296289848e-05,
+      "loss": 0.9782,
+      "step": 13052
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7675459908395181,
+      "learning_rate": 1.6225604180045e-05,
+      "loss": 0.9296,
+      "step": 13053
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6531082201481975,
+      "learning_rate": 1.6225007027552715e-05,
+      "loss": 0.9259,
+      "step": 13054
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.07097106147256,
+      "learning_rate": 1.6224409838816465e-05,
+      "loss": 0.9082,
+      "step": 13055
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5143726987039057,
+      "learning_rate": 1.6223812613839734e-05,
+      "loss": 0.8648,
+      "step": 13056
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.762293214123383,
+      "learning_rate": 1.6223215352625994e-05,
+      "loss": 0.9038,
+      "step": 13057
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9024817437462282,
+      "learning_rate": 1.6222618055178723e-05,
+      "loss": 0.9924,
+      "step": 13058
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5470304141638587,
+      "learning_rate": 1.62220207215014e-05,
+      "loss": 0.8758,
+      "step": 13059
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4385100629924454,
+      "learning_rate": 1.6221423351597505e-05,
+      "loss": 0.9849,
+      "step": 13060
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7503216071065433,
+      "learning_rate": 1.622082594547051e-05,
+      "loss": 0.8254,
+      "step": 13061
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7604725904241698,
+      "learning_rate": 1.62202285031239e-05,
+      "loss": 0.9132,
+      "step": 13062
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.130722153819299,
+      "learning_rate": 1.6219631024561148e-05,
+      "loss": 0.8321,
+      "step": 13063
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.476919389694691,
+      "learning_rate": 1.6219033509785738e-05,
+      "loss": 0.7869,
+      "step": 13064
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7480070483210017,
+      "learning_rate": 1.6218435958801143e-05,
+      "loss": 0.8915,
+      "step": 13065
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7539630107573347,
+      "learning_rate": 1.6217838371610847e-05,
+      "loss": 0.8739,
+      "step": 13066
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7693478059030262,
+      "learning_rate": 1.6217240748218325e-05,
+      "loss": 0.7282,
+      "step": 13067
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7155707388610388,
+      "learning_rate": 1.621664308862706e-05,
+      "loss": 0.7713,
+      "step": 13068
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.82516146191424,
+      "learning_rate": 1.621604539284053e-05,
+      "loss": 0.9238,
+      "step": 13069
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7298483968408864,
+      "learning_rate": 1.6215447660862216e-05,
+      "loss": 0.8855,
+      "step": 13070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5654645091650077,
+      "learning_rate": 1.6214849892695602e-05,
+      "loss": 0.9966,
+      "step": 13071
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.603168978426852,
+      "learning_rate": 1.6214252088344164e-05,
+      "loss": 0.8775,
+      "step": 13072
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7284815961039386,
+      "learning_rate": 1.621365424781138e-05,
+      "loss": 0.8423,
+      "step": 13073
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.78332292006112,
+      "learning_rate": 1.6213056371100732e-05,
+      "loss": 0.8408,
+      "step": 13074
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5980871720499938,
+      "learning_rate": 1.6212458458215706e-05,
+      "loss": 0.8705,
+      "step": 13075
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.727769510291179,
+      "learning_rate": 1.621186050915978e-05,
+      "loss": 0.9112,
+      "step": 13076
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4929702693180353,
+      "learning_rate": 1.6211262523936432e-05,
+      "loss": 1.0061,
+      "step": 13077
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7421100937809475,
+      "learning_rate": 1.6210664502549152e-05,
+      "loss": 0.922,
+      "step": 13078
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8759039738704082,
+      "learning_rate": 1.6210066445001413e-05,
+      "loss": 0.9677,
+      "step": 13079
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7055830061907684,
+      "learning_rate": 1.6209468351296703e-05,
+      "loss": 0.9096,
+      "step": 13080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.626732931901177,
+      "learning_rate": 1.6208870221438503e-05,
+      "loss": 0.8308,
+      "step": 13081
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6384686148354732,
+      "learning_rate": 1.6208272055430292e-05,
+      "loss": 0.7891,
+      "step": 13082
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7808779541140722,
+      "learning_rate": 1.620767385327556e-05,
+      "loss": 0.9044,
+      "step": 13083
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.407272016887305,
+      "learning_rate": 1.620707561497778e-05,
+      "loss": 0.8976,
+      "step": 13084
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0362415974044508,
+      "learning_rate": 1.6206477340540444e-05,
+      "loss": 0.757,
+      "step": 13085
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8122774003728105,
+      "learning_rate": 1.6205879029967032e-05,
+      "loss": 0.7547,
+      "step": 13086
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9328528821652728,
+      "learning_rate": 1.620528068326103e-05,
+      "loss": 0.8555,
+      "step": 13087
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9322856847482444,
+      "learning_rate": 1.6204682300425915e-05,
+      "loss": 0.8439,
+      "step": 13088
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8139784353871415,
+      "learning_rate": 1.6204083881465175e-05,
+      "loss": 0.8834,
+      "step": 13089
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7530759075672364,
+      "learning_rate": 1.62034854263823e-05,
+      "loss": 0.9374,
+      "step": 13090
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8251794373764325,
+      "learning_rate": 1.620288693518076e-05,
+      "loss": 0.8565,
+      "step": 13091
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.958121927543762,
+      "learning_rate": 1.6202288407864056e-05,
+      "loss": 0.7751,
+      "step": 13092
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8653131994317413,
+      "learning_rate": 1.620168984443566e-05,
+      "loss": 0.9264,
+      "step": 13093
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.886563324995957,
+      "learning_rate": 1.6201091244899067e-05,
+      "loss": 0.8448,
+      "step": 13094
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4610024344706227,
+      "learning_rate": 1.6200492609257758e-05,
+      "loss": 1.0365,
+      "step": 13095
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.639035820790745,
+      "learning_rate": 1.6199893937515214e-05,
+      "loss": 0.9556,
+      "step": 13096
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.560092614161879,
+      "learning_rate": 1.6199295229674927e-05,
+      "loss": 0.9376,
+      "step": 13097
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1072922951310873,
+      "learning_rate": 1.6198696485740378e-05,
+      "loss": 0.911,
+      "step": 13098
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6945428193251073,
+      "learning_rate": 1.619809770571506e-05,
+      "loss": 0.929,
+      "step": 13099
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.686739243483214,
+      "learning_rate": 1.619749888960245e-05,
+      "loss": 1.032,
+      "step": 13100
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7323498236490023,
+      "learning_rate": 1.6196900037406043e-05,
+      "loss": 1.0104,
+      "step": 13101
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9304102210759615,
+      "learning_rate": 1.619630114912932e-05,
+      "loss": 0.9138,
+      "step": 13102
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0151979926353216,
+      "learning_rate": 1.6195702224775768e-05,
+      "loss": 0.8872,
+      "step": 13103
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7896562252389236,
+      "learning_rate": 1.619510326434888e-05,
+      "loss": 0.6705,
+      "step": 13104
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5751597291598334,
+      "learning_rate": 1.619450426785214e-05,
+      "loss": 0.7836,
+      "step": 13105
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6935856998344325,
+      "learning_rate": 1.619390523528903e-05,
+      "loss": 0.9333,
+      "step": 13106
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.565548489624836,
+      "learning_rate": 1.619330616666305e-05,
+      "loss": 0.9333,
+      "step": 13107
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7442826782256293,
+      "learning_rate": 1.6192707061977676e-05,
+      "loss": 0.8958,
+      "step": 13108
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5178074184064005,
+      "learning_rate": 1.61921079212364e-05,
+      "loss": 0.8741,
+      "step": 13109
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.419255648050619,
+      "learning_rate": 1.6191508744442713e-05,
+      "loss": 0.8151,
+      "step": 13110
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8382938424721154,
+      "learning_rate": 1.6190909531600103e-05,
+      "loss": 1.0125,
+      "step": 13111
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6930675463090352,
+      "learning_rate": 1.6190310282712056e-05,
+      "loss": 0.8438,
+      "step": 13112
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.758012264990463,
+      "learning_rate": 1.6189710997782063e-05,
+      "loss": 0.9945,
+      "step": 13113
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6489168315274059,
+      "learning_rate": 1.6189111676813613e-05,
+      "loss": 0.8915,
+      "step": 13114
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6201353384138517,
+      "learning_rate": 1.6188512319810197e-05,
+      "loss": 0.964,
+      "step": 13115
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2709750711675327,
+      "learning_rate": 1.6187912926775303e-05,
+      "loss": 0.9896,
+      "step": 13116
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5918681577616876,
+      "learning_rate": 1.618731349771242e-05,
+      "loss": 0.8518,
+      "step": 13117
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4016996563474677,
+      "learning_rate": 1.6186714032625036e-05,
+      "loss": 0.9345,
+      "step": 13118
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6767687160876703,
+      "learning_rate": 1.6186114531516646e-05,
+      "loss": 0.9288,
+      "step": 13119
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7724028803936551,
+      "learning_rate": 1.6185514994390738e-05,
+      "loss": 0.9395,
+      "step": 13120
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3065113108521955,
+      "learning_rate": 1.6184915421250805e-05,
+      "loss": 0.8438,
+      "step": 13121
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.620327057830951,
+      "learning_rate": 1.6184315812100337e-05,
+      "loss": 0.8606,
+      "step": 13122
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6000672553369557,
+      "learning_rate": 1.6183716166942824e-05,
+      "loss": 0.9252,
+      "step": 13123
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.736637931992919,
+      "learning_rate": 1.6183116485781755e-05,
+      "loss": 0.9319,
+      "step": 13124
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7083948041095631,
+      "learning_rate": 1.6182516768620627e-05,
+      "loss": 0.9697,
+      "step": 13125
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8213453154868677,
+      "learning_rate": 1.618191701546293e-05,
+      "loss": 0.6962,
+      "step": 13126
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.811050377370658,
+      "learning_rate": 1.6181317226312154e-05,
+      "loss": 0.7835,
+      "step": 13127
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6636889902786476,
+      "learning_rate": 1.618071740117179e-05,
+      "loss": 0.8511,
+      "step": 13128
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3720445998758972,
+      "learning_rate": 1.6180117540045338e-05,
+      "loss": 0.892,
+      "step": 13129
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8311450643445321,
+      "learning_rate": 1.617951764293628e-05,
+      "loss": 0.7964,
+      "step": 13130
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7750570194425908,
+      "learning_rate": 1.6178917709848113e-05,
+      "loss": 0.8541,
+      "step": 13131
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6785642016844768,
+      "learning_rate": 1.6178317740784333e-05,
+      "loss": 0.9167,
+      "step": 13132
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5737882830653613,
+      "learning_rate": 1.6177717735748433e-05,
+      "loss": 0.9706,
+      "step": 13133
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7550734200580098,
+      "learning_rate": 1.61771176947439e-05,
+      "loss": 0.8898,
+      "step": 13134
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1482307356939425,
+      "learning_rate": 1.6176517617774234e-05,
+      "loss": 1.0092,
+      "step": 13135
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7939183539172039,
+      "learning_rate": 1.617591750484293e-05,
+      "loss": 0.6566,
+      "step": 13136
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4671235493597496,
+      "learning_rate": 1.6175317355953477e-05,
+      "loss": 0.9155,
+      "step": 13137
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9125864750128936,
+      "learning_rate": 1.6174717171109372e-05,
+      "loss": 0.9333,
+      "step": 13138
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7365844898217306,
+      "learning_rate": 1.6174116950314106e-05,
+      "loss": 0.905,
+      "step": 13139
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4890132983564217,
+      "learning_rate": 1.6173516693571178e-05,
+      "loss": 0.8195,
+      "step": 13140
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8185549462568613,
+      "learning_rate": 1.617291640088408e-05,
+      "loss": 0.9375,
+      "step": 13141
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1095736493536648,
+      "learning_rate": 1.617231607225631e-05,
+      "loss": 0.9134,
+      "step": 13142
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7276037482753351,
+      "learning_rate": 1.6171715707691366e-05,
+      "loss": 0.7537,
+      "step": 13143
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5559954428558305,
+      "learning_rate": 1.6171115307192734e-05,
+      "loss": 0.8861,
+      "step": 13144
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7631297869730121,
+      "learning_rate": 1.6170514870763912e-05,
+      "loss": 0.7115,
+      "step": 13145
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0714261303532457,
+      "learning_rate": 1.6169914398408405e-05,
+      "loss": 0.9279,
+      "step": 13146
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.502976081931748,
+      "learning_rate": 1.6169313890129702e-05,
+      "loss": 0.9956,
+      "step": 13147
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.480456896511305,
+      "learning_rate": 1.61687133459313e-05,
+      "loss": 0.9005,
+      "step": 13148
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8470676788570923,
+      "learning_rate": 1.6168112765816692e-05,
+      "loss": 0.9056,
+      "step": 13149
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6292319632155996,
+      "learning_rate": 1.616751214978938e-05,
+      "loss": 0.8836,
+      "step": 13150
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.581888923711087,
+      "learning_rate": 1.6166911497852864e-05,
+      "loss": 0.9141,
+      "step": 13151
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7018881214775703,
+      "learning_rate": 1.6166310810010633e-05,
+      "loss": 0.9454,
+      "step": 13152
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.049492694451238,
+      "learning_rate": 1.616571008626619e-05,
+      "loss": 0.9489,
+      "step": 13153
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8185236329173037,
+      "learning_rate": 1.616510932662303e-05,
+      "loss": 0.8079,
+      "step": 13154
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6680185721433438,
+      "learning_rate": 1.6164508531084653e-05,
+      "loss": 0.9912,
+      "step": 13155
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.854416534809547,
+      "learning_rate": 1.616390769965455e-05,
+      "loss": 0.7815,
+      "step": 13156
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6043334592999325,
+      "learning_rate": 1.6163306832336232e-05,
+      "loss": 0.9388,
+      "step": 13157
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6786709884678057,
+      "learning_rate": 1.616270592913319e-05,
+      "loss": 0.9666,
+      "step": 13158
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.537271883547843,
+      "learning_rate": 1.616210499004892e-05,
+      "loss": 0.8877,
+      "step": 13159
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6117299701192762,
+      "learning_rate": 1.6161504015086923e-05,
+      "loss": 0.8793,
+      "step": 13160
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9162259574264444,
+      "learning_rate": 1.61609030042507e-05,
+      "loss": 0.9666,
+      "step": 13161
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.944210891406398,
+      "learning_rate": 1.616030195754375e-05,
+      "loss": 0.9106,
+      "step": 13162
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6745444683688444,
+      "learning_rate": 1.615970087496957e-05,
+      "loss": 0.9517,
+      "step": 13163
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7906989637433208,
+      "learning_rate": 1.6159099756531668e-05,
+      "loss": 1.003,
+      "step": 13164
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.601570099174393,
+      "learning_rate": 1.615849860223353e-05,
+      "loss": 0.8526,
+      "step": 13165
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6110978981832946,
+      "learning_rate": 1.6157897412078668e-05,
+      "loss": 0.8675,
+      "step": 13166
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1069011117446994,
+      "learning_rate": 1.615729618607057e-05,
+      "loss": 0.9509,
+      "step": 13167
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3903787455069638,
+      "learning_rate": 1.615669492421275e-05,
+      "loss": 0.8119,
+      "step": 13168
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5399347129267849,
+      "learning_rate": 1.6156093626508706e-05,
+      "loss": 0.9505,
+      "step": 13169
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6112872943798477,
+      "learning_rate": 1.615549229296193e-05,
+      "loss": 0.8704,
+      "step": 13170
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.724362607684154,
+      "learning_rate": 1.6154890923575933e-05,
+      "loss": 0.9526,
+      "step": 13171
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9140590400992963,
+      "learning_rate": 1.6154289518354208e-05,
+      "loss": 0.9305,
+      "step": 13172
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4403477495334365,
+      "learning_rate": 1.6153688077300265e-05,
+      "loss": 0.8303,
+      "step": 13173
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7292945335813463,
+      "learning_rate": 1.6153086600417602e-05,
+      "loss": 0.9258,
+      "step": 13174
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6464839303514112,
+      "learning_rate": 1.615248508770972e-05,
+      "loss": 1.0256,
+      "step": 13175
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7648847397793652,
+      "learning_rate": 1.615188353918012e-05,
+      "loss": 0.9924,
+      "step": 13176
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9970799162279347,
+      "learning_rate": 1.6151281954832307e-05,
+      "loss": 0.9433,
+      "step": 13177
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8881052678612682,
+      "learning_rate": 1.6150680334669785e-05,
+      "loss": 0.9202,
+      "step": 13178
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5703515618332748,
+      "learning_rate": 1.615007867869605e-05,
+      "loss": 0.9754,
+      "step": 13179
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8879511186112798,
+      "learning_rate": 1.6149476986914617e-05,
+      "loss": 0.7923,
+      "step": 13180
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0737783285521223,
+      "learning_rate": 1.6148875259328978e-05,
+      "loss": 0.9181,
+      "step": 13181
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.246992395185522,
+      "learning_rate": 1.614827349594264e-05,
+      "loss": 0.9208,
+      "step": 13182
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8200040837249783,
+      "learning_rate": 1.614767169675911e-05,
+      "loss": 0.7526,
+      "step": 13183
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8279702453843718,
+      "learning_rate": 1.6147069861781887e-05,
+      "loss": 0.7568,
+      "step": 13184
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7501310787932265,
+      "learning_rate": 1.614646799101448e-05,
+      "loss": 0.845,
+      "step": 13185
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6710921311618643,
+      "learning_rate": 1.6145866084460384e-05,
+      "loss": 0.8136,
+      "step": 13186
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5585775093115928,
+      "learning_rate": 1.6145264142123114e-05,
+      "loss": 0.9143,
+      "step": 13187
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5792111912648827,
+      "learning_rate": 1.6144662164006172e-05,
+      "loss": 0.8931,
+      "step": 13188
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7308831939800895,
+      "learning_rate": 1.614406015011306e-05,
+      "loss": 0.9664,
+      "step": 13189
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9537985799956523,
+      "learning_rate": 1.6143458100447283e-05,
+      "loss": 0.8527,
+      "step": 13190
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5914072899456282,
+      "learning_rate": 1.614285601501235e-05,
+      "loss": 1.0104,
+      "step": 13191
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4685061539398367,
+      "learning_rate": 1.6142253893811766e-05,
+      "loss": 0.8458,
+      "step": 13192
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3030026945869686,
+      "learning_rate": 1.6141651736849032e-05,
+      "loss": 0.8785,
+      "step": 13193
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4761392522288443,
+      "learning_rate": 1.6141049544127657e-05,
+      "loss": 0.8983,
+      "step": 13194
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.568544751988776,
+      "learning_rate": 1.6140447315651152e-05,
+      "loss": 0.9046,
+      "step": 13195
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.96482484806754,
+      "learning_rate": 1.6139845051423012e-05,
+      "loss": 0.9918,
+      "step": 13196
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7601446365756849,
+      "learning_rate": 1.6139242751446754e-05,
+      "loss": 0.7328,
+      "step": 13197
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9900229159381144,
+      "learning_rate": 1.6138640415725882e-05,
+      "loss": 0.8423,
+      "step": 13198
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6173161190092422,
+      "learning_rate": 1.61380380442639e-05,
+      "loss": 1.0026,
+      "step": 13199
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6885591409233816,
+      "learning_rate": 1.613743563706432e-05,
+      "loss": 0.9708,
+      "step": 13200
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.662740520379899,
+      "learning_rate": 1.6136833194130642e-05,
+      "loss": 0.9038,
+      "step": 13201
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6173361193863058,
+      "learning_rate": 1.6136230715466384e-05,
+      "loss": 0.9625,
+      "step": 13202
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7881834916913248,
+      "learning_rate": 1.6135628201075042e-05,
+      "loss": 0.9332,
+      "step": 13203
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7691774533672646,
+      "learning_rate": 1.6135025650960137e-05,
+      "loss": 0.9144,
+      "step": 13204
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7162918762868817,
+      "learning_rate": 1.6134423065125164e-05,
+      "loss": 0.8086,
+      "step": 13205
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4210130229322235,
+      "learning_rate": 1.613382044357364e-05,
+      "loss": 0.8475,
+      "step": 13206
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8209379114126774,
+      "learning_rate": 1.613321778630907e-05,
+      "loss": 0.9266,
+      "step": 13207
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0738425709414527,
+      "learning_rate": 1.613261509333497e-05,
+      "loss": 1.0155,
+      "step": 13208
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5065441515433802,
+      "learning_rate": 1.613201236465484e-05,
+      "loss": 0.9997,
+      "step": 13209
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7881431589314758,
+      "learning_rate": 1.6131409600272186e-05,
+      "loss": 0.9183,
+      "step": 13210
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8135355499497894,
+      "learning_rate": 1.6130806800190532e-05,
+      "loss": 1.0713,
+      "step": 13211
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5984578648058643,
+      "learning_rate": 1.613020396441338e-05,
+      "loss": 0.8283,
+      "step": 13212
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7809898973339607,
+      "learning_rate": 1.612960109294424e-05,
+      "loss": 0.7101,
+      "step": 13213
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1577832454093504,
+      "learning_rate": 1.6128998185786616e-05,
+      "loss": 1.0207,
+      "step": 13214
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6831953640389221,
+      "learning_rate": 1.6128395242944027e-05,
+      "loss": 0.826,
+      "step": 13215
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.691952252917136,
+      "learning_rate": 1.6127792264419983e-05,
+      "loss": 1.0,
+      "step": 13216
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6711760182340405,
+      "learning_rate": 1.6127189250217988e-05,
+      "loss": 0.9205,
+      "step": 13217
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4933689899060842,
+      "learning_rate": 1.612658620034156e-05,
+      "loss": 0.9353,
+      "step": 13218
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.760236412456105,
+      "learning_rate": 1.612598311479421e-05,
+      "loss": 0.8435,
+      "step": 13219
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.631818471836646,
+      "learning_rate": 1.6125379993579448e-05,
+      "loss": 0.8575,
+      "step": 13220
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0311622922357637,
+      "learning_rate": 1.6124776836700778e-05,
+      "loss": 0.8675,
+      "step": 13221
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7575016474710796,
+      "learning_rate": 1.612417364416172e-05,
+      "loss": 0.7881,
+      "step": 13222
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6175992063777636,
+      "learning_rate": 1.6123570415965785e-05,
+      "loss": 0.8931,
+      "step": 13223
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6912126481958063,
+      "learning_rate": 1.6122967152116483e-05,
+      "loss": 1.016,
+      "step": 13224
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.856902378771833,
+      "learning_rate": 1.6122363852617333e-05,
+      "loss": 0.8795,
+      "step": 13225
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5768980397498795,
+      "learning_rate": 1.6121760517471837e-05,
+      "loss": 0.8089,
+      "step": 13226
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8345755406118338,
+      "learning_rate": 1.6121157146683514e-05,
+      "loss": 0.9221,
+      "step": 13227
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0577706682906216,
+      "learning_rate": 1.6120553740255877e-05,
+      "loss": 0.792,
+      "step": 13228
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6491184300523818,
+      "learning_rate": 1.6119950298192438e-05,
+      "loss": 0.8718,
+      "step": 13229
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7427398795307796,
+      "learning_rate": 1.6119346820496712e-05,
+      "loss": 0.9581,
+      "step": 13230
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8425879634718882,
+      "learning_rate": 1.6118743307172212e-05,
+      "loss": 0.7246,
+      "step": 13231
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.709073720061115,
+      "learning_rate": 1.611813975822245e-05,
+      "loss": 0.8756,
+      "step": 13232
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6800033766832752,
+      "learning_rate": 1.611753617365094e-05,
+      "loss": 0.7862,
+      "step": 13233
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2619519797581553,
+      "learning_rate": 1.61169325534612e-05,
+      "loss": 0.9008,
+      "step": 13234
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5705830900954363,
+      "learning_rate": 1.611632889765674e-05,
+      "loss": 0.7907,
+      "step": 13235
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7196481221493602,
+      "learning_rate": 1.6115725206241078e-05,
+      "loss": 1.0995,
+      "step": 13236
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4665058025520712,
+      "learning_rate": 1.6115121479217724e-05,
+      "loss": 0.9248,
+      "step": 13237
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8326364903171333,
+      "learning_rate": 1.6114517716590205e-05,
+      "loss": 0.9668,
+      "step": 13238
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5257789329012466,
+      "learning_rate": 1.6113913918362023e-05,
+      "loss": 0.8899,
+      "step": 13239
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7042591838094623,
+      "learning_rate": 1.6113310084536698e-05,
+      "loss": 0.8291,
+      "step": 13240
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4929150187489053,
+      "learning_rate": 1.6112706215117748e-05,
+      "loss": 0.9783,
+      "step": 13241
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5453400060361588,
+      "learning_rate": 1.6112102310108686e-05,
+      "loss": 1.0101,
+      "step": 13242
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7686592829722547,
+      "learning_rate": 1.611149836951303e-05,
+      "loss": 0.891,
+      "step": 13243
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8007040089587553,
+      "learning_rate": 1.6110894393334293e-05,
+      "loss": 0.7521,
+      "step": 13244
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7710216744096279,
+      "learning_rate": 1.6110290381575996e-05,
+      "loss": 0.8848,
+      "step": 13245
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6440904459698937,
+      "learning_rate": 1.6109686334241655e-05,
+      "loss": 0.8903,
+      "step": 13246
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5602393347851304,
+      "learning_rate": 1.6109082251334786e-05,
+      "loss": 0.8869,
+      "step": 13247
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.626730422708558,
+      "learning_rate": 1.6108478132858903e-05,
+      "loss": 0.8424,
+      "step": 13248
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6667015666873048,
+      "learning_rate": 1.610787397881753e-05,
+      "loss": 0.8893,
+      "step": 13249
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.914995803142586,
+      "learning_rate": 1.610726978921418e-05,
+      "loss": 0.8465,
+      "step": 13250
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7045210186819273,
+      "learning_rate": 1.610666556405237e-05,
+      "loss": 0.857,
+      "step": 13251
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8828400962227336,
+      "learning_rate": 1.6106061303335622e-05,
+      "loss": 1.0755,
+      "step": 13252
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6177380525663065,
+      "learning_rate": 1.610545700706745e-05,
+      "loss": 0.8889,
+      "step": 13253
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3987514958155798,
+      "learning_rate": 1.6104852675251373e-05,
+      "loss": 0.889,
+      "step": 13254
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6281622307304735,
+      "learning_rate": 1.6104248307890915e-05,
+      "loss": 0.8557,
+      "step": 13255
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3928319041798756,
+      "learning_rate": 1.6103643904989588e-05,
+      "loss": 0.8813,
+      "step": 13256
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6138846826480775,
+      "learning_rate": 1.6103039466550916e-05,
+      "loss": 0.9165,
+      "step": 13257
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8831731581734337,
+      "learning_rate": 1.6102434992578415e-05,
+      "loss": 0.9037,
+      "step": 13258
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2755000050802088,
+      "learning_rate": 1.6101830483075603e-05,
+      "loss": 1.0219,
+      "step": 13259
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7599413737332976,
+      "learning_rate": 1.6101225938046007e-05,
+      "loss": 0.8865,
+      "step": 13260
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4953635511965673,
+      "learning_rate": 1.610062135749314e-05,
+      "loss": 0.9668,
+      "step": 13261
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9314469177146796,
+      "learning_rate": 1.610001674142052e-05,
+      "loss": 0.9436,
+      "step": 13262
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5172714430589447,
+      "learning_rate": 1.6099412089831676e-05,
+      "loss": 0.8491,
+      "step": 13263
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8982798345377168,
+      "learning_rate": 1.6098807402730124e-05,
+      "loss": 0.8211,
+      "step": 13264
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6045661812055538,
+      "learning_rate": 1.6098202680119382e-05,
+      "loss": 0.9061,
+      "step": 13265
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8250923714187968,
+      "learning_rate": 1.6097597922002974e-05,
+      "loss": 0.9204,
+      "step": 13266
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.62612308756287,
+      "learning_rate": 1.609699312838442e-05,
+      "loss": 0.9171,
+      "step": 13267
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.936944374728112,
+      "learning_rate": 1.609638829926724e-05,
+      "loss": 0.9145,
+      "step": 13268
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5865176824927074,
+      "learning_rate": 1.6095783434654963e-05,
+      "loss": 0.8763,
+      "step": 13269
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6844488496286534,
+      "learning_rate": 1.60951785345511e-05,
+      "loss": 0.8385,
+      "step": 13270
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3689997187463323,
+      "learning_rate": 1.609457359895918e-05,
+      "loss": 0.8313,
+      "step": 13271
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6071962941021058,
+      "learning_rate": 1.6093968627882723e-05,
+      "loss": 0.9323,
+      "step": 13272
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8301207693364494,
+      "learning_rate": 1.609336362132525e-05,
+      "loss": 1.0201,
+      "step": 13273
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.724749651459163,
+      "learning_rate": 1.6092758579290286e-05,
+      "loss": 0.8343,
+      "step": 13274
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5614439726829197,
+      "learning_rate": 1.6092153501781352e-05,
+      "loss": 0.8321,
+      "step": 13275
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7996900459061546,
+      "learning_rate": 1.6091548388801975e-05,
+      "loss": 0.7514,
+      "step": 13276
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6274661125722134,
+      "learning_rate": 1.609094324035567e-05,
+      "loss": 0.9683,
+      "step": 13277
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.759062532799867,
+      "learning_rate": 1.6090338056445965e-05,
+      "loss": 0.8902,
+      "step": 13278
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6657011156350767,
+      "learning_rate": 1.6089732837076387e-05,
+      "loss": 0.9094,
+      "step": 13279
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9410636478310417,
+      "learning_rate": 1.6089127582250453e-05,
+      "loss": 0.9709,
+      "step": 13280
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7809431894831464,
+      "learning_rate": 1.6088522291971695e-05,
+      "loss": 0.9648,
+      "step": 13281
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.096263461792773,
+      "learning_rate": 1.608791696624363e-05,
+      "loss": 0.9014,
+      "step": 13282
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9203555391144562,
+      "learning_rate": 1.608731160506978e-05,
+      "loss": 0.9834,
+      "step": 13283
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7602556852570705,
+      "learning_rate": 1.6086706208453683e-05,
+      "loss": 0.7268,
+      "step": 13284
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6491755426319548,
+      "learning_rate": 1.608610077639885e-05,
+      "loss": 0.8849,
+      "step": 13285
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7365320674118994,
+      "learning_rate": 1.6085495308908813e-05,
+      "loss": 0.8002,
+      "step": 13286
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7859309537430208,
+      "learning_rate": 1.60848898059871e-05,
+      "loss": 0.7246,
+      "step": 13287
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8205940656716356,
+      "learning_rate": 1.6084284267637228e-05,
+      "loss": 0.7909,
+      "step": 13288
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7710806768589046,
+      "learning_rate": 1.6083678693862725e-05,
+      "loss": 0.7411,
+      "step": 13289
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8248284156843195,
+      "learning_rate": 1.608307308466712e-05,
+      "loss": 0.918,
+      "step": 13290
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.211659088028223,
+      "learning_rate": 1.608246744005394e-05,
+      "loss": 0.9448,
+      "step": 13291
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7340525887339576,
+      "learning_rate": 1.6081861760026708e-05,
+      "loss": 0.6975,
+      "step": 13292
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5464027631944817,
+      "learning_rate": 1.608125604458895e-05,
+      "loss": 0.991,
+      "step": 13293
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9482616063873508,
+      "learning_rate": 1.6080650293744192e-05,
+      "loss": 0.8924,
+      "step": 13294
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9400322954582039,
+      "learning_rate": 1.6080044507495967e-05,
+      "loss": 0.9101,
+      "step": 13295
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5443059635311034,
+      "learning_rate": 1.6079438685847796e-05,
+      "loss": 0.9548,
+      "step": 13296
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.473346892564918,
+      "learning_rate": 1.607883282880321e-05,
+      "loss": 0.9854,
+      "step": 13297
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5276968233410058,
+      "learning_rate": 1.607822693636573e-05,
+      "loss": 0.8413,
+      "step": 13298
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.59955823579898,
+      "learning_rate": 1.607762100853889e-05,
+      "loss": 0.8932,
+      "step": 13299
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8623664928884804,
+      "learning_rate": 1.607701504532622e-05,
+      "loss": 0.7622,
+      "step": 13300
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.988420001083008,
+      "learning_rate": 1.6076409046731243e-05,
+      "loss": 0.899,
+      "step": 13301
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9637074684180642,
+      "learning_rate": 1.607580301275749e-05,
+      "loss": 0.887,
+      "step": 13302
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.818402534732853,
+      "learning_rate": 1.6075196943408485e-05,
+      "loss": 0.8723,
+      "step": 13303
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5541644614005266,
+      "learning_rate": 1.607459083868776e-05,
+      "loss": 0.7462,
+      "step": 13304
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.343420358261279,
+      "learning_rate": 1.6073984698598845e-05,
+      "loss": 0.8613,
+      "step": 13305
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6294763968249693,
+      "learning_rate": 1.6073378523145272e-05,
+      "loss": 0.8941,
+      "step": 13306
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0483174521730727,
+      "learning_rate": 1.607277231233056e-05,
+      "loss": 0.9864,
+      "step": 13307
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2415659843750144,
+      "learning_rate": 1.607216606615825e-05,
+      "loss": 0.9914,
+      "step": 13308
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.633998954843783,
+      "learning_rate": 1.6071559784631862e-05,
+      "loss": 0.7604,
+      "step": 13309
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6405276314510466,
+      "learning_rate": 1.6070953467754935e-05,
+      "loss": 0.9509,
+      "step": 13310
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5173043991029533,
+      "learning_rate": 1.6070347115530994e-05,
+      "loss": 0.8618,
+      "step": 13311
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.661471939139421,
+      "learning_rate": 1.606974072796357e-05,
+      "loss": 0.8062,
+      "step": 13312
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6153600445948488,
+      "learning_rate": 1.6069134305056193e-05,
+      "loss": 0.8935,
+      "step": 13313
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.681054704507441,
+      "learning_rate": 1.6068527846812398e-05,
+      "loss": 0.8478,
+      "step": 13314
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5671545104548545,
+      "learning_rate": 1.6067921353235707e-05,
+      "loss": 0.9075,
+      "step": 13315
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.84443156906302,
+      "learning_rate": 1.606731482432966e-05,
+      "loss": 1.0455,
+      "step": 13316
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5632183543894451,
+      "learning_rate": 1.6066708260097782e-05,
+      "loss": 0.9763,
+      "step": 13317
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.645897750662952,
+      "learning_rate": 1.6066101660543612e-05,
+      "loss": 0.9889,
+      "step": 13318
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5556578467296387,
+      "learning_rate": 1.6065495025670675e-05,
+      "loss": 0.8747,
+      "step": 13319
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5549161475702407,
+      "learning_rate": 1.606488835548251e-05,
+      "loss": 0.956,
+      "step": 13320
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.581651867692216,
+      "learning_rate": 1.6064281649982637e-05,
+      "loss": 0.9974,
+      "step": 13321
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5195812092863243,
+      "learning_rate": 1.6063674909174604e-05,
+      "loss": 0.953,
+      "step": 13322
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6127317962614665,
+      "learning_rate": 1.606306813306193e-05,
+      "loss": 0.867,
+      "step": 13323
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.811212617436318,
+      "learning_rate": 1.606246132164816e-05,
+      "loss": 0.8787,
+      "step": 13324
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7371298095763215,
+      "learning_rate": 1.606185447493682e-05,
+      "loss": 1.0252,
+      "step": 13325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7590565729541174,
+      "learning_rate": 1.606124759293144e-05,
+      "loss": 0.873,
+      "step": 13326
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4485351409849743,
+      "learning_rate": 1.606064067563556e-05,
+      "loss": 0.8752,
+      "step": 13327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6953712671697816,
+      "learning_rate": 1.6060033723052707e-05,
+      "loss": 1.0418,
+      "step": 13328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7865306645039245,
+      "learning_rate": 1.6059426735186426e-05,
+      "loss": 0.8658,
+      "step": 13329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6892519040843235,
+      "learning_rate": 1.6058819712040237e-05,
+      "loss": 0.9901,
+      "step": 13330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.631484565924258,
+      "learning_rate": 1.6058212653617688e-05,
+      "loss": 0.9249,
+      "step": 13331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5845094686795937,
+      "learning_rate": 1.60576055599223e-05,
+      "loss": 0.8882,
+      "step": 13332
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6551820786233,
+      "learning_rate": 1.605699843095762e-05,
+      "loss": 0.8367,
+      "step": 13333
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.808260203899369,
+      "learning_rate": 1.6056391266727178e-05,
+      "loss": 0.7964,
+      "step": 13334
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.900690240286219,
+      "learning_rate": 1.6055784067234507e-05,
+      "loss": 0.8832,
+      "step": 13335
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5813967611649076,
+      "learning_rate": 1.6055176832483143e-05,
+      "loss": 0.8602,
+      "step": 13336
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.73262570529994,
+      "learning_rate": 1.6054569562476622e-05,
+      "loss": 0.9876,
+      "step": 13337
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8342748317533613,
+      "learning_rate": 1.6053962257218483e-05,
+      "loss": 0.7487,
+      "step": 13338
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7942304435721074,
+      "learning_rate": 1.6053354916712258e-05,
+      "loss": 0.7541,
+      "step": 13339
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7613651864543356,
+      "learning_rate": 1.6052747540961483e-05,
+      "loss": 0.8245,
+      "step": 13340
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.653932329472362,
+      "learning_rate": 1.60521401299697e-05,
+      "loss": 0.8791,
+      "step": 13341
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.749597440455945,
+      "learning_rate": 1.6051532683740433e-05,
+      "loss": 0.9285,
+      "step": 13342
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.619083434699407,
+      "learning_rate": 1.6050925202277232e-05,
+      "loss": 0.8711,
+      "step": 13343
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.665823545226649,
+      "learning_rate": 1.6050317685583627e-05,
+      "loss": 0.9256,
+      "step": 13344
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6549305817459397,
+      "learning_rate": 1.604971013366316e-05,
+      "loss": 0.7939,
+      "step": 13345
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5436159595600545,
+      "learning_rate": 1.6049102546519363e-05,
+      "loss": 0.8315,
+      "step": 13346
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5409651472923895,
+      "learning_rate": 1.6048494924155776e-05,
+      "loss": 0.8257,
+      "step": 13347
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6593481144154314,
+      "learning_rate": 1.6047887266575936e-05,
+      "loss": 0.9802,
+      "step": 13348
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7694757527975953,
+      "learning_rate": 1.6047279573783382e-05,
+      "loss": 0.8926,
+      "step": 13349
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7431632683555813,
+      "learning_rate": 1.6046671845781653e-05,
+      "loss": 0.9369,
+      "step": 13350
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6330455988740082,
+      "learning_rate": 1.6046064082574284e-05,
+      "loss": 0.9397,
+      "step": 13351
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8898838720076774,
+      "learning_rate": 1.604545628416482e-05,
+      "loss": 0.9395,
+      "step": 13352
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5165553757387444,
+      "learning_rate": 1.6044848450556788e-05,
+      "loss": 0.8242,
+      "step": 13353
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7431652994324169,
+      "learning_rate": 1.6044240581753742e-05,
+      "loss": 0.7261,
+      "step": 13354
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4200559119769107,
+      "learning_rate": 1.604363267775921e-05,
+      "loss": 0.8121,
+      "step": 13355
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4644292480805525,
+      "learning_rate": 1.604302473857674e-05,
+      "loss": 0.7627,
+      "step": 13356
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9955933857712527,
+      "learning_rate": 1.604241676420986e-05,
+      "loss": 0.7342,
+      "step": 13357
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3785084254665207,
+      "learning_rate": 1.604180875466212e-05,
+      "loss": 0.8803,
+      "step": 13358
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6470521588162206,
+      "learning_rate": 1.6041200709937058e-05,
+      "loss": 0.8969,
+      "step": 13359
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1229901701014398,
+      "learning_rate": 1.604059263003821e-05,
+      "loss": 1.0489,
+      "step": 13360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3618853080664444,
+      "learning_rate": 1.603998451496912e-05,
+      "loss": 0.8875,
+      "step": 13361
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7662832002329785,
+      "learning_rate": 1.6039376364733328e-05,
+      "loss": 0.8578,
+      "step": 13362
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5470870888899282,
+      "learning_rate": 1.6038768179334373e-05,
+      "loss": 0.779,
+      "step": 13363
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9106739518944265,
+      "learning_rate": 1.60381599587758e-05,
+      "loss": 0.8048,
+      "step": 13364
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7510492318115267,
+      "learning_rate": 1.6037551703061148e-05,
+      "loss": 0.7348,
+      "step": 13365
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0502936453385496,
+      "learning_rate": 1.603694341219396e-05,
+      "loss": 1.0093,
+      "step": 13366
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.378198795497905,
+      "learning_rate": 1.6036335086177772e-05,
+      "loss": 0.9116,
+      "step": 13367
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.828148681453363,
+      "learning_rate": 1.6035726725016134e-05,
+      "loss": 0.7724,
+      "step": 13368
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8074717380875374,
+      "learning_rate": 1.6035118328712583e-05,
+      "loss": 0.8814,
+      "step": 13369
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5217614329453124,
+      "learning_rate": 1.603450989727066e-05,
+      "loss": 0.9842,
+      "step": 13370
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7588037306240506,
+      "learning_rate": 1.603390143069391e-05,
+      "loss": 0.7638,
+      "step": 13371
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3944064861577092,
+      "learning_rate": 1.603329292898588e-05,
+      "loss": 0.8551,
+      "step": 13372
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4679451822215033,
+      "learning_rate": 1.6032684392150108e-05,
+      "loss": 0.9154,
+      "step": 13373
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7243785877816236,
+      "learning_rate": 1.6032075820190134e-05,
+      "loss": 0.7205,
+      "step": 13374
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.005926340077844,
+      "learning_rate": 1.6031467213109505e-05,
+      "loss": 0.9122,
+      "step": 13375
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5926958629979546,
+      "learning_rate": 1.6030858570911768e-05,
+      "loss": 0.7921,
+      "step": 13376
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5722599653251736,
+      "learning_rate": 1.603024989360046e-05,
+      "loss": 0.8054,
+      "step": 13377
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8016328412615519,
+      "learning_rate": 1.602964118117913e-05,
+      "loss": 1.0011,
+      "step": 13378
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7594969264779652,
+      "learning_rate": 1.6029032433651317e-05,
+      "loss": 0.7446,
+      "step": 13379
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.746840272902509,
+      "learning_rate": 1.6028423651020572e-05,
+      "loss": 1.0151,
+      "step": 13380
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.365154808831526,
+      "learning_rate": 1.6027814833290435e-05,
+      "loss": 0.8328,
+      "step": 13381
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.004391471518366,
+      "learning_rate": 1.602720598046445e-05,
+      "loss": 0.7144,
+      "step": 13382
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7131504713131975,
+      "learning_rate": 1.6026597092546164e-05,
+      "loss": 0.9234,
+      "step": 13383
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5781583091288172,
+      "learning_rate": 1.6025988169539122e-05,
+      "loss": 0.9162,
+      "step": 13384
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8072543816049662,
+      "learning_rate": 1.602537921144687e-05,
+      "loss": 0.9937,
+      "step": 13385
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5919148127367138,
+      "learning_rate": 1.602477021827295e-05,
+      "loss": 1.0434,
+      "step": 13386
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.017710655512892,
+      "learning_rate": 1.6024161190020913e-05,
+      "loss": 0.9942,
+      "step": 13387
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6285596858429956,
+      "learning_rate": 1.60235521266943e-05,
+      "loss": 0.9254,
+      "step": 13388
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.622626445932437,
+      "learning_rate": 1.6022943028296663e-05,
+      "loss": 0.9136,
+      "step": 13389
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5472572234602382,
+      "learning_rate": 1.6022333894831543e-05,
+      "loss": 0.8979,
+      "step": 13390
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9086336329980822,
+      "learning_rate": 1.6021724726302487e-05,
+      "loss": 0.8954,
+      "step": 13391
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8059807106741306,
+      "learning_rate": 1.6021115522713047e-05,
+      "loss": 0.7409,
+      "step": 13392
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7232018028458413,
+      "learning_rate": 1.6020506284066763e-05,
+      "loss": 0.9859,
+      "step": 13393
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8562161043816554,
+      "learning_rate": 1.6019897010367186e-05,
+      "loss": 0.8551,
+      "step": 13394
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6287156262116131,
+      "learning_rate": 1.6019287701617863e-05,
+      "loss": 1.0456,
+      "step": 13395
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.77196931396854,
+      "learning_rate": 1.6018678357822343e-05,
+      "loss": 0.7477,
+      "step": 13396
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5266794064165297,
+      "learning_rate": 1.601806897898417e-05,
+      "loss": 0.9289,
+      "step": 13397
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7592544127685966,
+      "learning_rate": 1.6017459565106896e-05,
+      "loss": 0.7127,
+      "step": 13398
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.678555476204215,
+      "learning_rate": 1.601685011619406e-05,
+      "loss": 0.9025,
+      "step": 13399
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4508830337131273,
+      "learning_rate": 1.6016240632249224e-05,
+      "loss": 0.8512,
+      "step": 13400
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7887996291078568,
+      "learning_rate": 1.601563111327593e-05,
+      "loss": 0.7647,
+      "step": 13401
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.589544327545359,
+      "learning_rate": 1.601502155927773e-05,
+      "loss": 0.9084,
+      "step": 13402
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.642888786356929,
+      "learning_rate": 1.6014411970258166e-05,
+      "loss": 0.9578,
+      "step": 13403
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7574307121804846,
+      "learning_rate": 1.601380234622079e-05,
+      "loss": 0.9105,
+      "step": 13404
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.779055699359544,
+      "learning_rate": 1.601319268716916e-05,
+      "loss": 0.685,
+      "step": 13405
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9764671847801587,
+      "learning_rate": 1.6012582993106812e-05,
+      "loss": 0.9779,
+      "step": 13406
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.734465013052186,
+      "learning_rate": 1.6011973264037305e-05,
+      "loss": 0.748,
+      "step": 13407
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7873690963837311,
+      "learning_rate": 1.6011363499964185e-05,
+      "loss": 0.6957,
+      "step": 13408
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5011300628458633,
+      "learning_rate": 1.6010753700891003e-05,
+      "loss": 0.7453,
+      "step": 13409
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5151279740145744,
+      "learning_rate": 1.601014386682131e-05,
+      "loss": 0.8646,
+      "step": 13410
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7450611452591143,
+      "learning_rate": 1.600953399775866e-05,
+      "loss": 0.9554,
+      "step": 13411
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6733052318253177,
+      "learning_rate": 1.60089240937066e-05,
+      "loss": 0.9076,
+      "step": 13412
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7322940827665572,
+      "learning_rate": 1.6008314154668676e-05,
+      "loss": 0.8259,
+      "step": 13413
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5269844545960565,
+      "learning_rate": 1.6007704180648448e-05,
+      "loss": 0.8759,
+      "step": 13414
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5606777856469733,
+      "learning_rate": 1.6007094171649463e-05,
+      "loss": 0.8409,
+      "step": 13415
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.472967162790429,
+      "learning_rate": 1.6006484127675278e-05,
+      "loss": 0.8364,
+      "step": 13416
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7702490072138806,
+      "learning_rate": 1.6005874048729437e-05,
+      "loss": 1.0055,
+      "step": 13417
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7856131387055063,
+      "learning_rate": 1.6005263934815498e-05,
+      "loss": 0.9191,
+      "step": 13418
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7453246168138188,
+      "learning_rate": 1.600465378593701e-05,
+      "loss": 0.8636,
+      "step": 13419
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4243568495012946,
+      "learning_rate": 1.6004043602097527e-05,
+      "loss": 0.9136,
+      "step": 13420
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3922190277606887,
+      "learning_rate": 1.6003433383300603e-05,
+      "loss": 0.9348,
+      "step": 13421
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0552901305073865,
+      "learning_rate": 1.6002823129549784e-05,
+      "loss": 0.8123,
+      "step": 13422
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7635820644049047,
+      "learning_rate": 1.600221284084863e-05,
+      "loss": 0.86,
+      "step": 13423
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0581594805683756,
+      "learning_rate": 1.6001602517200698e-05,
+      "loss": 0.8904,
+      "step": 13424
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7335822694137153,
+      "learning_rate": 1.600099215860953e-05,
+      "loss": 0.8392,
+      "step": 13425
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.635223574092222,
+      "learning_rate": 1.600038176507869e-05,
+      "loss": 0.8207,
+      "step": 13426
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5558946945008645,
+      "learning_rate": 1.5999771336611726e-05,
+      "loss": 0.7771,
+      "step": 13427
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.511949802185267,
+      "learning_rate": 1.5999160873212193e-05,
+      "loss": 0.8577,
+      "step": 13428
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9929621142734784,
+      "learning_rate": 1.599855037488365e-05,
+      "loss": 0.8554,
+      "step": 13429
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.768993006750124,
+      "learning_rate": 1.599793984162964e-05,
+      "loss": 0.9713,
+      "step": 13430
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.788134939057076,
+      "learning_rate": 1.599732927345373e-05,
+      "loss": 1.0287,
+      "step": 13431
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5483537621520642,
+      "learning_rate": 1.599671867035947e-05,
+      "loss": 0.9189,
+      "step": 13432
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6784166010893615,
+      "learning_rate": 1.5996108032350414e-05,
+      "loss": 0.9178,
+      "step": 13433
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9087569499539683,
+      "learning_rate": 1.599549735943012e-05,
+      "loss": 0.8643,
+      "step": 13434
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1728938230814223,
+      "learning_rate": 1.599488665160214e-05,
+      "loss": 0.9782,
+      "step": 13435
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8545760299425171,
+      "learning_rate": 1.5994275908870037e-05,
+      "loss": 0.9991,
+      "step": 13436
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5723337470795573,
+      "learning_rate": 1.599366513123736e-05,
+      "loss": 0.984,
+      "step": 13437
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.712669478409587,
+      "learning_rate": 1.5993054318707665e-05,
+      "loss": 0.9429,
+      "step": 13438
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.522930618203426,
+      "learning_rate": 1.599244347128451e-05,
+      "loss": 0.883,
+      "step": 13439
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6358639092847436,
+      "learning_rate": 1.5991832588971454e-05,
+      "loss": 0.9016,
+      "step": 13440
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8218632608050265,
+      "learning_rate": 1.5991221671772053e-05,
+      "loss": 0.8304,
+      "step": 13441
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.464969583181455,
+      "learning_rate": 1.5990610719689858e-05,
+      "loss": 0.8943,
+      "step": 13442
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7474850455165176,
+      "learning_rate": 1.5989999732728434e-05,
+      "loss": 0.9346,
+      "step": 13443
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5901175772314533,
+      "learning_rate": 1.5989388710891334e-05,
+      "loss": 0.8041,
+      "step": 13444
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8871909865058767,
+      "learning_rate": 1.5988777654182118e-05,
+      "loss": 0.8911,
+      "step": 13445
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.154193137089203,
+      "learning_rate": 1.5988166562604343e-05,
+      "loss": 0.8379,
+      "step": 13446
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4681374879001075,
+      "learning_rate": 1.5987555436161564e-05,
+      "loss": 0.8164,
+      "step": 13447
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6086391742248811,
+      "learning_rate": 1.5986944274857344e-05,
+      "loss": 0.8831,
+      "step": 13448
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7199546465993996,
+      "learning_rate": 1.598633307869524e-05,
+      "loss": 1.0196,
+      "step": 13449
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8306698677364601,
+      "learning_rate": 1.5985721847678806e-05,
+      "loss": 1.1304,
+      "step": 13450
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.767759774717876,
+      "learning_rate": 1.59851105818116e-05,
+      "loss": 0.9178,
+      "step": 13451
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.544065948794481,
+      "learning_rate": 1.5984499281097194e-05,
+      "loss": 1.0395,
+      "step": 13452
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7479982494372086,
+      "learning_rate": 1.5983887945539138e-05,
+      "loss": 0.969,
+      "step": 13453
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5453251352772364,
+      "learning_rate": 1.5983276575140986e-05,
+      "loss": 0.935,
+      "step": 13454
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6062215214911684,
+      "learning_rate": 1.598266516990631e-05,
+      "loss": 0.8981,
+      "step": 13455
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7774682460041626,
+      "learning_rate": 1.5982053729838657e-05,
+      "loss": 0.7712,
+      "step": 13456
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.275373110990287,
+      "learning_rate": 1.59814422549416e-05,
+      "loss": 0.8863,
+      "step": 13457
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5312633517243381,
+      "learning_rate": 1.5980830745218686e-05,
+      "loss": 0.8756,
+      "step": 13458
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6060146274413125,
+      "learning_rate": 1.5980219200673488e-05,
+      "loss": 0.9806,
+      "step": 13459
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5363587516359358,
+      "learning_rate": 1.5979607621309557e-05,
+      "loss": 0.8512,
+      "step": 13460
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.900572636892893,
+      "learning_rate": 1.597899600713046e-05,
+      "loss": 0.8598,
+      "step": 13461
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.66181844483198,
+      "learning_rate": 1.5978384358139752e-05,
+      "loss": 0.9152,
+      "step": 13462
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.710561945778161,
+      "learning_rate": 1.5977772674340996e-05,
+      "loss": 0.957,
+      "step": 13463
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6362946140967365,
+      "learning_rate": 1.5977160955737758e-05,
+      "loss": 0.8634,
+      "step": 13464
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.780344601342769,
+      "learning_rate": 1.59765492023336e-05,
+      "loss": 0.7219,
+      "step": 13465
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4821850143904798,
+      "learning_rate": 1.5975937414132076e-05,
+      "loss": 0.9055,
+      "step": 13466
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5106117503772571,
+      "learning_rate": 1.5975325591136753e-05,
+      "loss": 1.0359,
+      "step": 13467
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8934820316253094,
+      "learning_rate": 1.5974713733351192e-05,
+      "loss": 0.8385,
+      "step": 13468
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9025305355593176,
+      "learning_rate": 1.5974101840778954e-05,
+      "loss": 0.8056,
+      "step": 13469
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8935744544057904,
+      "learning_rate": 1.597348991342361e-05,
+      "loss": 0.9837,
+      "step": 13470
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.62117530290152,
+      "learning_rate": 1.597287795128871e-05,
+      "loss": 0.9398,
+      "step": 13471
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.978636016247216,
+      "learning_rate": 1.597226595437783e-05,
+      "loss": 0.9277,
+      "step": 13472
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5954270374613702,
+      "learning_rate": 1.597165392269452e-05,
+      "loss": 0.8985,
+      "step": 13473
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8324802534327997,
+      "learning_rate": 1.5971041856242356e-05,
+      "loss": 0.8942,
+      "step": 13474
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7441890511030798,
+      "learning_rate": 1.5970429755024893e-05,
+      "loss": 0.7247,
+      "step": 13475
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5626451260999725,
+      "learning_rate": 1.59698176190457e-05,
+      "loss": 0.9563,
+      "step": 13476
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5716451021254205,
+      "learning_rate": 1.5969205448308335e-05,
+      "loss": 0.904,
+      "step": 13477
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7839771693028731,
+      "learning_rate": 1.596859324281637e-05,
+      "loss": 0.7449,
+      "step": 13478
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.567938432150312,
+      "learning_rate": 1.5967981002573362e-05,
+      "loss": 0.9616,
+      "step": 13479
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7612691071398883,
+      "learning_rate": 1.5967368727582882e-05,
+      "loss": 0.7351,
+      "step": 13480
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5524804509714432,
+      "learning_rate": 1.5966756417848492e-05,
+      "loss": 0.9666,
+      "step": 13481
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6028982691342712,
+      "learning_rate": 1.5966144073373755e-05,
+      "loss": 0.8571,
+      "step": 13482
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8462071631569217,
+      "learning_rate": 1.596553169416224e-05,
+      "loss": 0.7582,
+      "step": 13483
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4630928460813437,
+      "learning_rate": 1.596491928021751e-05,
+      "loss": 0.7992,
+      "step": 13484
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5381522150367741,
+      "learning_rate": 1.5964306831543132e-05,
+      "loss": 0.9163,
+      "step": 13485
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4568518800338779,
+      "learning_rate": 1.5963694348142672e-05,
+      "loss": 0.8583,
+      "step": 13486
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8384962337235918,
+      "learning_rate": 1.5963081830019692e-05,
+      "loss": 0.8012,
+      "step": 13487
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6293280596587916,
+      "learning_rate": 1.5962469277177766e-05,
+      "loss": 0.867,
+      "step": 13488
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6831986804324943,
+      "learning_rate": 1.5961856689620453e-05,
+      "loss": 0.864,
+      "step": 13489
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.721296959880215,
+      "learning_rate": 1.5961244067351326e-05,
+      "loss": 0.8346,
+      "step": 13490
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7421776551884617,
+      "learning_rate": 1.5960631410373947e-05,
+      "loss": 0.7369,
+      "step": 13491
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5523136286932622,
+      "learning_rate": 1.5960018718691886e-05,
+      "loss": 0.9926,
+      "step": 13492
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6291850376425165,
+      "learning_rate": 1.5959405992308705e-05,
+      "loss": 0.8601,
+      "step": 13493
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5977708331246043,
+      "learning_rate": 1.595879323122798e-05,
+      "loss": 0.997,
+      "step": 13494
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5647434733575458,
+      "learning_rate": 1.5958180435453273e-05,
+      "loss": 0.9786,
+      "step": 13495
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8559014289118556,
+      "learning_rate": 1.5957567604988154e-05,
+      "loss": 0.8601,
+      "step": 13496
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7947245062300372,
+      "learning_rate": 1.595695473983619e-05,
+      "loss": 0.8804,
+      "step": 13497
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6152932875313173,
+      "learning_rate": 1.5956341840000948e-05,
+      "loss": 0.9348,
+      "step": 13498
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6472021266286245,
+      "learning_rate": 1.5955728905486e-05,
+      "loss": 0.9057,
+      "step": 13499
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0480417187944098,
+      "learning_rate": 1.595511593629491e-05,
+      "loss": 0.8849,
+      "step": 13500
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7694687675549632,
+      "learning_rate": 1.595450293243125e-05,
+      "loss": 0.9101,
+      "step": 13501
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.742858382679007,
+      "learning_rate": 1.5953889893898592e-05,
+      "loss": 0.9746,
+      "step": 13502
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9822967156236335,
+      "learning_rate": 1.5953276820700502e-05,
+      "loss": 0.8386,
+      "step": 13503
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5581056837824736,
+      "learning_rate": 1.5952663712840542e-05,
+      "loss": 0.881,
+      "step": 13504
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.581183048089411,
+      "learning_rate": 1.59520505703223e-05,
+      "loss": 0.9079,
+      "step": 13505
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.723641851647317,
+      "learning_rate": 1.595143739314933e-05,
+      "loss": 1.0199,
+      "step": 13506
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.912187187919737,
+      "learning_rate": 1.5950824181325206e-05,
+      "loss": 1.0396,
+      "step": 13507
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.578982876677069,
+      "learning_rate": 1.5950210934853502e-05,
+      "loss": 0.9858,
+      "step": 13508
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5822739382611948,
+      "learning_rate": 1.5949597653737785e-05,
+      "loss": 0.8522,
+      "step": 13509
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1034416083295677,
+      "learning_rate": 1.594898433798163e-05,
+      "loss": 1.0145,
+      "step": 13510
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.559219991604783,
+      "learning_rate": 1.59483709875886e-05,
+      "loss": 0.7992,
+      "step": 13511
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8940040659093074,
+      "learning_rate": 1.5947757602562273e-05,
+      "loss": 0.9308,
+      "step": 13512
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7722359946068087,
+      "learning_rate": 1.5947144182906217e-05,
+      "loss": 0.963,
+      "step": 13513
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5525832706791118,
+      "learning_rate": 1.5946530728624005e-05,
+      "loss": 0.899,
+      "step": 13514
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7524491045159107,
+      "learning_rate": 1.594591723971921e-05,
+      "loss": 0.8087,
+      "step": 13515
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6485377304610578,
+      "learning_rate": 1.5945303716195403e-05,
+      "loss": 1.005,
+      "step": 13516
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4137236920490737,
+      "learning_rate": 1.5944690158056154e-05,
+      "loss": 0.8403,
+      "step": 13517
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4838438410713253,
+      "learning_rate": 1.5944076565305032e-05,
+      "loss": 1.0205,
+      "step": 13518
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7230040214969404,
+      "learning_rate": 1.5943462937945624e-05,
+      "loss": 0.9838,
+      "step": 13519
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1329564528040574,
+      "learning_rate": 1.5942849275981488e-05,
+      "loss": 0.8988,
+      "step": 13520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1338921047128494,
+      "learning_rate": 1.59422355794162e-05,
+      "loss": 0.9203,
+      "step": 13521
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2167341047749853,
+      "learning_rate": 1.5941621848253343e-05,
+      "loss": 0.9038,
+      "step": 13522
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6137186346634107,
+      "learning_rate": 1.5941008082496473e-05,
+      "loss": 0.876,
+      "step": 13523
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.401843502254692,
+      "learning_rate": 1.594039428214918e-05,
+      "loss": 0.8813,
+      "step": 13524
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.223547001832459,
+      "learning_rate": 1.5939780447215024e-05,
+      "loss": 0.8191,
+      "step": 13525
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5182193342305594,
+      "learning_rate": 1.5939166577697593e-05,
+      "loss": 0.8731,
+      "step": 13526
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9213840525043946,
+      "learning_rate": 1.593855267360045e-05,
+      "loss": 0.8499,
+      "step": 13527
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3855197626672313,
+      "learning_rate": 1.5937938734927174e-05,
+      "loss": 0.7966,
+      "step": 13528
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.586169955190888,
+      "learning_rate": 1.593732476168134e-05,
+      "loss": 0.9046,
+      "step": 13529
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.607318109813841,
+      "learning_rate": 1.5936710753866516e-05,
+      "loss": 0.8731,
+      "step": 13530
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8645636311611824,
+      "learning_rate": 1.5936096711486287e-05,
+      "loss": 0.7651,
+      "step": 13531
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.792981714693356,
+      "learning_rate": 1.5935482634544227e-05,
+      "loss": 0.7252,
+      "step": 13532
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9299064603220903,
+      "learning_rate": 1.5934868523043903e-05,
+      "loss": 0.7762,
+      "step": 13533
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.588611171702064,
+      "learning_rate": 1.5934254376988895e-05,
+      "loss": 0.8954,
+      "step": 13534
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6371871134004057,
+      "learning_rate": 1.5933640196382782e-05,
+      "loss": 0.9406,
+      "step": 13535
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.716695800787938,
+      "learning_rate": 1.5933025981229136e-05,
+      "loss": 0.9599,
+      "step": 13536
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.527950497812951,
+      "learning_rate": 1.5932411731531535e-05,
+      "loss": 0.9081,
+      "step": 13537
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8642802130663931,
+      "learning_rate": 1.5931797447293553e-05,
+      "loss": 0.7456,
+      "step": 13538
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.739806296496257,
+      "learning_rate": 1.593118312851877e-05,
+      "loss": 0.7052,
+      "step": 13539
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8338981495540654,
+      "learning_rate": 1.5930568775210762e-05,
+      "loss": 0.9031,
+      "step": 13540
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.676820712271961,
+      "learning_rate": 1.5929954387373103e-05,
+      "loss": 0.85,
+      "step": 13541
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0601614496985414,
+      "learning_rate": 1.5929339965009373e-05,
+      "loss": 0.9436,
+      "step": 13542
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5279542244955868,
+      "learning_rate": 1.592872550812315e-05,
+      "loss": 0.7694,
+      "step": 13543
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7850199441600783,
+      "learning_rate": 1.5928111016718005e-05,
+      "loss": 0.6871,
+      "step": 13544
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7979227715780117,
+      "learning_rate": 1.5927496490797526e-05,
+      "loss": 0.95,
+      "step": 13545
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8016752628926154,
+      "learning_rate": 1.5926881930365285e-05,
+      "loss": 0.9134,
+      "step": 13546
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.66047494886655,
+      "learning_rate": 1.5926267335424857e-05,
+      "loss": 0.7172,
+      "step": 13547
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7415776103414659,
+      "learning_rate": 1.5925652705979828e-05,
+      "loss": 0.7438,
+      "step": 13548
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.863079298456774,
+      "learning_rate": 1.592503804203377e-05,
+      "loss": 0.9065,
+      "step": 13549
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.99032491397535,
+      "learning_rate": 1.592442334359027e-05,
+      "loss": 0.8282,
+      "step": 13550
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8891097170139537,
+      "learning_rate": 1.59238086106529e-05,
+      "loss": 0.8621,
+      "step": 13551
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6446172171178954,
+      "learning_rate": 1.592319384322524e-05,
+      "loss": 0.9785,
+      "step": 13552
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.984887099225787,
+      "learning_rate": 1.5922579041310866e-05,
+      "loss": 0.8666,
+      "step": 13553
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8360754299854736,
+      "learning_rate": 1.5921964204913366e-05,
+      "loss": 0.9881,
+      "step": 13554
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.769430326853215,
+      "learning_rate": 1.5921349334036316e-05,
+      "loss": 0.8942,
+      "step": 13555
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4988137695993626,
+      "learning_rate": 1.5920734428683295e-05,
+      "loss": 0.8701,
+      "step": 13556
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6071063723498025,
+      "learning_rate": 1.5920119488857885e-05,
+      "loss": 0.9281,
+      "step": 13557
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5275580324166407,
+      "learning_rate": 1.591950451456366e-05,
+      "loss": 0.8203,
+      "step": 13558
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8024013846927383,
+      "learning_rate": 1.591888950580421e-05,
+      "loss": 0.9525,
+      "step": 13559
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9539132445996945,
+      "learning_rate": 1.591827446258311e-05,
+      "loss": 0.8755,
+      "step": 13560
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6214419626849463,
+      "learning_rate": 1.5917659384903944e-05,
+      "loss": 0.9202,
+      "step": 13561
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5194877730759273,
+      "learning_rate": 1.5917044272770288e-05,
+      "loss": 0.9149,
+      "step": 13562
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0970589142834064,
+      "learning_rate": 1.5916429126185732e-05,
+      "loss": 0.8732,
+      "step": 13563
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8339084115758453,
+      "learning_rate": 1.5915813945153847e-05,
+      "loss": 0.8572,
+      "step": 13564
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.354509503990565,
+      "learning_rate": 1.5915198729678224e-05,
+      "loss": 0.8531,
+      "step": 13565
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7591600536151517,
+      "learning_rate": 1.591458347976244e-05,
+      "loss": 0.9844,
+      "step": 13566
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5238899711951661,
+      "learning_rate": 1.5913968195410077e-05,
+      "loss": 0.9181,
+      "step": 13567
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6447761168689945,
+      "learning_rate": 1.5913352876624722e-05,
+      "loss": 0.9367,
+      "step": 13568
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6098675612085274,
+      "learning_rate": 1.591273752340995e-05,
+      "loss": 0.8242,
+      "step": 13569
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.492509029804116,
+      "learning_rate": 1.5912122135769356e-05,
+      "loss": 0.8595,
+      "step": 13570
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.849967988667861,
+      "learning_rate": 1.5911506713706507e-05,
+      "loss": 0.8363,
+      "step": 13571
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5526922461092176,
+      "learning_rate": 1.5910891257225e-05,
+      "loss": 0.9189,
+      "step": 13572
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.64252514119755,
+      "learning_rate": 1.5910275766328408e-05,
+      "loss": 0.9268,
+      "step": 13573
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.650931549622106,
+      "learning_rate": 1.5909660241020323e-05,
+      "loss": 1.0375,
+      "step": 13574
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7770086347406181,
+      "learning_rate": 1.5909044681304324e-05,
+      "loss": 0.7344,
+      "step": 13575
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.769931920991025,
+      "learning_rate": 1.5908429087183998e-05,
+      "loss": 0.7931,
+      "step": 13576
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.840758760774042,
+      "learning_rate": 1.5907813458662923e-05,
+      "loss": 1.0472,
+      "step": 13577
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7858349890798608,
+      "learning_rate": 1.590719779574469e-05,
+      "loss": 0.9391,
+      "step": 13578
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7075641226461635,
+      "learning_rate": 1.590658209843288e-05,
+      "loss": 0.8575,
+      "step": 13579
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7596976907100706,
+      "learning_rate": 1.590596636673108e-05,
+      "loss": 0.7372,
+      "step": 13580
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6901546478305494,
+      "learning_rate": 1.5905350600642876e-05,
+      "loss": 0.9001,
+      "step": 13581
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.76370997662356,
+      "learning_rate": 1.590473480017185e-05,
+      "loss": 0.7153,
+      "step": 13582
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7629000893852174,
+      "learning_rate": 1.5904118965321585e-05,
+      "loss": 0.9337,
+      "step": 13583
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6017204022925804,
+      "learning_rate": 1.5903503096095674e-05,
+      "loss": 0.8817,
+      "step": 13584
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5341879355124206,
+      "learning_rate": 1.5902887192497698e-05,
+      "loss": 0.8257,
+      "step": 13585
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7517236586475258,
+      "learning_rate": 1.5902271254531245e-05,
+      "loss": 0.7085,
+      "step": 13586
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.07475987070494,
+      "learning_rate": 1.59016552821999e-05,
+      "loss": 0.9021,
+      "step": 13587
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4267496565132658,
+      "learning_rate": 1.5901039275507247e-05,
+      "loss": 0.8373,
+      "step": 13588
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.064201823641732,
+      "learning_rate": 1.5900423234456877e-05,
+      "loss": 0.7752,
+      "step": 13589
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5065351238750906,
+      "learning_rate": 1.5899807159052373e-05,
+      "loss": 0.9971,
+      "step": 13590
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2039962012852103,
+      "learning_rate": 1.5899191049297328e-05,
+      "loss": 0.8627,
+      "step": 13591
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6833409814530496,
+      "learning_rate": 1.589857490519532e-05,
+      "loss": 0.848,
+      "step": 13592
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.640664854800265,
+      "learning_rate": 1.5897958726749947e-05,
+      "loss": 0.8467,
+      "step": 13593
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7587058179936754,
+      "learning_rate": 1.589734251396479e-05,
+      "loss": 0.9113,
+      "step": 13594
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6859987944180723,
+      "learning_rate": 1.5896726266843433e-05,
+      "loss": 0.9949,
+      "step": 13595
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5654720545626155,
+      "learning_rate": 1.589610998538947e-05,
+      "loss": 0.8276,
+      "step": 13596
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5998833561988166,
+      "learning_rate": 1.589549366960649e-05,
+      "loss": 0.8026,
+      "step": 13597
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6541958719868042,
+      "learning_rate": 1.589487731949808e-05,
+      "loss": 1.0014,
+      "step": 13598
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8518475835733683,
+      "learning_rate": 1.5894260935067828e-05,
+      "loss": 0.908,
+      "step": 13599
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.781433423012512,
+      "learning_rate": 1.589364451631932e-05,
+      "loss": 0.7582,
+      "step": 13600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.817981924816172,
+      "learning_rate": 1.589302806325615e-05,
+      "loss": 0.9395,
+      "step": 13601
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5685608899785408,
+      "learning_rate": 1.5892411575881906e-05,
+      "loss": 0.7952,
+      "step": 13602
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6158589547053848,
+      "learning_rate": 1.5891795054200173e-05,
+      "loss": 1.0438,
+      "step": 13603
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.54290153226478,
+      "learning_rate": 1.5891178498214545e-05,
+      "loss": 1.0629,
+      "step": 13604
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8639246291681393,
+      "learning_rate": 1.5890561907928612e-05,
+      "loss": 0.9222,
+      "step": 13605
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5532979256300272,
+      "learning_rate": 1.5889945283345963e-05,
+      "loss": 0.8856,
+      "step": 13606
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.677467866548449,
+      "learning_rate": 1.5889328624470186e-05,
+      "loss": 0.9659,
+      "step": 13607
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5907855950317524,
+      "learning_rate": 1.5888711931304874e-05,
+      "loss": 0.9513,
+      "step": 13608
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5234151129422622,
+      "learning_rate": 1.5888095203853614e-05,
+      "loss": 0.8886,
+      "step": 13609
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.879791349935586,
+      "learning_rate": 1.5887478442120007e-05,
+      "loss": 0.9856,
+      "step": 13610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7787401246224976,
+      "learning_rate": 1.588686164610763e-05,
+      "loss": 0.7366,
+      "step": 13611
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6297887740139898,
+      "learning_rate": 1.5886244815820082e-05,
+      "loss": 0.9219,
+      "step": 13612
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7994576861924428,
+      "learning_rate": 1.5885627951260954e-05,
+      "loss": 0.7761,
+      "step": 13613
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.731715465529374,
+      "learning_rate": 1.5885011052433834e-05,
+      "loss": 0.8783,
+      "step": 13614
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4492747856169572,
+      "learning_rate": 1.588439411934232e-05,
+      "loss": 0.8971,
+      "step": 13615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2001798312435787,
+      "learning_rate": 1.5883777151990002e-05,
+      "loss": 0.7396,
+      "step": 13616
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8627400384641197,
+      "learning_rate": 1.588316015038047e-05,
+      "loss": 0.9113,
+      "step": 13617
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6449298793167226,
+      "learning_rate": 1.5882543114517313e-05,
+      "loss": 0.8782,
+      "step": 13618
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8608554166377673,
+      "learning_rate": 1.588192604440413e-05,
+      "loss": 0.9453,
+      "step": 13619
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.574278400344219,
+      "learning_rate": 1.588130894004451e-05,
+      "loss": 0.8258,
+      "step": 13620
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.798836949750184,
+      "learning_rate": 1.5880691801442047e-05,
+      "loss": 0.8938,
+      "step": 13621
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8252828771490186,
+      "learning_rate": 1.5880074628600338e-05,
+      "loss": 0.8606,
+      "step": 13622
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5036319961953004,
+      "learning_rate": 1.587945742152297e-05,
+      "loss": 0.8053,
+      "step": 13623
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.609677167037763,
+      "learning_rate": 1.587884018021354e-05,
+      "loss": 0.9815,
+      "step": 13624
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6994669789626917,
+      "learning_rate": 1.5878222904675643e-05,
+      "loss": 0.8702,
+      "step": 13625
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5654340743659623,
+      "learning_rate": 1.587760559491287e-05,
+      "loss": 0.9306,
+      "step": 13626
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.489435651061182,
+      "learning_rate": 1.5876988250928812e-05,
+      "loss": 0.9067,
+      "step": 13627
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.714191529854627,
+      "learning_rate": 1.5876370872727072e-05,
+      "loss": 0.9146,
+      "step": 13628
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5460931875326405,
+      "learning_rate": 1.5875753460311245e-05,
+      "loss": 0.7289,
+      "step": 13629
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.876922471153885,
+      "learning_rate": 1.5875136013684917e-05,
+      "loss": 0.8181,
+      "step": 13630
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6176707830158996,
+      "learning_rate": 1.5874518532851688e-05,
+      "loss": 0.8131,
+      "step": 13631
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8042778924073586,
+      "learning_rate": 1.5873901017815146e-05,
+      "loss": 0.9498,
+      "step": 13632
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9251756059552647,
+      "learning_rate": 1.58732834685789e-05,
+      "loss": 0.9116,
+      "step": 13633
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.479179350528352,
+      "learning_rate": 1.5872665885146535e-05,
+      "loss": 0.8496,
+      "step": 13634
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.624680373134981,
+      "learning_rate": 1.587204826752165e-05,
+      "loss": 0.8217,
+      "step": 13635
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6500670971210563,
+      "learning_rate": 1.5871430615707843e-05,
+      "loss": 1.0292,
+      "step": 13636
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5185494888736446,
+      "learning_rate": 1.5870812929708706e-05,
+      "loss": 0.8435,
+      "step": 13637
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8183180345474688,
+      "learning_rate": 1.587019520952784e-05,
+      "loss": 0.9562,
+      "step": 13638
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.929981324019612,
+      "learning_rate": 1.5869577455168833e-05,
+      "loss": 0.8602,
+      "step": 13639
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8158995791008815,
+      "learning_rate": 1.5868959666635297e-05,
+      "loss": 0.6825,
+      "step": 13640
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5612917908906045,
+      "learning_rate": 1.5868341843930812e-05,
+      "loss": 0.9149,
+      "step": 13641
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5677829832279415,
+      "learning_rate": 1.5867723987058986e-05,
+      "loss": 0.9781,
+      "step": 13642
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4642056227176752,
+      "learning_rate": 1.586710609602341e-05,
+      "loss": 0.9477,
+      "step": 13643
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.589081644496577,
+      "learning_rate": 1.586648817082769e-05,
+      "loss": 0.9133,
+      "step": 13644
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7644662784144165,
+      "learning_rate": 1.5865870211475417e-05,
+      "loss": 0.7564,
+      "step": 13645
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6616290289256113,
+      "learning_rate": 1.5865252217970187e-05,
+      "loss": 0.8053,
+      "step": 13646
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5906774066347669,
+      "learning_rate": 1.5864634190315608e-05,
+      "loss": 0.8953,
+      "step": 13647
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4682793864554755,
+      "learning_rate": 1.5864016128515267e-05,
+      "loss": 0.9081,
+      "step": 13648
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3556019228223382,
+      "learning_rate": 1.586339803257277e-05,
+      "loss": 0.9165,
+      "step": 13649
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.021494563769265,
+      "learning_rate": 1.586277990249171e-05,
+      "loss": 0.8735,
+      "step": 13650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6713854956615453,
+      "learning_rate": 1.586216173827569e-05,
+      "loss": 0.971,
+      "step": 13651
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0402729404018354,
+      "learning_rate": 1.5861543539928313e-05,
+      "loss": 0.8554,
+      "step": 13652
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6016079298705395,
+      "learning_rate": 1.586092530745317e-05,
+      "loss": 0.9228,
+      "step": 13653
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.620653602763669,
+      "learning_rate": 1.5860307040853867e-05,
+      "loss": 0.859,
+      "step": 13654
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5562719260843747,
+      "learning_rate": 1.5859688740133998e-05,
+      "loss": 0.9814,
+      "step": 13655
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.715242620323764,
+      "learning_rate": 1.5859070405297168e-05,
+      "loss": 1.0096,
+      "step": 13656
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8053804726795781,
+      "learning_rate": 1.5858452036346973e-05,
+      "loss": 0.7163,
+      "step": 13657
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8114439511790155,
+      "learning_rate": 1.5857833633287015e-05,
+      "loss": 1.0732,
+      "step": 13658
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.653365302146199,
+      "learning_rate": 1.58572151961209e-05,
+      "loss": 0.7811,
+      "step": 13659
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.239580673122166,
+      "learning_rate": 1.585659672485222e-05,
+      "loss": 0.871,
+      "step": 13660
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5759943639785148,
+      "learning_rate": 1.5855978219484582e-05,
+      "loss": 1.02,
+      "step": 13661
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.688580370088432,
+      "learning_rate": 1.5855359680021583e-05,
+      "loss": 0.8989,
+      "step": 13662
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.59525890328904,
+      "learning_rate": 1.5854741106466825e-05,
+      "loss": 0.9535,
+      "step": 13663
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5964528955516557,
+      "learning_rate": 1.585412249882391e-05,
+      "loss": 0.9441,
+      "step": 13664
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4618305446502449,
+      "learning_rate": 1.5853503857096446e-05,
+      "loss": 0.9672,
+      "step": 13665
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9867126977895289,
+      "learning_rate": 1.5852885181288023e-05,
+      "loss": 0.9054,
+      "step": 13666
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5550421852776486,
+      "learning_rate": 1.5852266471402256e-05,
+      "loss": 0.9395,
+      "step": 13667
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.775619814720764,
+      "learning_rate": 1.5851647727442735e-05,
+      "loss": 0.9127,
+      "step": 13668
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7967153611257416,
+      "learning_rate": 1.585102894941307e-05,
+      "loss": 0.8676,
+      "step": 13669
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7301977176562664,
+      "learning_rate": 1.5850410137316865e-05,
+      "loss": 0.7583,
+      "step": 13670
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7993096832055329,
+      "learning_rate": 1.5849791291157717e-05,
+      "loss": 0.9197,
+      "step": 13671
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6866875473327685,
+      "learning_rate": 1.584917241093923e-05,
+      "loss": 0.7781,
+      "step": 13672
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6749309546762319,
+      "learning_rate": 1.5848553496665013e-05,
+      "loss": 0.9441,
+      "step": 13673
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.135114762762634,
+      "learning_rate": 1.5847934548338665e-05,
+      "loss": 0.9143,
+      "step": 13674
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6684911409457683,
+      "learning_rate": 1.584731556596379e-05,
+      "loss": 0.9901,
+      "step": 13675
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.736403887439804,
+      "learning_rate": 1.5846696549543993e-05,
+      "loss": 0.7191,
+      "step": 13676
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.89514169960206,
+      "learning_rate": 1.5846077499082883e-05,
+      "loss": 0.923,
+      "step": 13677
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5432663212989195,
+      "learning_rate": 1.5845458414584048e-05,
+      "loss": 0.9392,
+      "step": 13678
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5251064074866965,
+      "learning_rate": 1.5844839296051115e-05,
+      "loss": 0.9979,
+      "step": 13679
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7317743721071486,
+      "learning_rate": 1.5844220143487667e-05,
+      "loss": 0.9003,
+      "step": 13680
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.163901977473921,
+      "learning_rate": 1.5843600956897324e-05,
+      "loss": 0.9159,
+      "step": 13681
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7786349094759617,
+      "learning_rate": 1.5842981736283686e-05,
+      "loss": 0.7833,
+      "step": 13682
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6561408018156158,
+      "learning_rate": 1.5842362481650356e-05,
+      "loss": 0.8956,
+      "step": 13683
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6167225310728175,
+      "learning_rate": 1.5841743193000945e-05,
+      "loss": 0.8478,
+      "step": 13684
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6639307353230064,
+      "learning_rate": 1.5841123870339055e-05,
+      "loss": 1.0032,
+      "step": 13685
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7272137592003907,
+      "learning_rate": 1.5840504513668294e-05,
+      "loss": 0.8744,
+      "step": 13686
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.74609048189227,
+      "learning_rate": 1.5839885122992263e-05,
+      "loss": 0.9286,
+      "step": 13687
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6502102423173246,
+      "learning_rate": 1.5839265698314575e-05,
+      "loss": 0.8235,
+      "step": 13688
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4670357262859957,
+      "learning_rate": 1.583864623963883e-05,
+      "loss": 0.8135,
+      "step": 13689
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4577082096712564,
+      "learning_rate": 1.5838026746968643e-05,
+      "loss": 0.7764,
+      "step": 13690
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6021260689762653,
+      "learning_rate": 1.5837407220307612e-05,
+      "loss": 0.8935,
+      "step": 13691
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4530780744842997,
+      "learning_rate": 1.583678765965935e-05,
+      "loss": 1.0052,
+      "step": 13692
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.676965995042231,
+      "learning_rate": 1.5836168065027463e-05,
+      "loss": 0.9303,
+      "step": 13693
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.759784154818888,
+      "learning_rate": 1.5835548436415558e-05,
+      "loss": 0.9983,
+      "step": 13694
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7179651503093698,
+      "learning_rate": 1.583492877382724e-05,
+      "loss": 0.8644,
+      "step": 13695
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5857734150827005,
+      "learning_rate": 1.583430907726612e-05,
+      "loss": 0.9513,
+      "step": 13696
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.65037579494506,
+      "learning_rate": 1.5833689346735807e-05,
+      "loss": 0.9494,
+      "step": 13697
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5714084636259884,
+      "learning_rate": 1.5833069582239905e-05,
+      "loss": 0.9561,
+      "step": 13698
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8721457401454056,
+      "learning_rate": 1.5832449783782027e-05,
+      "loss": 0.8736,
+      "step": 13699
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8107732757910873,
+      "learning_rate": 1.583182995136578e-05,
+      "loss": 0.8456,
+      "step": 13700
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8814542265530578,
+      "learning_rate": 1.5831210084994775e-05,
+      "loss": 0.8305,
+      "step": 13701
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5931623110777298,
+      "learning_rate": 1.5830590184672613e-05,
+      "loss": 1.0316,
+      "step": 13702
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5513159621794326,
+      "learning_rate": 1.5829970250402916e-05,
+      "loss": 0.8353,
+      "step": 13703
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7390997211131516,
+      "learning_rate": 1.5829350282189278e-05,
+      "loss": 0.91,
+      "step": 13704
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4990698899999249,
+      "learning_rate": 1.582873028003532e-05,
+      "loss": 1.0827,
+      "step": 13705
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6265555227911406,
+      "learning_rate": 1.582811024394465e-05,
+      "loss": 0.8978,
+      "step": 13706
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.707814406373472,
+      "learning_rate": 1.5827490173920882e-05,
+      "loss": 0.956,
+      "step": 13707
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9276516613636447,
+      "learning_rate": 1.5826870069967615e-05,
+      "loss": 0.9179,
+      "step": 13708
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5920538769410009,
+      "learning_rate": 1.5826249932088467e-05,
+      "loss": 0.9534,
+      "step": 13709
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.678824524697494,
+      "learning_rate": 1.5825629760287052e-05,
+      "loss": 0.8758,
+      "step": 13710
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0655190289284344,
+      "learning_rate": 1.582500955456697e-05,
+      "loss": 0.7919,
+      "step": 13711
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6513947214087612,
+      "learning_rate": 1.582438931493184e-05,
+      "loss": 0.9023,
+      "step": 13712
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3212850959657323,
+      "learning_rate": 1.5823769041385272e-05,
+      "loss": 0.8638,
+      "step": 13713
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5444251802007547,
+      "learning_rate": 1.5823148733930876e-05,
+      "loss": 0.8127,
+      "step": 13714
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6627050309441638,
+      "learning_rate": 1.5822528392572266e-05,
+      "loss": 0.8858,
+      "step": 13715
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7870133240122303,
+      "learning_rate": 1.5821908017313052e-05,
+      "loss": 0.791,
+      "step": 13716
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9971463191589645,
+      "learning_rate": 1.5821287608156847e-05,
+      "loss": 0.9186,
+      "step": 13717
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.863307796713767,
+      "learning_rate": 1.582066716510726e-05,
+      "loss": 0.8087,
+      "step": 13718
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7991836179883403,
+      "learning_rate": 1.582004668816791e-05,
+      "loss": 0.9111,
+      "step": 13719
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7371013574146492,
+      "learning_rate": 1.58194261773424e-05,
+      "loss": 0.7246,
+      "step": 13720
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8216872716977124,
+      "learning_rate": 1.581880563263435e-05,
+      "loss": 0.7482,
+      "step": 13721
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.68392327705266,
+      "learning_rate": 1.5818185054047376e-05,
+      "loss": 0.9726,
+      "step": 13722
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7973031189319546,
+      "learning_rate": 1.5817564441585083e-05,
+      "loss": 0.8076,
+      "step": 13723
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7081958406446742,
+      "learning_rate": 1.581694379525109e-05,
+      "loss": 0.8669,
+      "step": 13724
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5917483540867332,
+      "learning_rate": 1.5816323115049005e-05,
+      "loss": 0.7894,
+      "step": 13725
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4822133033807896,
+      "learning_rate": 1.5815702400982447e-05,
+      "loss": 0.8631,
+      "step": 13726
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7788647125930113,
+      "learning_rate": 1.5815081653055028e-05,
+      "loss": 0.89,
+      "step": 13727
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6447218896449296,
+      "learning_rate": 1.5814460871270364e-05,
+      "loss": 0.8767,
+      "step": 13728
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5695795670584307,
+      "learning_rate": 1.5813840055632067e-05,
+      "loss": 0.8477,
+      "step": 13729
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7875015814332265,
+      "learning_rate": 1.5813219206143755e-05,
+      "loss": 0.8644,
+      "step": 13730
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.886663572260288,
+      "learning_rate": 1.5812598322809036e-05,
+      "loss": 0.9192,
+      "step": 13731
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.653261701735098,
+      "learning_rate": 1.581197740563153e-05,
+      "loss": 0.8352,
+      "step": 13732
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5892036468569941,
+      "learning_rate": 1.5811356454614853e-05,
+      "loss": 0.93,
+      "step": 13733
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8152430242014004,
+      "learning_rate": 1.5810735469762623e-05,
+      "loss": 0.7606,
+      "step": 13734
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.759532098022246,
+      "learning_rate": 1.5810114451078446e-05,
+      "loss": 0.9051,
+      "step": 13735
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7687905088123259,
+      "learning_rate": 1.5809493398565945e-05,
+      "loss": 0.7961,
+      "step": 13736
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.803665535004781,
+      "learning_rate": 1.5808872312228735e-05,
+      "loss": 0.7668,
+      "step": 13737
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.15964350787866,
+      "learning_rate": 1.5808251192070428e-05,
+      "loss": 0.9778,
+      "step": 13738
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5904481183044263,
+      "learning_rate": 1.5807630038094646e-05,
+      "loss": 0.8408,
+      "step": 13739
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8474016208339599,
+      "learning_rate": 1.5807008850305003e-05,
+      "loss": 0.9183,
+      "step": 13740
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8122993295015106,
+      "learning_rate": 1.580638762870512e-05,
+      "loss": 0.8772,
+      "step": 13741
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7849178418006518,
+      "learning_rate": 1.5805766373298605e-05,
+      "loss": 0.6883,
+      "step": 13742
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3932918882716518,
+      "learning_rate": 1.580514508408908e-05,
+      "loss": 0.8964,
+      "step": 13743
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9976466670757995,
+      "learning_rate": 1.580452376108016e-05,
+      "loss": 0.8966,
+      "step": 13744
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6081958322972658,
+      "learning_rate": 1.580390240427547e-05,
+      "loss": 0.9046,
+      "step": 13745
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6997729432059583,
+      "learning_rate": 1.5803281013678623e-05,
+      "loss": 1.1025,
+      "step": 13746
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6510860785728876,
+      "learning_rate": 1.5802659589293234e-05,
+      "loss": 0.8534,
+      "step": 13747
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5808528337750982,
+      "learning_rate": 1.5802038131122926e-05,
+      "loss": 0.8605,
+      "step": 13748
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.080238757932534,
+      "learning_rate": 1.5801416639171314e-05,
+      "loss": 0.8701,
+      "step": 13749
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.714639652399211,
+      "learning_rate": 1.5800795113442017e-05,
+      "loss": 0.8653,
+      "step": 13750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9165947887107944,
+      "learning_rate": 1.5800173553938654e-05,
+      "loss": 0.8684,
+      "step": 13751
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.096959174179512,
+      "learning_rate": 1.5799551960664846e-05,
+      "loss": 0.8245,
+      "step": 13752
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6938852836824616,
+      "learning_rate": 1.579893033362421e-05,
+      "loss": 0.8197,
+      "step": 13753
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7457134906984302,
+      "learning_rate": 1.5798308672820363e-05,
+      "loss": 0.8741,
+      "step": 13754
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5838520802000633,
+      "learning_rate": 1.579768697825693e-05,
+      "loss": 0.8034,
+      "step": 13755
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8117958026295901,
+      "learning_rate": 1.5797065249937528e-05,
+      "loss": 0.7935,
+      "step": 13756
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7765546333496396,
+      "learning_rate": 1.5796443487865774e-05,
+      "loss": 0.9839,
+      "step": 13757
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4985245014885824,
+      "learning_rate": 1.5795821692045294e-05,
+      "loss": 0.9088,
+      "step": 13758
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5834589813459452,
+      "learning_rate": 1.5795199862479706e-05,
+      "loss": 0.9078,
+      "step": 13759
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.536900115961813,
+      "learning_rate": 1.5794577999172628e-05,
+      "loss": 0.9022,
+      "step": 13760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.539372214604095,
+      "learning_rate": 1.5793956102127682e-05,
+      "loss": 0.9025,
+      "step": 13761
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9697056232799481,
+      "learning_rate": 1.579333417134849e-05,
+      "loss": 1.0324,
+      "step": 13762
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.782210180504509,
+      "learning_rate": 1.579271220683867e-05,
+      "loss": 0.8995,
+      "step": 13763
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8224680177821702,
+      "learning_rate": 1.5792090208601847e-05,
+      "loss": 0.8436,
+      "step": 13764
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4982728756957524,
+      "learning_rate": 1.5791468176641642e-05,
+      "loss": 1.0597,
+      "step": 13765
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.411481054690309,
+      "learning_rate": 1.5790846110961676e-05,
+      "loss": 0.9116,
+      "step": 13766
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5281619030998215,
+      "learning_rate": 1.5790224011565566e-05,
+      "loss": 0.841,
+      "step": 13767
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8122070219326991,
+      "learning_rate": 1.5789601878456946e-05,
+      "loss": 0.8291,
+      "step": 13768
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4961403142369083,
+      "learning_rate": 1.5788979711639427e-05,
+      "loss": 0.8368,
+      "step": 13769
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5920617115832398,
+      "learning_rate": 1.5788357511116633e-05,
+      "loss": 0.9137,
+      "step": 13770
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4561849909349878,
+      "learning_rate": 1.5787735276892194e-05,
+      "loss": 0.8653,
+      "step": 13771
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.597809252678881,
+      "learning_rate": 1.5787113008969725e-05,
+      "loss": 0.8947,
+      "step": 13772
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7526251226646956,
+      "learning_rate": 1.5786490707352852e-05,
+      "loss": 0.9326,
+      "step": 13773
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4932024922658595,
+      "learning_rate": 1.57858683720452e-05,
+      "loss": 0.8796,
+      "step": 13774
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8676379211957883,
+      "learning_rate": 1.578524600305039e-05,
+      "loss": 0.9207,
+      "step": 13775
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5380727609947982,
+      "learning_rate": 1.578462360037204e-05,
+      "loss": 0.9329,
+      "step": 13776
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7612149007643005,
+      "learning_rate": 1.5784001164013788e-05,
+      "loss": 0.9245,
+      "step": 13777
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.078356703836752,
+      "learning_rate": 1.578337869397925e-05,
+      "loss": 0.8974,
+      "step": 13778
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5099126747576117,
+      "learning_rate": 1.5782756190272047e-05,
+      "loss": 0.8827,
+      "step": 13779
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.877312127230408,
+      "learning_rate": 1.5782133652895804e-05,
+      "loss": 0.8405,
+      "step": 13780
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.898942716878687,
+      "learning_rate": 1.5781511081854152e-05,
+      "loss": 0.858,
+      "step": 13781
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.923665248700819,
+      "learning_rate": 1.5780888477150712e-05,
+      "loss": 0.8838,
+      "step": 13782
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1253240614395645,
+      "learning_rate": 1.5780265838789107e-05,
+      "loss": 0.865,
+      "step": 13783
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5835528713116886,
+      "learning_rate": 1.5779643166772966e-05,
+      "loss": 0.7728,
+      "step": 13784
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9455618021754506,
+      "learning_rate": 1.5779020461105913e-05,
+      "loss": 0.9207,
+      "step": 13785
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9101146189777771,
+      "learning_rate": 1.5778397721791574e-05,
+      "loss": 0.8266,
+      "step": 13786
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.626431511688116,
+      "learning_rate": 1.5777774948833574e-05,
+      "loss": 1.0059,
+      "step": 13787
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7282360456613208,
+      "learning_rate": 1.5777152142235535e-05,
+      "loss": 0.9341,
+      "step": 13788
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6612550767148204,
+      "learning_rate": 1.5776529302001093e-05,
+      "loss": 0.7983,
+      "step": 13789
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6962884726129395,
+      "learning_rate": 1.5775906428133866e-05,
+      "loss": 0.9772,
+      "step": 13790
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.603403106522325,
+      "learning_rate": 1.5775283520637483e-05,
+      "loss": 0.932,
+      "step": 13791
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4833923962685407,
+      "learning_rate": 1.577466057951557e-05,
+      "loss": 0.8548,
+      "step": 13792
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9328116209915043,
+      "learning_rate": 1.5774037604771756e-05,
+      "loss": 0.88,
+      "step": 13793
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6426913683090665,
+      "learning_rate": 1.5773414596409666e-05,
+      "loss": 0.9129,
+      "step": 13794
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0974083224542013,
+      "learning_rate": 1.577279155443293e-05,
+      "loss": 0.9811,
+      "step": 13795
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8000254457654188,
+      "learning_rate": 1.5772168478845174e-05,
+      "loss": 0.8332,
+      "step": 13796
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8276839042043822,
+      "learning_rate": 1.5771545369650026e-05,
+      "loss": 0.7064,
+      "step": 13797
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6711173076195813,
+      "learning_rate": 1.577092222685111e-05,
+      "loss": 1.0022,
+      "step": 13798
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.710462010457876,
+      "learning_rate": 1.577029905045206e-05,
+      "loss": 0.8364,
+      "step": 13799
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8537639752326012,
+      "learning_rate": 1.5769675840456503e-05,
+      "loss": 0.882,
+      "step": 13800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.70892756258674,
+      "learning_rate": 1.5769052596868065e-05,
+      "loss": 0.933,
+      "step": 13801
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.691756937802293,
+      "learning_rate": 1.5768429319690376e-05,
+      "loss": 0.9076,
+      "step": 13802
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7983331496649866,
+      "learning_rate": 1.5767806008927068e-05,
+      "loss": 0.8757,
+      "step": 13803
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.338054897559529,
+      "learning_rate": 1.5767182664581767e-05,
+      "loss": 0.8864,
+      "step": 13804
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7569751329514371,
+      "learning_rate": 1.57665592866581e-05,
+      "loss": 0.9747,
+      "step": 13805
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7702278534303906,
+      "learning_rate": 1.5765935875159703e-05,
+      "loss": 0.7725,
+      "step": 13806
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7902222709084095,
+      "learning_rate": 1.5765312430090197e-05,
+      "loss": 0.7238,
+      "step": 13807
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.634766775629164,
+      "learning_rate": 1.5764688951453216e-05,
+      "loss": 0.7996,
+      "step": 13808
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7260280229302956,
+      "learning_rate": 1.57640654392524e-05,
+      "loss": 0.8467,
+      "step": 13809
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6318831582666569,
+      "learning_rate": 1.5763441893491364e-05,
+      "loss": 0.9246,
+      "step": 13810
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.656944794380074,
+      "learning_rate": 1.5762818314173743e-05,
+      "loss": 0.8089,
+      "step": 13811
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6142101236139,
+      "learning_rate": 1.576219470130317e-05,
+      "loss": 0.9086,
+      "step": 13812
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7610412378971125,
+      "learning_rate": 1.576157105488328e-05,
+      "loss": 0.9898,
+      "step": 13813
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.726581356254563,
+      "learning_rate": 1.5760947374917693e-05,
+      "loss": 0.821,
+      "step": 13814
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7385354323532027,
+      "learning_rate": 1.576032366141005e-05,
+      "loss": 0.7076,
+      "step": 13815
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6575333020788325,
+      "learning_rate": 1.5759699914363977e-05,
+      "loss": 0.9743,
+      "step": 13816
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7787518455416538,
+      "learning_rate": 1.575907613378311e-05,
+      "loss": 0.8189,
+      "step": 13817
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9329140574674823,
+      "learning_rate": 1.5758452319671075e-05,
+      "loss": 0.9075,
+      "step": 13818
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6603599083149356,
+      "learning_rate": 1.5757828472031512e-05,
+      "loss": 0.9186,
+      "step": 13819
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9268381739268903,
+      "learning_rate": 1.5757204590868042e-05,
+      "loss": 0.7789,
+      "step": 13820
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7081529602392531,
+      "learning_rate": 1.575658067618431e-05,
+      "loss": 0.8563,
+      "step": 13821
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8631471576695207,
+      "learning_rate": 1.5755956727983943e-05,
+      "loss": 0.7825,
+      "step": 13822
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.627039406481778,
+      "learning_rate": 1.5755332746270574e-05,
+      "loss": 0.9401,
+      "step": 13823
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6232838592509244,
+      "learning_rate": 1.5754708731047833e-05,
+      "loss": 0.9173,
+      "step": 13824
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4477208281689609,
+      "learning_rate": 1.5754084682319356e-05,
+      "loss": 0.9491,
+      "step": 13825
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4851601996435586,
+      "learning_rate": 1.5753460600088777e-05,
+      "loss": 0.9125,
+      "step": 13826
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6637551587619444,
+      "learning_rate": 1.5752836484359726e-05,
+      "loss": 0.9555,
+      "step": 13827
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1256018719197045,
+      "learning_rate": 1.575221233513584e-05,
+      "loss": 0.9147,
+      "step": 13828
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5803996829522744,
+      "learning_rate": 1.5751588152420757e-05,
+      "loss": 0.8698,
+      "step": 13829
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.640231945540418,
+      "learning_rate": 1.5750963936218104e-05,
+      "loss": 1.0088,
+      "step": 13830
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9444022589151646,
+      "learning_rate": 1.575033968653152e-05,
+      "loss": 0.9537,
+      "step": 13831
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7805004959386757,
+      "learning_rate": 1.5749715403364635e-05,
+      "loss": 0.8342,
+      "step": 13832
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.93142427808232,
+      "learning_rate": 1.574909108672109e-05,
+      "loss": 0.8163,
+      "step": 13833
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6057806821593477,
+      "learning_rate": 1.5748466736604514e-05,
+      "loss": 0.9055,
+      "step": 13834
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8001142071132914,
+      "learning_rate": 1.574784235301855e-05,
+      "loss": 0.9223,
+      "step": 13835
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7558950095361769,
+      "learning_rate": 1.574721793596682e-05,
+      "loss": 0.7072,
+      "step": 13836
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9777688987678714,
+      "learning_rate": 1.5746593485452973e-05,
+      "loss": 0.7923,
+      "step": 13837
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7878930899121175,
+      "learning_rate": 1.5745969001480637e-05,
+      "loss": 0.8088,
+      "step": 13838
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7023294928464074,
+      "learning_rate": 1.574534448405345e-05,
+      "loss": 0.779,
+      "step": 13839
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6144640397813752,
+      "learning_rate": 1.5744719933175053e-05,
+      "loss": 0.8736,
+      "step": 13840
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5964713760746139,
+      "learning_rate": 1.5744095348849073e-05,
+      "loss": 0.7967,
+      "step": 13841
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5852249320480551,
+      "learning_rate": 1.5743470731079152e-05,
+      "loss": 0.9387,
+      "step": 13842
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.421003739786463,
+      "learning_rate": 1.5742846079868926e-05,
+      "loss": 0.7991,
+      "step": 13843
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.660364973011331,
+      "learning_rate": 1.5742221395222033e-05,
+      "loss": 0.8788,
+      "step": 13844
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8285820867985215,
+      "learning_rate": 1.574159667714211e-05,
+      "loss": 0.77,
+      "step": 13845
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6819839607659688,
+      "learning_rate": 1.574097192563279e-05,
+      "loss": 0.8812,
+      "step": 13846
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.659878117056468,
+      "learning_rate": 1.574034714069772e-05,
+      "loss": 0.9926,
+      "step": 13847
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.650887429374874,
+      "learning_rate": 1.5739722322340524e-05,
+      "loss": 0.8222,
+      "step": 13848
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5485406192321542,
+      "learning_rate": 1.5739097470564853e-05,
+      "loss": 1.0459,
+      "step": 13849
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.403390604057964,
+      "learning_rate": 1.5738472585374334e-05,
+      "loss": 0.9139,
+      "step": 13850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5315156686869997,
+      "learning_rate": 1.5737847666772615e-05,
+      "loss": 0.9785,
+      "step": 13851
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6109494149872983,
+      "learning_rate": 1.573722271476333e-05,
+      "loss": 0.9417,
+      "step": 13852
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7171007300165606,
+      "learning_rate": 1.573659772935012e-05,
+      "loss": 0.734,
+      "step": 13853
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.980885044632398,
+      "learning_rate": 1.5735972710536623e-05,
+      "loss": 0.8342,
+      "step": 13854
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6886070062252814,
+      "learning_rate": 1.5735347658326472e-05,
+      "loss": 0.809,
+      "step": 13855
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7658116565678001,
+      "learning_rate": 1.5734722572723313e-05,
+      "loss": 0.8316,
+      "step": 13856
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.831880436082236,
+      "learning_rate": 1.573409745373078e-05,
+      "loss": 0.727,
+      "step": 13857
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.793675906791444,
+      "learning_rate": 1.573347230135252e-05,
+      "loss": 0.9251,
+      "step": 13858
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.895648604389666,
+      "learning_rate": 1.573284711559217e-05,
+      "loss": 0.8642,
+      "step": 13859
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6282792170589762,
+      "learning_rate": 1.5732221896453368e-05,
+      "loss": 0.9382,
+      "step": 13860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6424143716625441,
+      "learning_rate": 1.5731596643939755e-05,
+      "loss": 0.956,
+      "step": 13861
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6725294932376145,
+      "learning_rate": 1.5730971358054972e-05,
+      "loss": 0.8593,
+      "step": 13862
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6655635077897337,
+      "learning_rate": 1.5730346038802657e-05,
+      "loss": 0.8888,
+      "step": 13863
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.538150907574795,
+      "learning_rate": 1.5729720686186457e-05,
+      "loss": 0.8999,
+      "step": 13864
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6709397045327103,
+      "learning_rate": 1.5729095300210006e-05,
+      "loss": 0.9976,
+      "step": 13865
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5508477092562338,
+      "learning_rate": 1.572846988087695e-05,
+      "loss": 1.0278,
+      "step": 13866
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8034248611510252,
+      "learning_rate": 1.572784442819093e-05,
+      "loss": 0.7763,
+      "step": 13867
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6829361356959516,
+      "learning_rate": 1.5727218942155588e-05,
+      "loss": 0.8752,
+      "step": 13868
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6519284158598184,
+      "learning_rate": 1.5726593422774558e-05,
+      "loss": 0.8245,
+      "step": 13869
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9030541131806957,
+      "learning_rate": 1.572596787005149e-05,
+      "loss": 1.0415,
+      "step": 13870
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.583313786924927,
+      "learning_rate": 1.5725342283990027e-05,
+      "loss": 0.803,
+      "step": 13871
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4068311739148736,
+      "learning_rate": 1.572471666459381e-05,
+      "loss": 0.9065,
+      "step": 13872
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7960847759166556,
+      "learning_rate": 1.5724091011866474e-05,
+      "loss": 0.7898,
+      "step": 13873
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7912961820111625,
+      "learning_rate": 1.572346532581167e-05,
+      "loss": 0.7636,
+      "step": 13874
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0174323597576493,
+      "learning_rate": 1.5722839606433043e-05,
+      "loss": 0.758,
+      "step": 13875
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5819286034428683,
+      "learning_rate": 1.5722213853734227e-05,
+      "loss": 0.8303,
+      "step": 13876
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6365765793680591,
+      "learning_rate": 1.572158806771887e-05,
+      "loss": 0.866,
+      "step": 13877
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.872320140311925,
+      "learning_rate": 1.572096224839062e-05,
+      "loss": 0.8019,
+      "step": 13878
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8266747199599376,
+      "learning_rate": 1.5720336395753117e-05,
+      "loss": 0.9963,
+      "step": 13879
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8261223493550136,
+      "learning_rate": 1.571971050981e-05,
+      "loss": 0.8618,
+      "step": 13880
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.567458764171586,
+      "learning_rate": 1.571908459056492e-05,
+      "loss": 0.9302,
+      "step": 13881
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2751550350638454,
+      "learning_rate": 1.571845863802152e-05,
+      "loss": 0.8221,
+      "step": 13882
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6164214585051395,
+      "learning_rate": 1.571783265218344e-05,
+      "loss": 0.963,
+      "step": 13883
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9004858969457155,
+      "learning_rate": 1.5717206633054328e-05,
+      "loss": 0.8746,
+      "step": 13884
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8908398986141257,
+      "learning_rate": 1.5716580580637833e-05,
+      "loss": 0.8476,
+      "step": 13885
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8287311060167195,
+      "learning_rate": 1.5715954494937595e-05,
+      "loss": 0.8424,
+      "step": 13886
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8124865834457922,
+      "learning_rate": 1.5715328375957256e-05,
+      "loss": 0.8702,
+      "step": 13887
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7913964978087757,
+      "learning_rate": 1.5714702223700472e-05,
+      "loss": 0.7497,
+      "step": 13888
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.732317752602688,
+      "learning_rate": 1.571407603817088e-05,
+      "loss": 0.7916,
+      "step": 13889
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5970826178114625,
+      "learning_rate": 1.571344981937213e-05,
+      "loss": 0.9633,
+      "step": 13890
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5170062076102884,
+      "learning_rate": 1.571282356730786e-05,
+      "loss": 0.91,
+      "step": 13891
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6608872397157415,
+      "learning_rate": 1.5712197281981727e-05,
+      "loss": 0.8876,
+      "step": 13892
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5211358741491239,
+      "learning_rate": 1.5711570963397376e-05,
+      "loss": 0.9019,
+      "step": 13893
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3864418194683337,
+      "learning_rate": 1.5710944611558445e-05,
+      "loss": 0.9175,
+      "step": 13894
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8213380857656845,
+      "learning_rate": 1.5710318226468587e-05,
+      "loss": 0.9593,
+      "step": 13895
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.760625195828172,
+      "learning_rate": 1.570969180813145e-05,
+      "loss": 0.8778,
+      "step": 13896
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8781691774693516,
+      "learning_rate": 1.570906535655068e-05,
+      "loss": 0.9195,
+      "step": 13897
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5806811088238881,
+      "learning_rate": 1.5708438871729925e-05,
+      "loss": 1.0677,
+      "step": 13898
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7246617399230573,
+      "learning_rate": 1.570781235367283e-05,
+      "loss": 0.8077,
+      "step": 13899
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5609918683591628,
+      "learning_rate": 1.5707185802383045e-05,
+      "loss": 0.849,
+      "step": 13900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7693227195307863,
+      "learning_rate": 1.5706559217864214e-05,
+      "loss": 1.011,
+      "step": 13901
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8348422430176294,
+      "learning_rate": 1.5705932600119993e-05,
+      "loss": 1.0323,
+      "step": 13902
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7675164354484604,
+      "learning_rate": 1.5705305949154025e-05,
+      "loss": 0.7562,
+      "step": 13903
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5568039312160644,
+      "learning_rate": 1.570467926496996e-05,
+      "loss": 0.8602,
+      "step": 13904
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6806513310243352,
+      "learning_rate": 1.5704052547571445e-05,
+      "loss": 0.9396,
+      "step": 13905
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6055497191641286,
+      "learning_rate": 1.570342579696213e-05,
+      "loss": 0.8642,
+      "step": 13906
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7830828937604066,
+      "learning_rate": 1.5702799013145665e-05,
+      "loss": 0.7158,
+      "step": 13907
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5765262236870692,
+      "learning_rate": 1.57021721961257e-05,
+      "loss": 0.7584,
+      "step": 13908
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5533169075399023,
+      "learning_rate": 1.570154534590588e-05,
+      "loss": 0.9301,
+      "step": 13909
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7305814321413577,
+      "learning_rate": 1.570091846248986e-05,
+      "loss": 0.9703,
+      "step": 13910
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.86069810773456,
+      "learning_rate": 1.570029154588129e-05,
+      "loss": 0.9451,
+      "step": 13911
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.83810739607255,
+      "learning_rate": 1.5699664596083814e-05,
+      "loss": 0.7494,
+      "step": 13912
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6761605650006381,
+      "learning_rate": 1.5699037613101087e-05,
+      "loss": 0.7491,
+      "step": 13913
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.733899907339525,
+      "learning_rate": 1.569841059693676e-05,
+      "loss": 1.1002,
+      "step": 13914
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.677876067419758,
+      "learning_rate": 1.5697783547594482e-05,
+      "loss": 0.8203,
+      "step": 13915
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0587030786978944,
+      "learning_rate": 1.5697156465077903e-05,
+      "loss": 0.8321,
+      "step": 13916
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8094525221200333,
+      "learning_rate": 1.5696529349390676e-05,
+      "loss": 0.8785,
+      "step": 13917
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6649619847437283,
+      "learning_rate": 1.569590220053645e-05,
+      "loss": 0.9335,
+      "step": 13918
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8652994072303575,
+      "learning_rate": 1.5695275018518885e-05,
+      "loss": 1.0268,
+      "step": 13919
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9522874506341186,
+      "learning_rate": 1.5694647803341618e-05,
+      "loss": 0.7894,
+      "step": 13920
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.84504512118992,
+      "learning_rate": 1.5694020555008314e-05,
+      "loss": 0.8286,
+      "step": 13921
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.127964304510153,
+      "learning_rate": 1.5693393273522618e-05,
+      "loss": 0.9219,
+      "step": 13922
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0208535044154634,
+      "learning_rate": 1.5692765958888182e-05,
+      "loss": 0.8509,
+      "step": 13923
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5967370683161541,
+      "learning_rate": 1.569213861110866e-05,
+      "loss": 0.9364,
+      "step": 13924
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9027523660703727,
+      "learning_rate": 1.5691511230187706e-05,
+      "loss": 0.8562,
+      "step": 13925
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7325624776089734,
+      "learning_rate": 1.5690883816128973e-05,
+      "loss": 0.9143,
+      "step": 13926
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.005659139916472,
+      "learning_rate": 1.569025636893611e-05,
+      "loss": 0.9466,
+      "step": 13927
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.648415141309247,
+      "learning_rate": 1.5689628888612774e-05,
+      "loss": 0.8043,
+      "step": 13928
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7344837691928467,
+      "learning_rate": 1.568900137516262e-05,
+      "loss": 0.8189,
+      "step": 13929
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.479663837601574,
+      "learning_rate": 1.5688373828589296e-05,
+      "loss": 0.899,
+      "step": 13930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8808037818041132,
+      "learning_rate": 1.568774624889646e-05,
+      "loss": 0.7569,
+      "step": 13931
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.483762314208912,
+      "learning_rate": 1.5687118636087764e-05,
+      "loss": 0.9066,
+      "step": 13932
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3995486201092315,
+      "learning_rate": 1.5686490990166864e-05,
+      "loss": 0.8262,
+      "step": 13933
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6799502116873093,
+      "learning_rate": 1.568586331113741e-05,
+      "loss": 0.8881,
+      "step": 13934
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8818353055112576,
+      "learning_rate": 1.5685235599003065e-05,
+      "loss": 0.8168,
+      "step": 13935
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6792316080071688,
+      "learning_rate": 1.5684607853767474e-05,
+      "loss": 0.8941,
+      "step": 13936
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7954834137905938,
+      "learning_rate": 1.5683980075434297e-05,
+      "loss": 0.7859,
+      "step": 13937
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2718823667542334,
+      "learning_rate": 1.5683352264007188e-05,
+      "loss": 0.9257,
+      "step": 13938
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6638568469652697,
+      "learning_rate": 1.5682724419489805e-05,
+      "loss": 0.9637,
+      "step": 13939
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7382889383794298,
+      "learning_rate": 1.5682096541885797e-05,
+      "loss": 0.9494,
+      "step": 13940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9714685429182675,
+      "learning_rate": 1.568146863119883e-05,
+      "loss": 0.9669,
+      "step": 13941
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6949035084167348,
+      "learning_rate": 1.568084068743255e-05,
+      "loss": 0.8152,
+      "step": 13942
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5763403681216905,
+      "learning_rate": 1.568021271059062e-05,
+      "loss": 0.9825,
+      "step": 13943
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.900654469922631,
+      "learning_rate": 1.567958470067669e-05,
+      "loss": 0.9849,
+      "step": 13944
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6808475050047482,
+      "learning_rate": 1.567895665769442e-05,
+      "loss": 0.8924,
+      "step": 13945
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8861466142602088,
+      "learning_rate": 1.5678328581647467e-05,
+      "loss": 0.8666,
+      "step": 13946
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.497146780779032,
+      "learning_rate": 1.567770047253949e-05,
+      "loss": 0.8563,
+      "step": 13947
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.598335586898376,
+      "learning_rate": 1.5677072330374138e-05,
+      "loss": 0.9686,
+      "step": 13948
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7308965585324942,
+      "learning_rate": 1.567644415515508e-05,
+      "loss": 0.8682,
+      "step": 13949
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9056388477626518,
+      "learning_rate": 1.567581594688596e-05,
+      "loss": 0.9112,
+      "step": 13950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7370827333497127,
+      "learning_rate": 1.567518770557045e-05,
+      "loss": 0.9493,
+      "step": 13951
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.960666028476967,
+      "learning_rate": 1.5674559431212193e-05,
+      "loss": 0.8762,
+      "step": 13952
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0144462828709986,
+      "learning_rate": 1.5673931123814862e-05,
+      "loss": 0.941,
+      "step": 13953
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6907549948609408,
+      "learning_rate": 1.5673302783382104e-05,
+      "loss": 1.0299,
+      "step": 13954
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6846817049570253,
+      "learning_rate": 1.5672674409917575e-05,
+      "loss": 0.8414,
+      "step": 13955
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6555164685341948,
+      "learning_rate": 1.5672046003424948e-05,
+      "loss": 0.8388,
+      "step": 13956
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7231331975480275,
+      "learning_rate": 1.5671417563907872e-05,
+      "loss": 0.9519,
+      "step": 13957
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6545386959587094,
+      "learning_rate": 1.5670789091370005e-05,
+      "loss": 0.9035,
+      "step": 13958
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6441551028421626,
+      "learning_rate": 1.567016058581501e-05,
+      "loss": 0.913,
+      "step": 13959
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6042678503347676,
+      "learning_rate": 1.5669532047246546e-05,
+      "loss": 0.9392,
+      "step": 13960
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8409350238038555,
+      "learning_rate": 1.566890347566827e-05,
+      "loss": 1.0627,
+      "step": 13961
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.63026226794191,
+      "learning_rate": 1.5668274871083844e-05,
+      "loss": 0.8536,
+      "step": 13962
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1359101496733444,
+      "learning_rate": 1.5667646233496925e-05,
+      "loss": 0.9681,
+      "step": 13963
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5568426059421079,
+      "learning_rate": 1.566701756291118e-05,
+      "loss": 0.9823,
+      "step": 13964
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5211248078540585,
+      "learning_rate": 1.566638885933026e-05,
+      "loss": 0.8872,
+      "step": 13965
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4862796644001823,
+      "learning_rate": 1.566576012275783e-05,
+      "loss": 0.9251,
+      "step": 13966
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5767713511345094,
+      "learning_rate": 1.5665131353197555e-05,
+      "loss": 0.9234,
+      "step": 13967
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6193603097974623,
+      "learning_rate": 1.5664502550653088e-05,
+      "loss": 0.822,
+      "step": 13968
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4665287322504033,
+      "learning_rate": 1.566387371512809e-05,
+      "loss": 0.7934,
+      "step": 13969
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.571414745461934,
+      "learning_rate": 1.566324484662624e-05,
+      "loss": 0.9914,
+      "step": 13970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.81421222467984,
+      "learning_rate": 1.5662615945151173e-05,
+      "loss": 0.9933,
+      "step": 13971
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7697584657192622,
+      "learning_rate": 1.566198701070657e-05,
+      "loss": 0.9413,
+      "step": 13972
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.680008979046094,
+      "learning_rate": 1.566135804329608e-05,
+      "loss": 0.8992,
+      "step": 13973
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9836459060494311,
+      "learning_rate": 1.5660729042923374e-05,
+      "loss": 0.881,
+      "step": 13974
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.641824602395643,
+      "learning_rate": 1.566010000959211e-05,
+      "loss": 0.855,
+      "step": 13975
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4529801486147114,
+      "learning_rate": 1.5659470943305956e-05,
+      "loss": 1.0221,
+      "step": 13976
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.922955332688156,
+      "learning_rate": 1.5658841844068568e-05,
+      "loss": 0.9574,
+      "step": 13977
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6353507089809116,
+      "learning_rate": 1.565821271188361e-05,
+      "loss": 1.0467,
+      "step": 13978
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6086446383873105,
+      "learning_rate": 1.5657583546754744e-05,
+      "loss": 1.0691,
+      "step": 13979
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0376731585671344,
+      "learning_rate": 1.565695434868564e-05,
+      "loss": 0.9664,
+      "step": 13980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7341065611023794,
+      "learning_rate": 1.5656325117679957e-05,
+      "loss": 0.9978,
+      "step": 13981
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5897098216921213,
+      "learning_rate": 1.5655695853741353e-05,
+      "loss": 0.8368,
+      "step": 13982
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6380505355182442,
+      "learning_rate": 1.5655066556873502e-05,
+      "loss": 1.0256,
+      "step": 13983
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.028540325474674,
+      "learning_rate": 1.565443722708006e-05,
+      "loss": 0.8047,
+      "step": 13984
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6770460914453114,
+      "learning_rate": 1.5653807864364696e-05,
+      "loss": 0.843,
+      "step": 13985
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8021090946763687,
+      "learning_rate": 1.5653178468731076e-05,
+      "loss": 0.8778,
+      "step": 13986
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3760253209852165,
+      "learning_rate": 1.5652549040182857e-05,
+      "loss": 0.9465,
+      "step": 13987
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5926256046519858,
+      "learning_rate": 1.5651919578723708e-05,
+      "loss": 0.8409,
+      "step": 13988
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7941725257315615,
+      "learning_rate": 1.5651290084357295e-05,
+      "loss": 0.9885,
+      "step": 13989
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5768787264688633,
+      "learning_rate": 1.5650660557087284e-05,
+      "loss": 0.883,
+      "step": 13990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9994406505561115,
+      "learning_rate": 1.5650030996917334e-05,
+      "loss": 1.0232,
+      "step": 13991
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6188010655199747,
+      "learning_rate": 1.5649401403851115e-05,
+      "loss": 0.8501,
+      "step": 13992
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7745425356494315,
+      "learning_rate": 1.5648771777892296e-05,
+      "loss": 0.6984,
+      "step": 13993
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5405119982452364,
+      "learning_rate": 1.564814211904454e-05,
+      "loss": 0.9548,
+      "step": 13994
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.192715749717499,
+      "learning_rate": 1.5647512427311508e-05,
+      "loss": 0.9324,
+      "step": 13995
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9844760601964522,
+      "learning_rate": 1.564688270269687e-05,
+      "loss": 0.8101,
+      "step": 13996
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8150614804070188,
+      "learning_rate": 1.5646252945204298e-05,
+      "loss": 0.7058,
+      "step": 13997
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5082273361583454,
+      "learning_rate": 1.5645623154837454e-05,
+      "loss": 0.9257,
+      "step": 13998
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7501717861078885,
+      "learning_rate": 1.56449933316e-05,
+      "loss": 0.9366,
+      "step": 13999
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5453479301963022,
+      "learning_rate": 1.5644363475495614e-05,
+      "loss": 0.9411,
+      "step": 14000
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7388285069434964,
+      "learning_rate": 1.564373358652795e-05,
+      "loss": 0.7033,
+      "step": 14001
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7513500712572524,
+      "learning_rate": 1.5643103664700687e-05,
+      "loss": 0.6865,
+      "step": 14002
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6963084964539366,
+      "learning_rate": 1.564247371001748e-05,
+      "loss": 0.9597,
+      "step": 14003
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6479409840872559,
+      "learning_rate": 1.5641843722482015e-05,
+      "loss": 0.7695,
+      "step": 14004
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5278531480258244,
+      "learning_rate": 1.5641213702097947e-05,
+      "loss": 0.9163,
+      "step": 14005
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8007337509040622,
+      "learning_rate": 1.5640583648868943e-05,
+      "loss": 0.8471,
+      "step": 14006
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8880816807598524,
+      "learning_rate": 1.5639953562798676e-05,
+      "loss": 1.0099,
+      "step": 14007
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6830690879779826,
+      "learning_rate": 1.563932344389082e-05,
+      "loss": 0.9544,
+      "step": 14008
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6307936817203887,
+      "learning_rate": 1.563869329214903e-05,
+      "loss": 0.913,
+      "step": 14009
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7546193547052664,
+      "learning_rate": 1.5638063107576983e-05,
+      "loss": 0.8369,
+      "step": 14010
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6533065477833015,
+      "learning_rate": 1.5637432890178354e-05,
+      "loss": 0.8803,
+      "step": 14011
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6382111669108783,
+      "learning_rate": 1.5636802639956798e-05,
+      "loss": 1.0409,
+      "step": 14012
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8860452761434383,
+      "learning_rate": 1.5636172356915995e-05,
+      "loss": 0.9301,
+      "step": 14013
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5619445592120622,
+      "learning_rate": 1.5635542041059615e-05,
+      "loss": 0.7081,
+      "step": 14014
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.045490586003302,
+      "learning_rate": 1.5634911692391323e-05,
+      "loss": 0.8949,
+      "step": 14015
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6423043778333768,
+      "learning_rate": 1.5634281310914792e-05,
+      "loss": 0.8783,
+      "step": 14016
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.543894371479287,
+      "learning_rate": 1.563365089663369e-05,
+      "loss": 0.9431,
+      "step": 14017
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5805312814980694,
+      "learning_rate": 1.5633020449551694e-05,
+      "loss": 0.8522,
+      "step": 14018
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.220770356462019,
+      "learning_rate": 1.5632389969672462e-05,
+      "loss": 0.9145,
+      "step": 14019
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5849421977329679,
+      "learning_rate": 1.5631759456999675e-05,
+      "loss": 0.856,
+      "step": 14020
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8178885659919728,
+      "learning_rate": 1.5631128911537007e-05,
+      "loss": 0.8723,
+      "step": 14021
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5176510433366168,
+      "learning_rate": 1.5630498333288115e-05,
+      "loss": 1.0014,
+      "step": 14022
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7789728039531727,
+      "learning_rate": 1.5629867722256684e-05,
+      "loss": 0.9573,
+      "step": 14023
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7875154454402072,
+      "learning_rate": 1.5629237078446378e-05,
+      "loss": 0.7334,
+      "step": 14024
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6276245405968393,
+      "learning_rate": 1.5628606401860875e-05,
+      "loss": 0.8615,
+      "step": 14025
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6702538235311823,
+      "learning_rate": 1.562797569250384e-05,
+      "loss": 0.9677,
+      "step": 14026
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9622370243052527,
+      "learning_rate": 1.562734495037895e-05,
+      "loss": 0.7693,
+      "step": 14027
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7898155901845574,
+      "learning_rate": 1.5626714175489877e-05,
+      "loss": 0.9756,
+      "step": 14028
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6336732394264248,
+      "learning_rate": 1.5626083367840296e-05,
+      "loss": 0.9425,
+      "step": 14029
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8073512329385111,
+      "learning_rate": 1.562545252743387e-05,
+      "loss": 0.7753,
+      "step": 14030
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.660065649063772,
+      "learning_rate": 1.562482165427428e-05,
+      "loss": 0.8518,
+      "step": 14031
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.007058047989511,
+      "learning_rate": 1.56241907483652e-05,
+      "loss": 0.8104,
+      "step": 14032
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.046374208207114,
+      "learning_rate": 1.56235598097103e-05,
+      "loss": 0.9094,
+      "step": 14033
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6025235225203271,
+      "learning_rate": 1.5622928838313254e-05,
+      "loss": 0.9151,
+      "step": 14034
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.641631373646644,
+      "learning_rate": 1.5622297834177738e-05,
+      "loss": 0.8624,
+      "step": 14035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6655782135017159,
+      "learning_rate": 1.562166679730742e-05,
+      "loss": 0.9774,
+      "step": 14036
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6281018033501387,
+      "learning_rate": 1.5621035727705982e-05,
+      "loss": 0.848,
+      "step": 14037
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.00039778429475,
+      "learning_rate": 1.562040462537709e-05,
+      "loss": 0.8845,
+      "step": 14038
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8251533993538992,
+      "learning_rate": 1.5619773490324427e-05,
+      "loss": 0.7043,
+      "step": 14039
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.282586552294611,
+      "learning_rate": 1.561914232255166e-05,
+      "loss": 0.8967,
+      "step": 14040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.133713910140083,
+      "learning_rate": 1.5618511122062475e-05,
+      "loss": 0.8734,
+      "step": 14041
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3946896283496597,
+      "learning_rate": 1.5617879888860533e-05,
+      "loss": 0.9262,
+      "step": 14042
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9438218823532294,
+      "learning_rate": 1.5617248622949515e-05,
+      "loss": 0.9124,
+      "step": 14043
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6952452982250439,
+      "learning_rate": 1.56166173243331e-05,
+      "loss": 0.837,
+      "step": 14044
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.774845829612289,
+      "learning_rate": 1.5615985993014956e-05,
+      "loss": 0.8823,
+      "step": 14045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7866228782059647,
+      "learning_rate": 1.561535462899877e-05,
+      "loss": 0.6649,
+      "step": 14046
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6294710424732584,
+      "learning_rate": 1.561472323228821e-05,
+      "loss": 0.8483,
+      "step": 14047
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7700328437964267,
+      "learning_rate": 1.561409180288695e-05,
+      "loss": 0.7282,
+      "step": 14048
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7104875157092971,
+      "learning_rate": 1.5613460340798674e-05,
+      "loss": 0.9638,
+      "step": 14049
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4340514131109932,
+      "learning_rate": 1.5612828846027052e-05,
+      "loss": 0.8767,
+      "step": 14050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7453477283412229,
+      "learning_rate": 1.5612197318575767e-05,
+      "loss": 0.7067,
+      "step": 14051
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5055058393986396,
+      "learning_rate": 1.561156575844849e-05,
+      "loss": 0.9496,
+      "step": 14052
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.013283751977717,
+      "learning_rate": 1.56109341656489e-05,
+      "loss": 0.8459,
+      "step": 14053
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9462656229893105,
+      "learning_rate": 1.5610302540180674e-05,
+      "loss": 0.8593,
+      "step": 14054
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.149209518074263,
+      "learning_rate": 1.5609670882047492e-05,
+      "loss": 0.9306,
+      "step": 14055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6292564992472196,
+      "learning_rate": 1.5609039191253032e-05,
+      "loss": 0.7678,
+      "step": 14056
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7935578532996108,
+      "learning_rate": 1.560840746780097e-05,
+      "loss": 0.9265,
+      "step": 14057
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.04735415586232,
+      "learning_rate": 1.560777571169498e-05,
+      "loss": 0.8705,
+      "step": 14058
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.770599905165935,
+      "learning_rate": 1.5607143922938747e-05,
+      "loss": 0.8759,
+      "step": 14059
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7742598292706753,
+      "learning_rate": 1.5606512101535947e-05,
+      "loss": 0.8781,
+      "step": 14060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5716529008791895,
+      "learning_rate": 1.560588024749026e-05,
+      "loss": 0.8447,
+      "step": 14061
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7412244858203967,
+      "learning_rate": 1.5605248360805358e-05,
+      "loss": 0.8003,
+      "step": 14062
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.523742139828796,
+      "learning_rate": 1.560461644148493e-05,
+      "loss": 0.8864,
+      "step": 14063
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8314968110052126,
+      "learning_rate": 1.560398448953265e-05,
+      "loss": 0.8178,
+      "step": 14064
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.66149236381922,
+      "learning_rate": 1.56033525049522e-05,
+      "loss": 0.8758,
+      "step": 14065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8372508379403292,
+      "learning_rate": 1.560272048774725e-05,
+      "loss": 1.0438,
+      "step": 14066
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4053552306689299,
+      "learning_rate": 1.5602088437921497e-05,
+      "loss": 0.8136,
+      "step": 14067
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5991866706549827,
+      "learning_rate": 1.5601456355478606e-05,
+      "loss": 0.8108,
+      "step": 14068
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6531137355964205,
+      "learning_rate": 1.5600824240422266e-05,
+      "loss": 0.9773,
+      "step": 14069
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6002757720316483,
+      "learning_rate": 1.5600192092756154e-05,
+      "loss": 0.8808,
+      "step": 14070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7517619942742144,
+      "learning_rate": 1.5599559912483945e-05,
+      "loss": 0.897,
+      "step": 14071
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5037409156828108,
+      "learning_rate": 1.559892769960933e-05,
+      "loss": 0.8773,
+      "step": 14072
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7829142935181287,
+      "learning_rate": 1.5598295454135985e-05,
+      "loss": 0.7261,
+      "step": 14073
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7665326953891507,
+      "learning_rate": 1.5597663176067596e-05,
+      "loss": 0.7073,
+      "step": 14074
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7854613350751011,
+      "learning_rate": 1.5597030865407828e-05,
+      "loss": 0.7974,
+      "step": 14075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.637392664609617,
+      "learning_rate": 1.5596398522160385e-05,
+      "loss": 0.9421,
+      "step": 14076
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7532536018937235,
+      "learning_rate": 1.5595766146328934e-05,
+      "loss": 0.976,
+      "step": 14077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8987239874556163,
+      "learning_rate": 1.5595133737917163e-05,
+      "loss": 0.9913,
+      "step": 14078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2316730619244245,
+      "learning_rate": 1.559450129692875e-05,
+      "loss": 0.8855,
+      "step": 14079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9265960449869215,
+      "learning_rate": 1.5593868823367378e-05,
+      "loss": 0.9763,
+      "step": 14080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6051373793467285,
+      "learning_rate": 1.5593236317236733e-05,
+      "loss": 0.7552,
+      "step": 14081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.668365702834658,
+      "learning_rate": 1.5592603778540493e-05,
+      "loss": 0.8821,
+      "step": 14082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0275238275024474,
+      "learning_rate": 1.5591971207282345e-05,
+      "loss": 0.8525,
+      "step": 14083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.11252448300382,
+      "learning_rate": 1.5591338603465972e-05,
+      "loss": 0.9826,
+      "step": 14084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.508362868619728,
+      "learning_rate": 1.5590705967095053e-05,
+      "loss": 0.9188,
+      "step": 14085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.62804575028605,
+      "learning_rate": 1.5590073298173275e-05,
+      "loss": 0.8224,
+      "step": 14086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4795866791093844,
+      "learning_rate": 1.558944059670432e-05,
+      "loss": 0.8155,
+      "step": 14087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4684170694137115,
+      "learning_rate": 1.5588807862691874e-05,
+      "loss": 0.8539,
+      "step": 14088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0021131758270596,
+      "learning_rate": 1.5588175096139614e-05,
+      "loss": 0.7924,
+      "step": 14089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5615830364783538,
+      "learning_rate": 1.5587542297051233e-05,
+      "loss": 0.9499,
+      "step": 14090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.678662579024652,
+      "learning_rate": 1.5586909465430414e-05,
+      "loss": 0.8807,
+      "step": 14091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5359357432089362,
+      "learning_rate": 1.5586276601280833e-05,
+      "loss": 0.7242,
+      "step": 14092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.148664514715343,
+      "learning_rate": 1.558564370460619e-05,
+      "loss": 0.9501,
+      "step": 14093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6271972811678166,
+      "learning_rate": 1.5585010775410153e-05,
+      "loss": 0.9332,
+      "step": 14094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4550190123129254,
+      "learning_rate": 1.5584377813696417e-05,
+      "loss": 0.8816,
+      "step": 14095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.43439973786584,
+      "learning_rate": 1.5583744819468667e-05,
+      "loss": 0.8184,
+      "step": 14096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8258116305267096,
+      "learning_rate": 1.5583111792730587e-05,
+      "loss": 0.9949,
+      "step": 14097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4952024173962981,
+      "learning_rate": 1.558247873348586e-05,
+      "loss": 0.8428,
+      "step": 14098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5791387312352132,
+      "learning_rate": 1.558184564173818e-05,
+      "loss": 0.9066,
+      "step": 14099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5146980236023655,
+      "learning_rate": 1.5581212517491224e-05,
+      "loss": 0.965,
+      "step": 14100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.675505911442009,
+      "learning_rate": 1.5580579360748676e-05,
+      "loss": 0.9161,
+      "step": 14101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.55858939265536,
+      "learning_rate": 1.5579946171514237e-05,
+      "loss": 0.843,
+      "step": 14102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0939101833153315,
+      "learning_rate": 1.557931294979158e-05,
+      "loss": 0.7808,
+      "step": 14103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7572973235499278,
+      "learning_rate": 1.5578679695584398e-05,
+      "loss": 0.7138,
+      "step": 14104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1689670309444047,
+      "learning_rate": 1.5578046408896378e-05,
+      "loss": 0.8311,
+      "step": 14105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6172982044599429,
+      "learning_rate": 1.5577413089731204e-05,
+      "loss": 0.8306,
+      "step": 14106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.101725359005353,
+      "learning_rate": 1.5576779738092566e-05,
+      "loss": 0.9381,
+      "step": 14107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7345488457526719,
+      "learning_rate": 1.557614635398415e-05,
+      "loss": 1.0598,
+      "step": 14108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5330222226589632,
+      "learning_rate": 1.5575512937409648e-05,
+      "loss": 0.8417,
+      "step": 14109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6330031212106384,
+      "learning_rate": 1.5574879488372738e-05,
+      "loss": 0.8614,
+      "step": 14110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.960199946075964,
+      "learning_rate": 1.5574246006877118e-05,
+      "loss": 0.9606,
+      "step": 14111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6293446115506522,
+      "learning_rate": 1.5573612492926472e-05,
+      "loss": 0.8681,
+      "step": 14112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5558828766406267,
+      "learning_rate": 1.5572978946524493e-05,
+      "loss": 0.8633,
+      "step": 14113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5614128510138816,
+      "learning_rate": 1.5572345367674864e-05,
+      "loss": 0.8877,
+      "step": 14114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5047912932638483,
+      "learning_rate": 1.5571711756381274e-05,
+      "loss": 0.9441,
+      "step": 14115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7450676282358786,
+      "learning_rate": 1.557107811264742e-05,
+      "loss": 0.8715,
+      "step": 14116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6659170710980078,
+      "learning_rate": 1.557044443647698e-05,
+      "loss": 0.7839,
+      "step": 14117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7107861730365104,
+      "learning_rate": 1.5569810727873652e-05,
+      "loss": 0.922,
+      "step": 14118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7679929651686446,
+      "learning_rate": 1.556917698684112e-05,
+      "loss": 0.8918,
+      "step": 14119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7444753993863293,
+      "learning_rate": 1.5568543213383076e-05,
+      "loss": 0.8635,
+      "step": 14120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5135221568212216,
+      "learning_rate": 1.5567909407503213e-05,
+      "loss": 1.0279,
+      "step": 14121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8455255095258823,
+      "learning_rate": 1.5567275569205216e-05,
+      "loss": 0.7398,
+      "step": 14122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.655661179953462,
+      "learning_rate": 1.556664169849278e-05,
+      "loss": 0.86,
+      "step": 14123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8413938576790261,
+      "learning_rate": 1.5566007795369594e-05,
+      "loss": 0.7689,
+      "step": 14124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5560041019013073,
+      "learning_rate": 1.5565373859839348e-05,
+      "loss": 0.8327,
+      "step": 14125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0669514444958645,
+      "learning_rate": 1.5564739891905734e-05,
+      "loss": 0.835,
+      "step": 14126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7088460485641768,
+      "learning_rate": 1.556410589157244e-05,
+      "loss": 0.8542,
+      "step": 14127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7211220277068708,
+      "learning_rate": 1.5563471858843163e-05,
+      "loss": 0.6879,
+      "step": 14128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4712942166738534,
+      "learning_rate": 1.556283779372159e-05,
+      "loss": 0.8859,
+      "step": 14129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8398557302098597,
+      "learning_rate": 1.5562203696211416e-05,
+      "loss": 0.9167,
+      "step": 14130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7404450888490695,
+      "learning_rate": 1.5561569566316327e-05,
+      "loss": 0.6915,
+      "step": 14131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9421037110946944,
+      "learning_rate": 1.5560935404040023e-05,
+      "loss": 0.9505,
+      "step": 14132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.717354237508155,
+      "learning_rate": 1.556030120938619e-05,
+      "loss": 0.9057,
+      "step": 14133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.860363994094408,
+      "learning_rate": 1.5559666982358525e-05,
+      "loss": 0.8942,
+      "step": 14134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.102211256013119,
+      "learning_rate": 1.555903272296072e-05,
+      "loss": 0.8836,
+      "step": 14135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7530063460637249,
+      "learning_rate": 1.5558398431196465e-05,
+      "loss": 0.845,
+      "step": 14136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6198832398390728,
+      "learning_rate": 1.5557764107069454e-05,
+      "loss": 0.8749,
+      "step": 14137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8515734029652093,
+      "learning_rate": 1.555712975058338e-05,
+      "loss": 0.9301,
+      "step": 14138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.583111285281978,
+      "learning_rate": 1.5556495361741936e-05,
+      "loss": 0.9281,
+      "step": 14139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.501803316252246,
+      "learning_rate": 1.555586094054882e-05,
+      "loss": 0.8672,
+      "step": 14140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9320779885108643,
+      "learning_rate": 1.5555226487007723e-05,
+      "loss": 0.9567,
+      "step": 14141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.782037524443305,
+      "learning_rate": 1.5554592001122335e-05,
+      "loss": 0.8768,
+      "step": 14142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5371326534669776,
+      "learning_rate": 1.5553957482896356e-05,
+      "loss": 0.9789,
+      "step": 14143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4262695501251441,
+      "learning_rate": 1.5553322932333477e-05,
+      "loss": 0.8675,
+      "step": 14144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8074336870419963,
+      "learning_rate": 1.5552688349437396e-05,
+      "loss": 0.8697,
+      "step": 14145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6282515432909996,
+      "learning_rate": 1.5552053734211805e-05,
+      "loss": 0.8716,
+      "step": 14146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7040457728689236,
+      "learning_rate": 1.55514190866604e-05,
+      "loss": 0.9159,
+      "step": 14147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7584967863453728,
+      "learning_rate": 1.5550784406786875e-05,
+      "loss": 0.7163,
+      "step": 14148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9549437383453725,
+      "learning_rate": 1.5550149694594924e-05,
+      "loss": 0.8574,
+      "step": 14149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.907427380744378,
+      "learning_rate": 1.5549514950088243e-05,
+      "loss": 0.8677,
+      "step": 14150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.763616847734174,
+      "learning_rate": 1.554888017327053e-05,
+      "loss": 0.9591,
+      "step": 14151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6154158581370168,
+      "learning_rate": 1.5548245364145486e-05,
+      "loss": 0.7966,
+      "step": 14152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7867595503006746,
+      "learning_rate": 1.5547610522716796e-05,
+      "loss": 0.7527,
+      "step": 14153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8705449634870772,
+      "learning_rate": 1.5546975648988157e-05,
+      "loss": 0.8582,
+      "step": 14154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.646627447364968,
+      "learning_rate": 1.5546340742963278e-05,
+      "loss": 0.9158,
+      "step": 14155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1032023838069787,
+      "learning_rate": 1.554570580464584e-05,
+      "loss": 1.0315,
+      "step": 14156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.630913529614064,
+      "learning_rate": 1.5545070834039556e-05,
+      "loss": 0.8365,
+      "step": 14157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5555957123934694,
+      "learning_rate": 1.5544435831148107e-05,
+      "loss": 0.9444,
+      "step": 14158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8852706075024646,
+      "learning_rate": 1.55438007959752e-05,
+      "loss": 0.7861,
+      "step": 14159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.523751563396392,
+      "learning_rate": 1.554316572852453e-05,
+      "loss": 0.7362,
+      "step": 14160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7463721464239507,
+      "learning_rate": 1.554253062879979e-05,
+      "loss": 0.7567,
+      "step": 14161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5539735238972483,
+      "learning_rate": 1.5541895496804686e-05,
+      "loss": 0.8444,
+      "step": 14162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.780716750947693,
+      "learning_rate": 1.5541260332542906e-05,
+      "loss": 0.9684,
+      "step": 14163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6019062282692644,
+      "learning_rate": 1.554062513601816e-05,
+      "loss": 0.9098,
+      "step": 14164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6946825958545424,
+      "learning_rate": 1.553998990723414e-05,
+      "loss": 0.8239,
+      "step": 14165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7150866929318194,
+      "learning_rate": 1.5539354646194544e-05,
+      "loss": 0.868,
+      "step": 14166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7957768270729972,
+      "learning_rate": 1.553871935290307e-05,
+      "loss": 0.8232,
+      "step": 14167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6684503067091527,
+      "learning_rate": 1.553808402736342e-05,
+      "loss": 0.8897,
+      "step": 14168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7550972255354722,
+      "learning_rate": 1.553744866957929e-05,
+      "loss": 0.7921,
+      "step": 14169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5823742324741554,
+      "learning_rate": 1.553681327955438e-05,
+      "loss": 0.9757,
+      "step": 14170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9224936907901842,
+      "learning_rate": 1.5536177857292392e-05,
+      "loss": 0.8555,
+      "step": 14171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.736833611329528,
+      "learning_rate": 1.5535542402797027e-05,
+      "loss": 0.8814,
+      "step": 14172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6544342190718524,
+      "learning_rate": 1.5534906916071976e-05,
+      "loss": 1.0501,
+      "step": 14173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7713974189191236,
+      "learning_rate": 1.5534271397120947e-05,
+      "loss": 0.7225,
+      "step": 14174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4187267946864817,
+      "learning_rate": 1.553363584594764e-05,
+      "loss": 0.8952,
+      "step": 14175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9478784425354938,
+      "learning_rate": 1.553300026255575e-05,
+      "loss": 0.9003,
+      "step": 14176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7561981744808535,
+      "learning_rate": 1.553236464694898e-05,
+      "loss": 0.9478,
+      "step": 14177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.485299634386292,
+      "learning_rate": 1.5531728999131034e-05,
+      "loss": 0.9275,
+      "step": 14178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7795456981503367,
+      "learning_rate": 1.553109331910561e-05,
+      "loss": 0.7066,
+      "step": 14179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4733977102225253,
+      "learning_rate": 1.553045760687641e-05,
+      "loss": 0.8424,
+      "step": 14180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7729473240880945,
+      "learning_rate": 1.5529821862447133e-05,
+      "loss": 0.7863,
+      "step": 14181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0930543060277342,
+      "learning_rate": 1.5529186085821485e-05,
+      "loss": 0.8625,
+      "step": 14182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8100821097583886,
+      "learning_rate": 1.552855027700316e-05,
+      "loss": 0.7611,
+      "step": 14183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4332587593724249,
+      "learning_rate": 1.552791443599587e-05,
+      "loss": 0.9623,
+      "step": 14184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8436930107332201,
+      "learning_rate": 1.5527278562803313e-05,
+      "loss": 0.9893,
+      "step": 14185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5716551284789972,
+      "learning_rate": 1.552664265742919e-05,
+      "loss": 0.9054,
+      "step": 14186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5951010858074393,
+      "learning_rate": 1.5526006719877198e-05,
+      "loss": 0.9834,
+      "step": 14187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7916140870756787,
+      "learning_rate": 1.552537075015105e-05,
+      "loss": 0.8016,
+      "step": 14188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6645920326081958,
+      "learning_rate": 1.5524734748254442e-05,
+      "loss": 0.9031,
+      "step": 14189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5626818879253166,
+      "learning_rate": 1.552409871419108e-05,
+      "loss": 0.8213,
+      "step": 14190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7891966337578615,
+      "learning_rate": 1.5523462647964667e-05,
+      "loss": 0.7224,
+      "step": 14191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7833766314754714,
+      "learning_rate": 1.5522826549578908e-05,
+      "loss": 0.701,
+      "step": 14192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7864826959063523,
+      "learning_rate": 1.55221904190375e-05,
+      "loss": 0.7145,
+      "step": 14193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.746621471966308,
+      "learning_rate": 1.5521554256344152e-05,
+      "loss": 0.8435,
+      "step": 14194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5660377946108006,
+      "learning_rate": 1.552091806150257e-05,
+      "loss": 0.8923,
+      "step": 14195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6734491371152869,
+      "learning_rate": 1.5520281834516452e-05,
+      "loss": 1.0026,
+      "step": 14196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1253240858830544,
+      "learning_rate": 1.5519645575389508e-05,
+      "loss": 0.986,
+      "step": 14197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5187624624369183,
+      "learning_rate": 1.5519009284125438e-05,
+      "loss": 0.8519,
+      "step": 14198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.756217387661434,
+      "learning_rate": 1.5518372960727947e-05,
+      "loss": 0.7487,
+      "step": 14199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7222737012035927,
+      "learning_rate": 1.5517736605200745e-05,
+      "loss": 0.8976,
+      "step": 14200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.625058737345759,
+      "learning_rate": 1.5517100217547532e-05,
+      "loss": 0.8021,
+      "step": 14201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9417724910489556,
+      "learning_rate": 1.5516463797772017e-05,
+      "loss": 0.8079,
+      "step": 14202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0720990408215973,
+      "learning_rate": 1.55158273458779e-05,
+      "loss": 1.0122,
+      "step": 14203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8650664887311519,
+      "learning_rate": 1.551519086186889e-05,
+      "loss": 0.8738,
+      "step": 14204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.629481840070093,
+      "learning_rate": 1.5514554345748694e-05,
+      "loss": 0.8948,
+      "step": 14205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6620746992859878,
+      "learning_rate": 1.551391779752102e-05,
+      "loss": 0.9633,
+      "step": 14206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8199048672609526,
+      "learning_rate": 1.5513281217189565e-05,
+      "loss": 0.9247,
+      "step": 14207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9570787808553,
+      "learning_rate": 1.5512644604758046e-05,
+      "loss": 0.9175,
+      "step": 14208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5358939721788092,
+      "learning_rate": 1.551200796023016e-05,
+      "loss": 0.8479,
+      "step": 14209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5296059195865825,
+      "learning_rate": 1.5511371283609622e-05,
+      "loss": 0.874,
+      "step": 14210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6380123346843263,
+      "learning_rate": 1.5510734574900137e-05,
+      "loss": 0.9649,
+      "step": 14211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8331354368916606,
+      "learning_rate": 1.5510097834105406e-05,
+      "loss": 0.923,
+      "step": 14212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9217264780067969,
+      "learning_rate": 1.5509461061229147e-05,
+      "loss": 0.9912,
+      "step": 14213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8151203189172764,
+      "learning_rate": 1.5508824256275055e-05,
+      "loss": 0.9949,
+      "step": 14214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.66653079439913,
+      "learning_rate": 1.550818741924685e-05,
+      "loss": 0.8796,
+      "step": 14215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.345083008829194,
+      "learning_rate": 1.550755055014823e-05,
+      "loss": 0.904,
+      "step": 14216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.596320772715889,
+      "learning_rate": 1.550691364898291e-05,
+      "loss": 0.9685,
+      "step": 14217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4744765322102475,
+      "learning_rate": 1.5506276715754592e-05,
+      "loss": 0.8731,
+      "step": 14218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7020777904750566,
+      "learning_rate": 1.5505639750466986e-05,
+      "loss": 1.0263,
+      "step": 14219
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5075482578906576,
+      "learning_rate": 1.5505002753123807e-05,
+      "loss": 1.0246,
+      "step": 14220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7132997085485395,
+      "learning_rate": 1.550436572372876e-05,
+      "loss": 1.0114,
+      "step": 14221
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.464354322786369,
+      "learning_rate": 1.550372866228555e-05,
+      "loss": 0.7618,
+      "step": 14222
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6844783012686275,
+      "learning_rate": 1.550309156879789e-05,
+      "loss": 0.9629,
+      "step": 14223
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7004357770933465,
+      "learning_rate": 1.5502454443269485e-05,
+      "loss": 0.9241,
+      "step": 14224
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7898578075227785,
+      "learning_rate": 1.5501817285704055e-05,
+      "loss": 0.8013,
+      "step": 14225
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5846523459830069,
+      "learning_rate": 1.55011800961053e-05,
+      "loss": 0.9595,
+      "step": 14226
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.095025601110617,
+      "learning_rate": 1.550054287447693e-05,
+      "loss": 0.847,
+      "step": 14227
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.214504563100106,
+      "learning_rate": 1.5499905620822662e-05,
+      "loss": 0.8153,
+      "step": 14228
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7817561450269468,
+      "learning_rate": 1.54992683351462e-05,
+      "loss": 0.8445,
+      "step": 14229
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5508259796005426,
+      "learning_rate": 1.5498631017451256e-05,
+      "loss": 0.8269,
+      "step": 14230
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5521408213465102,
+      "learning_rate": 1.5497993667741545e-05,
+      "loss": 0.796,
+      "step": 14231
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5898955404670447,
+      "learning_rate": 1.549735628602077e-05,
+      "loss": 0.9593,
+      "step": 14232
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3767740255325374,
+      "learning_rate": 1.5496718872292648e-05,
+      "loss": 0.8158,
+      "step": 14233
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.032167785612915,
+      "learning_rate": 1.549608142656089e-05,
+      "loss": 0.9756,
+      "step": 14234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5597897548600195,
+      "learning_rate": 1.5495443948829203e-05,
+      "loss": 0.8718,
+      "step": 14235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.117374938248296,
+      "learning_rate": 1.5494806439101306e-05,
+      "loss": 0.8793,
+      "step": 14236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5714905585852552,
+      "learning_rate": 1.5494168897380905e-05,
+      "loss": 0.8786,
+      "step": 14237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0995518297388056,
+      "learning_rate": 1.5493531323671707e-05,
+      "loss": 0.8285,
+      "step": 14238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6531818009767145,
+      "learning_rate": 1.5492893717977438e-05,
+      "loss": 0.8738,
+      "step": 14239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5917396305108864,
+      "learning_rate": 1.54922560803018e-05,
+      "loss": 0.8399,
+      "step": 14240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5536570093085647,
+      "learning_rate": 1.5491618410648507e-05,
+      "loss": 0.8245,
+      "step": 14241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.748159931615251,
+      "learning_rate": 1.549098070902127e-05,
+      "loss": 0.97,
+      "step": 14242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6009428213446266,
+      "learning_rate": 1.5490342975423814e-05,
+      "loss": 0.924,
+      "step": 14243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.707042083501947,
+      "learning_rate": 1.5489705209859838e-05,
+      "loss": 0.9531,
+      "step": 14244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6195690472377362,
+      "learning_rate": 1.548906741233306e-05,
+      "loss": 0.9362,
+      "step": 14245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4826801020406772,
+      "learning_rate": 1.5488429582847194e-05,
+      "loss": 0.8566,
+      "step": 14246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3865939352369547,
+      "learning_rate": 1.5487791721405953e-05,
+      "loss": 0.8015,
+      "step": 14247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2180830028659817,
+      "learning_rate": 1.548715382801305e-05,
+      "loss": 0.9771,
+      "step": 14248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.895152830575918,
+      "learning_rate": 1.54865159026722e-05,
+      "loss": 0.9745,
+      "step": 14249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8871553134493884,
+      "learning_rate": 1.548587794538712e-05,
+      "loss": 0.8675,
+      "step": 14250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.820136789177871,
+      "learning_rate": 1.5485239956161523e-05,
+      "loss": 0.9158,
+      "step": 14251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7995570934017234,
+      "learning_rate": 1.548460193499912e-05,
+      "loss": 0.7758,
+      "step": 14252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8309741857157582,
+      "learning_rate": 1.548396388190363e-05,
+      "loss": 0.9532,
+      "step": 14253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6391438278340063,
+      "learning_rate": 1.5483325796878764e-05,
+      "loss": 0.8414,
+      "step": 14254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6796383136829118,
+      "learning_rate": 1.548268767992824e-05,
+      "loss": 0.9012,
+      "step": 14255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5112938599741248,
+      "learning_rate": 1.5482049531055772e-05,
+      "loss": 0.7773,
+      "step": 14256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5531422305443563,
+      "learning_rate": 1.5481411350265075e-05,
+      "loss": 0.9506,
+      "step": 14257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.617853099380045,
+      "learning_rate": 1.5480773137559867e-05,
+      "loss": 0.9592,
+      "step": 14258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7893324670360422,
+      "learning_rate": 1.5480134892943865e-05,
+      "loss": 0.8864,
+      "step": 14259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8043905231782476,
+      "learning_rate": 1.5479496616420782e-05,
+      "loss": 0.7405,
+      "step": 14260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.599990598688629,
+      "learning_rate": 1.5478858307994333e-05,
+      "loss": 0.8198,
+      "step": 14261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.542109655664625,
+      "learning_rate": 1.5478219967668236e-05,
+      "loss": 0.7981,
+      "step": 14262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6298043589355304,
+      "learning_rate": 1.547758159544621e-05,
+      "loss": 0.9297,
+      "step": 14263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.56142683595922,
+      "learning_rate": 1.5476943191331965e-05,
+      "loss": 0.8689,
+      "step": 14264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4194421092126686,
+      "learning_rate": 1.5476304755329227e-05,
+      "loss": 0.9563,
+      "step": 14265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9812291926613574,
+      "learning_rate": 1.547566628744171e-05,
+      "loss": 0.8772,
+      "step": 14266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3901900444347906,
+      "learning_rate": 1.5475027787673126e-05,
+      "loss": 0.9651,
+      "step": 14267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8276860866534115,
+      "learning_rate": 1.54743892560272e-05,
+      "loss": 1.0507,
+      "step": 14268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6706429170265262,
+      "learning_rate": 1.5473750692507647e-05,
+      "loss": 0.9438,
+      "step": 14269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.737532700443121,
+      "learning_rate": 1.547311209711818e-05,
+      "loss": 0.7564,
+      "step": 14270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7041666817383045,
+      "learning_rate": 1.547247346986253e-05,
+      "loss": 0.8199,
+      "step": 14271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5059898829527185,
+      "learning_rate": 1.54718348107444e-05,
+      "loss": 0.8686,
+      "step": 14272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9211424429635666,
+      "learning_rate": 1.5471196119767515e-05,
+      "loss": 0.9995,
+      "step": 14273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6985256919573586,
+      "learning_rate": 1.5470557396935597e-05,
+      "loss": 0.7715,
+      "step": 14274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5404972768334138,
+      "learning_rate": 1.5469918642252363e-05,
+      "loss": 0.9029,
+      "step": 14275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5810237326531111,
+      "learning_rate": 1.5469279855721527e-05,
+      "loss": 0.9489,
+      "step": 14276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5230973819629492,
+      "learning_rate": 1.5468641037346816e-05,
+      "loss": 0.9192,
+      "step": 14277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1183786477527278,
+      "learning_rate": 1.5468002187131944e-05,
+      "loss": 0.8545,
+      "step": 14278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6703938020736613,
+      "learning_rate": 1.546736330508063e-05,
+      "loss": 0.8867,
+      "step": 14279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7313851811173331,
+      "learning_rate": 1.5466724391196598e-05,
+      "loss": 0.7879,
+      "step": 14280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6792842071724905,
+      "learning_rate": 1.5466085445483568e-05,
+      "loss": 0.9243,
+      "step": 14281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8228937703121668,
+      "learning_rate": 1.5465446467945254e-05,
+      "loss": 0.9904,
+      "step": 14282
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7874157017354269,
+      "learning_rate": 1.5464807458585383e-05,
+      "loss": 0.731,
+      "step": 14283
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8542485832617903,
+      "learning_rate": 1.546416841740767e-05,
+      "loss": 0.9343,
+      "step": 14284
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.381248648101603,
+      "learning_rate": 1.5463529344415842e-05,
+      "loss": 0.8432,
+      "step": 14285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9134004796883797,
+      "learning_rate": 1.5462890239613613e-05,
+      "loss": 0.6821,
+      "step": 14286
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7598739834777777,
+      "learning_rate": 1.546225110300471e-05,
+      "loss": 0.7319,
+      "step": 14287
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8154637620879484,
+      "learning_rate": 1.5461611934592852e-05,
+      "loss": 1.0614,
+      "step": 14288
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6694390386489557,
+      "learning_rate": 1.546097273438176e-05,
+      "loss": 0.9075,
+      "step": 14289
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6971555624322157,
+      "learning_rate": 1.5460333502375155e-05,
+      "loss": 1.0263,
+      "step": 14290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6777820457221253,
+      "learning_rate": 1.5459694238576757e-05,
+      "loss": 1.0206,
+      "step": 14291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7425708206343388,
+      "learning_rate": 1.5459054942990298e-05,
+      "loss": 0.7278,
+      "step": 14292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0310150292842173,
+      "learning_rate": 1.5458415615619486e-05,
+      "loss": 0.9085,
+      "step": 14293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5923436118154772,
+      "learning_rate": 1.545777625646805e-05,
+      "loss": 0.9058,
+      "step": 14294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.623925386655662,
+      "learning_rate": 1.5457136865539716e-05,
+      "loss": 0.9137,
+      "step": 14295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5104413744439193,
+      "learning_rate": 1.5456497442838202e-05,
+      "loss": 0.8797,
+      "step": 14296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4106582775532246,
+      "learning_rate": 1.5455857988367236e-05,
+      "loss": 0.9286,
+      "step": 14297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.001065357400254,
+      "learning_rate": 1.5455218502130533e-05,
+      "loss": 0.911,
+      "step": 14298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.805862323143483,
+      "learning_rate": 1.5454578984131822e-05,
+      "loss": 0.7512,
+      "step": 14299
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5187559948433609,
+      "learning_rate": 1.5453939434374824e-05,
+      "loss": 0.868,
+      "step": 14300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6458379865247792,
+      "learning_rate": 1.5453299852863265e-05,
+      "loss": 0.8682,
+      "step": 14301
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5960953594004528,
+      "learning_rate": 1.5452660239600873e-05,
+      "loss": 0.8343,
+      "step": 14302
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6421902528660006,
+      "learning_rate": 1.545202059459136e-05,
+      "loss": 0.9701,
+      "step": 14303
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.761393519168777,
+      "learning_rate": 1.5451380917838462e-05,
+      "loss": 0.8459,
+      "step": 14304
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9143596140009003,
+      "learning_rate": 1.5450741209345894e-05,
+      "loss": 0.9045,
+      "step": 14305
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5478388481565604,
+      "learning_rate": 1.5450101469117386e-05,
+      "loss": 0.8726,
+      "step": 14306
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.768458234269401,
+      "learning_rate": 1.544946169715666e-05,
+      "loss": 0.8873,
+      "step": 14307
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5117431731524464,
+      "learning_rate": 1.5448821893467444e-05,
+      "loss": 0.8675,
+      "step": 14308
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8854721720456362,
+      "learning_rate": 1.5448182058053466e-05,
+      "loss": 0.8912,
+      "step": 14309
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6371126119435842,
+      "learning_rate": 1.5447542190918444e-05,
+      "loss": 0.8519,
+      "step": 14310
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.681610177983008,
+      "learning_rate": 1.5446902292066106e-05,
+      "loss": 0.8222,
+      "step": 14311
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7554787258773057,
+      "learning_rate": 1.5446262361500178e-05,
+      "loss": 0.9514,
+      "step": 14312
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1016381932397383,
+      "learning_rate": 1.544562239922439e-05,
+      "loss": 0.8627,
+      "step": 14313
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7164836687250733,
+      "learning_rate": 1.544498240524246e-05,
+      "loss": 0.9034,
+      "step": 14314
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.959058337708678,
+      "learning_rate": 1.544434237955812e-05,
+      "loss": 0.8661,
+      "step": 14315
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5679203786101186,
+      "learning_rate": 1.5443702322175095e-05,
+      "loss": 0.891,
+      "step": 14316
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5791566888674111,
+      "learning_rate": 1.544306223309711e-05,
+      "loss": 0.9914,
+      "step": 14317
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5856995116269383,
+      "learning_rate": 1.54424221123279e-05,
+      "loss": 0.993,
+      "step": 14318
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6003277123491235,
+      "learning_rate": 1.544178195987118e-05,
+      "loss": 0.9548,
+      "step": 14319
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6235648318496987,
+      "learning_rate": 1.5441141775730685e-05,
+      "loss": 0.956,
+      "step": 14320
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7700423943753028,
+      "learning_rate": 1.5440501559910135e-05,
+      "loss": 0.8825,
+      "step": 14321
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0430219960271496,
+      "learning_rate": 1.5439861312413267e-05,
+      "loss": 0.8866,
+      "step": 14322
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9801404267354943,
+      "learning_rate": 1.5439221033243804e-05,
+      "loss": 0.9658,
+      "step": 14323
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6702879460192606,
+      "learning_rate": 1.5438580722405473e-05,
+      "loss": 0.8208,
+      "step": 14324
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1724825524164584,
+      "learning_rate": 1.5437940379902e-05,
+      "loss": 0.869,
+      "step": 14325
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.71064303633954,
+      "learning_rate": 1.5437300005737122e-05,
+      "loss": 0.9968,
+      "step": 14326
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.600018780136102,
+      "learning_rate": 1.543665959991456e-05,
+      "loss": 0.9449,
+      "step": 14327
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0003655836712384,
+      "learning_rate": 1.5436019162438044e-05,
+      "loss": 1.0175,
+      "step": 14328
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.62978778799343,
+      "learning_rate": 1.5435378693311305e-05,
+      "loss": 0.9012,
+      "step": 14329
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8384827187902718,
+      "learning_rate": 1.5434738192538067e-05,
+      "loss": 0.7612,
+      "step": 14330
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6299820922260166,
+      "learning_rate": 1.5434097660122066e-05,
+      "loss": 0.7694,
+      "step": 14331
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7781336982408813,
+      "learning_rate": 1.5433457096067027e-05,
+      "loss": 0.9902,
+      "step": 14332
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8885859287249942,
+      "learning_rate": 1.543281650037668e-05,
+      "loss": 0.8658,
+      "step": 14333
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.188195966085785,
+      "learning_rate": 1.5432175873054758e-05,
+      "loss": 0.8807,
+      "step": 14334
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6653847809348425,
+      "learning_rate": 1.543153521410498e-05,
+      "loss": 0.9905,
+      "step": 14335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3698550280896633,
+      "learning_rate": 1.5430894523531094e-05,
+      "loss": 0.8506,
+      "step": 14336
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8889762744435439,
+      "learning_rate": 1.5430253801336816e-05,
+      "loss": 1.0001,
+      "step": 14337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8657094726262684,
+      "learning_rate": 1.5429613047525885e-05,
+      "loss": 0.8871,
+      "step": 14338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6081953088358714,
+      "learning_rate": 1.5428972262102027e-05,
+      "loss": 1.0218,
+      "step": 14339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6011400267240825,
+      "learning_rate": 1.542833144506897e-05,
+      "loss": 0.8278,
+      "step": 14340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9291183697441554,
+      "learning_rate": 1.542769059643045e-05,
+      "loss": 0.875,
+      "step": 14341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8568931651547993,
+      "learning_rate": 1.5427049716190196e-05,
+      "loss": 0.8488,
+      "step": 14342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4471717958015515,
+      "learning_rate": 1.5426408804351943e-05,
+      "loss": 0.9859,
+      "step": 14343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6297587465490186,
+      "learning_rate": 1.542576786091942e-05,
+      "loss": 0.9493,
+      "step": 14344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.728281390219261,
+      "learning_rate": 1.5425126885896357e-05,
+      "loss": 0.9686,
+      "step": 14345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9686424238634461,
+      "learning_rate": 1.5424485879286488e-05,
+      "loss": 0.9022,
+      "step": 14346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5799877200895118,
+      "learning_rate": 1.5423844841093545e-05,
+      "loss": 0.8676,
+      "step": 14347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.783828897232682,
+      "learning_rate": 1.5423203771321263e-05,
+      "loss": 0.814,
+      "step": 14348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5065756056319988,
+      "learning_rate": 1.5422562669973367e-05,
+      "loss": 0.9478,
+      "step": 14349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7502774678616597,
+      "learning_rate": 1.5421921537053596e-05,
+      "loss": 0.7641,
+      "step": 14350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.089677310422773,
+      "learning_rate": 1.5421280372565683e-05,
+      "loss": 0.8142,
+      "step": 14351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.502964286958993,
+      "learning_rate": 1.5420639176513358e-05,
+      "loss": 0.8488,
+      "step": 14352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7121493078927084,
+      "learning_rate": 1.5419997948900358e-05,
+      "loss": 1.0568,
+      "step": 14353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7243166021219845,
+      "learning_rate": 1.5419356689730408e-05,
+      "loss": 0.9182,
+      "step": 14354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.613225814599189,
+      "learning_rate": 1.5418715399007253e-05,
+      "loss": 0.7831,
+      "step": 14355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7650133185308903,
+      "learning_rate": 1.541807407673462e-05,
+      "loss": 0.7191,
+      "step": 14356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.347708934410573,
+      "learning_rate": 1.5417432722916244e-05,
+      "loss": 0.9648,
+      "step": 14357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6198942582999007,
+      "learning_rate": 1.541679133755586e-05,
+      "loss": 0.8584,
+      "step": 14358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.580562883440322,
+      "learning_rate": 1.54161499206572e-05,
+      "loss": 0.8228,
+      "step": 14359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.034331854641393,
+      "learning_rate": 1.5415508472224002e-05,
+      "loss": 0.8286,
+      "step": 14360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5578381605118967,
+      "learning_rate": 1.541486699226e-05,
+      "loss": 1.0854,
+      "step": 14361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5728819441025268,
+      "learning_rate": 1.541422548076893e-05,
+      "loss": 0.9471,
+      "step": 14362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7012422214304568,
+      "learning_rate": 1.5413583937754517e-05,
+      "loss": 0.6709,
+      "step": 14363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4964904030571662,
+      "learning_rate": 1.5412942363220512e-05,
+      "loss": 0.9186,
+      "step": 14364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6172167924301866,
+      "learning_rate": 1.541230075717064e-05,
+      "loss": 0.8522,
+      "step": 14365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0301866385523586,
+      "learning_rate": 1.5411659119608643e-05,
+      "loss": 0.7901,
+      "step": 14366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7399514863246843,
+      "learning_rate": 1.541101745053825e-05,
+      "loss": 0.9348,
+      "step": 14367
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6419577878677178,
+      "learning_rate": 1.5410375749963197e-05,
+      "loss": 0.9238,
+      "step": 14368
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.554557911611679,
+      "learning_rate": 1.540973401788723e-05,
+      "loss": 0.8983,
+      "step": 14369
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7336170865225995,
+      "learning_rate": 1.5409092254314076e-05,
+      "loss": 0.8875,
+      "step": 14370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7252941267207287,
+      "learning_rate": 1.5408450459247474e-05,
+      "loss": 0.9027,
+      "step": 14371
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7896718078283955,
+      "learning_rate": 1.5407808632691162e-05,
+      "loss": 0.9965,
+      "step": 14372
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.353679880246221,
+      "learning_rate": 1.5407166774648877e-05,
+      "loss": 0.833,
+      "step": 14373
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.739812427202209,
+      "learning_rate": 1.5406524885124352e-05,
+      "loss": 0.9581,
+      "step": 14374
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.694413958939424,
+      "learning_rate": 1.540588296412133e-05,
+      "loss": 0.996,
+      "step": 14375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7986888614463208,
+      "learning_rate": 1.5405241011643544e-05,
+      "loss": 0.9321,
+      "step": 14376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.67246132347466,
+      "learning_rate": 1.5404599027694734e-05,
+      "loss": 0.8912,
+      "step": 14377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7615995445906203,
+      "learning_rate": 1.540395701227864e-05,
+      "loss": 0.791,
+      "step": 14378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7698807737946531,
+      "learning_rate": 1.5403314965398995e-05,
+      "loss": 0.7678,
+      "step": 14379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9109761245471069,
+      "learning_rate": 1.540267288705954e-05,
+      "loss": 0.9982,
+      "step": 14380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4642640196869148,
+      "learning_rate": 1.5402030777264014e-05,
+      "loss": 0.9099,
+      "step": 14381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1638265377651082,
+      "learning_rate": 1.5401388636016154e-05,
+      "loss": 0.9231,
+      "step": 14382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.036548617270788,
+      "learning_rate": 1.54007464633197e-05,
+      "loss": 0.9667,
+      "step": 14383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8229889034964548,
+      "learning_rate": 1.5400104259178386e-05,
+      "loss": 0.7968,
+      "step": 14384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.653115823268349,
+      "learning_rate": 1.539946202359596e-05,
+      "loss": 0.9428,
+      "step": 14385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5536695104832912,
+      "learning_rate": 1.5398819756576155e-05,
+      "loss": 0.9602,
+      "step": 14386
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7682924122271757,
+      "learning_rate": 1.539817745812271e-05,
+      "loss": 0.7058,
+      "step": 14387
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8480895501881917,
+      "learning_rate": 1.539753512823937e-05,
+      "loss": 0.7131,
+      "step": 14388
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5272413198324013,
+      "learning_rate": 1.539689276692987e-05,
+      "loss": 0.8559,
+      "step": 14389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5686971642039922,
+      "learning_rate": 1.5396250374197953e-05,
+      "loss": 0.6989,
+      "step": 14390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7875307911080844,
+      "learning_rate": 1.5395607950047357e-05,
+      "loss": 0.9622,
+      "step": 14391
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9192192649970425,
+      "learning_rate": 1.5394965494481824e-05,
+      "loss": 0.9649,
+      "step": 14392
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8255205070726579,
+      "learning_rate": 1.5394323007505094e-05,
+      "loss": 0.7431,
+      "step": 14393
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2076794330660703,
+      "learning_rate": 1.5393680489120907e-05,
+      "loss": 0.8551,
+      "step": 14394
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7685397810820328,
+      "learning_rate": 1.5393037939333004e-05,
+      "loss": 0.743,
+      "step": 14395
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.592760010574413,
+      "learning_rate": 1.5392395358145125e-05,
+      "loss": 0.9345,
+      "step": 14396
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.52275802946615,
+      "learning_rate": 1.5391752745561022e-05,
+      "loss": 0.8496,
+      "step": 14397
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.639127630611902,
+      "learning_rate": 1.539111010158442e-05,
+      "loss": 0.8627,
+      "step": 14398
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0240682400803345,
+      "learning_rate": 1.539046742621907e-05,
+      "loss": 0.8917,
+      "step": 14399
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8655385276832814,
+      "learning_rate": 1.538982471946871e-05,
+      "loss": 0.9381,
+      "step": 14400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7812429995893719,
+      "learning_rate": 1.5389181981337085e-05,
+      "loss": 0.6833,
+      "step": 14401
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8443374952231486,
+      "learning_rate": 1.5388539211827934e-05,
+      "loss": 0.9351,
+      "step": 14402
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7391153637528335,
+      "learning_rate": 1.538789641094501e-05,
+      "loss": 0.7137,
+      "step": 14403
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4897380738267545,
+      "learning_rate": 1.538725357869204e-05,
+      "loss": 0.9234,
+      "step": 14404
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.488069226221125,
+      "learning_rate": 1.5386610715072773e-05,
+      "loss": 0.9642,
+      "step": 14405
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8045488071766316,
+      "learning_rate": 1.5385967820090958e-05,
+      "loss": 0.8165,
+      "step": 14406
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9538709471148765,
+      "learning_rate": 1.538532489375033e-05,
+      "loss": 0.7478,
+      "step": 14407
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.853132279318537,
+      "learning_rate": 1.5384681936054636e-05,
+      "loss": 0.9346,
+      "step": 14408
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2459229188462313,
+      "learning_rate": 1.5384038947007617e-05,
+      "loss": 0.9263,
+      "step": 14409
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7353440847843529,
+      "learning_rate": 1.5383395926613026e-05,
+      "loss": 0.8511,
+      "step": 14410
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6006571136458831,
+      "learning_rate": 1.5382752874874595e-05,
+      "loss": 0.9744,
+      "step": 14411
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7121482272859054,
+      "learning_rate": 1.5382109791796073e-05,
+      "loss": 0.8838,
+      "step": 14412
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6542457628580594,
+      "learning_rate": 1.53814666773812e-05,
+      "loss": 0.9309,
+      "step": 14413
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.116812483629149,
+      "learning_rate": 1.5380823531633727e-05,
+      "loss": 0.8542,
+      "step": 14414
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.895598673826127,
+      "learning_rate": 1.5380180354557398e-05,
+      "loss": 0.8693,
+      "step": 14415
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5611842252523627,
+      "learning_rate": 1.5379537146155954e-05,
+      "loss": 0.9199,
+      "step": 14416
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4144310629978967,
+      "learning_rate": 1.5378893906433142e-05,
+      "loss": 0.9368,
+      "step": 14417
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4478603445532217,
+      "learning_rate": 1.5378250635392707e-05,
+      "loss": 0.9142,
+      "step": 14418
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.976103891507378,
+      "learning_rate": 1.5377607333038392e-05,
+      "loss": 0.9228,
+      "step": 14419
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6053993706884626,
+      "learning_rate": 1.5376963999373945e-05,
+      "loss": 0.8934,
+      "step": 14420
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8769602416942612,
+      "learning_rate": 1.5376320634403112e-05,
+      "loss": 0.9579,
+      "step": 14421
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6794169087932698,
+      "learning_rate": 1.537567723812964e-05,
+      "loss": 0.885,
+      "step": 14422
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7944279942457584,
+      "learning_rate": 1.537503381055727e-05,
+      "loss": 0.8751,
+      "step": 14423
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6296754091208099,
+      "learning_rate": 1.5374390351689752e-05,
+      "loss": 0.7918,
+      "step": 14424
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7441260390550164,
+      "learning_rate": 1.5373746861530835e-05,
+      "loss": 0.9196,
+      "step": 14425
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7883227203412089,
+      "learning_rate": 1.5373103340084262e-05,
+      "loss": 0.756,
+      "step": 14426
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6111029923006066,
+      "learning_rate": 1.537245978735378e-05,
+      "loss": 0.9681,
+      "step": 14427
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6651043073324727,
+      "learning_rate": 1.5371816203343132e-05,
+      "loss": 0.9273,
+      "step": 14428
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6652563151923785,
+      "learning_rate": 1.5371172588056076e-05,
+      "loss": 0.931,
+      "step": 14429
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8109684188915336,
+      "learning_rate": 1.5370528941496346e-05,
+      "loss": 0.7297,
+      "step": 14430
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1442426641406147,
+      "learning_rate": 1.53698852636677e-05,
+      "loss": 0.7499,
+      "step": 14431
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7080543578792309,
+      "learning_rate": 1.536924155457388e-05,
+      "loss": 0.9086,
+      "step": 14432
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7922466509944406,
+      "learning_rate": 1.536859781421864e-05,
+      "loss": 0.7324,
+      "step": 14433
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6730645904462025,
+      "learning_rate": 1.536795404260572e-05,
+      "loss": 0.8393,
+      "step": 14434
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7418785906361067,
+      "learning_rate": 1.536731023973887e-05,
+      "loss": 0.865,
+      "step": 14435
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0670549832885596,
+      "learning_rate": 1.5366666405621846e-05,
+      "loss": 0.9238,
+      "step": 14436
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8206601893498062,
+      "learning_rate": 1.5366022540258388e-05,
+      "loss": 0.8165,
+      "step": 14437
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7428852202380323,
+      "learning_rate": 1.536537864365225e-05,
+      "loss": 0.9068,
+      "step": 14438
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1195185913927275,
+      "learning_rate": 1.5364734715807176e-05,
+      "loss": 0.9163,
+      "step": 14439
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8060362620797773,
+      "learning_rate": 1.536409075672692e-05,
+      "loss": 0.7724,
+      "step": 14440
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7575422378757948,
+      "learning_rate": 1.5363446766415226e-05,
+      "loss": 0.7432,
+      "step": 14441
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.021963363244152,
+      "learning_rate": 1.536280274487585e-05,
+      "loss": 0.9267,
+      "step": 14442
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.550191092528743,
+      "learning_rate": 1.536215869211254e-05,
+      "loss": 0.9137,
+      "step": 14443
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7276985295570988,
+      "learning_rate": 1.5361514608129043e-05,
+      "loss": 0.9035,
+      "step": 14444
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.758436520184717,
+      "learning_rate": 1.536087049292911e-05,
+      "loss": 0.8306,
+      "step": 14445
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9285371208335444,
+      "learning_rate": 1.5360226346516493e-05,
+      "loss": 0.9542,
+      "step": 14446
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.858184676998969,
+      "learning_rate": 1.535958216889494e-05,
+      "loss": 0.8981,
+      "step": 14447
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6701425919244541,
+      "learning_rate": 1.5358937960068204e-05,
+      "loss": 0.8647,
+      "step": 14448
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7032900880130017,
+      "learning_rate": 1.535829372004003e-05,
+      "loss": 0.86,
+      "step": 14449
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.651003988584073,
+      "learning_rate": 1.5357649448814177e-05,
+      "loss": 0.845,
+      "step": 14450
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6926478835882979,
+      "learning_rate": 1.5357005146394394e-05,
+      "loss": 0.9201,
+      "step": 14451
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8617159634926523,
+      "learning_rate": 1.535636081278443e-05,
+      "loss": 0.8135,
+      "step": 14452
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.413208352949969,
+      "learning_rate": 1.5355716447988038e-05,
+      "loss": 0.9011,
+      "step": 14453
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.613711500549203,
+      "learning_rate": 1.535507205200897e-05,
+      "loss": 0.9002,
+      "step": 14454
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4624011084100377,
+      "learning_rate": 1.5354427624850978e-05,
+      "loss": 0.8815,
+      "step": 14455
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.530837633107697,
+      "learning_rate": 1.5353783166517806e-05,
+      "loss": 0.953,
+      "step": 14456
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6173541513270595,
+      "learning_rate": 1.535313867701322e-05,
+      "loss": 0.7692,
+      "step": 14457
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6170505883103157,
+      "learning_rate": 1.5352494156340963e-05,
+      "loss": 1.0363,
+      "step": 14458
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5051310333643706,
+      "learning_rate": 1.5351849604504794e-05,
+      "loss": 0.8557,
+      "step": 14459
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4941313100505758,
+      "learning_rate": 1.5351205021508457e-05,
+      "loss": 0.9656,
+      "step": 14460
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.627196815741778,
+      "learning_rate": 1.5350560407355714e-05,
+      "loss": 0.9933,
+      "step": 14461
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6745516030781427,
+      "learning_rate": 1.5349915762050316e-05,
+      "loss": 0.9928,
+      "step": 14462
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8284667860730408,
+      "learning_rate": 1.5349271085596013e-05,
+      "loss": 0.9062,
+      "step": 14463
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.757845682023038,
+      "learning_rate": 1.534862637799656e-05,
+      "loss": 0.8972,
+      "step": 14464
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5743526460292614,
+      "learning_rate": 1.534798163925571e-05,
+      "loss": 1.0232,
+      "step": 14465
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0613898098755423,
+      "learning_rate": 1.5347336869377215e-05,
+      "loss": 0.8431,
+      "step": 14466
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8204550384476366,
+      "learning_rate": 1.5346692068364834e-05,
+      "loss": 0.7565,
+      "step": 14467
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7967944359891608,
+      "learning_rate": 1.534604723622232e-05,
+      "loss": 0.9419,
+      "step": 14468
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8619710419766302,
+      "learning_rate": 1.5345402372953427e-05,
+      "loss": 0.9222,
+      "step": 14469
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8061984824718311,
+      "learning_rate": 1.534475747856191e-05,
+      "loss": 0.7204,
+      "step": 14470
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.815518313563905,
+      "learning_rate": 1.534411255305152e-05,
+      "loss": 0.8919,
+      "step": 14471
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.892328226577549,
+      "learning_rate": 1.5343467596426015e-05,
+      "loss": 0.9372,
+      "step": 14472
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.769618412771343,
+      "learning_rate": 1.5342822608689153e-05,
+      "loss": 0.9571,
+      "step": 14473
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8053423040778753,
+      "learning_rate": 1.5342177589844682e-05,
+      "loss": 0.8937,
+      "step": 14474
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8682111038064766,
+      "learning_rate": 1.5341532539896364e-05,
+      "loss": 0.9105,
+      "step": 14475
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7509701053863576,
+      "learning_rate": 1.5340887458847954e-05,
+      "loss": 0.7263,
+      "step": 14476
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.794163506540931,
+      "learning_rate": 1.5340242346703204e-05,
+      "loss": 0.7743,
+      "step": 14477
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7590386263699553,
+      "learning_rate": 1.5339597203465876e-05,
+      "loss": 0.9011,
+      "step": 14478
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7314459045290003,
+      "learning_rate": 1.5338952029139718e-05,
+      "loss": 0.8784,
+      "step": 14479
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6066725873764758,
+      "learning_rate": 1.5338306823728494e-05,
+      "loss": 0.8267,
+      "step": 14480
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5830789058971868,
+      "learning_rate": 1.5337661587235952e-05,
+      "loss": 0.8482,
+      "step": 14481
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5736348073597408,
+      "learning_rate": 1.533701631966586e-05,
+      "loss": 0.8763,
+      "step": 14482
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7781439403853874,
+      "learning_rate": 1.533637102102197e-05,
+      "loss": 0.9033,
+      "step": 14483
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.645447349993965,
+      "learning_rate": 1.5335725691308035e-05,
+      "loss": 0.9115,
+      "step": 14484
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6051455185809307,
+      "learning_rate": 1.5335080330527818e-05,
+      "loss": 0.8443,
+      "step": 14485
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5645321183771381,
+      "learning_rate": 1.5334434938685073e-05,
+      "loss": 0.9234,
+      "step": 14486
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6167355495240143,
+      "learning_rate": 1.5333789515783558e-05,
+      "loss": 0.8483,
+      "step": 14487
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0739219096113466,
+      "learning_rate": 1.5333144061827033e-05,
+      "loss": 0.8946,
+      "step": 14488
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7964608427979785,
+      "learning_rate": 1.5332498576819256e-05,
+      "loss": 0.8864,
+      "step": 14489
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.744685747002442,
+      "learning_rate": 1.5331853060763985e-05,
+      "loss": 0.9413,
+      "step": 14490
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5746880061087063,
+      "learning_rate": 1.5331207513664973e-05,
+      "loss": 0.9097,
+      "step": 14491
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5946871436143222,
+      "learning_rate": 1.5330561935525988e-05,
+      "loss": 0.9572,
+      "step": 14492
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6976081625922015,
+      "learning_rate": 1.532991632635078e-05,
+      "loss": 1.0216,
+      "step": 14493
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5889086402861863,
+      "learning_rate": 1.5329270686143112e-05,
+      "loss": 0.9429,
+      "step": 14494
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8169876884611982,
+      "learning_rate": 1.5328625014906746e-05,
+      "loss": 0.7569,
+      "step": 14495
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6115567978519882,
+      "learning_rate": 1.5327979312645433e-05,
+      "loss": 1.0085,
+      "step": 14496
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.635690178329006,
+      "learning_rate": 1.5327333579362945e-05,
+      "loss": 0.868,
+      "step": 14497
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6018339853237697,
+      "learning_rate": 1.5326687815063025e-05,
+      "loss": 0.8785,
+      "step": 14498
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4611830791826426,
+      "learning_rate": 1.532604201974945e-05,
+      "loss": 0.8775,
+      "step": 14499
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.587798239145319,
+      "learning_rate": 1.5325396193425967e-05,
+      "loss": 0.8915,
+      "step": 14500
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8259393924185796,
+      "learning_rate": 1.5324750336096347e-05,
+      "loss": 0.7475,
+      "step": 14501
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7847727219837816,
+      "learning_rate": 1.5324104447764342e-05,
+      "loss": 0.7473,
+      "step": 14502
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.699500516040813,
+      "learning_rate": 1.5323458528433713e-05,
+      "loss": 0.9747,
+      "step": 14503
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5945124696704847,
+      "learning_rate": 1.5322812578108224e-05,
+      "loss": 0.8324,
+      "step": 14504
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8193988599702209,
+      "learning_rate": 1.5322166596791636e-05,
+      "loss": 0.7269,
+      "step": 14505
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7619028163620569,
+      "learning_rate": 1.532152058448771e-05,
+      "loss": 0.6689,
+      "step": 14506
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.111815059529134,
+      "learning_rate": 1.5320874541200202e-05,
+      "loss": 0.9008,
+      "step": 14507
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7783953369302103,
+      "learning_rate": 1.5320228466932885e-05,
+      "loss": 0.8043,
+      "step": 14508
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7125198610076537,
+      "learning_rate": 1.5319582361689508e-05,
+      "loss": 1.0337,
+      "step": 14509
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6911887563219945,
+      "learning_rate": 1.531893622547384e-05,
+      "loss": 0.9165,
+      "step": 14510
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5716556854062327,
+      "learning_rate": 1.531829005828964e-05,
+      "loss": 1.0353,
+      "step": 14511
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.580839537093289,
+      "learning_rate": 1.5317643860140675e-05,
+      "loss": 0.9944,
+      "step": 14512
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7687067990223662,
+      "learning_rate": 1.53169976310307e-05,
+      "loss": 0.9289,
+      "step": 14513
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7227138207386412,
+      "learning_rate": 1.5316351370963484e-05,
+      "loss": 0.8806,
+      "step": 14514
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5043699495403708,
+      "learning_rate": 1.5315705079942787e-05,
+      "loss": 0.9163,
+      "step": 14515
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.978331455639636,
+      "learning_rate": 1.5315058757972373e-05,
+      "loss": 1.0126,
+      "step": 14516
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.509506444340939,
+      "learning_rate": 1.5314412405056e-05,
+      "loss": 0.8425,
+      "step": 14517
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8069273057830895,
+      "learning_rate": 1.531376602119744e-05,
+      "loss": 0.936,
+      "step": 14518
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0778238261985242,
+      "learning_rate": 1.531311960640045e-05,
+      "loss": 0.9776,
+      "step": 14519
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6037263150660255,
+      "learning_rate": 1.5312473160668796e-05,
+      "loss": 1.0012,
+      "step": 14520
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7630646760305284,
+      "learning_rate": 1.5311826684006243e-05,
+      "loss": 0.8944,
+      "step": 14521
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.465638159851386,
+      "learning_rate": 1.5311180176416548e-05,
+      "loss": 0.9192,
+      "step": 14522
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6268699930007422,
+      "learning_rate": 1.5310533637903483e-05,
+      "loss": 0.9938,
+      "step": 14523
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6355455874857825,
+      "learning_rate": 1.530988706847081e-05,
+      "loss": 0.9103,
+      "step": 14524
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9647111780135418,
+      "learning_rate": 1.5309240468122294e-05,
+      "loss": 0.9052,
+      "step": 14525
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4553189542187481,
+      "learning_rate": 1.53085938368617e-05,
+      "loss": 0.9163,
+      "step": 14526
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.863739761107783,
+      "learning_rate": 1.5307947174692795e-05,
+      "loss": 0.8102,
+      "step": 14527
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6120606050624198,
+      "learning_rate": 1.5307300481619334e-05,
+      "loss": 0.8077,
+      "step": 14528
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5471666752432656,
+      "learning_rate": 1.5306653757645095e-05,
+      "loss": 0.8928,
+      "step": 14529
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5601148017830113,
+      "learning_rate": 1.5306007002773834e-05,
+      "loss": 0.9008,
+      "step": 14530
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8104658750447524,
+      "learning_rate": 1.5305360217009324e-05,
+      "loss": 0.7013,
+      "step": 14531
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.221168301985867,
+      "learning_rate": 1.530471340035532e-05,
+      "loss": 0.8689,
+      "step": 14532
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6886595873784942,
+      "learning_rate": 1.5304066552815604e-05,
+      "loss": 0.9216,
+      "step": 14533
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6538950522420268,
+      "learning_rate": 1.530341967439393e-05,
+      "loss": 0.7743,
+      "step": 14534
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.76644486848168,
+      "learning_rate": 1.5302772765094065e-05,
+      "loss": 0.8743,
+      "step": 14535
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9836013720707473,
+      "learning_rate": 1.530212582491978e-05,
+      "loss": 0.9833,
+      "step": 14536
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8057125867661978,
+      "learning_rate": 1.5301478853874838e-05,
+      "loss": 0.9603,
+      "step": 14537
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.693187389197692,
+      "learning_rate": 1.5300831851963013e-05,
+      "loss": 0.8193,
+      "step": 14538
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6393421947954832,
+      "learning_rate": 1.530018481918806e-05,
+      "loss": 0.9106,
+      "step": 14539
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8021690768977877,
+      "learning_rate": 1.529953775555376e-05,
+      "loss": 0.7759,
+      "step": 14540
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7546927092737185,
+      "learning_rate": 1.529889066106387e-05,
+      "loss": 0.7174,
+      "step": 14541
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6331437833024867,
+      "learning_rate": 1.5298243535722162e-05,
+      "loss": 0.8631,
+      "step": 14542
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5494591595977727,
+      "learning_rate": 1.52975963795324e-05,
+      "loss": 0.8355,
+      "step": 14543
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.854637376220249,
+      "learning_rate": 1.5296949192498358e-05,
+      "loss": 0.8435,
+      "step": 14544
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4736489321700448,
+      "learning_rate": 1.5296301974623794e-05,
+      "loss": 0.8563,
+      "step": 14545
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6214946935915984,
+      "learning_rate": 1.529565472591249e-05,
+      "loss": 0.9143,
+      "step": 14546
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5714900291756237,
+      "learning_rate": 1.5295007446368207e-05,
+      "loss": 0.8582,
+      "step": 14547
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9728529649677018,
+      "learning_rate": 1.5294360135994712e-05,
+      "loss": 0.9233,
+      "step": 14548
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4445077783697668,
+      "learning_rate": 1.5293712794795777e-05,
+      "loss": 1.02,
+      "step": 14549
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5051549595642706,
+      "learning_rate": 1.529306542277517e-05,
+      "loss": 0.8007,
+      "step": 14550
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6857196722844934,
+      "learning_rate": 1.529241801993666e-05,
+      "loss": 0.8755,
+      "step": 14551
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6571820011418092,
+      "learning_rate": 1.5291770586284018e-05,
+      "loss": 0.9782,
+      "step": 14552
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.904761941940558,
+      "learning_rate": 1.5291123121821013e-05,
+      "loss": 0.7833,
+      "step": 14553
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6900269273137705,
+      "learning_rate": 1.529047562655141e-05,
+      "loss": 0.8834,
+      "step": 14554
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7565937448150142,
+      "learning_rate": 1.528982810047899e-05,
+      "loss": 0.8748,
+      "step": 14555
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7848249515327772,
+      "learning_rate": 1.528918054360751e-05,
+      "loss": 0.7286,
+      "step": 14556
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4365994289641795,
+      "learning_rate": 1.528853295594075e-05,
+      "loss": 0.8288,
+      "step": 14557
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.595629366888284,
+      "learning_rate": 1.5287885337482472e-05,
+      "loss": 0.939,
+      "step": 14558
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.557996250758025,
+      "learning_rate": 1.5287237688236455e-05,
+      "loss": 1.0039,
+      "step": 14559
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.677966771222756,
+      "learning_rate": 1.5286590008206465e-05,
+      "loss": 0.9084,
+      "step": 14560
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7504632234966171,
+      "learning_rate": 1.5285942297396277e-05,
+      "loss": 0.7272,
+      "step": 14561
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9152012607509559,
+      "learning_rate": 1.5285294555809658e-05,
+      "loss": 0.8079,
+      "step": 14562
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.53684299580162,
+      "learning_rate": 1.528464678345038e-05,
+      "loss": 0.7985,
+      "step": 14563
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6536374607349027,
+      "learning_rate": 1.5283998980322214e-05,
+      "loss": 0.9024,
+      "step": 14564
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5476658406585098,
+      "learning_rate": 1.5283351146428933e-05,
+      "loss": 0.9201,
+      "step": 14565
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6358789840653396,
+      "learning_rate": 1.5282703281774312e-05,
+      "loss": 0.8917,
+      "step": 14566
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.077178562413389,
+      "learning_rate": 1.5282055386362116e-05,
+      "loss": 0.9927,
+      "step": 14567
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5837355646607352,
+      "learning_rate": 1.5281407460196126e-05,
+      "loss": 0.8162,
+      "step": 14568
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.224287621937967,
+      "learning_rate": 1.528075950328011e-05,
+      "loss": 0.7782,
+      "step": 14569
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.668506746399886,
+      "learning_rate": 1.5280111515617835e-05,
+      "loss": 0.9258,
+      "step": 14570
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0100121531975637,
+      "learning_rate": 1.527946349721308e-05,
+      "loss": 0.8998,
+      "step": 14571
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.753900069752312,
+      "learning_rate": 1.5278815448069622e-05,
+      "loss": 0.8806,
+      "step": 14572
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4700289565866422,
+      "learning_rate": 1.5278167368191224e-05,
+      "loss": 0.9406,
+      "step": 14573
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6892375573888314,
+      "learning_rate": 1.5277519257581668e-05,
+      "loss": 0.9258,
+      "step": 14574
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4927150241047051,
+      "learning_rate": 1.527687111624472e-05,
+      "loss": 0.9843,
+      "step": 14575
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7598469486971573,
+      "learning_rate": 1.527622294418416e-05,
+      "loss": 0.7501,
+      "step": 14576
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5555431619966213,
+      "learning_rate": 1.5275574741403763e-05,
+      "loss": 0.8514,
+      "step": 14577
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6759073796231105,
+      "learning_rate": 1.52749265079073e-05,
+      "loss": 0.9207,
+      "step": 14578
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6254947272763334,
+      "learning_rate": 1.527427824369854e-05,
+      "loss": 0.8732,
+      "step": 14579
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5446485154250935,
+      "learning_rate": 1.5273629948781264e-05,
+      "loss": 1.0133,
+      "step": 14580
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3608458868519424,
+      "learning_rate": 1.527298162315924e-05,
+      "loss": 0.9088,
+      "step": 14581
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5807576837685249,
+      "learning_rate": 1.5272333266836256e-05,
+      "loss": 0.8594,
+      "step": 14582
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6281106126014533,
+      "learning_rate": 1.5271684879816074e-05,
+      "loss": 1.1067,
+      "step": 14583
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4891016086362865,
+      "learning_rate": 1.5271036462102472e-05,
+      "loss": 0.8875,
+      "step": 14584
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5908046615568636,
+      "learning_rate": 1.5270388013699232e-05,
+      "loss": 0.8859,
+      "step": 14585
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6454769635174078,
+      "learning_rate": 1.526973953461012e-05,
+      "loss": 0.9319,
+      "step": 14586
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7553160406938113,
+      "learning_rate": 1.526909102483892e-05,
+      "loss": 0.8972,
+      "step": 14587
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.443433933060828,
+      "learning_rate": 1.52684424843894e-05,
+      "loss": 0.9504,
+      "step": 14588
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.69035232730275,
+      "learning_rate": 1.5267793913265344e-05,
+      "loss": 0.9012,
+      "step": 14589
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7546033702937391,
+      "learning_rate": 1.526714531147052e-05,
+      "loss": 0.9409,
+      "step": 14590
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7576683839689344,
+      "learning_rate": 1.526649667900871e-05,
+      "loss": 0.8503,
+      "step": 14591
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6072877853548502,
+      "learning_rate": 1.5265848015883687e-05,
+      "loss": 0.8624,
+      "step": 14592
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6944994024642992,
+      "learning_rate": 1.5265199322099235e-05,
+      "loss": 0.8946,
+      "step": 14593
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7816249061856244,
+      "learning_rate": 1.5264550597659122e-05,
+      "loss": 0.7355,
+      "step": 14594
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5380059253226266,
+      "learning_rate": 1.526390184256713e-05,
+      "loss": 0.9897,
+      "step": 14595
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7276701868626986,
+      "learning_rate": 1.5263253056827032e-05,
+      "loss": 0.694,
+      "step": 14596
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.48675163915659,
+      "learning_rate": 1.526260424044261e-05,
+      "loss": 0.995,
+      "step": 14597
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6982200527133364,
+      "learning_rate": 1.5261955393417643e-05,
+      "loss": 0.8799,
+      "step": 14598
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8048002345360032,
+      "learning_rate": 1.5261306515755904e-05,
+      "loss": 0.9398,
+      "step": 14599
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9555104828099803,
+      "learning_rate": 1.5260657607461167e-05,
+      "loss": 0.9992,
+      "step": 14600
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4645383806619996,
+      "learning_rate": 1.5260008668537225e-05,
+      "loss": 0.8992,
+      "step": 14601
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0433544324027695,
+      "learning_rate": 1.525935969898784e-05,
+      "loss": 0.7785,
+      "step": 14602
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5914629014135435,
+      "learning_rate": 1.52587106988168e-05,
+      "loss": 0.8471,
+      "step": 14603
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7350796936934623,
+      "learning_rate": 1.5258061668027884e-05,
+      "loss": 0.8495,
+      "step": 14604
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5707724604878734,
+      "learning_rate": 1.5257412606624868e-05,
+      "loss": 0.9282,
+      "step": 14605
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7259616095265182,
+      "learning_rate": 1.525676351461153e-05,
+      "loss": 0.9025,
+      "step": 14606
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7400577715915782,
+      "learning_rate": 1.525611439199165e-05,
+      "loss": 0.8648,
+      "step": 14607
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6960292873681937,
+      "learning_rate": 1.5255465238769009e-05,
+      "loss": 0.8099,
+      "step": 14608
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9079995689786076,
+      "learning_rate": 1.5254816054947385e-05,
+      "loss": 0.9375,
+      "step": 14609
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.632322045880217,
+      "learning_rate": 1.525416684053056e-05,
+      "loss": 0.7621,
+      "step": 14610
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.146209373619463,
+      "learning_rate": 1.5253517595522308e-05,
+      "loss": 1.0284,
+      "step": 14611
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6648516898078325,
+      "learning_rate": 1.5252868319926418e-05,
+      "loss": 0.8348,
+      "step": 14612
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6869960415169678,
+      "learning_rate": 1.5252219013746663e-05,
+      "loss": 0.9185,
+      "step": 14613
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5617680104799077,
+      "learning_rate": 1.5251569676986826e-05,
+      "loss": 0.9186,
+      "step": 14614
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8296778968120309,
+      "learning_rate": 1.5250920309650688e-05,
+      "loss": 0.9278,
+      "step": 14615
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.85333739180657,
+      "learning_rate": 1.5250270911742031e-05,
+      "loss": 0.9254,
+      "step": 14616
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4313164333788766,
+      "learning_rate": 1.5249621483264632e-05,
+      "loss": 0.8999,
+      "step": 14617
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6623256389230467,
+      "learning_rate": 1.5248972024222277e-05,
+      "loss": 1.0203,
+      "step": 14618
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4730242164653833,
+      "learning_rate": 1.5248322534618742e-05,
+      "loss": 0.8255,
+      "step": 14619
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6857863880711323,
+      "learning_rate": 1.5247673014457815e-05,
+      "loss": 0.778,
+      "step": 14620
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8680276267390443,
+      "learning_rate": 1.5247023463743274e-05,
+      "loss": 0.9217,
+      "step": 14621
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.587221089629546,
+      "learning_rate": 1.5246373882478899e-05,
+      "loss": 1.004,
+      "step": 14622
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.600814504762379,
+      "learning_rate": 1.5245724270668476e-05,
+      "loss": 1.0136,
+      "step": 14623
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5793878835590562,
+      "learning_rate": 1.5245074628315786e-05,
+      "loss": 0.8356,
+      "step": 14624
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0557728457294706,
+      "learning_rate": 1.5244424955424611e-05,
+      "loss": 0.9254,
+      "step": 14625
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5961316701923063,
+      "learning_rate": 1.5243775251998729e-05,
+      "loss": 0.9568,
+      "step": 14626
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2459574582768065,
+      "learning_rate": 1.5243125518041935e-05,
+      "loss": 0.9005,
+      "step": 14627
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.250859771208239,
+      "learning_rate": 1.5242475753558e-05,
+      "loss": 0.8589,
+      "step": 14628
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5710038502458952,
+      "learning_rate": 1.5241825958550712e-05,
+      "loss": 0.8632,
+      "step": 14629
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5929249438760427,
+      "learning_rate": 1.524117613302385e-05,
+      "loss": 0.8537,
+      "step": 14630
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3156353779397265,
+      "learning_rate": 1.5240526276981207e-05,
+      "loss": 0.938,
+      "step": 14631
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6085644910151378,
+      "learning_rate": 1.5239876390426558e-05,
+      "loss": 0.8334,
+      "step": 14632
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5754722649693813,
+      "learning_rate": 1.523922647336369e-05,
+      "loss": 0.897,
+      "step": 14633
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7843148797418129,
+      "learning_rate": 1.523857652579639e-05,
+      "loss": 0.802,
+      "step": 14634
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7611696507341426,
+      "learning_rate": 1.5237926547728432e-05,
+      "loss": 0.7246,
+      "step": 14635
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7098161435348354,
+      "learning_rate": 1.5237276539163613e-05,
+      "loss": 0.8931,
+      "step": 14636
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6243009579061267,
+      "learning_rate": 1.5236626500105706e-05,
+      "loss": 0.871,
+      "step": 14637
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7779811913785096,
+      "learning_rate": 1.523597643055851e-05,
+      "loss": 0.7346,
+      "step": 14638
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6841637873443729,
+      "learning_rate": 1.5235326330525794e-05,
+      "loss": 0.914,
+      "step": 14639
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6557845490171699,
+      "learning_rate": 1.5234676200011356e-05,
+      "loss": 0.8484,
+      "step": 14640
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8493888227135584,
+      "learning_rate": 1.5234026039018974e-05,
+      "loss": 0.9027,
+      "step": 14641
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8415954409271765,
+      "learning_rate": 1.5233375847552436e-05,
+      "loss": 0.7289,
+      "step": 14642
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7365115633714798,
+      "learning_rate": 1.5232725625615525e-05,
+      "loss": 0.9279,
+      "step": 14643
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.602039311351594,
+      "learning_rate": 1.523207537321203e-05,
+      "loss": 0.9137,
+      "step": 14644
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6782557025467442,
+      "learning_rate": 1.5231425090345737e-05,
+      "loss": 0.8807,
+      "step": 14645
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0867383063970557,
+      "learning_rate": 1.5230774777020428e-05,
+      "loss": 0.9227,
+      "step": 14646
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7046829267605697,
+      "learning_rate": 1.5230124433239894e-05,
+      "loss": 0.9675,
+      "step": 14647
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5765622020830794,
+      "learning_rate": 1.5229474059007921e-05,
+      "loss": 0.9609,
+      "step": 14648
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.789338786840464,
+      "learning_rate": 1.5228823654328292e-05,
+      "loss": 0.7552,
+      "step": 14649
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6169126403990193,
+      "learning_rate": 1.5228173219204795e-05,
+      "loss": 0.8392,
+      "step": 14650
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8525538750472668,
+      "learning_rate": 1.5227522753641222e-05,
+      "loss": 0.7579,
+      "step": 14651
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9766612916408575,
+      "learning_rate": 1.5226872257641354e-05,
+      "loss": 0.8907,
+      "step": 14652
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8084894909394986,
+      "learning_rate": 1.5226221731208984e-05,
+      "loss": 0.7455,
+      "step": 14653
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6983853245072036,
+      "learning_rate": 1.5225571174347893e-05,
+      "loss": 0.8667,
+      "step": 14654
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.627796998112716,
+      "learning_rate": 1.5224920587061873e-05,
+      "loss": 1.0157,
+      "step": 14655
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5961797606013326,
+      "learning_rate": 1.5224269969354713e-05,
+      "loss": 0.8797,
+      "step": 14656
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.710289729703263,
+      "learning_rate": 1.52236193212302e-05,
+      "loss": 0.9929,
+      "step": 14657
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4949362419211558,
+      "learning_rate": 1.5222968642692118e-05,
+      "loss": 0.9256,
+      "step": 14658
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6081309096412384,
+      "learning_rate": 1.5222317933744262e-05,
+      "loss": 0.9054,
+      "step": 14659
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5133238187485445,
+      "learning_rate": 1.5221667194390412e-05,
+      "loss": 1.0006,
+      "step": 14660
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6814563545173997,
+      "learning_rate": 1.522101642463437e-05,
+      "loss": 0.7936,
+      "step": 14661
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8191595551427826,
+      "learning_rate": 1.5220365624479914e-05,
+      "loss": 0.8607,
+      "step": 14662
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.67237723814577,
+      "learning_rate": 1.5219714793930838e-05,
+      "loss": 0.7819,
+      "step": 14663
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.722011427231119,
+      "learning_rate": 1.5219063932990927e-05,
+      "loss": 0.8819,
+      "step": 14664
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.819079429786089,
+      "learning_rate": 1.5218413041663976e-05,
+      "loss": 0.7739,
+      "step": 14665
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6994139236206924,
+      "learning_rate": 1.521776211995377e-05,
+      "loss": 0.9486,
+      "step": 14666
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8358623497501279,
+      "learning_rate": 1.5217111167864102e-05,
+      "loss": 0.8414,
+      "step": 14667
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.524218740151944,
+      "learning_rate": 1.5216460185398761e-05,
+      "loss": 0.9473,
+      "step": 14668
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.38944968372484,
+      "learning_rate": 1.521580917256154e-05,
+      "loss": 0.8276,
+      "step": 14669
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.957776517232428,
+      "learning_rate": 1.5215158129356222e-05,
+      "loss": 0.8137,
+      "step": 14670
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.642186970944182,
+      "learning_rate": 1.5214507055786605e-05,
+      "loss": 0.8696,
+      "step": 14671
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6849762125038386,
+      "learning_rate": 1.5213855951856475e-05,
+      "loss": 0.8066,
+      "step": 14672
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7326211136987567,
+      "learning_rate": 1.5213204817569627e-05,
+      "loss": 0.6998,
+      "step": 14673
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.691122301315896,
+      "learning_rate": 1.521255365292985e-05,
+      "loss": 0.942,
+      "step": 14674
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.497870308700998,
+      "learning_rate": 1.5211902457940934e-05,
+      "loss": 0.9155,
+      "step": 14675
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5979660458186709,
+      "learning_rate": 1.5211251232606674e-05,
+      "loss": 0.9051,
+      "step": 14676
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.842626213595112,
+      "learning_rate": 1.5210599976930857e-05,
+      "loss": 0.7976,
+      "step": 14677
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7576046951508474,
+      "learning_rate": 1.5209948690917282e-05,
+      "loss": 0.8094,
+      "step": 14678
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6792763098334955,
+      "learning_rate": 1.5209297374569732e-05,
+      "loss": 0.7508,
+      "step": 14679
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7547591938177822,
+      "learning_rate": 1.5208646027892005e-05,
+      "loss": 0.9874,
+      "step": 14680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5754447488313015,
+      "learning_rate": 1.520799465088789e-05,
+      "loss": 0.8249,
+      "step": 14681
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.58636835061238,
+      "learning_rate": 1.5207343243561182e-05,
+      "loss": 0.7418,
+      "step": 14682
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8455769060197889,
+      "learning_rate": 1.5206691805915674e-05,
+      "loss": 0.915,
+      "step": 14683
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7679020498727946,
+      "learning_rate": 1.5206040337955157e-05,
+      "loss": 0.9411,
+      "step": 14684
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5245483502028199,
+      "learning_rate": 1.5205388839683428e-05,
+      "loss": 0.9504,
+      "step": 14685
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6488945270279765,
+      "learning_rate": 1.5204737311104271e-05,
+      "loss": 0.9325,
+      "step": 14686
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7231331198889549,
+      "learning_rate": 1.5204085752221491e-05,
+      "loss": 0.7397,
+      "step": 14687
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8138546853098272,
+      "learning_rate": 1.5203434163038872e-05,
+      "loss": 0.8624,
+      "step": 14688
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5015739514087771,
+      "learning_rate": 1.5202782543560214e-05,
+      "loss": 0.9225,
+      "step": 14689
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5541886768004514,
+      "learning_rate": 1.520213089378931e-05,
+      "loss": 0.7971,
+      "step": 14690
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5823844146808688,
+      "learning_rate": 1.5201479213729954e-05,
+      "loss": 0.9082,
+      "step": 14691
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6454614550331617,
+      "learning_rate": 1.5200827503385935e-05,
+      "loss": 0.8205,
+      "step": 14692
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6066507901943803,
+      "learning_rate": 1.5200175762761055e-05,
+      "loss": 0.9138,
+      "step": 14693
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4347369601259603,
+      "learning_rate": 1.5199523991859107e-05,
+      "loss": 0.9067,
+      "step": 14694
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.928697352421878,
+      "learning_rate": 1.519887219068388e-05,
+      "loss": 0.8934,
+      "step": 14695
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.892572509099243,
+      "learning_rate": 1.5198220359239173e-05,
+      "loss": 0.8893,
+      "step": 14696
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5293783850151264,
+      "learning_rate": 1.5197568497528785e-05,
+      "loss": 0.9933,
+      "step": 14697
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6454131480856033,
+      "learning_rate": 1.5196916605556504e-05,
+      "loss": 0.9102,
+      "step": 14698
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9180514920113299,
+      "learning_rate": 1.5196264683326131e-05,
+      "loss": 0.9248,
+      "step": 14699
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.880529903070316,
+      "learning_rate": 1.519561273084146e-05,
+      "loss": 0.939,
+      "step": 14700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7715196985486679,
+      "learning_rate": 1.5194960748106287e-05,
+      "loss": 0.7251,
+      "step": 14701
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8839127604476846,
+      "learning_rate": 1.5194308735124406e-05,
+      "loss": 0.8778,
+      "step": 14702
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6087426214058307,
+      "learning_rate": 1.5193656691899618e-05,
+      "loss": 0.9208,
+      "step": 14703
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7441477492089243,
+      "learning_rate": 1.5193004618435713e-05,
+      "loss": 0.8784,
+      "step": 14704
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.36883245258511,
+      "learning_rate": 1.5192352514736495e-05,
+      "loss": 0.9162,
+      "step": 14705
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7990455471094635,
+      "learning_rate": 1.5191700380805754e-05,
+      "loss": 0.7736,
+      "step": 14706
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5785425300389102,
+      "learning_rate": 1.5191048216647291e-05,
+      "loss": 0.8899,
+      "step": 14707
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9404984639517642,
+      "learning_rate": 1.5190396022264903e-05,
+      "loss": 0.9124,
+      "step": 14708
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5662362143043058,
+      "learning_rate": 1.5189743797662382e-05,
+      "loss": 0.9508,
+      "step": 14709
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9535259632612174,
+      "learning_rate": 1.5189091542843533e-05,
+      "loss": 1.024,
+      "step": 14710
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.1508647345957375,
+      "learning_rate": 1.5188439257812148e-05,
+      "loss": 0.9293,
+      "step": 14711
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1015297227292313,
+      "learning_rate": 1.5187786942572032e-05,
+      "loss": 0.9099,
+      "step": 14712
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9194114618158393,
+      "learning_rate": 1.518713459712697e-05,
+      "loss": 0.9547,
+      "step": 14713
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8612013527363003,
+      "learning_rate": 1.5186482221480774e-05,
+      "loss": 0.7534,
+      "step": 14714
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8391852065548056,
+      "learning_rate": 1.5185829815637236e-05,
+      "loss": 0.9119,
+      "step": 14715
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9808182278114006,
+      "learning_rate": 1.5185177379600153e-05,
+      "loss": 1.001,
+      "step": 14716
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7240708600264076,
+      "learning_rate": 1.5184524913373326e-05,
+      "loss": 0.9577,
+      "step": 14717
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7651726049635701,
+      "learning_rate": 1.5183872416960554e-05,
+      "loss": 0.6925,
+      "step": 14718
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8132800204184774,
+      "learning_rate": 1.5183219890365638e-05,
+      "loss": 0.916,
+      "step": 14719
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6450155432538975,
+      "learning_rate": 1.5182567333592368e-05,
+      "loss": 0.8577,
+      "step": 14720
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7897894754770522,
+      "learning_rate": 1.5181914746644557e-05,
+      "loss": 0.7579,
+      "step": 14721
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4011475256728931,
+      "learning_rate": 1.5181262129525996e-05,
+      "loss": 0.993,
+      "step": 14722
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5343644577913234,
+      "learning_rate": 1.5180609482240485e-05,
+      "loss": 0.8649,
+      "step": 14723
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5637832667612528,
+      "learning_rate": 1.5179956804791826e-05,
+      "loss": 0.9452,
+      "step": 14724
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.130405299656281,
+      "learning_rate": 1.517930409718382e-05,
+      "loss": 0.9527,
+      "step": 14725
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5385365389409684,
+      "learning_rate": 1.5178651359420264e-05,
+      "loss": 1.0143,
+      "step": 14726
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0378639467603628,
+      "learning_rate": 1.5177998591504963e-05,
+      "loss": 0.846,
+      "step": 14727
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6770347796474852,
+      "learning_rate": 1.517734579344171e-05,
+      "loss": 0.8473,
+      "step": 14728
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7624730189597304,
+      "learning_rate": 1.5176692965234316e-05,
+      "loss": 0.9382,
+      "step": 14729
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5766667956470573,
+      "learning_rate": 1.5176040106886574e-05,
+      "loss": 0.8862,
+      "step": 14730
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2157787694331903,
+      "learning_rate": 1.5175387218402287e-05,
+      "loss": 0.8686,
+      "step": 14731
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5348475784202482,
+      "learning_rate": 1.5174734299785259e-05,
+      "loss": 0.871,
+      "step": 14732
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6374478173395743,
+      "learning_rate": 1.5174081351039287e-05,
+      "loss": 0.8263,
+      "step": 14733
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7083868901387305,
+      "learning_rate": 1.5173428372168177e-05,
+      "loss": 0.9552,
+      "step": 14734
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6968933265849393,
+      "learning_rate": 1.5172775363175732e-05,
+      "loss": 0.8519,
+      "step": 14735
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7849951158092526,
+      "learning_rate": 1.5172122324065748e-05,
+      "loss": 0.7216,
+      "step": 14736
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5067810094087426,
+      "learning_rate": 1.5171469254842031e-05,
+      "loss": 0.9861,
+      "step": 14737
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2296774013056435,
+      "learning_rate": 1.517081615550838e-05,
+      "loss": 0.9523,
+      "step": 14738
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4876154081829518,
+      "learning_rate": 1.5170163026068602e-05,
+      "loss": 0.9384,
+      "step": 14739
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.664154800531975,
+      "learning_rate": 1.5169509866526501e-05,
+      "loss": 0.7921,
+      "step": 14740
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5700579861861923,
+      "learning_rate": 1.5168856676885874e-05,
+      "loss": 1.004,
+      "step": 14741
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7168718809585772,
+      "learning_rate": 1.5168203457150527e-05,
+      "loss": 1.036,
+      "step": 14742
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7306826494746785,
+      "learning_rate": 1.5167550207324263e-05,
+      "loss": 0.8463,
+      "step": 14743
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8475869564423664,
+      "learning_rate": 1.516689692741089e-05,
+      "loss": 0.783,
+      "step": 14744
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6610989692091087,
+      "learning_rate": 1.5166243617414202e-05,
+      "loss": 0.8421,
+      "step": 14745
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9253421182064534,
+      "learning_rate": 1.5165590277338011e-05,
+      "loss": 0.9637,
+      "step": 14746
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5028439238645253,
+      "learning_rate": 1.5164936907186116e-05,
+      "loss": 0.8469,
+      "step": 14747
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.221850027704142,
+      "learning_rate": 1.5164283506962326e-05,
+      "loss": 0.8948,
+      "step": 14748
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5439259236395322,
+      "learning_rate": 1.516363007667044e-05,
+      "loss": 0.894,
+      "step": 14749
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5723917150879172,
+      "learning_rate": 1.5162976616314264e-05,
+      "loss": 0.9046,
+      "step": 14750
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6466559161328722,
+      "learning_rate": 1.5162323125897604e-05,
+      "loss": 0.9223,
+      "step": 14751
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.540466462532851,
+      "learning_rate": 1.5161669605424268e-05,
+      "loss": 0.8721,
+      "step": 14752
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6359700867971452,
+      "learning_rate": 1.5161016054898055e-05,
+      "loss": 0.7729,
+      "step": 14753
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6141417487089762,
+      "learning_rate": 1.5160362474322771e-05,
+      "loss": 1.0134,
+      "step": 14754
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.059646572984295,
+      "learning_rate": 1.5159708863702225e-05,
+      "loss": 0.985,
+      "step": 14755
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7014487906282572,
+      "learning_rate": 1.515905522304022e-05,
+      "loss": 0.9642,
+      "step": 14756
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6979452259341266,
+      "learning_rate": 1.5158401552340562e-05,
+      "loss": 0.8648,
+      "step": 14757
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5201113522178173,
+      "learning_rate": 1.5157747851607055e-05,
+      "loss": 0.8506,
+      "step": 14758
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6457203565949068,
+      "learning_rate": 1.5157094120843512e-05,
+      "loss": 0.8548,
+      "step": 14759
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6963000810193383,
+      "learning_rate": 1.515644036005373e-05,
+      "loss": 0.972,
+      "step": 14760
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.84770302572807,
+      "learning_rate": 1.5155786569241521e-05,
+      "loss": 0.8028,
+      "step": 14761
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.77319135408749,
+      "learning_rate": 1.515513274841069e-05,
+      "loss": 0.8999,
+      "step": 14762
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0130357953768976,
+      "learning_rate": 1.5154478897565046e-05,
+      "loss": 0.8428,
+      "step": 14763
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.541340814315082,
+      "learning_rate": 1.515382501670839e-05,
+      "loss": 0.8951,
+      "step": 14764
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6043821969745595,
+      "learning_rate": 1.5153171105844537e-05,
+      "loss": 0.9486,
+      "step": 14765
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8792071923812208,
+      "learning_rate": 1.5152517164977289e-05,
+      "loss": 0.8804,
+      "step": 14766
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7214604335743473,
+      "learning_rate": 1.5151863194110455e-05,
+      "loss": 0.9587,
+      "step": 14767
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6294489928381328,
+      "learning_rate": 1.515120919324784e-05,
+      "loss": 0.9126,
+      "step": 14768
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5236431200466465,
+      "learning_rate": 1.5150555162393254e-05,
+      "loss": 1.0114,
+      "step": 14769
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6400670110473146,
+      "learning_rate": 1.514990110155051e-05,
+      "loss": 0.9617,
+      "step": 14770
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5653055822149033,
+      "learning_rate": 1.5149247010723409e-05,
+      "loss": 0.7619,
+      "step": 14771
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5077525086231924,
+      "learning_rate": 1.5148592889915763e-05,
+      "loss": 0.8619,
+      "step": 14772
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5089925154488433,
+      "learning_rate": 1.5147938739131375e-05,
+      "loss": 0.9352,
+      "step": 14773
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6750376223448002,
+      "learning_rate": 1.5147284558374062e-05,
+      "loss": 0.9117,
+      "step": 14774
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4814055634090277,
+      "learning_rate": 1.5146630347647628e-05,
+      "loss": 0.9607,
+      "step": 14775
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5901638822694444,
+      "learning_rate": 1.5145976106955882e-05,
+      "loss": 0.9013,
+      "step": 14776
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6490143526088619,
+      "learning_rate": 1.5145321836302635e-05,
+      "loss": 1.0202,
+      "step": 14777
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8491516528004368,
+      "learning_rate": 1.5144667535691694e-05,
+      "loss": 0.8136,
+      "step": 14778
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8210242793885651,
+      "learning_rate": 1.5144013205126868e-05,
+      "loss": 0.9371,
+      "step": 14779
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5223847493113931,
+      "learning_rate": 1.5143358844611973e-05,
+      "loss": 0.9411,
+      "step": 14780
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.779216560362211,
+      "learning_rate": 1.5142704454150812e-05,
+      "loss": 0.7449,
+      "step": 14781
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.646308913245149,
+      "learning_rate": 1.5142050033747197e-05,
+      "loss": 0.9602,
+      "step": 14782
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.255567808410833,
+      "learning_rate": 1.5141395583404939e-05,
+      "loss": 0.8356,
+      "step": 14783
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8287164191762177,
+      "learning_rate": 1.5140741103127848e-05,
+      "loss": 0.7246,
+      "step": 14784
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0154982206443117,
+      "learning_rate": 1.5140086592919735e-05,
+      "loss": 0.9081,
+      "step": 14785
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6373816688424716,
+      "learning_rate": 1.513943205278441e-05,
+      "loss": 0.9131,
+      "step": 14786
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7746679849875656,
+      "learning_rate": 1.5138777482725688e-05,
+      "loss": 0.7312,
+      "step": 14787
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.697465260217285,
+      "learning_rate": 1.5138122882747373e-05,
+      "loss": 0.9838,
+      "step": 14788
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8352903803685976,
+      "learning_rate": 1.513746825285328e-05,
+      "loss": 0.74,
+      "step": 14789
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5308240054963367,
+      "learning_rate": 1.5136813593047222e-05,
+      "loss": 0.886,
+      "step": 14790
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6167462643842534,
+      "learning_rate": 1.5136158903333007e-05,
+      "loss": 0.9207,
+      "step": 14791
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0878779280847874,
+      "learning_rate": 1.5135504183714448e-05,
+      "loss": 1.0937,
+      "step": 14792
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.975987853900475,
+      "learning_rate": 1.5134849434195362e-05,
+      "loss": 0.8621,
+      "step": 14793
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7176638171636915,
+      "learning_rate": 1.5134194654779555e-05,
+      "loss": 0.8442,
+      "step": 14794
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7560151063560142,
+      "learning_rate": 1.5133539845470839e-05,
+      "loss": 0.695,
+      "step": 14795
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5964268313194927,
+      "learning_rate": 1.5132885006273032e-05,
+      "loss": 0.9809,
+      "step": 14796
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7842770127218683,
+      "learning_rate": 1.5132230137189941e-05,
+      "loss": 0.8901,
+      "step": 14797
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.687739018739146,
+      "learning_rate": 1.5131575238225385e-05,
+      "loss": 0.8963,
+      "step": 14798
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.482793989835164,
+      "learning_rate": 1.5130920309383167e-05,
+      "loss": 0.9182,
+      "step": 14799
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5588543168130764,
+      "learning_rate": 1.5130265350667111e-05,
+      "loss": 0.9773,
+      "step": 14800
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.808615214480386,
+      "learning_rate": 1.5129610362081027e-05,
+      "loss": 0.7365,
+      "step": 14801
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7943438158477898,
+      "learning_rate": 1.5128955343628724e-05,
+      "loss": 0.91,
+      "step": 14802
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8528543556458144,
+      "learning_rate": 1.5128300295314021e-05,
+      "loss": 0.7571,
+      "step": 14803
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5868177193409918,
+      "learning_rate": 1.512764521714073e-05,
+      "loss": 0.8445,
+      "step": 14804
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7402261298217903,
+      "learning_rate": 1.5126990109112665e-05,
+      "loss": 0.7446,
+      "step": 14805
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8156850094634955,
+      "learning_rate": 1.512633497123364e-05,
+      "loss": 0.9756,
+      "step": 14806
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6516987113857016,
+      "learning_rate": 1.5125679803507472e-05,
+      "loss": 1.0002,
+      "step": 14807
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5508581227724154,
+      "learning_rate": 1.512502460593797e-05,
+      "loss": 0.8496,
+      "step": 14808
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7856103539423243,
+      "learning_rate": 1.5124369378528954e-05,
+      "loss": 0.8526,
+      "step": 14809
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5345369924507308,
+      "learning_rate": 1.512371412128424e-05,
+      "loss": 1.022,
+      "step": 14810
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7861748857956503,
+      "learning_rate": 1.5123058834207633e-05,
+      "loss": 0.9015,
+      "step": 14811
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.661996926263367,
+      "learning_rate": 1.5122403517302964e-05,
+      "loss": 0.8822,
+      "step": 14812
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6100184591370008,
+      "learning_rate": 1.512174817057403e-05,
+      "loss": 0.8228,
+      "step": 14813
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6868658789435462,
+      "learning_rate": 1.5121092794024668e-05,
+      "loss": 0.8796,
+      "step": 14814
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6153387872149787,
+      "learning_rate": 1.5120437387658675e-05,
+      "loss": 0.9962,
+      "step": 14815
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5332026508013181,
+      "learning_rate": 1.5119781951479877e-05,
+      "loss": 0.9913,
+      "step": 14816
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4718743650878017,
+      "learning_rate": 1.5119126485492086e-05,
+      "loss": 0.8442,
+      "step": 14817
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8048813995186586,
+      "learning_rate": 1.5118470989699119e-05,
+      "loss": 0.7416,
+      "step": 14818
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6776485939343644,
+      "learning_rate": 1.5117815464104797e-05,
+      "loss": 0.844,
+      "step": 14819
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7078127952461017,
+      "learning_rate": 1.5117159908712929e-05,
+      "loss": 0.8863,
+      "step": 14820
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.67431043557134,
+      "learning_rate": 1.511650432352734e-05,
+      "loss": 0.7567,
+      "step": 14821
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7794572968755316,
+      "learning_rate": 1.511584870855184e-05,
+      "loss": 0.9131,
+      "step": 14822
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0294518901474117,
+      "learning_rate": 1.5115193063790253e-05,
+      "loss": 0.9744,
+      "step": 14823
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.518638887382421,
+      "learning_rate": 1.5114537389246388e-05,
+      "loss": 0.9346,
+      "step": 14824
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6070874908619264,
+      "learning_rate": 1.511388168492407e-05,
+      "loss": 1.0214,
+      "step": 14825
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7257531353886013,
+      "learning_rate": 1.511322595082711e-05,
+      "loss": 0.8172,
+      "step": 14826
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8076378934701361,
+      "learning_rate": 1.5112570186959334e-05,
+      "loss": 0.7855,
+      "step": 14827
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4826421818591295,
+      "learning_rate": 1.5111914393324553e-05,
+      "loss": 0.9224,
+      "step": 14828
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8534335337685297,
+      "learning_rate": 1.511125856992659e-05,
+      "loss": 0.8195,
+      "step": 14829
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6367128938153923,
+      "learning_rate": 1.511060271676926e-05,
+      "loss": 0.9453,
+      "step": 14830
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9606028621764857,
+      "learning_rate": 1.5109946833856386e-05,
+      "loss": 0.8359,
+      "step": 14831
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.802195893715436,
+      "learning_rate": 1.510929092119178e-05,
+      "loss": 0.817,
+      "step": 14832
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7608607764648693,
+      "learning_rate": 1.5108634978779267e-05,
+      "loss": 0.7396,
+      "step": 14833
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7376264996914552,
+      "learning_rate": 1.5107979006622662e-05,
+      "loss": 0.8733,
+      "step": 14834
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6162363897074137,
+      "learning_rate": 1.5107323004725786e-05,
+      "loss": 0.9289,
+      "step": 14835
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.819506695236646,
+      "learning_rate": 1.510666697309246e-05,
+      "loss": 0.7892,
+      "step": 14836
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3276812234760644,
+      "learning_rate": 1.5106010911726501e-05,
+      "loss": 0.9003,
+      "step": 14837
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4811379658475665,
+      "learning_rate": 1.5105354820631732e-05,
+      "loss": 0.9397,
+      "step": 14838
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8084643260646662,
+      "learning_rate": 1.5104698699811968e-05,
+      "loss": 0.9426,
+      "step": 14839
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.985994569240437,
+      "learning_rate": 1.5104042549271033e-05,
+      "loss": 0.7639,
+      "step": 14840
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5932554592453274,
+      "learning_rate": 1.5103386369012746e-05,
+      "loss": 0.9296,
+      "step": 14841
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4800447177407658,
+      "learning_rate": 1.510273015904093e-05,
+      "loss": 0.9182,
+      "step": 14842
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7521631957870814,
+      "learning_rate": 1.5102073919359402e-05,
+      "loss": 0.9119,
+      "step": 14843
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4530445058336638,
+      "learning_rate": 1.5101417649971986e-05,
+      "loss": 0.8854,
+      "step": 14844
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6594998263494347,
+      "learning_rate": 1.51007613508825e-05,
+      "loss": 1.0324,
+      "step": 14845
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.743784980020901,
+      "learning_rate": 1.5100105022094766e-05,
+      "loss": 0.8162,
+      "step": 14846
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.508380529094082,
+      "learning_rate": 1.5099448663612605e-05,
+      "loss": 0.8544,
+      "step": 14847
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4054069081835348,
+      "learning_rate": 1.5098792275439842e-05,
+      "loss": 0.8467,
+      "step": 14848
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.520848584315485,
+      "learning_rate": 1.5098135857580296e-05,
+      "loss": 1.0202,
+      "step": 14849
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.863353496145302,
+      "learning_rate": 1.5097479410037787e-05,
+      "loss": 0.8614,
+      "step": 14850
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7470358317389854,
+      "learning_rate": 1.509682293281614e-05,
+      "loss": 0.925,
+      "step": 14851
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0729604415416336,
+      "learning_rate": 1.5096166425919176e-05,
+      "loss": 0.881,
+      "step": 14852
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9269006395113135,
+      "learning_rate": 1.5095509889350716e-05,
+      "loss": 0.7726,
+      "step": 14853
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6183175366965796,
+      "learning_rate": 1.5094853323114587e-05,
+      "loss": 0.8512,
+      "step": 14854
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.739188731938509,
+      "learning_rate": 1.5094196727214607e-05,
+      "loss": 0.8843,
+      "step": 14855
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.905816796204148,
+      "learning_rate": 1.5093540101654602e-05,
+      "loss": 1.028,
+      "step": 14856
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5851471244233013,
+      "learning_rate": 1.5092883446438396e-05,
+      "loss": 0.8933,
+      "step": 14857
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.224065965784221,
+      "learning_rate": 1.5092226761569805e-05,
+      "loss": 0.8247,
+      "step": 14858
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.213778693828256,
+      "learning_rate": 1.5091570047052666e-05,
+      "loss": 0.9749,
+      "step": 14859
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.580702961497319,
+      "learning_rate": 1.5090913302890786e-05,
+      "loss": 0.9546,
+      "step": 14860
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5873507809712963,
+      "learning_rate": 1.5090256529088e-05,
+      "loss": 0.8695,
+      "step": 14861
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5574405058556269,
+      "learning_rate": 1.5089599725648128e-05,
+      "loss": 0.9613,
+      "step": 14862
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5677969314923212,
+      "learning_rate": 1.5088942892574995e-05,
+      "loss": 0.8685,
+      "step": 14863
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.737632582750886,
+      "learning_rate": 1.5088286029872427e-05,
+      "loss": 0.8548,
+      "step": 14864
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8437293433390018,
+      "learning_rate": 1.5087629137544246e-05,
+      "loss": 0.8217,
+      "step": 14865
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.657833503818368,
+      "learning_rate": 1.508697221559428e-05,
+      "loss": 0.9829,
+      "step": 14866
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.542702616207925,
+      "learning_rate": 1.5086315264026345e-05,
+      "loss": 0.9214,
+      "step": 14867
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8686539682601035,
+      "learning_rate": 1.5085658282844278e-05,
+      "loss": 1.0039,
+      "step": 14868
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.776314898219905,
+      "learning_rate": 1.5085001272051893e-05,
+      "loss": 0.8919,
+      "step": 14869
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3419323762866946,
+      "learning_rate": 1.5084344231653027e-05,
+      "loss": 0.9912,
+      "step": 14870
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8048657378886093,
+      "learning_rate": 1.5083687161651495e-05,
+      "loss": 0.8767,
+      "step": 14871
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4656221445749134,
+      "learning_rate": 1.508303006205113e-05,
+      "loss": 0.7705,
+      "step": 14872
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5215355172211151,
+      "learning_rate": 1.5082372932855753e-05,
+      "loss": 0.9094,
+      "step": 14873
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7853828770028041,
+      "learning_rate": 1.5081715774069191e-05,
+      "loss": 0.7483,
+      "step": 14874
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4654959424427558,
+      "learning_rate": 1.5081058585695272e-05,
+      "loss": 1.0384,
+      "step": 14875
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9698781576229316,
+      "learning_rate": 1.508040136773782e-05,
+      "loss": 0.8079,
+      "step": 14876
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8331432343630727,
+      "learning_rate": 1.5079744120200663e-05,
+      "loss": 1.0064,
+      "step": 14877
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6559880451266087,
+      "learning_rate": 1.5079086843087626e-05,
+      "loss": 0.9375,
+      "step": 14878
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7502102494498553,
+      "learning_rate": 1.507842953640254e-05,
+      "loss": 0.7582,
+      "step": 14879
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7131294590203776,
+      "learning_rate": 1.507777220014923e-05,
+      "loss": 0.8991,
+      "step": 14880
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9830743083380975,
+      "learning_rate": 1.5077114834331524e-05,
+      "loss": 0.8036,
+      "step": 14881
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0719356397864495,
+      "learning_rate": 1.5076457438953242e-05,
+      "loss": 0.8836,
+      "step": 14882
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9031187009463697,
+      "learning_rate": 1.5075800014018223e-05,
+      "loss": 0.8077,
+      "step": 14883
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8407815711285587,
+      "learning_rate": 1.5075142559530287e-05,
+      "loss": 0.7876,
+      "step": 14884
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.484447231229328,
+      "learning_rate": 1.5074485075493263e-05,
+      "loss": 0.9563,
+      "step": 14885
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6696362006900276,
+      "learning_rate": 1.5073827561910983e-05,
+      "loss": 0.8361,
+      "step": 14886
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.60265187461225,
+      "learning_rate": 1.5073170018787272e-05,
+      "loss": 0.9447,
+      "step": 14887
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7768968986053241,
+      "learning_rate": 1.5072512446125958e-05,
+      "loss": 0.7067,
+      "step": 14888
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7759506870276777,
+      "learning_rate": 1.5071854843930873e-05,
+      "loss": 0.7143,
+      "step": 14889
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5193819868158775,
+      "learning_rate": 1.5071197212205842e-05,
+      "loss": 0.8676,
+      "step": 14890
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4739845847178867,
+      "learning_rate": 1.5070539550954696e-05,
+      "loss": 0.8807,
+      "step": 14891
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8029158418134548,
+      "learning_rate": 1.5069881860181261e-05,
+      "loss": 0.941,
+      "step": 14892
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.212018495845884,
+      "learning_rate": 1.5069224139889372e-05,
+      "loss": 0.9256,
+      "step": 14893
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6644263511077244,
+      "learning_rate": 1.506856639008285e-05,
+      "loss": 0.8012,
+      "step": 14894
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7770415897182027,
+      "learning_rate": 1.5067908610765536e-05,
+      "loss": 0.7406,
+      "step": 14895
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5621370656610412,
+      "learning_rate": 1.506725080194125e-05,
+      "loss": 0.9948,
+      "step": 14896
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2651327754862334,
+      "learning_rate": 1.5066592963613826e-05,
+      "loss": 0.8897,
+      "step": 14897
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.682127033590401,
+      "learning_rate": 1.5065935095787093e-05,
+      "loss": 0.8428,
+      "step": 14898
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7141053123664671,
+      "learning_rate": 1.5065277198464885e-05,
+      "loss": 0.9487,
+      "step": 14899
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6908751184916118,
+      "learning_rate": 1.5064619271651027e-05,
+      "loss": 0.8963,
+      "step": 14900
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7013420943383744,
+      "learning_rate": 1.5063961315349352e-05,
+      "loss": 0.9497,
+      "step": 14901
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7774595842664023,
+      "learning_rate": 1.5063303329563693e-05,
+      "loss": 0.9052,
+      "step": 14902
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.12776217568858,
+      "learning_rate": 1.5062645314297877e-05,
+      "loss": 0.853,
+      "step": 14903
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6949975488485303,
+      "learning_rate": 1.5061987269555735e-05,
+      "loss": 0.8834,
+      "step": 14904
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6604004037711502,
+      "learning_rate": 1.5061329195341105e-05,
+      "loss": 0.8842,
+      "step": 14905
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.644709225183442,
+      "learning_rate": 1.5060671091657813e-05,
+      "loss": 0.9678,
+      "step": 14906
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6954207149590135,
+      "learning_rate": 1.506001295850969e-05,
+      "loss": 0.9662,
+      "step": 14907
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7971245134238603,
+      "learning_rate": 1.5059354795900571e-05,
+      "loss": 0.8326,
+      "step": 14908
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6577797507611673,
+      "learning_rate": 1.5058696603834287e-05,
+      "loss": 0.8804,
+      "step": 14909
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6154511954784954,
+      "learning_rate": 1.5058038382314673e-05,
+      "loss": 0.8517,
+      "step": 14910
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9237419865840635,
+      "learning_rate": 1.505738013134555e-05,
+      "loss": 0.8131,
+      "step": 14911
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9050864533840263,
+      "learning_rate": 1.5056721850930766e-05,
+      "loss": 0.8017,
+      "step": 14912
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6748705907289172,
+      "learning_rate": 1.505606354107414e-05,
+      "loss": 0.8095,
+      "step": 14913
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6841351505417426,
+      "learning_rate": 1.5055405201779518e-05,
+      "loss": 0.7638,
+      "step": 14914
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6782187993030548,
+      "learning_rate": 1.5054746833050721e-05,
+      "loss": 0.846,
+      "step": 14915
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.704593253180078,
+      "learning_rate": 1.505408843489159e-05,
+      "loss": 0.9694,
+      "step": 14916
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6694970452187932,
+      "learning_rate": 1.5053430007305958e-05,
+      "loss": 0.8752,
+      "step": 14917
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5671941610804876,
+      "learning_rate": 1.5052771550297651e-05,
+      "loss": 0.9361,
+      "step": 14918
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7919687030835925,
+      "learning_rate": 1.5052113063870514e-05,
+      "loss": 0.8539,
+      "step": 14919
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4111178991375866,
+      "learning_rate": 1.5051454548028373e-05,
+      "loss": 0.8449,
+      "step": 14920
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5215742066476399,
+      "learning_rate": 1.5050796002775065e-05,
+      "loss": 0.7911,
+      "step": 14921
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5589795292095225,
+      "learning_rate": 1.5050137428114422e-05,
+      "loss": 0.903,
+      "step": 14922
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.566176387423712,
+      "learning_rate": 1.5049478824050282e-05,
+      "loss": 0.9568,
+      "step": 14923
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.608220546376197,
+      "learning_rate": 1.5048820190586478e-05,
+      "loss": 1.0178,
+      "step": 14924
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.529182326797057,
+      "learning_rate": 1.5048161527726843e-05,
+      "loss": 0.8554,
+      "step": 14925
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8367682194022248,
+      "learning_rate": 1.5047502835475211e-05,
+      "loss": 0.9355,
+      "step": 14926
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.603417597845634,
+      "learning_rate": 1.5046844113835426e-05,
+      "loss": 0.9761,
+      "step": 14927
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.057362752935415,
+      "learning_rate": 1.504618536281131e-05,
+      "loss": 0.9526,
+      "step": 14928
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5985869119134624,
+      "learning_rate": 1.5045526582406708e-05,
+      "loss": 0.8498,
+      "step": 14929
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.822108149943226,
+      "learning_rate": 1.5044867772625455e-05,
+      "loss": 0.9578,
+      "step": 14930
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8965795800188476,
+      "learning_rate": 1.5044208933471383e-05,
+      "loss": 0.9026,
+      "step": 14931
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.657350729110526,
+      "learning_rate": 1.5043550064948327e-05,
+      "loss": 0.7632,
+      "step": 14932
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7706459678489983,
+      "learning_rate": 1.504289116706013e-05,
+      "loss": 0.9481,
+      "step": 14933
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8173491890679239,
+      "learning_rate": 1.5042232239810618e-05,
+      "loss": 0.9095,
+      "step": 14934
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8813606335890116,
+      "learning_rate": 1.5041573283203641e-05,
+      "loss": 0.9362,
+      "step": 14935
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8343839292668374,
+      "learning_rate": 1.5040914297243023e-05,
+      "loss": 0.9381,
+      "step": 14936
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.638680500280905,
+      "learning_rate": 1.5040255281932608e-05,
+      "loss": 0.958,
+      "step": 14937
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8440318287783795,
+      "learning_rate": 1.5039596237276234e-05,
+      "loss": 0.8042,
+      "step": 14938
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5108940414066023,
+      "learning_rate": 1.5038937163277729e-05,
+      "loss": 0.8336,
+      "step": 14939
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7689494158025446,
+      "learning_rate": 1.503827805994094e-05,
+      "loss": 0.793,
+      "step": 14940
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6373446320782865,
+      "learning_rate": 1.50376189272697e-05,
+      "loss": 0.9046,
+      "step": 14941
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6415416363288464,
+      "learning_rate": 1.5036959765267852e-05,
+      "loss": 0.8671,
+      "step": 14942
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.954530939447006,
+      "learning_rate": 1.5036300573939224e-05,
+      "loss": 0.8355,
+      "step": 14943
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5497096296174517,
+      "learning_rate": 1.5035641353287662e-05,
+      "loss": 0.8571,
+      "step": 14944
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.707510835955016,
+      "learning_rate": 1.5034982103317001e-05,
+      "loss": 0.8142,
+      "step": 14945
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6932815929676874,
+      "learning_rate": 1.5034322824031083e-05,
+      "loss": 0.989,
+      "step": 14946
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6919611079326613,
+      "learning_rate": 1.5033663515433738e-05,
+      "loss": 0.8858,
+      "step": 14947
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5247491043608916,
+      "learning_rate": 1.5033004177528816e-05,
+      "loss": 0.8999,
+      "step": 14948
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5257418679308057,
+      "learning_rate": 1.5032344810320148e-05,
+      "loss": 0.8159,
+      "step": 14949
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6489071978396457,
+      "learning_rate": 1.5031685413811574e-05,
+      "loss": 0.899,
+      "step": 14950
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8350002194588177,
+      "learning_rate": 1.5031025988006935e-05,
+      "loss": 1.0216,
+      "step": 14951
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6213924343927657,
+      "learning_rate": 1.503036653291007e-05,
+      "loss": 0.9908,
+      "step": 14952
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6060304020580538,
+      "learning_rate": 1.502970704852482e-05,
+      "loss": 0.8891,
+      "step": 14953
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8836468487768108,
+      "learning_rate": 1.502904753485502e-05,
+      "loss": 0.9051,
+      "step": 14954
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6541556137543394,
+      "learning_rate": 1.5028387991904517e-05,
+      "loss": 1.0335,
+      "step": 14955
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6256725292656524,
+      "learning_rate": 1.5027728419677146e-05,
+      "loss": 0.9285,
+      "step": 14956
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7423753454544986,
+      "learning_rate": 1.5027068818176747e-05,
+      "loss": 0.805,
+      "step": 14957
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4850687308721795,
+      "learning_rate": 1.502640918740716e-05,
+      "loss": 0.9426,
+      "step": 14958
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9011468234913935,
+      "learning_rate": 1.502574952737223e-05,
+      "loss": 0.868,
+      "step": 14959
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7836964788215954,
+      "learning_rate": 1.5025089838075797e-05,
+      "loss": 0.924,
+      "step": 14960
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6320565771060036,
+      "learning_rate": 1.50244301195217e-05,
+      "loss": 0.9224,
+      "step": 14961
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.572906457023171,
+      "learning_rate": 1.5023770371713778e-05,
+      "loss": 1.0543,
+      "step": 14962
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4322134845927408,
+      "learning_rate": 1.5023110594655875e-05,
+      "loss": 0.8447,
+      "step": 14963
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7225623532476197,
+      "learning_rate": 1.5022450788351828e-05,
+      "loss": 0.9175,
+      "step": 14964
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7819057262336172,
+      "learning_rate": 1.5021790952805489e-05,
+      "loss": 0.7793,
+      "step": 14965
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5607184268356245,
+      "learning_rate": 1.5021131088020688e-05,
+      "loss": 0.8707,
+      "step": 14966
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7768022660662024,
+      "learning_rate": 1.5020471194001274e-05,
+      "loss": 0.7649,
+      "step": 14967
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6028090132627502,
+      "learning_rate": 1.501981127075109e-05,
+      "loss": 0.9142,
+      "step": 14968
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6921590972169045,
+      "learning_rate": 1.501915131827397e-05,
+      "loss": 0.8863,
+      "step": 14969
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6377839257532731,
+      "learning_rate": 1.5018491336573768e-05,
+      "loss": 0.8412,
+      "step": 14970
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9575042416102009,
+      "learning_rate": 1.5017831325654316e-05,
+      "loss": 0.9182,
+      "step": 14971
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8606458908260386,
+      "learning_rate": 1.5017171285519463e-05,
+      "loss": 0.8898,
+      "step": 14972
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.612062442541705,
+      "learning_rate": 1.5016511216173052e-05,
+      "loss": 0.7745,
+      "step": 14973
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7000434847650323,
+      "learning_rate": 1.5015851117618923e-05,
+      "loss": 0.9762,
+      "step": 14974
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9016580200470696,
+      "learning_rate": 1.501519098986092e-05,
+      "loss": 0.9959,
+      "step": 14975
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.905874423860583,
+      "learning_rate": 1.5014530832902886e-05,
+      "loss": 0.8575,
+      "step": 14976
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.662950313010664,
+      "learning_rate": 1.501387064674867e-05,
+      "loss": 0.9519,
+      "step": 14977
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.842153432369058,
+      "learning_rate": 1.501321043140211e-05,
+      "loss": 0.9608,
+      "step": 14978
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6683224877805112,
+      "learning_rate": 1.5012550186867051e-05,
+      "loss": 1.0037,
+      "step": 14979
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7222572337967046,
+      "learning_rate": 1.501188991314734e-05,
+      "loss": 0.8364,
+      "step": 14980
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.621193344578533,
+      "learning_rate": 1.5011229610246815e-05,
+      "loss": 0.8944,
+      "step": 14981
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4568788292555774,
+      "learning_rate": 1.5010569278169327e-05,
+      "loss": 0.985,
+      "step": 14982
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.589115995942945,
+      "learning_rate": 1.5009908916918719e-05,
+      "loss": 0.8141,
+      "step": 14983
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.578394510813162,
+      "learning_rate": 1.5009248526498834e-05,
+      "loss": 1.0179,
+      "step": 14984
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6640354671676498,
+      "learning_rate": 1.500858810691352e-05,
+      "loss": 0.8644,
+      "step": 14985
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5526701838567254,
+      "learning_rate": 1.5007927658166619e-05,
+      "loss": 0.8923,
+      "step": 14986
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.653696594308284,
+      "learning_rate": 1.5007267180261975e-05,
+      "loss": 0.8542,
+      "step": 14987
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4548917967110138,
+      "learning_rate": 1.500660667320344e-05,
+      "loss": 0.7882,
+      "step": 14988
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6007198398157594,
+      "learning_rate": 1.5005946136994857e-05,
+      "loss": 0.7908,
+      "step": 14989
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6602277630241515,
+      "learning_rate": 1.5005285571640067e-05,
+      "loss": 0.8294,
+      "step": 14990
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5768252550931254,
+      "learning_rate": 1.5004624977142923e-05,
+      "loss": 0.7465,
+      "step": 14991
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9407335160137353,
+      "learning_rate": 1.5003964353507265e-05,
+      "loss": 0.8998,
+      "step": 14992
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.060630874452721,
+      "learning_rate": 1.5003303700736944e-05,
+      "loss": 0.9215,
+      "step": 14993
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.021674755824909,
+      "learning_rate": 1.5002643018835806e-05,
+      "loss": 0.9923,
+      "step": 14994
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7849319578415421,
+      "learning_rate": 1.5001982307807694e-05,
+      "loss": 0.7931,
+      "step": 14995
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4966083185732717,
+      "learning_rate": 1.5001321567656459e-05,
+      "loss": 0.8556,
+      "step": 14996
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7660523005362344,
+      "learning_rate": 1.5000660798385943e-05,
+      "loss": 0.7098,
+      "step": 14997
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6771420135230084,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.9318,
+      "step": 14998
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1693748786911073,
+      "learning_rate": 1.4999339172502477e-05,
+      "loss": 0.8304,
+      "step": 14999
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.562744885324379,
+      "learning_rate": 1.4998678315897213e-05,
+      "loss": 0.859,
+      "step": 15000
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.737551537497392,
+      "learning_rate": 1.4998017430188063e-05,
+      "loss": 0.7434,
+      "step": 15001
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8339824997220238,
+      "learning_rate": 1.4997356515378877e-05,
+      "loss": 0.9248,
+      "step": 15002
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6933860251131898,
+      "learning_rate": 1.4996695571473495e-05,
+      "loss": 0.8594,
+      "step": 15003
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4056983255139077,
+      "learning_rate": 1.4996034598475769e-05,
+      "loss": 0.875,
+      "step": 15004
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7212825801387623,
+      "learning_rate": 1.4995373596389549e-05,
+      "loss": 0.6954,
+      "step": 15005
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7071373047143914,
+      "learning_rate": 1.499471256521868e-05,
+      "loss": 0.7754,
+      "step": 15006
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7078035057015402,
+      "learning_rate": 1.4994051504967018e-05,
+      "loss": 0.9104,
+      "step": 15007
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4264569600140031,
+      "learning_rate": 1.4993390415638405e-05,
+      "loss": 0.8069,
+      "step": 15008
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7568613484594058,
+      "learning_rate": 1.4992729297236693e-05,
+      "loss": 0.7615,
+      "step": 15009
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.76310868835038,
+      "learning_rate": 1.4992068149765726e-05,
+      "loss": 0.8569,
+      "step": 15010
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8350625117532564,
+      "learning_rate": 1.499140697322936e-05,
+      "loss": 0.7726,
+      "step": 15011
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.595099443834224,
+      "learning_rate": 1.4990745767631445e-05,
+      "loss": 0.912,
+      "step": 15012
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3944260286598764,
+      "learning_rate": 1.4990084532975826e-05,
+      "loss": 0.9129,
+      "step": 15013
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5424203979853268,
+      "learning_rate": 1.4989423269266357e-05,
+      "loss": 1.0043,
+      "step": 15014
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5407206279385315,
+      "learning_rate": 1.4988761976506885e-05,
+      "loss": 0.9145,
+      "step": 15015
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4137212660135887,
+      "learning_rate": 1.4988100654701258e-05,
+      "loss": 0.9276,
+      "step": 15016
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0283439525056335,
+      "learning_rate": 1.4987439303853336e-05,
+      "loss": 0.8417,
+      "step": 15017
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.113857592558988,
+      "learning_rate": 1.4986777923966958e-05,
+      "loss": 0.9912,
+      "step": 15018
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8863007926341915,
+      "learning_rate": 1.4986116515045982e-05,
+      "loss": 0.8954,
+      "step": 15019
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5385909565738816,
+      "learning_rate": 1.4985455077094258e-05,
+      "loss": 1.065,
+      "step": 15020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6717924038173981,
+      "learning_rate": 1.4984793610115635e-05,
+      "loss": 0.9079,
+      "step": 15021
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7382302673857056,
+      "learning_rate": 1.4984132114113965e-05,
+      "loss": 0.7299,
+      "step": 15022
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.884922544896995,
+      "learning_rate": 1.4983470589093103e-05,
+      "loss": 0.8154,
+      "step": 15023
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9242995979692508,
+      "learning_rate": 1.4982809035056894e-05,
+      "loss": 0.9707,
+      "step": 15024
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0929015507587176,
+      "learning_rate": 1.4982147452009197e-05,
+      "loss": 1.0223,
+      "step": 15025
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6828261136644789,
+      "learning_rate": 1.4981485839953855e-05,
+      "loss": 0.9248,
+      "step": 15026
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4763332247793919,
+      "learning_rate": 1.498082419889473e-05,
+      "loss": 0.717,
+      "step": 15027
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.569516885100648,
+      "learning_rate": 1.4980162528835666e-05,
+      "loss": 0.9051,
+      "step": 15028
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7687693037112135,
+      "learning_rate": 1.4979500829780522e-05,
+      "loss": 0.8979,
+      "step": 15029
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5136190365213624,
+      "learning_rate": 1.4978839101733144e-05,
+      "loss": 0.9038,
+      "step": 15030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7416621418099334,
+      "learning_rate": 1.4978177344697394e-05,
+      "loss": 1.0228,
+      "step": 15031
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7044097088418209,
+      "learning_rate": 1.4977515558677116e-05,
+      "loss": 0.8559,
+      "step": 15032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7056299324413324,
+      "learning_rate": 1.4976853743676165e-05,
+      "loss": 0.9992,
+      "step": 15033
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5850650935065915,
+      "learning_rate": 1.4976191899698398e-05,
+      "loss": 0.8873,
+      "step": 15034
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.454159684380968,
+      "learning_rate": 1.4975530026747667e-05,
+      "loss": 0.9785,
+      "step": 15035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6011234143846444,
+      "learning_rate": 1.4974868124827823e-05,
+      "loss": 0.9411,
+      "step": 15036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7803400351460736,
+      "learning_rate": 1.4974206193942724e-05,
+      "loss": 0.682,
+      "step": 15037
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3926543559366964,
+      "learning_rate": 1.497354423409622e-05,
+      "loss": 0.8661,
+      "step": 15038
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5437949540455849,
+      "learning_rate": 1.4972882245292165e-05,
+      "loss": 0.8086,
+      "step": 15039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7289186371023884,
+      "learning_rate": 1.4972220227534422e-05,
+      "loss": 0.9187,
+      "step": 15040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7450407275750044,
+      "learning_rate": 1.4971558180826832e-05,
+      "loss": 0.8494,
+      "step": 15041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.831112299346572,
+      "learning_rate": 1.497089610517326e-05,
+      "loss": 0.9763,
+      "step": 15042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7716530566404232,
+      "learning_rate": 1.4970234000577555e-05,
+      "loss": 1.0252,
+      "step": 15043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4501476094760049,
+      "learning_rate": 1.4969571867043576e-05,
+      "loss": 0.9474,
+      "step": 15044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4661450569354293,
+      "learning_rate": 1.4968909704575171e-05,
+      "loss": 0.9299,
+      "step": 15045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6844046847321372,
+      "learning_rate": 1.4968247513176209e-05,
+      "loss": 0.8479,
+      "step": 15046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4586745237569787,
+      "learning_rate": 1.4967585292850531e-05,
+      "loss": 1.0189,
+      "step": 15047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7580455443089201,
+      "learning_rate": 1.4966923043602002e-05,
+      "loss": 0.7161,
+      "step": 15048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1489250863770155,
+      "learning_rate": 1.4966260765434472e-05,
+      "loss": 0.8447,
+      "step": 15049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6146613580539315,
+      "learning_rate": 1.4965598458351797e-05,
+      "loss": 0.8752,
+      "step": 15050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8062541583595322,
+      "learning_rate": 1.4964936122357843e-05,
+      "loss": 0.9658,
+      "step": 15051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6706273849215931,
+      "learning_rate": 1.4964273757456451e-05,
+      "loss": 0.9326,
+      "step": 15052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9027871056267087,
+      "learning_rate": 1.4963611363651494e-05,
+      "loss": 0.9479,
+      "step": 15053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5688803356170378,
+      "learning_rate": 1.4962948940946814e-05,
+      "loss": 0.87,
+      "step": 15054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5045635102911734,
+      "learning_rate": 1.4962286489346276e-05,
+      "loss": 0.8418,
+      "step": 15055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4964047850495554,
+      "learning_rate": 1.4961624008853735e-05,
+      "loss": 0.9217,
+      "step": 15056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.940875866526283,
+      "learning_rate": 1.4960961499473048e-05,
+      "loss": 0.8785,
+      "step": 15057
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.702332028832766,
+      "learning_rate": 1.4960298961208073e-05,
+      "loss": 0.923,
+      "step": 15058
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6250466893776254,
+      "learning_rate": 1.4959636394062666e-05,
+      "loss": 0.861,
+      "step": 15059
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.654133181234428,
+      "learning_rate": 1.4958973798040685e-05,
+      "loss": 0.9473,
+      "step": 15060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.786930295360957,
+      "learning_rate": 1.4958311173145991e-05,
+      "loss": 0.7648,
+      "step": 15061
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5365454684249527,
+      "learning_rate": 1.495764851938244e-05,
+      "loss": 0.9639,
+      "step": 15062
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6248682257483098,
+      "learning_rate": 1.4956985836753887e-05,
+      "loss": 0.8962,
+      "step": 15063
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8300076520505116,
+      "learning_rate": 1.4956323125264193e-05,
+      "loss": 0.7685,
+      "step": 15064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7857663272718218,
+      "learning_rate": 1.4955660384917218e-05,
+      "loss": 0.7102,
+      "step": 15065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3009273922539517,
+      "learning_rate": 1.495499761571682e-05,
+      "loss": 1.1125,
+      "step": 15066
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.718174956485731,
+      "learning_rate": 1.4954334817666855e-05,
+      "loss": 1.0085,
+      "step": 15067
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.644902788411928,
+      "learning_rate": 1.4953671990771189e-05,
+      "loss": 0.8263,
+      "step": 15068
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7427310781184342,
+      "learning_rate": 1.4953009135033671e-05,
+      "loss": 0.879,
+      "step": 15069
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5716551863846246,
+      "learning_rate": 1.495234625045817e-05,
+      "loss": 0.8244,
+      "step": 15070
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9771115357180897,
+      "learning_rate": 1.4951683337048536e-05,
+      "loss": 0.8948,
+      "step": 15071
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.068602535739321,
+      "learning_rate": 1.495102039480864e-05,
+      "loss": 0.8283,
+      "step": 15072
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5821471197801316,
+      "learning_rate": 1.495035742374233e-05,
+      "loss": 0.8249,
+      "step": 15073
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6236270394532941,
+      "learning_rate": 1.4949694423853478e-05,
+      "loss": 0.9381,
+      "step": 15074
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7311192727860014,
+      "learning_rate": 1.4949031395145932e-05,
+      "loss": 0.8985,
+      "step": 15075
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.797735359206613,
+      "learning_rate": 1.4948368337623562e-05,
+      "loss": 0.877,
+      "step": 15076
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8586243833286162,
+      "learning_rate": 1.4947705251290223e-05,
+      "loss": 0.9576,
+      "step": 15077
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4793955844753839,
+      "learning_rate": 1.4947042136149778e-05,
+      "loss": 0.8947,
+      "step": 15078
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7328149349870967,
+      "learning_rate": 1.494637899220609e-05,
+      "loss": 0.8995,
+      "step": 15079
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6108410479115278,
+      "learning_rate": 1.4945715819463013e-05,
+      "loss": 0.9198,
+      "step": 15080
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1335756417029206,
+      "learning_rate": 1.4945052617924415e-05,
+      "loss": 0.9335,
+      "step": 15081
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8531344554938918,
+      "learning_rate": 1.4944389387594156e-05,
+      "loss": 0.9029,
+      "step": 15082
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.768290486502422,
+      "learning_rate": 1.4943726128476097e-05,
+      "loss": 0.8774,
+      "step": 15083
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6497480859273288,
+      "learning_rate": 1.4943062840574096e-05,
+      "loss": 0.9112,
+      "step": 15084
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6155936186416497,
+      "learning_rate": 1.4942399523892021e-05,
+      "loss": 0.9332,
+      "step": 15085
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6620546177763054,
+      "learning_rate": 1.494173617843373e-05,
+      "loss": 0.8461,
+      "step": 15086
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5774676441868236,
+      "learning_rate": 1.4941072804203089e-05,
+      "loss": 0.9052,
+      "step": 15087
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.539657584928778,
+      "learning_rate": 1.4940409401203955e-05,
+      "loss": 0.9777,
+      "step": 15088
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9192906901837383,
+      "learning_rate": 1.4939745969440194e-05,
+      "loss": 0.8084,
+      "step": 15089
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7754743841633146,
+      "learning_rate": 1.4939082508915668e-05,
+      "loss": 0.7239,
+      "step": 15090
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7663542783738942,
+      "learning_rate": 1.4938419019634243e-05,
+      "loss": 0.7202,
+      "step": 15091
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7818210661110832,
+      "learning_rate": 1.4937755501599774e-05,
+      "loss": 0.9094,
+      "step": 15092
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7531525654473202,
+      "learning_rate": 1.4937091954816131e-05,
+      "loss": 0.7665,
+      "step": 15093
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6639871486135702,
+      "learning_rate": 1.4936428379287176e-05,
+      "loss": 0.87,
+      "step": 15094
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6299620243266766,
+      "learning_rate": 1.4935764775016775e-05,
+      "loss": 0.8618,
+      "step": 15095
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7448346833960469,
+      "learning_rate": 1.4935101142008783e-05,
+      "loss": 0.8085,
+      "step": 15096
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.693813189741158,
+      "learning_rate": 1.4934437480267075e-05,
+      "loss": 0.9978,
+      "step": 15097
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8772923367267422,
+      "learning_rate": 1.4933773789795508e-05,
+      "loss": 0.9213,
+      "step": 15098
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7822087266271207,
+      "learning_rate": 1.4933110070597945e-05,
+      "loss": 0.8826,
+      "step": 15099
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5517713333365168,
+      "learning_rate": 1.4932446322678255e-05,
+      "loss": 0.8496,
+      "step": 15100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.509618580043183,
+      "learning_rate": 1.4931782546040302e-05,
+      "loss": 0.9779,
+      "step": 15101
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9975898322257626,
+      "learning_rate": 1.4931118740687952e-05,
+      "loss": 0.8793,
+      "step": 15102
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.016077938316293,
+      "learning_rate": 1.4930454906625062e-05,
+      "loss": 0.8878,
+      "step": 15103
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5537454518024818,
+      "learning_rate": 1.4929791043855506e-05,
+      "loss": 1.0042,
+      "step": 15104
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6558380250867422,
+      "learning_rate": 1.4929127152383148e-05,
+      "loss": 0.9783,
+      "step": 15105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.150582223823744,
+      "learning_rate": 1.4928463232211848e-05,
+      "loss": 0.9293,
+      "step": 15106
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6123169483201527,
+      "learning_rate": 1.4927799283345474e-05,
+      "loss": 0.8759,
+      "step": 15107
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5519670729237809,
+      "learning_rate": 1.4927135305787895e-05,
+      "loss": 0.9021,
+      "step": 15108
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.678630347606631,
+      "learning_rate": 1.492647129954297e-05,
+      "loss": 0.9265,
+      "step": 15109
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7781354704939953,
+      "learning_rate": 1.4925807264614573e-05,
+      "loss": 0.9,
+      "step": 15110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.09555099335694,
+      "learning_rate": 1.4925143201006566e-05,
+      "loss": 0.8654,
+      "step": 15111
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6549762473458,
+      "learning_rate": 1.4924479108722816e-05,
+      "loss": 0.9933,
+      "step": 15112
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7246889670493133,
+      "learning_rate": 1.4923814987767191e-05,
+      "loss": 0.9097,
+      "step": 15113
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5145998333019803,
+      "learning_rate": 1.4923150838143555e-05,
+      "loss": 0.9011,
+      "step": 15114
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9000821857817098,
+      "learning_rate": 1.4922486659855774e-05,
+      "loss": 1.0359,
+      "step": 15115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7723374890425154,
+      "learning_rate": 1.492182245290772e-05,
+      "loss": 0.8295,
+      "step": 15116
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.067858697888986,
+      "learning_rate": 1.4921158217303257e-05,
+      "loss": 0.8365,
+      "step": 15117
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2857758088811355,
+      "learning_rate": 1.4920493953046254e-05,
+      "loss": 0.8988,
+      "step": 15118
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5639134587706753,
+      "learning_rate": 1.4919829660140577e-05,
+      "loss": 0.8718,
+      "step": 15119
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6810696016032112,
+      "learning_rate": 1.491916533859009e-05,
+      "loss": 0.905,
+      "step": 15120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7563392215179698,
+      "learning_rate": 1.491850098839867e-05,
+      "loss": 0.8576,
+      "step": 15121
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4975602834931125,
+      "learning_rate": 1.4917836609570176e-05,
+      "loss": 0.8675,
+      "step": 15122
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8598895488966626,
+      "learning_rate": 1.4917172202108485e-05,
+      "loss": 0.7698,
+      "step": 15123
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.505227616882281,
+      "learning_rate": 1.4916507766017455e-05,
+      "loss": 0.8043,
+      "step": 15124
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.074484457489421,
+      "learning_rate": 1.4915843301300968e-05,
+      "loss": 0.8568,
+      "step": 15125
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9583686070348572,
+      "learning_rate": 1.491517880796288e-05,
+      "loss": 0.8099,
+      "step": 15126
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.622707952079682,
+      "learning_rate": 1.4914514286007064e-05,
+      "loss": 0.8636,
+      "step": 15127
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.733842776880295,
+      "learning_rate": 1.4913849735437392e-05,
+      "loss": 0.8929,
+      "step": 15128
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7500730022595599,
+      "learning_rate": 1.491318515625773e-05,
+      "loss": 0.945,
+      "step": 15129
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.120285878392847,
+      "learning_rate": 1.491252054847195e-05,
+      "loss": 0.8839,
+      "step": 15130
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6497542943048609,
+      "learning_rate": 1.4911855912083916e-05,
+      "loss": 0.8714,
+      "step": 15131
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7871345565612669,
+      "learning_rate": 1.4911191247097504e-05,
+      "loss": 0.7284,
+      "step": 15132
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4434087202289239,
+      "learning_rate": 1.4910526553516585e-05,
+      "loss": 0.9152,
+      "step": 15133
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1581121014612816,
+      "learning_rate": 1.4909861831345022e-05,
+      "loss": 0.8449,
+      "step": 15134
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7905286217303433,
+      "learning_rate": 1.490919708058669e-05,
+      "loss": 0.8385,
+      "step": 15135
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.791871675264175,
+      "learning_rate": 1.4908532301245459e-05,
+      "loss": 0.9486,
+      "step": 15136
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7790399732422584,
+      "learning_rate": 1.4907867493325199e-05,
+      "loss": 0.8881,
+      "step": 15137
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6868238441092613,
+      "learning_rate": 1.490720265682978e-05,
+      "loss": 0.8806,
+      "step": 15138
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2520965278782765,
+      "learning_rate": 1.4906537791763073e-05,
+      "loss": 0.9006,
+      "step": 15139
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0230121767912785,
+      "learning_rate": 1.490587289812895e-05,
+      "loss": 0.8123,
+      "step": 15140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8021706430017468,
+      "learning_rate": 1.4905207975931282e-05,
+      "loss": 0.8024,
+      "step": 15141
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5622591888856052,
+      "learning_rate": 1.4904543025173943e-05,
+      "loss": 0.9535,
+      "step": 15142
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7441793735624906,
+      "learning_rate": 1.4903878045860798e-05,
+      "loss": 0.8866,
+      "step": 15143
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5826288418415462,
+      "learning_rate": 1.4903213037995725e-05,
+      "loss": 0.9514,
+      "step": 15144
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3595696599921019,
+      "learning_rate": 1.4902548001582591e-05,
+      "loss": 0.9202,
+      "step": 15145
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6061813330726138,
+      "learning_rate": 1.4901882936625272e-05,
+      "loss": 0.9654,
+      "step": 15146
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.578556135105932,
+      "learning_rate": 1.4901217843127636e-05,
+      "loss": 0.9142,
+      "step": 15147
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9682686711145732,
+      "learning_rate": 1.4900552721093559e-05,
+      "loss": 0.8248,
+      "step": 15148
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.570050020964227,
+      "learning_rate": 1.4899887570526916e-05,
+      "loss": 0.8352,
+      "step": 15149
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4869734233713903,
+      "learning_rate": 1.4899222391431572e-05,
+      "loss": 0.7983,
+      "step": 15150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5826925775760614,
+      "learning_rate": 1.4898557183811409e-05,
+      "loss": 0.8839,
+      "step": 15151
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.681215119841362,
+      "learning_rate": 1.489789194767029e-05,
+      "loss": 0.9456,
+      "step": 15152
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.486260406341878,
+      "learning_rate": 1.4897226683012096e-05,
+      "loss": 0.8074,
+      "step": 15153
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6351252469744892,
+      "learning_rate": 1.4896561389840696e-05,
+      "loss": 0.7554,
+      "step": 15154
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7162121067016547,
+      "learning_rate": 1.4895896068159967e-05,
+      "loss": 0.9856,
+      "step": 15155
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.543712223081444,
+      "learning_rate": 1.4895230717973783e-05,
+      "loss": 0.9421,
+      "step": 15156
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5864052732854257,
+      "learning_rate": 1.4894565339286014e-05,
+      "loss": 0.8206,
+      "step": 15157
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5627127203645506,
+      "learning_rate": 1.4893899932100534e-05,
+      "loss": 0.9969,
+      "step": 15158
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7924810054486962,
+      "learning_rate": 1.4893234496421222e-05,
+      "loss": 0.7426,
+      "step": 15159
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6247274793271436,
+      "learning_rate": 1.4892569032251949e-05,
+      "loss": 0.9548,
+      "step": 15160
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6209736631981553,
+      "learning_rate": 1.4891903539596588e-05,
+      "loss": 0.9306,
+      "step": 15161
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6848101175185004,
+      "learning_rate": 1.489123801845902e-05,
+      "loss": 0.9955,
+      "step": 15162
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8020521113962521,
+      "learning_rate": 1.4890572468843112e-05,
+      "loss": 0.6854,
+      "step": 15163
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6058465840684928,
+      "learning_rate": 1.4889906890752745e-05,
+      "loss": 0.7073,
+      "step": 15164
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.634880272080761,
+      "learning_rate": 1.488924128419179e-05,
+      "loss": 0.9831,
+      "step": 15165
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.875202662472723,
+      "learning_rate": 1.4888575649164125e-05,
+      "loss": 0.8613,
+      "step": 15166
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6054762031193357,
+      "learning_rate": 1.4887909985673626e-05,
+      "loss": 0.9118,
+      "step": 15167
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7876132868494625,
+      "learning_rate": 1.4887244293724167e-05,
+      "loss": 0.7551,
+      "step": 15168
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.092195284189093,
+      "learning_rate": 1.4886578573319623e-05,
+      "loss": 0.8955,
+      "step": 15169
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7528442822407386,
+      "learning_rate": 1.4885912824463875e-05,
+      "loss": 0.8504,
+      "step": 15170
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7734488506278179,
+      "learning_rate": 1.4885247047160792e-05,
+      "loss": 0.8964,
+      "step": 15171
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6122990204818655,
+      "learning_rate": 1.4884581241414257e-05,
+      "loss": 0.9462,
+      "step": 15172
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.921561972088282,
+      "learning_rate": 1.488391540722814e-05,
+      "loss": 0.8837,
+      "step": 15173
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.95516945510566,
+      "learning_rate": 1.4883249544606326e-05,
+      "loss": 0.899,
+      "step": 15174
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8886217169637516,
+      "learning_rate": 1.4882583653552683e-05,
+      "loss": 0.9562,
+      "step": 15175
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7632772062747497,
+      "learning_rate": 1.4881917734071094e-05,
+      "loss": 0.6947,
+      "step": 15176
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4370913426786174,
+      "learning_rate": 1.4881251786165436e-05,
+      "loss": 0.8909,
+      "step": 15177
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7402183369283892,
+      "learning_rate": 1.488058580983958e-05,
+      "loss": 0.8082,
+      "step": 15178
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.1757174233407595,
+      "learning_rate": 1.4879919805097412e-05,
+      "loss": 0.8449,
+      "step": 15179
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7123868412374463,
+      "learning_rate": 1.4879253771942805e-05,
+      "loss": 0.8817,
+      "step": 15180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6177057300129019,
+      "learning_rate": 1.4878587710379639e-05,
+      "loss": 0.9852,
+      "step": 15181
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.690445989859892,
+      "learning_rate": 1.4877921620411789e-05,
+      "loss": 0.8323,
+      "step": 15182
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.560395740701512,
+      "learning_rate": 1.4877255502043136e-05,
+      "loss": 0.932,
+      "step": 15183
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.011752967621589,
+      "learning_rate": 1.4876589355277555e-05,
+      "loss": 0.9766,
+      "step": 15184
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.910866445967978,
+      "learning_rate": 1.487592318011893e-05,
+      "loss": 0.9049,
+      "step": 15185
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.507038570763582,
+      "learning_rate": 1.4875256976571135e-05,
+      "loss": 0.9451,
+      "step": 15186
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7195124536199287,
+      "learning_rate": 1.487459074463805e-05,
+      "loss": 0.8472,
+      "step": 15187
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.621879356219424,
+      "learning_rate": 1.4873924484323557e-05,
+      "loss": 0.8353,
+      "step": 15188
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8012733395636785,
+      "learning_rate": 1.487325819563153e-05,
+      "loss": 0.764,
+      "step": 15189
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.645630129242604,
+      "learning_rate": 1.4872591878565852e-05,
+      "loss": 0.8855,
+      "step": 15190
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.164564487904711,
+      "learning_rate": 1.4871925533130402e-05,
+      "loss": 0.8539,
+      "step": 15191
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5594370704780742,
+      "learning_rate": 1.4871259159329056e-05,
+      "loss": 0.7881,
+      "step": 15192
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9071937547439997,
+      "learning_rate": 1.48705927571657e-05,
+      "loss": 0.8588,
+      "step": 15193
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6683990319571358,
+      "learning_rate": 1.4869926326644208e-05,
+      "loss": 0.7805,
+      "step": 15194
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7683644903099495,
+      "learning_rate": 1.486925986776847e-05,
+      "loss": 0.7346,
+      "step": 15195
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8085289490637937,
+      "learning_rate": 1.4868593380542351e-05,
+      "loss": 0.8782,
+      "step": 15196
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7972372929012869,
+      "learning_rate": 1.4867926864969744e-05,
+      "loss": 0.7507,
+      "step": 15197
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.738890948461524,
+      "learning_rate": 1.4867260321054526e-05,
+      "loss": 0.7163,
+      "step": 15198
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6296891009103955,
+      "learning_rate": 1.4866593748800578e-05,
+      "loss": 0.9277,
+      "step": 15199
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5393443873252288,
+      "learning_rate": 1.4865927148211779e-05,
+      "loss": 0.8861,
+      "step": 15200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6873042376439051,
+      "learning_rate": 1.486526051929201e-05,
+      "loss": 0.9393,
+      "step": 15201
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5394078678211047,
+      "learning_rate": 1.4864593862045159e-05,
+      "loss": 0.7953,
+      "step": 15202
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4875211642300772,
+      "learning_rate": 1.4863927176475097e-05,
+      "loss": 0.7966,
+      "step": 15203
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4893547166502654,
+      "learning_rate": 1.4863260462585715e-05,
+      "loss": 0.9854,
+      "step": 15204
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9824164417208832,
+      "learning_rate": 1.4862593720380886e-05,
+      "loss": 0.9339,
+      "step": 15205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4666026353443185,
+      "learning_rate": 1.4861926949864502e-05,
+      "loss": 0.9208,
+      "step": 15206
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8770376470419456,
+      "learning_rate": 1.486126015104044e-05,
+      "loss": 0.7999,
+      "step": 15207
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.6609703339319575,
+      "learning_rate": 1.486059332391258e-05,
+      "loss": 1.015,
+      "step": 15208
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7419809826234336,
+      "learning_rate": 1.4859926468484807e-05,
+      "loss": 0.9898,
+      "step": 15209
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8586636844694568,
+      "learning_rate": 1.4859259584761002e-05,
+      "loss": 0.7885,
+      "step": 15210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7389200920061967,
+      "learning_rate": 1.4858592672745053e-05,
+      "loss": 0.9671,
+      "step": 15211
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.799408539634994,
+      "learning_rate": 1.4857925732440839e-05,
+      "loss": 0.8737,
+      "step": 15212
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.837548602278294,
+      "learning_rate": 1.4857258763852241e-05,
+      "loss": 1.0333,
+      "step": 15213
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.654185498863667,
+      "learning_rate": 1.4856591766983145e-05,
+      "loss": 0.9376,
+      "step": 15214
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7001803532945556,
+      "learning_rate": 1.4855924741837435e-05,
+      "loss": 0.8615,
+      "step": 15215
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9106484580846552,
+      "learning_rate": 1.4855257688418991e-05,
+      "loss": 0.8872,
+      "step": 15216
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9488955815857876,
+      "learning_rate": 1.4854590606731702e-05,
+      "loss": 0.9557,
+      "step": 15217
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6903089772076596,
+      "learning_rate": 1.485392349677945e-05,
+      "loss": 0.8936,
+      "step": 15218
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6895577321163961,
+      "learning_rate": 1.485325635856612e-05,
+      "loss": 0.8689,
+      "step": 15219
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5618624413518063,
+      "learning_rate": 1.4852589192095592e-05,
+      "loss": 0.8213,
+      "step": 15220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6580694358703372,
+      "learning_rate": 1.4851921997371756e-05,
+      "loss": 0.9214,
+      "step": 15221
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.958868011421843,
+      "learning_rate": 1.4851254774398492e-05,
+      "loss": 0.9764,
+      "step": 15222
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6327908802326552,
+      "learning_rate": 1.4850587523179689e-05,
+      "loss": 0.8977,
+      "step": 15223
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6786813384531794,
+      "learning_rate": 1.4849920243719227e-05,
+      "loss": 0.9447,
+      "step": 15224
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.036566082264147,
+      "learning_rate": 1.4849252936020998e-05,
+      "loss": 0.9285,
+      "step": 15225
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6640576898737292,
+      "learning_rate": 1.4848585600088876e-05,
+      "loss": 0.8609,
+      "step": 15226
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8132565152654296,
+      "learning_rate": 1.484791823592676e-05,
+      "loss": 0.8775,
+      "step": 15227
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6960004125370587,
+      "learning_rate": 1.4847250843538527e-05,
+      "loss": 0.8003,
+      "step": 15228
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8510801506840482,
+      "learning_rate": 1.4846583422928067e-05,
+      "loss": 0.94,
+      "step": 15229
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8215064712402755,
+      "learning_rate": 1.484591597409926e-05,
+      "loss": 0.8498,
+      "step": 15230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7731434714256875,
+      "learning_rate": 1.4845248497055997e-05,
+      "loss": 0.7864,
+      "step": 15231
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7539794993098912,
+      "learning_rate": 1.4844580991802166e-05,
+      "loss": 0.8603,
+      "step": 15232
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7420476687743982,
+      "learning_rate": 1.4843913458341646e-05,
+      "loss": 0.9642,
+      "step": 15233
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7632830627485099,
+      "learning_rate": 1.4843245896678333e-05,
+      "loss": 0.6991,
+      "step": 15234
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5781990284681218,
+      "learning_rate": 1.4842578306816108e-05,
+      "loss": 0.8927,
+      "step": 15235
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6509629713926741,
+      "learning_rate": 1.4841910688758858e-05,
+      "loss": 0.9353,
+      "step": 15236
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.413095507205453,
+      "learning_rate": 1.4841243042510472e-05,
+      "loss": 0.9308,
+      "step": 15237
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5879928928665097,
+      "learning_rate": 1.4840575368074835e-05,
+      "loss": 0.8736,
+      "step": 15238
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1084151627344108,
+      "learning_rate": 1.4839907665455837e-05,
+      "loss": 1.021,
+      "step": 15239
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7282829025462312,
+      "learning_rate": 1.4839239934657364e-05,
+      "loss": 0.8458,
+      "step": 15240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4857792812193558,
+      "learning_rate": 1.4838572175683304e-05,
+      "loss": 0.7738,
+      "step": 15241
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5506682329605013,
+      "learning_rate": 1.4837904388537545e-05,
+      "loss": 0.859,
+      "step": 15242
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.690842001792969,
+      "learning_rate": 1.4837236573223973e-05,
+      "loss": 0.9319,
+      "step": 15243
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0573640973884575,
+      "learning_rate": 1.483656872974648e-05,
+      "loss": 0.78,
+      "step": 15244
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1227828105909556,
+      "learning_rate": 1.4835900858108951e-05,
+      "loss": 0.98,
+      "step": 15245
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.545360959732883,
+      "learning_rate": 1.4835232958315278e-05,
+      "loss": 0.9152,
+      "step": 15246
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.513934047378893,
+      "learning_rate": 1.4834565030369348e-05,
+      "loss": 0.9279,
+      "step": 15247
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5544550419163674,
+      "learning_rate": 1.4833897074275049e-05,
+      "loss": 0.9219,
+      "step": 15248
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7819209886548173,
+      "learning_rate": 1.4833229090036273e-05,
+      "loss": 1.0559,
+      "step": 15249
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6674859359443535,
+      "learning_rate": 1.4832561077656905e-05,
+      "loss": 0.8939,
+      "step": 15250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.553712186162484,
+      "learning_rate": 1.4831893037140836e-05,
+      "loss": 0.9833,
+      "step": 15251
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0702324056511405,
+      "learning_rate": 1.4831224968491955e-05,
+      "loss": 0.8704,
+      "step": 15252
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.706915504716248,
+      "learning_rate": 1.4830556871714157e-05,
+      "loss": 0.8436,
+      "step": 15253
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7315565200682763,
+      "learning_rate": 1.4829888746811323e-05,
+      "loss": 0.8152,
+      "step": 15254
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4610073069057208,
+      "learning_rate": 1.482922059378735e-05,
+      "loss": 0.9109,
+      "step": 15255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.52997672292505,
+      "learning_rate": 1.4828552412646126e-05,
+      "loss": 0.9228,
+      "step": 15256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6221394878391715,
+      "learning_rate": 1.482788420339154e-05,
+      "loss": 0.9027,
+      "step": 15257
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5376803715870755,
+      "learning_rate": 1.4827215966027483e-05,
+      "loss": 0.9387,
+      "step": 15258
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5563184582094594,
+      "learning_rate": 1.482654770055785e-05,
+      "loss": 0.8708,
+      "step": 15259
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6434746957305397,
+      "learning_rate": 1.4825879406986525e-05,
+      "loss": 0.9038,
+      "step": 15260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4459842591552718,
+      "learning_rate": 1.4825211085317404e-05,
+      "loss": 0.9067,
+      "step": 15261
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8839295544005137,
+      "learning_rate": 1.4824542735554375e-05,
+      "loss": 0.9031,
+      "step": 15262
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5793500751274747,
+      "learning_rate": 1.482387435770133e-05,
+      "loss": 0.9283,
+      "step": 15263
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5056939040695188,
+      "learning_rate": 1.4823205951762164e-05,
+      "loss": 0.9255,
+      "step": 15264
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4476740283544756,
+      "learning_rate": 1.4822537517740763e-05,
+      "loss": 0.9157,
+      "step": 15265
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3145805777367507,
+      "learning_rate": 1.4821869055641023e-05,
+      "loss": 0.8994,
+      "step": 15266
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6268509542609633,
+      "learning_rate": 1.4821200565466835e-05,
+      "loss": 0.9556,
+      "step": 15267
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1437942275727884,
+      "learning_rate": 1.4820532047222091e-05,
+      "loss": 0.9381,
+      "step": 15268
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5592302520516386,
+      "learning_rate": 1.4819863500910683e-05,
+      "loss": 0.9274,
+      "step": 15269
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.592690087323082,
+      "learning_rate": 1.4819194926536504e-05,
+      "loss": 0.8409,
+      "step": 15270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6114108689874749,
+      "learning_rate": 1.4818526324103445e-05,
+      "loss": 0.8024,
+      "step": 15271
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2622411983607575,
+      "learning_rate": 1.4817857693615406e-05,
+      "loss": 1.0481,
+      "step": 15272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.144141135428692,
+      "learning_rate": 1.4817189035076265e-05,
+      "loss": 0.8203,
+      "step": 15273
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8023955804792975,
+      "learning_rate": 1.4816520348489932e-05,
+      "loss": 0.7454,
+      "step": 15274
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0215021608037853,
+      "learning_rate": 1.4815851633860288e-05,
+      "loss": 0.8816,
+      "step": 15275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6797489888627337,
+      "learning_rate": 1.4815182891191234e-05,
+      "loss": 0.8287,
+      "step": 15276
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6441622343626408,
+      "learning_rate": 1.4814514120486657e-05,
+      "loss": 0.9804,
+      "step": 15277
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8643873656253915,
+      "learning_rate": 1.481384532175046e-05,
+      "loss": 0.8759,
+      "step": 15278
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8286392719213211,
+      "learning_rate": 1.481317649498653e-05,
+      "loss": 0.8678,
+      "step": 15279
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5431593996353998,
+      "learning_rate": 1.481250764019876e-05,
+      "loss": 1.0343,
+      "step": 15280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1140777452982675,
+      "learning_rate": 1.4811838757391049e-05,
+      "loss": 0.9781,
+      "step": 15281
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6230627961581434,
+      "learning_rate": 1.4811169846567286e-05,
+      "loss": 0.9137,
+      "step": 15282
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.698077174872185,
+      "learning_rate": 1.4810500907731374e-05,
+      "loss": 0.7777,
+      "step": 15283
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.799667063166338,
+      "learning_rate": 1.4809831940887196e-05,
+      "loss": 0.9145,
+      "step": 15284
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4694563892984747,
+      "learning_rate": 1.4809162946038659e-05,
+      "loss": 0.8127,
+      "step": 15285
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6641578955839522,
+      "learning_rate": 1.4808493923189652e-05,
+      "loss": 0.9105,
+      "step": 15286
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7681261276304234,
+      "learning_rate": 1.4807824872344072e-05,
+      "loss": 0.9122,
+      "step": 15287
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7000945700995367,
+      "learning_rate": 1.4807155793505811e-05,
+      "loss": 0.91,
+      "step": 15288
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5170740334540582,
+      "learning_rate": 1.480648668667877e-05,
+      "loss": 0.7871,
+      "step": 15289
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4575935273101193,
+      "learning_rate": 1.4805817551866839e-05,
+      "loss": 0.9897,
+      "step": 15290
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.778332766804997,
+      "learning_rate": 1.4805148389073918e-05,
+      "loss": 0.9064,
+      "step": 15291
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7382582307514285,
+      "learning_rate": 1.48044791983039e-05,
+      "loss": 0.8281,
+      "step": 15292
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7605459357908337,
+      "learning_rate": 1.4803809979560688e-05,
+      "loss": 0.9539,
+      "step": 15293
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5423629732369801,
+      "learning_rate": 1.480314073284817e-05,
+      "loss": 0.9253,
+      "step": 15294
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8315780455807145,
+      "learning_rate": 1.4802471458170245e-05,
+      "loss": 0.8279,
+      "step": 15295
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8777853362090686,
+      "learning_rate": 1.480180215553081e-05,
+      "loss": 1.0133,
+      "step": 15296
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6966452542355048,
+      "learning_rate": 1.4801132824933767e-05,
+      "loss": 0.8455,
+      "step": 15297
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6031666316473976,
+      "learning_rate": 1.4800463466383004e-05,
+      "loss": 0.947,
+      "step": 15298
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.573100963809534,
+      "learning_rate": 1.4799794079882426e-05,
+      "loss": 0.88,
+      "step": 15299
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7804603282877374,
+      "learning_rate": 1.4799124665435927e-05,
+      "loss": 0.9176,
+      "step": 15300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5397440563759286,
+      "learning_rate": 1.4798455223047401e-05,
+      "loss": 0.8656,
+      "step": 15301
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.557367304168527,
+      "learning_rate": 1.4797785752720755e-05,
+      "loss": 0.8942,
+      "step": 15302
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1766038167871047,
+      "learning_rate": 1.4797116254459878e-05,
+      "loss": 0.8648,
+      "step": 15303
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5560589697582772,
+      "learning_rate": 1.4796446728268674e-05,
+      "loss": 0.8761,
+      "step": 15304
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7408390594434582,
+      "learning_rate": 1.4795777174151033e-05,
+      "loss": 0.9819,
+      "step": 15305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6322206991911894,
+      "learning_rate": 1.4795107592110866e-05,
+      "loss": 0.9646,
+      "step": 15306
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5949481641622654,
+      "learning_rate": 1.4794437982152061e-05,
+      "loss": 0.7196,
+      "step": 15307
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7751508254175328,
+      "learning_rate": 1.4793768344278521e-05,
+      "loss": 1.0335,
+      "step": 15308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.72783444945655,
+      "learning_rate": 1.4793098678494145e-05,
+      "loss": 0.9046,
+      "step": 15309
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0841502218162784,
+      "learning_rate": 1.479242898480283e-05,
+      "loss": 0.9327,
+      "step": 15310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8922774135154399,
+      "learning_rate": 1.4791759263208474e-05,
+      "loss": 0.9837,
+      "step": 15311
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5565083576537817,
+      "learning_rate": 1.479108951371498e-05,
+      "loss": 0.8949,
+      "step": 15312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7156433774717799,
+      "learning_rate": 1.4790419736326248e-05,
+      "loss": 0.967,
+      "step": 15313
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7736351322563283,
+      "learning_rate": 1.4789749931046173e-05,
+      "loss": 0.7328,
+      "step": 15314
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7970990263860043,
+      "learning_rate": 1.4789080097878659e-05,
+      "loss": 0.8568,
+      "step": 15315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0956134000630597,
+      "learning_rate": 1.4788410236827605e-05,
+      "loss": 0.8581,
+      "step": 15316
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.502637633118364,
+      "learning_rate": 1.4787740347896908e-05,
+      "loss": 0.9187,
+      "step": 15317
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7414231299763185,
+      "learning_rate": 1.4787070431090474e-05,
+      "loss": 0.8563,
+      "step": 15318
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1072861956011253,
+      "learning_rate": 1.4786400486412198e-05,
+      "loss": 0.7683,
+      "step": 15319
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9055995331080788,
+      "learning_rate": 1.4785730513865983e-05,
+      "loss": 0.851,
+      "step": 15320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6676939600028384,
+      "learning_rate": 1.4785060513455732e-05,
+      "loss": 0.7845,
+      "step": 15321
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.684084216176167,
+      "learning_rate": 1.4784390485185342e-05,
+      "loss": 0.9465,
+      "step": 15322
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6983758879999322,
+      "learning_rate": 1.4783720429058719e-05,
+      "loss": 0.8641,
+      "step": 15323
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9249979469321556,
+      "learning_rate": 1.4783050345079755e-05,
+      "loss": 0.9785,
+      "step": 15324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0623872533357726,
+      "learning_rate": 1.4782380233252365e-05,
+      "loss": 0.7507,
+      "step": 15325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4482359625189969,
+      "learning_rate": 1.4781710093580437e-05,
+      "loss": 0.8176,
+      "step": 15326
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.788269864360806,
+      "learning_rate": 1.4781039926067884e-05,
+      "loss": 0.9633,
+      "step": 15327
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.784951015530248,
+      "learning_rate": 1.4780369730718595e-05,
+      "loss": 0.9115,
+      "step": 15328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.664693043694664,
+      "learning_rate": 1.477969950753649e-05,
+      "loss": 0.8221,
+      "step": 15329
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6808720846423413,
+      "learning_rate": 1.4779029256525457e-05,
+      "loss": 0.816,
+      "step": 15330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6172476258775443,
+      "learning_rate": 1.47783589776894e-05,
+      "loss": 0.8286,
+      "step": 15331
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5734462783620333,
+      "learning_rate": 1.4777688671032226e-05,
+      "loss": 0.9327,
+      "step": 15332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5363184003009331,
+      "learning_rate": 1.4777018336557835e-05,
+      "loss": 0.8695,
+      "step": 15333
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8170813370520549,
+      "learning_rate": 1.4776347974270135e-05,
+      "loss": 0.8737,
+      "step": 15334
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7473563412048971,
+      "learning_rate": 1.477567758417302e-05,
+      "loss": 0.8444,
+      "step": 15335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8600251540676878,
+      "learning_rate": 1.4775007166270402e-05,
+      "loss": 0.9565,
+      "step": 15336
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5757590186951487,
+      "learning_rate": 1.4774336720566178e-05,
+      "loss": 0.8416,
+      "step": 15337
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9758647825270526,
+      "learning_rate": 1.4773666247064255e-05,
+      "loss": 0.9162,
+      "step": 15338
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7613883735216975,
+      "learning_rate": 1.4772995745768538e-05,
+      "loss": 0.7626,
+      "step": 15339
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4596071242360815,
+      "learning_rate": 1.4772325216682926e-05,
+      "loss": 0.8921,
+      "step": 15340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0875957956518594,
+      "learning_rate": 1.4771654659811326e-05,
+      "loss": 0.9422,
+      "step": 15341
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0566744590747144,
+      "learning_rate": 1.4770984075157641e-05,
+      "loss": 0.9217,
+      "step": 15342
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4770273299365337,
+      "learning_rate": 1.4770313462725777e-05,
+      "loss": 0.8683,
+      "step": 15343
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5414875547797273,
+      "learning_rate": 1.476964282251964e-05,
+      "loss": 0.8852,
+      "step": 15344
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7238722421484234,
+      "learning_rate": 1.4768972154543129e-05,
+      "loss": 0.8534,
+      "step": 15345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5412792462221052,
+      "learning_rate": 1.4768301458800154e-05,
+      "loss": 0.835,
+      "step": 15346
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5439530270220576,
+      "learning_rate": 1.4767630735294618e-05,
+      "loss": 0.7355,
+      "step": 15347
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6201380579732578,
+      "learning_rate": 1.4766959984030425e-05,
+      "loss": 0.8657,
+      "step": 15348
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8146596378883626,
+      "learning_rate": 1.4766289205011483e-05,
+      "loss": 0.809,
+      "step": 15349
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8684171222063024,
+      "learning_rate": 1.4765618398241697e-05,
+      "loss": 0.9293,
+      "step": 15350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6163476888456458,
+      "learning_rate": 1.4764947563724974e-05,
+      "loss": 0.9019,
+      "step": 15351
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7597572265796644,
+      "learning_rate": 1.4764276701465212e-05,
+      "loss": 0.9123,
+      "step": 15352
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5714819865202077,
+      "learning_rate": 1.4763605811466328e-05,
+      "loss": 0.8892,
+      "step": 15353
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7093858505130535,
+      "learning_rate": 1.4762934893732218e-05,
+      "loss": 0.6902,
+      "step": 15354
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5243714850590306,
+      "learning_rate": 1.47622639482668e-05,
+      "loss": 0.8043,
+      "step": 15355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8562463536719738,
+      "learning_rate": 1.4761592975073965e-05,
+      "loss": 0.7232,
+      "step": 15356
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8981222732842221,
+      "learning_rate": 1.4760921974157636e-05,
+      "loss": 0.7321,
+      "step": 15357
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8887847124368502,
+      "learning_rate": 1.4760250945521708e-05,
+      "loss": 0.8299,
+      "step": 15358
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5519618934485777,
+      "learning_rate": 1.4759579889170093e-05,
+      "loss": 0.9559,
+      "step": 15359
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6295106380329805,
+      "learning_rate": 1.4758908805106696e-05,
+      "loss": 0.925,
+      "step": 15360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5809878357959681,
+      "learning_rate": 1.4758237693335424e-05,
+      "loss": 1.0382,
+      "step": 15361
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0084416485037075,
+      "learning_rate": 1.475756655386019e-05,
+      "loss": 0.9162,
+      "step": 15362
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6255458339676347,
+      "learning_rate": 1.4756895386684893e-05,
+      "loss": 0.8258,
+      "step": 15363
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7359284460003462,
+      "learning_rate": 1.4756224191813449e-05,
+      "loss": 0.7128,
+      "step": 15364
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4786400611125998,
+      "learning_rate": 1.4755552969249758e-05,
+      "loss": 1.0299,
+      "step": 15365
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.403465180851822,
+      "learning_rate": 1.4754881718997735e-05,
+      "loss": 0.9405,
+      "step": 15366
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7729035566704021,
+      "learning_rate": 1.4754210441061283e-05,
+      "loss": 0.8322,
+      "step": 15367
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8352914417879624,
+      "learning_rate": 1.4753539135444314e-05,
+      "loss": 0.899,
+      "step": 15368
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5193025504338569,
+      "learning_rate": 1.4752867802150735e-05,
+      "loss": 0.9518,
+      "step": 15369
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6988546001305056,
+      "learning_rate": 1.4752196441184454e-05,
+      "loss": 0.8623,
+      "step": 15370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7586818974971998,
+      "learning_rate": 1.4751525052549383e-05,
+      "loss": 0.8525,
+      "step": 15371
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9686297272654882,
+      "learning_rate": 1.475085363624943e-05,
+      "loss": 0.8104,
+      "step": 15372
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.733893422119452,
+      "learning_rate": 1.4750182192288497e-05,
+      "loss": 0.9102,
+      "step": 15373
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6856394844504228,
+      "learning_rate": 1.4749510720670506e-05,
+      "loss": 0.9284,
+      "step": 15374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9018078091205306,
+      "learning_rate": 1.4748839221399354e-05,
+      "loss": 0.9425,
+      "step": 15375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8755991552436444,
+      "learning_rate": 1.4748167694478963e-05,
+      "loss": 0.9592,
+      "step": 15376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5566724513113699,
+      "learning_rate": 1.4747496139913229e-05,
+      "loss": 0.7786,
+      "step": 15377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7314686102961758,
+      "learning_rate": 1.4746824557706076e-05,
+      "loss": 0.8972,
+      "step": 15378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8170779363848646,
+      "learning_rate": 1.4746152947861405e-05,
+      "loss": 0.7687,
+      "step": 15379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.114368429742621,
+      "learning_rate": 1.4745481310383127e-05,
+      "loss": 0.8609,
+      "step": 15380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7862086160232729,
+      "learning_rate": 1.4744809645275157e-05,
+      "loss": 0.8745,
+      "step": 15381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6099359072116035,
+      "learning_rate": 1.4744137952541402e-05,
+      "loss": 0.8392,
+      "step": 15382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7594134286609199,
+      "learning_rate": 1.4743466232185775e-05,
+      "loss": 0.7639,
+      "step": 15383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5450992350930148,
+      "learning_rate": 1.4742794484212184e-05,
+      "loss": 0.9977,
+      "step": 15384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5661116260381378,
+      "learning_rate": 1.4742122708624544e-05,
+      "loss": 0.923,
+      "step": 15385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.591289439931777,
+      "learning_rate": 1.4741450905426763e-05,
+      "loss": 0.7448,
+      "step": 15386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7656877290160522,
+      "learning_rate": 1.4740779074622756e-05,
+      "loss": 0.9004,
+      "step": 15387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.031012514392135,
+      "learning_rate": 1.4740107216216428e-05,
+      "loss": 0.7979,
+      "step": 15388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6434610153699407,
+      "learning_rate": 1.4739435330211697e-05,
+      "loss": 0.8757,
+      "step": 15389
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6346742368819283,
+      "learning_rate": 1.4738763416612476e-05,
+      "loss": 0.8825,
+      "step": 15390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7642598386634858,
+      "learning_rate": 1.473809147542267e-05,
+      "loss": 0.9284,
+      "step": 15391
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5586821523476766,
+      "learning_rate": 1.4737419506646197e-05,
+      "loss": 0.8527,
+      "step": 15392
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3328485187633596,
+      "learning_rate": 1.4736747510286968e-05,
+      "loss": 0.8966,
+      "step": 15393
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5729291205629927,
+      "learning_rate": 1.4736075486348896e-05,
+      "loss": 0.8757,
+      "step": 15394
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9966125714510143,
+      "learning_rate": 1.4735403434835891e-05,
+      "loss": 0.9929,
+      "step": 15395
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5180394618475535,
+      "learning_rate": 1.473473135575187e-05,
+      "loss": 0.9478,
+      "step": 15396
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.881837034262262,
+      "learning_rate": 1.4734059249100743e-05,
+      "loss": 0.8789,
+      "step": 15397
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.762113013686714,
+      "learning_rate": 1.4733387114886424e-05,
+      "loss": 0.8386,
+      "step": 15398
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6053072830130717,
+      "learning_rate": 1.4732714953112828e-05,
+      "loss": 0.8333,
+      "step": 15399
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.777262400030218,
+      "learning_rate": 1.4732042763783866e-05,
+      "loss": 0.7659,
+      "step": 15400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5813663916799257,
+      "learning_rate": 1.4731370546903454e-05,
+      "loss": 0.9041,
+      "step": 15401
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5827197025401922,
+      "learning_rate": 1.4730698302475505e-05,
+      "loss": 0.9516,
+      "step": 15402
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5740149412421085,
+      "learning_rate": 1.473002603050393e-05,
+      "loss": 0.8591,
+      "step": 15403
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.507018741698635,
+      "learning_rate": 1.472935373099265e-05,
+      "loss": 0.8694,
+      "step": 15404
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.646995771062379,
+      "learning_rate": 1.472868140394557e-05,
+      "loss": 0.8793,
+      "step": 15405
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7862392177118401,
+      "learning_rate": 1.4728009049366617e-05,
+      "loss": 0.7242,
+      "step": 15406
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5297829839797232,
+      "learning_rate": 1.472733666725969e-05,
+      "loss": 0.9377,
+      "step": 15407
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5885525458674528,
+      "learning_rate": 1.472666425762872e-05,
+      "loss": 0.912,
+      "step": 15408
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8048607961848522,
+      "learning_rate": 1.4725991820477612e-05,
+      "loss": 0.7791,
+      "step": 15409
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.061288091220482,
+      "learning_rate": 1.4725319355810282e-05,
+      "loss": 0.9001,
+      "step": 15410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1053698432652483,
+      "learning_rate": 1.4724646863630646e-05,
+      "loss": 0.8207,
+      "step": 15411
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7724081933696503,
+      "learning_rate": 1.472397434394262e-05,
+      "loss": 0.8848,
+      "step": 15412
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6368225247981298,
+      "learning_rate": 1.4723301796750122e-05,
+      "loss": 0.9404,
+      "step": 15413
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6820730365823493,
+      "learning_rate": 1.4722629222057063e-05,
+      "loss": 0.8795,
+      "step": 15414
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8861140821143023,
+      "learning_rate": 1.4721956619867364e-05,
+      "loss": 1.0234,
+      "step": 15415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6872209632415567,
+      "learning_rate": 1.4721283990184938e-05,
+      "loss": 0.9717,
+      "step": 15416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6400472471115695,
+      "learning_rate": 1.4720611333013701e-05,
+      "loss": 0.9743,
+      "step": 15417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.550329258035968,
+      "learning_rate": 1.471993864835757e-05,
+      "loss": 0.9247,
+      "step": 15418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5740097105074629,
+      "learning_rate": 1.4719265936220464e-05,
+      "loss": 0.8525,
+      "step": 15419
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6232588604428635,
+      "learning_rate": 1.4718593196606294e-05,
+      "loss": 0.8099,
+      "step": 15420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7388162222002825,
+      "learning_rate": 1.4717920429518984e-05,
+      "loss": 0.9436,
+      "step": 15421
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5806059818208245,
+      "learning_rate": 1.4717247634962446e-05,
+      "loss": 0.9123,
+      "step": 15422
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6771768434705725,
+      "learning_rate": 1.4716574812940603e-05,
+      "loss": 0.7845,
+      "step": 15423
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6016564668552489,
+      "learning_rate": 1.4715901963457362e-05,
+      "loss": 0.8775,
+      "step": 15424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7527565709915482,
+      "learning_rate": 1.4715229086516649e-05,
+      "loss": 0.9117,
+      "step": 15425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.800744923576378,
+      "learning_rate": 1.471455618212238e-05,
+      "loss": 0.96,
+      "step": 15426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5629787470101626,
+      "learning_rate": 1.4713883250278471e-05,
+      "loss": 0.7993,
+      "step": 15427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6270206011926467,
+      "learning_rate": 1.4713210290988842e-05,
+      "loss": 0.9234,
+      "step": 15428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7869895611550627,
+      "learning_rate": 1.471253730425741e-05,
+      "loss": 0.774,
+      "step": 15429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5156036247096123,
+      "learning_rate": 1.4711864290088099e-05,
+      "loss": 0.8349,
+      "step": 15430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7974917236662062,
+      "learning_rate": 1.4711191248484812e-05,
+      "loss": 0.8947,
+      "step": 15431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5444234013032667,
+      "learning_rate": 1.4710518179451487e-05,
+      "loss": 0.849,
+      "step": 15432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5706896332387483,
+      "learning_rate": 1.470984508299203e-05,
+      "loss": 0.8446,
+      "step": 15433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8543532227427169,
+      "learning_rate": 1.4709171959110367e-05,
+      "loss": 0.9955,
+      "step": 15434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4786748360241517,
+      "learning_rate": 1.470849880781041e-05,
+      "loss": 0.887,
+      "step": 15435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4569291919358145,
+      "learning_rate": 1.4707825629096085e-05,
+      "loss": 1.0116,
+      "step": 15436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7436023864303376,
+      "learning_rate": 1.4707152422971307e-05,
+      "loss": 0.8544,
+      "step": 15437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9312389519533244,
+      "learning_rate": 1.4706479189439998e-05,
+      "loss": 0.892,
+      "step": 15438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.628139440144977,
+      "learning_rate": 1.4705805928506076e-05,
+      "loss": 0.8041,
+      "step": 15439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7974834686457255,
+      "learning_rate": 1.4705132640173463e-05,
+      "loss": 0.9054,
+      "step": 15440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7635522390076699,
+      "learning_rate": 1.4704459324446078e-05,
+      "loss": 0.9424,
+      "step": 15441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7429692448539231,
+      "learning_rate": 1.4703785981327841e-05,
+      "loss": 0.681,
+      "step": 15442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7584438962286886,
+      "learning_rate": 1.4703112610822674e-05,
+      "loss": 0.6822,
+      "step": 15443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7462179505539174,
+      "learning_rate": 1.4702439212934496e-05,
+      "loss": 0.6939,
+      "step": 15444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9857075459180609,
+      "learning_rate": 1.4701765787667228e-05,
+      "loss": 0.9426,
+      "step": 15445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5930544174070949,
+      "learning_rate": 1.4701092335024792e-05,
+      "loss": 0.9976,
+      "step": 15446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5627314883169734,
+      "learning_rate": 1.4700418855011105e-05,
+      "loss": 0.8409,
+      "step": 15447
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4017474606088975,
+      "learning_rate": 1.4699745347630095e-05,
+      "loss": 0.8067,
+      "step": 15448
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.784998092568329,
+      "learning_rate": 1.4699071812885677e-05,
+      "loss": 0.8785,
+      "step": 15449
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7149411016142686,
+      "learning_rate": 1.4698398250781776e-05,
+      "loss": 0.917,
+      "step": 15450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.617595497072666,
+      "learning_rate": 1.4697724661322312e-05,
+      "loss": 0.9526,
+      "step": 15451
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6230207598582553,
+      "learning_rate": 1.469705104451121e-05,
+      "loss": 0.7709,
+      "step": 15452
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7654732189471537,
+      "learning_rate": 1.469637740035239e-05,
+      "loss": 0.7705,
+      "step": 15453
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7935117553766298,
+      "learning_rate": 1.4695703728849771e-05,
+      "loss": 0.7084,
+      "step": 15454
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.615758790472847,
+      "learning_rate": 1.4695030030007282e-05,
+      "loss": 0.9228,
+      "step": 15455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4872715735420161,
+      "learning_rate": 1.4694356303828837e-05,
+      "loss": 0.9529,
+      "step": 15456
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8342294956542562,
+      "learning_rate": 1.4693682550318368e-05,
+      "loss": 0.8718,
+      "step": 15457
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8357264266781885,
+      "learning_rate": 1.4693008769479788e-05,
+      "loss": 0.9187,
+      "step": 15458
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.595345759186658,
+      "learning_rate": 1.4692334961317033e-05,
+      "loss": 0.8809,
+      "step": 15459
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1708623658394277,
+      "learning_rate": 1.4691661125834012e-05,
+      "loss": 0.9151,
+      "step": 15460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7728285489060058,
+      "learning_rate": 1.4690987263034655e-05,
+      "loss": 0.8769,
+      "step": 15461
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7104810341815788,
+      "learning_rate": 1.4690313372922887e-05,
+      "loss": 0.9002,
+      "step": 15462
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.49991588799856,
+      "learning_rate": 1.4689639455502629e-05,
+      "loss": 0.8556,
+      "step": 15463
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5725192438546816,
+      "learning_rate": 1.4688965510777803e-05,
+      "loss": 0.9356,
+      "step": 15464
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.602686832309674,
+      "learning_rate": 1.468829153875234e-05,
+      "loss": 0.8945,
+      "step": 15465
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8022657278166416,
+      "learning_rate": 1.4687617539430154e-05,
+      "loss": 0.7166,
+      "step": 15466
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4798378882599104,
+      "learning_rate": 1.4686943512815179e-05,
+      "loss": 0.8627,
+      "step": 15467
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6246754612437544,
+      "learning_rate": 1.4686269458911333e-05,
+      "loss": 0.9546,
+      "step": 15468
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5678993621720345,
+      "learning_rate": 1.4685595377722544e-05,
+      "loss": 0.8461,
+      "step": 15469
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6335345513643615,
+      "learning_rate": 1.4684921269252734e-05,
+      "loss": 0.9001,
+      "step": 15470
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1226873759688396,
+      "learning_rate": 1.468424713350583e-05,
+      "loss": 0.778,
+      "step": 15471
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7163430808375089,
+      "learning_rate": 1.4683572970485756e-05,
+      "loss": 0.8364,
+      "step": 15472
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6775372743846866,
+      "learning_rate": 1.4682898780196434e-05,
+      "loss": 0.8463,
+      "step": 15473
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5893806543237161,
+      "learning_rate": 1.4682224562641798e-05,
+      "loss": 0.7841,
+      "step": 15474
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7820209891672565,
+      "learning_rate": 1.4681550317825766e-05,
+      "loss": 0.7214,
+      "step": 15475
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.751847396775292,
+      "learning_rate": 1.4680876045752266e-05,
+      "loss": 0.8708,
+      "step": 15476
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.002628846761608,
+      "learning_rate": 1.4680201746425221e-05,
+      "loss": 0.9128,
+      "step": 15477
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6792751991221975,
+      "learning_rate": 1.4679527419848563e-05,
+      "loss": 0.9553,
+      "step": 15478
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.57933630010265,
+      "learning_rate": 1.4678853066026216e-05,
+      "loss": 0.9056,
+      "step": 15479
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7845995934926222,
+      "learning_rate": 1.4678178684962101e-05,
+      "loss": 1.0435,
+      "step": 15480
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4375640973533181,
+      "learning_rate": 1.4677504276660153e-05,
+      "loss": 0.7613,
+      "step": 15481
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5445851279500782,
+      "learning_rate": 1.4676829841124288e-05,
+      "loss": 0.9056,
+      "step": 15482
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.660835014702021,
+      "learning_rate": 1.4676155378358447e-05,
+      "loss": 0.9659,
+      "step": 15483
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3762991926013595,
+      "learning_rate": 1.4675480888366545e-05,
+      "loss": 0.8358,
+      "step": 15484
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.770128614961235,
+      "learning_rate": 1.4674806371152514e-05,
+      "loss": 0.934,
+      "step": 15485
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7572252823350479,
+      "learning_rate": 1.4674131826720278e-05,
+      "loss": 0.77,
+      "step": 15486
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.616342582047942,
+      "learning_rate": 1.4673457255073769e-05,
+      "loss": 0.9704,
+      "step": 15487
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.588574004603435,
+      "learning_rate": 1.4672782656216911e-05,
+      "loss": 0.9475,
+      "step": 15488
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5403600607402979,
+      "learning_rate": 1.4672108030153634e-05,
+      "loss": 0.8684,
+      "step": 15489
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4763437406223194,
+      "learning_rate": 1.4671433376887864e-05,
+      "loss": 0.948,
+      "step": 15490
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5625129099960748,
+      "learning_rate": 1.467075869642353e-05,
+      "loss": 0.8863,
+      "step": 15491
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.763337472973373,
+      "learning_rate": 1.4670083988764559e-05,
+      "loss": 0.9371,
+      "step": 15492
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6373560890860883,
+      "learning_rate": 1.4669409253914882e-05,
+      "loss": 0.8351,
+      "step": 15493
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5978308091156745,
+      "learning_rate": 1.4668734491878426e-05,
+      "loss": 0.9036,
+      "step": 15494
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6816445703071132,
+      "learning_rate": 1.4668059702659121e-05,
+      "loss": 1.0371,
+      "step": 15495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8260024814528872,
+      "learning_rate": 1.4667384886260892e-05,
+      "loss": 0.9089,
+      "step": 15496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7780794846232681,
+      "learning_rate": 1.466671004268767e-05,
+      "loss": 0.9249,
+      "step": 15497
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5147918860594398,
+      "learning_rate": 1.4666035171943388e-05,
+      "loss": 0.8793,
+      "step": 15498
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7432629190424198,
+      "learning_rate": 1.466536027403197e-05,
+      "loss": 0.8838,
+      "step": 15499
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.328920283850334,
+      "learning_rate": 1.4664685348957348e-05,
+      "loss": 0.858,
+      "step": 15500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5701498934461433,
+      "learning_rate": 1.4664010396723449e-05,
+      "loss": 0.8812,
+      "step": 15501
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.618511698223746,
+      "learning_rate": 1.466333541733421e-05,
+      "loss": 0.9503,
+      "step": 15502
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.786423907303789,
+      "learning_rate": 1.466266041079355e-05,
+      "loss": 0.7444,
+      "step": 15503
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6381008780741186,
+      "learning_rate": 1.466198537710541e-05,
+      "loss": 0.9387,
+      "step": 15504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5060423565587202,
+      "learning_rate": 1.4661310316273709e-05,
+      "loss": 0.7849,
+      "step": 15505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.522997130480112,
+      "learning_rate": 1.4660635228302388e-05,
+      "loss": 0.8061,
+      "step": 15506
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5131402743820688,
+      "learning_rate": 1.4659960113195371e-05,
+      "loss": 0.8467,
+      "step": 15507
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6634830180596216,
+      "learning_rate": 1.4659284970956593e-05,
+      "loss": 0.8245,
+      "step": 15508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6763057194730726,
+      "learning_rate": 1.4658609801589982e-05,
+      "loss": 0.7129,
+      "step": 15509
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7899485528511776,
+      "learning_rate": 1.465793460509947e-05,
+      "loss": 0.9324,
+      "step": 15510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.160807588750036,
+      "learning_rate": 1.4657259381488988e-05,
+      "loss": 0.8938,
+      "step": 15511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6938676321589765,
+      "learning_rate": 1.4656584130762467e-05,
+      "loss": 0.8595,
+      "step": 15512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.736839207063751,
+      "learning_rate": 1.4655908852923841e-05,
+      "loss": 0.921,
+      "step": 15513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7548897554488971,
+      "learning_rate": 1.4655233547977037e-05,
+      "loss": 0.9023,
+      "step": 15514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5646344687273936,
+      "learning_rate": 1.465455821592599e-05,
+      "loss": 0.9282,
+      "step": 15515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.596938993238797,
+      "learning_rate": 1.465388285677463e-05,
+      "loss": 0.9367,
+      "step": 15516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.597707137711614,
+      "learning_rate": 1.4653207470526893e-05,
+      "loss": 0.9675,
+      "step": 15517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.035981145062482,
+      "learning_rate": 1.465253205718671e-05,
+      "loss": 0.8442,
+      "step": 15518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5217483443823177,
+      "learning_rate": 1.4651856616758009e-05,
+      "loss": 0.7798,
+      "step": 15519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.156371320788601,
+      "learning_rate": 1.4651181149244728e-05,
+      "loss": 0.8847,
+      "step": 15520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.543593765847541,
+      "learning_rate": 1.46505056546508e-05,
+      "loss": 0.8423,
+      "step": 15521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6187000524565562,
+      "learning_rate": 1.4649830132980151e-05,
+      "loss": 0.8282,
+      "step": 15522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8870069831413037,
+      "learning_rate": 1.4649154584236723e-05,
+      "loss": 0.8728,
+      "step": 15523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7595510604822217,
+      "learning_rate": 1.464847900842444e-05,
+      "loss": 0.6991,
+      "step": 15524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5023420001191246,
+      "learning_rate": 1.4647803405547248e-05,
+      "loss": 0.9744,
+      "step": 15525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.064166732152063,
+      "learning_rate": 1.4647127775609067e-05,
+      "loss": 0.8787,
+      "step": 15526
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7838600560087812,
+      "learning_rate": 1.4646452118613838e-05,
+      "loss": 0.75,
+      "step": 15527
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8506700183682412,
+      "learning_rate": 1.4645776434565492e-05,
+      "loss": 0.8383,
+      "step": 15528
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6828013098036638,
+      "learning_rate": 1.4645100723467969e-05,
+      "loss": 0.8908,
+      "step": 15529
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9429628566561061,
+      "learning_rate": 1.4644424985325198e-05,
+      "loss": 0.8043,
+      "step": 15530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6416667594854077,
+      "learning_rate": 1.4643749220141113e-05,
+      "loss": 0.9075,
+      "step": 15531
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.554458814790981,
+      "learning_rate": 1.4643073427919652e-05,
+      "loss": 0.7274,
+      "step": 15532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7238939405592142,
+      "learning_rate": 1.4642397608664744e-05,
+      "loss": 0.9053,
+      "step": 15533
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4627529478669108,
+      "learning_rate": 1.4641721762380332e-05,
+      "loss": 0.9801,
+      "step": 15534
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7406725730653112,
+      "learning_rate": 1.4641045889070343e-05,
+      "loss": 1.0164,
+      "step": 15535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4631990812135414,
+      "learning_rate": 1.4640369988738717e-05,
+      "loss": 0.881,
+      "step": 15536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.225235716990252,
+      "learning_rate": 1.4639694061389384e-05,
+      "loss": 0.8361,
+      "step": 15537
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6406211009646248,
+      "learning_rate": 1.4639018107026288e-05,
+      "loss": 0.8463,
+      "step": 15538
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8076971899111005,
+      "learning_rate": 1.4638342125653358e-05,
+      "loss": 1.1037,
+      "step": 15539
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7520768605256991,
+      "learning_rate": 1.4637666117274531e-05,
+      "loss": 0.7423,
+      "step": 15540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7548088119556997,
+      "learning_rate": 1.4636990081893746e-05,
+      "loss": 1.0377,
+      "step": 15541
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6764651030813975,
+      "learning_rate": 1.4636314019514935e-05,
+      "loss": 0.8888,
+      "step": 15542
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9883355248798817,
+      "learning_rate": 1.4635637930142035e-05,
+      "loss": 1.035,
+      "step": 15543
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5146614411908483,
+      "learning_rate": 1.4634961813778982e-05,
+      "loss": 0.9667,
+      "step": 15544
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6463397782468199,
+      "learning_rate": 1.4634285670429717e-05,
+      "loss": 0.9792,
+      "step": 15545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.642976597928986,
+      "learning_rate": 1.4633609500098174e-05,
+      "loss": 0.8832,
+      "step": 15546
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.632342489922948,
+      "learning_rate": 1.4632933302788289e-05,
+      "loss": 0.865,
+      "step": 15547
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.779636804898961,
+      "learning_rate": 1.4632257078503997e-05,
+      "loss": 0.8745,
+      "step": 15548
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9751183626329656,
+      "learning_rate": 1.4631580827249242e-05,
+      "loss": 0.8521,
+      "step": 15549
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.05247194503324,
+      "learning_rate": 1.4630904549027954e-05,
+      "loss": 0.8097,
+      "step": 15550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7911285725395579,
+      "learning_rate": 1.4630228243844073e-05,
+      "loss": 0.7634,
+      "step": 15551
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.692577619998093,
+      "learning_rate": 1.462955191170154e-05,
+      "loss": 0.7321,
+      "step": 15552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.152952212597643,
+      "learning_rate": 1.4628875552604291e-05,
+      "loss": 0.8951,
+      "step": 15553
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9561704921324508,
+      "learning_rate": 1.4628199166556258e-05,
+      "loss": 0.784,
+      "step": 15554
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9309694092896945,
+      "learning_rate": 1.462752275356139e-05,
+      "loss": 0.8075,
+      "step": 15555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.118540219184963,
+      "learning_rate": 1.4626846313623615e-05,
+      "loss": 0.9338,
+      "step": 15556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9496731045373543,
+      "learning_rate": 1.4626169846746878e-05,
+      "loss": 0.9285,
+      "step": 15557
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.718533399887315,
+      "learning_rate": 1.4625493352935114e-05,
+      "loss": 1.0003,
+      "step": 15558
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4487583366366437,
+      "learning_rate": 1.4624816832192266e-05,
+      "loss": 0.8739,
+      "step": 15559
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.610991507449392,
+      "learning_rate": 1.4624140284522267e-05,
+      "loss": 0.9468,
+      "step": 15560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.839431796222523,
+      "learning_rate": 1.462346370992906e-05,
+      "loss": 0.8655,
+      "step": 15561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6104010269084783,
+      "learning_rate": 1.4622787108416585e-05,
+      "loss": 0.8437,
+      "step": 15562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8307459396188536,
+      "learning_rate": 1.4622110479988778e-05,
+      "loss": 0.7572,
+      "step": 15563
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5578928692274112,
+      "learning_rate": 1.4621433824649582e-05,
+      "loss": 0.9369,
+      "step": 15564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4290014592770341,
+      "learning_rate": 1.4620757142402933e-05,
+      "loss": 0.8749,
+      "step": 15565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4963625622904844,
+      "learning_rate": 1.4620080433252776e-05,
+      "loss": 0.8408,
+      "step": 15566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6209416006756123,
+      "learning_rate": 1.4619403697203045e-05,
+      "loss": 0.837,
+      "step": 15567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1153763431197916,
+      "learning_rate": 1.4618726934257687e-05,
+      "loss": 0.7796,
+      "step": 15568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.679732265229364,
+      "learning_rate": 1.4618050144420637e-05,
+      "loss": 1.103,
+      "step": 15569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6063713264675226,
+      "learning_rate": 1.4617373327695837e-05,
+      "loss": 0.8522,
+      "step": 15570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5663488624588848,
+      "learning_rate": 1.4616696484087228e-05,
+      "loss": 0.8523,
+      "step": 15571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7143656666822313,
+      "learning_rate": 1.4616019613598748e-05,
+      "loss": 0.9392,
+      "step": 15572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.774712215899061,
+      "learning_rate": 1.4615342716234342e-05,
+      "loss": 0.7298,
+      "step": 15573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4983575712552373,
+      "learning_rate": 1.461466579199795e-05,
+      "loss": 0.8315,
+      "step": 15574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8205079938084832,
+      "learning_rate": 1.4613988840893512e-05,
+      "loss": 1.0109,
+      "step": 15575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5256575705216735,
+      "learning_rate": 1.461331186292497e-05,
+      "loss": 0.7983,
+      "step": 15576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2150185160387807,
+      "learning_rate": 1.4612634858096266e-05,
+      "loss": 0.908,
+      "step": 15577
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8444605550378665,
+      "learning_rate": 1.4611957826411343e-05,
+      "loss": 0.9308,
+      "step": 15578
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.673390664608174,
+      "learning_rate": 1.461128076787414e-05,
+      "loss": 0.8192,
+      "step": 15579
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7979951225036115,
+      "learning_rate": 1.46106036824886e-05,
+      "loss": 0.9521,
+      "step": 15580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9444950397669167,
+      "learning_rate": 1.4609926570258665e-05,
+      "loss": 0.797,
+      "step": 15581
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.440976192646612,
+      "learning_rate": 1.460924943118828e-05,
+      "loss": 0.8781,
+      "step": 15582
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5206130254709427,
+      "learning_rate": 1.460857226528139e-05,
+      "loss": 0.9223,
+      "step": 15583
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.773184212077777,
+      "learning_rate": 1.4607895072541925e-05,
+      "loss": 0.872,
+      "step": 15584
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7883835391037883,
+      "learning_rate": 1.4607217852973841e-05,
+      "loss": 0.7996,
+      "step": 15585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6341751106440494,
+      "learning_rate": 1.460654060658107e-05,
+      "loss": 0.8801,
+      "step": 15586
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4995770678313949,
+      "learning_rate": 1.4605863333367567e-05,
+      "loss": 0.7165,
+      "step": 15587
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8380995369710995,
+      "learning_rate": 1.4605186033337266e-05,
+      "loss": 1.0175,
+      "step": 15588
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5931287811095132,
+      "learning_rate": 1.4604508706494119e-05,
+      "loss": 0.9811,
+      "step": 15589
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6270676433162736,
+      "learning_rate": 1.4603831352842061e-05,
+      "loss": 0.9793,
+      "step": 15590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8123245413022818,
+      "learning_rate": 1.460315397238504e-05,
+      "loss": 0.7673,
+      "step": 15591
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7088876646300466,
+      "learning_rate": 1.4602476565127e-05,
+      "loss": 0.9609,
+      "step": 15592
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7444423443000219,
+      "learning_rate": 1.4601799131071883e-05,
+      "loss": 0.7177,
+      "step": 15593
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6245880512654274,
+      "learning_rate": 1.4601121670223633e-05,
+      "loss": 0.9007,
+      "step": 15594
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7731728411759344,
+      "learning_rate": 1.4600444182586197e-05,
+      "loss": 1.002,
+      "step": 15595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8196683737900503,
+      "learning_rate": 1.4599766668163522e-05,
+      "loss": 0.8857,
+      "step": 15596
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7517657481932651,
+      "learning_rate": 1.4599089126959545e-05,
+      "loss": 0.7333,
+      "step": 15597
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6676424533384837,
+      "learning_rate": 1.4598411558978217e-05,
+      "loss": 0.8782,
+      "step": 15598
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6982070669079283,
+      "learning_rate": 1.4597733964223479e-05,
+      "loss": 0.9429,
+      "step": 15599
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1727487338573037,
+      "learning_rate": 1.459705634269928e-05,
+      "loss": 0.8581,
+      "step": 15600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7344085737284087,
+      "learning_rate": 1.4596378694409565e-05,
+      "loss": 0.691,
+      "step": 15601
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.765233473907464,
+      "learning_rate": 1.4595701019358275e-05,
+      "loss": 0.7127,
+      "step": 15602
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4989701685037722,
+      "learning_rate": 1.4595023317549362e-05,
+      "loss": 0.8549,
+      "step": 15603
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4621702773499636,
+      "learning_rate": 1.4594345588986767e-05,
+      "loss": 1.0042,
+      "step": 15604
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7728858013762852,
+      "learning_rate": 1.4593667833674437e-05,
+      "loss": 0.8122,
+      "step": 15605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8554628128032058,
+      "learning_rate": 1.4592990051616318e-05,
+      "loss": 0.9067,
+      "step": 15606
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8395650936032568,
+      "learning_rate": 1.4592312242816355e-05,
+      "loss": 0.8947,
+      "step": 15607
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.402200255506778,
+      "learning_rate": 1.45916344072785e-05,
+      "loss": 0.826,
+      "step": 15608
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.675054836898338,
+      "learning_rate": 1.4590956545006693e-05,
+      "loss": 0.9544,
+      "step": 15609
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6984194756807325,
+      "learning_rate": 1.4590278656004888e-05,
+      "loss": 0.8476,
+      "step": 15610
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.764868790591151,
+      "learning_rate": 1.4589600740277023e-05,
+      "loss": 0.8279,
+      "step": 15611
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.669173350064,
+      "learning_rate": 1.4588922797827052e-05,
+      "loss": 0.9167,
+      "step": 15612
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.157268782088542,
+      "learning_rate": 1.4588244828658918e-05,
+      "loss": 0.9217,
+      "step": 15613
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8068012265901392,
+      "learning_rate": 1.4587566832776569e-05,
+      "loss": 0.6784,
+      "step": 15614
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7815975176994534,
+      "learning_rate": 1.4586888810183956e-05,
+      "loss": 0.776,
+      "step": 15615
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7746919407419783,
+      "learning_rate": 1.4586210760885023e-05,
+      "loss": 0.709,
+      "step": 15616
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9565552054099433,
+      "learning_rate": 1.4585532684883718e-05,
+      "loss": 0.8094,
+      "step": 15617
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8323571195647688,
+      "learning_rate": 1.4584854582183992e-05,
+      "loss": 0.9397,
+      "step": 15618
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5029843541315382,
+      "learning_rate": 1.4584176452789791e-05,
+      "loss": 0.8821,
+      "step": 15619
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9257849559726514,
+      "learning_rate": 1.4583498296705063e-05,
+      "loss": 0.9823,
+      "step": 15620
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7683689015009751,
+      "learning_rate": 1.4582820113933755e-05,
+      "loss": 1.0203,
+      "step": 15621
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5332762510368791,
+      "learning_rate": 1.4582141904479819e-05,
+      "loss": 0.795,
+      "step": 15622
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.029293781823494,
+      "learning_rate": 1.4581463668347201e-05,
+      "loss": 0.9236,
+      "step": 15623
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8043017268580166,
+      "learning_rate": 1.4580785405539853e-05,
+      "loss": 0.8391,
+      "step": 15624
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.82629401189313,
+      "learning_rate": 1.4580107116061722e-05,
+      "loss": 0.9526,
+      "step": 15625
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5106361782684385,
+      "learning_rate": 1.4579428799916754e-05,
+      "loss": 0.9181,
+      "step": 15626
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5623576515518673,
+      "learning_rate": 1.4578750457108904e-05,
+      "loss": 0.8275,
+      "step": 15627
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9242540894242526,
+      "learning_rate": 1.4578072087642118e-05,
+      "loss": 0.8178,
+      "step": 15628
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6522174283449613,
+      "learning_rate": 1.4577393691520349e-05,
+      "loss": 0.9888,
+      "step": 15629
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.014303066183311,
+      "learning_rate": 1.4576715268747545e-05,
+      "loss": 0.9567,
+      "step": 15630
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8931525132786657,
+      "learning_rate": 1.4576036819327654e-05,
+      "loss": 0.8862,
+      "step": 15631
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9730865868669796,
+      "learning_rate": 1.4575358343264628e-05,
+      "loss": 0.9439,
+      "step": 15632
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6212600429898092,
+      "learning_rate": 1.4574679840562417e-05,
+      "loss": 0.8699,
+      "step": 15633
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.661442131141578,
+      "learning_rate": 1.4574001311224976e-05,
+      "loss": 0.8252,
+      "step": 15634
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7135667358224618,
+      "learning_rate": 1.4573322755256245e-05,
+      "loss": 0.8835,
+      "step": 15635
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4864135180863058,
+      "learning_rate": 1.4572644172660186e-05,
+      "loss": 0.9664,
+      "step": 15636
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.807609021902833,
+      "learning_rate": 1.457196556344074e-05,
+      "loss": 0.8987,
+      "step": 15637
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.524869827056044,
+      "learning_rate": 1.4571286927601868e-05,
+      "loss": 0.8864,
+      "step": 15638
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7883498345743948,
+      "learning_rate": 1.4570608265147512e-05,
+      "loss": 0.688,
+      "step": 15639
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5819262019723033,
+      "learning_rate": 1.4569929576081633e-05,
+      "loss": 0.859,
+      "step": 15640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4843727481583404,
+      "learning_rate": 1.4569250860408173e-05,
+      "loss": 0.8449,
+      "step": 15641
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8459859524889346,
+      "learning_rate": 1.4568572118131087e-05,
+      "loss": 0.8598,
+      "step": 15642
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6465053843442026,
+      "learning_rate": 1.456789334925433e-05,
+      "loss": 0.7621,
+      "step": 15643
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.730678181814876,
+      "learning_rate": 1.456721455378185e-05,
+      "loss": 0.6823,
+      "step": 15644
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4417696960383304,
+      "learning_rate": 1.4566535731717603e-05,
+      "loss": 0.899,
+      "step": 15645
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8450604345722503,
+      "learning_rate": 1.456585688306554e-05,
+      "loss": 0.9022,
+      "step": 15646
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.724912950074899,
+      "learning_rate": 1.456517800782961e-05,
+      "loss": 0.9841,
+      "step": 15647
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9289410734796824,
+      "learning_rate": 1.4564499106013768e-05,
+      "loss": 0.8226,
+      "step": 15648
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8061211401669293,
+      "learning_rate": 1.4563820177621968e-05,
+      "loss": 0.9354,
+      "step": 15649
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.701348252355846,
+      "learning_rate": 1.4563141222658163e-05,
+      "loss": 0.8351,
+      "step": 15650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9936089105368735,
+      "learning_rate": 1.4562462241126305e-05,
+      "loss": 0.9017,
+      "step": 15651
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.832664691255146,
+      "learning_rate": 1.4561783233030346e-05,
+      "loss": 0.8446,
+      "step": 15652
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7851827733807892,
+      "learning_rate": 1.4561104198374243e-05,
+      "loss": 0.8897,
+      "step": 15653
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3893407243386138,
+      "learning_rate": 1.4560425137161946e-05,
+      "loss": 0.8489,
+      "step": 15654
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7267546368666442,
+      "learning_rate": 1.4559746049397411e-05,
+      "loss": 0.9013,
+      "step": 15655
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.951915665757571,
+      "learning_rate": 1.455906693508459e-05,
+      "loss": 0.8878,
+      "step": 15656
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.603183913348036,
+      "learning_rate": 1.4558387794227438e-05,
+      "loss": 0.8244,
+      "step": 15657
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4849151729192633,
+      "learning_rate": 1.4557708626829907e-05,
+      "loss": 0.8614,
+      "step": 15658
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.760751223518135,
+      "learning_rate": 1.455702943289596e-05,
+      "loss": 0.7186,
+      "step": 15659
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6155293221603684,
+      "learning_rate": 1.4556350212429538e-05,
+      "loss": 0.8788,
+      "step": 15660
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8014175795923473,
+      "learning_rate": 1.4555670965434602e-05,
+      "loss": 0.7618,
+      "step": 15661
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9809865322442164,
+      "learning_rate": 1.4554991691915115e-05,
+      "loss": 0.9866,
+      "step": 15662
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.703889352789233,
+      "learning_rate": 1.4554312391875016e-05,
+      "loss": 0.9167,
+      "step": 15663
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.077414770407925,
+      "learning_rate": 1.4553633065318274e-05,
+      "loss": 1.0138,
+      "step": 15664
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0514027495621843,
+      "learning_rate": 1.4552953712248834e-05,
+      "loss": 0.9991,
+      "step": 15665
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.685614221084311,
+      "learning_rate": 1.4552274332670662e-05,
+      "loss": 0.8498,
+      "step": 15666
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7706016947855452,
+      "learning_rate": 1.4551594926587703e-05,
+      "loss": 0.7728,
+      "step": 15667
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8489852790693545,
+      "learning_rate": 1.4550915494003919e-05,
+      "loss": 0.7197,
+      "step": 15668
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.501079200593413,
+      "learning_rate": 1.4550236034923263e-05,
+      "loss": 0.9468,
+      "step": 15669
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6189990782549546,
+      "learning_rate": 1.4549556549349694e-05,
+      "loss": 0.8025,
+      "step": 15670
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6609041157933442,
+      "learning_rate": 1.4548877037287166e-05,
+      "loss": 0.8752,
+      "step": 15671
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9067937293092467,
+      "learning_rate": 1.4548197498739636e-05,
+      "loss": 0.8871,
+      "step": 15672
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5847263359769457,
+      "learning_rate": 1.4547517933711058e-05,
+      "loss": 0.8472,
+      "step": 15673
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.632624816396105,
+      "learning_rate": 1.4546838342205392e-05,
+      "loss": 0.9235,
+      "step": 15674
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5945435940683115,
+      "learning_rate": 1.4546158724226594e-05,
+      "loss": 0.758,
+      "step": 15675
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7921639303563298,
+      "learning_rate": 1.4545479079778622e-05,
+      "loss": 0.7433,
+      "step": 15676
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.555731498947656,
+      "learning_rate": 1.4544799408865428e-05,
+      "loss": 0.9094,
+      "step": 15677
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8802539164372931,
+      "learning_rate": 1.4544119711490976e-05,
+      "loss": 0.963,
+      "step": 15678
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7522190892366628,
+      "learning_rate": 1.4543439987659221e-05,
+      "loss": 0.7469,
+      "step": 15679
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2506528882705066,
+      "learning_rate": 1.4542760237374119e-05,
+      "loss": 0.9602,
+      "step": 15680
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4027644657145166,
+      "learning_rate": 1.4542080460639627e-05,
+      "loss": 0.9179,
+      "step": 15681
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6013127391714814,
+      "learning_rate": 1.4541400657459705e-05,
+      "loss": 0.7378,
+      "step": 15682
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6824257832624696,
+      "learning_rate": 1.4540720827838313e-05,
+      "loss": 0.7774,
+      "step": 15683
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8458027124817376,
+      "learning_rate": 1.4540040971779403e-05,
+      "loss": 0.8159,
+      "step": 15684
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5294249962495112,
+      "learning_rate": 1.4539361089286942e-05,
+      "loss": 0.9184,
+      "step": 15685
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4729184932930453,
+      "learning_rate": 1.4538681180364878e-05,
+      "loss": 0.8937,
+      "step": 15686
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0648544893195675,
+      "learning_rate": 1.4538001245017182e-05,
+      "loss": 0.9382,
+      "step": 15687
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4574851690967028,
+      "learning_rate": 1.4537321283247799e-05,
+      "loss": 0.8677,
+      "step": 15688
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.381621546818559,
+      "learning_rate": 1.4536641295060701e-05,
+      "loss": 0.903,
+      "step": 15689
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5444491742652529,
+      "learning_rate": 1.4535961280459838e-05,
+      "loss": 0.8394,
+      "step": 15690
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4511195986590704,
+      "learning_rate": 1.4535281239449173e-05,
+      "loss": 1.0254,
+      "step": 15691
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8128106358402465,
+      "learning_rate": 1.4534601172032663e-05,
+      "loss": 0.8399,
+      "step": 15692
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9180996372428358,
+      "learning_rate": 1.4533921078214271e-05,
+      "loss": 0.923,
+      "step": 15693
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5875356227577326,
+      "learning_rate": 1.4533240957997956e-05,
+      "loss": 0.9939,
+      "step": 15694
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5435977839255286,
+      "learning_rate": 1.4532560811387674e-05,
+      "loss": 0.8964,
+      "step": 15695
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7528634560745533,
+      "learning_rate": 1.453188063838739e-05,
+      "loss": 0.9009,
+      "step": 15696
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8045655693347227,
+      "learning_rate": 1.4531200439001063e-05,
+      "loss": 0.7706,
+      "step": 15697
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.659587635421772,
+      "learning_rate": 1.453052021323265e-05,
+      "loss": 1.0253,
+      "step": 15698
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6667689157038326,
+      "learning_rate": 1.4529839961086116e-05,
+      "loss": 0.8929,
+      "step": 15699
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5106316960979553,
+      "learning_rate": 1.452915968256542e-05,
+      "loss": 0.8824,
+      "step": 15700
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.577536532502556,
+      "learning_rate": 1.4528479377674521e-05,
+      "loss": 0.8739,
+      "step": 15701
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.024287467836584,
+      "learning_rate": 1.452779904641738e-05,
+      "loss": 0.9815,
+      "step": 15702
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6244549889895343,
+      "learning_rate": 1.4527118688797963e-05,
+      "loss": 0.812,
+      "step": 15703
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4997053435150256,
+      "learning_rate": 1.4526438304820225e-05,
+      "loss": 0.9079,
+      "step": 15704
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8121300446101272,
+      "learning_rate": 1.452575789448813e-05,
+      "loss": 0.987,
+      "step": 15705
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0166810901537846,
+      "learning_rate": 1.4525077457805646e-05,
+      "loss": 0.8155,
+      "step": 15706
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.543657635417688,
+      "learning_rate": 1.452439699477672e-05,
+      "loss": 0.869,
+      "step": 15707
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.551120061434262,
+      "learning_rate": 1.4523716505405327e-05,
+      "loss": 0.859,
+      "step": 15708
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7349304656906996,
+      "learning_rate": 1.4523035989695422e-05,
+      "loss": 1.0475,
+      "step": 15709
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6673895085339294,
+      "learning_rate": 1.452235544765097e-05,
+      "loss": 0.9139,
+      "step": 15710
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8748941825490333,
+      "learning_rate": 1.4521674879275933e-05,
+      "loss": 0.9507,
+      "step": 15711
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.566549658472085,
+      "learning_rate": 1.4520994284574275e-05,
+      "loss": 0.8601,
+      "step": 15712
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.66722135827576,
+      "learning_rate": 1.4520313663549957e-05,
+      "loss": 0.8802,
+      "step": 15713
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6973904492377498,
+      "learning_rate": 1.4519633016206938e-05,
+      "loss": 0.9488,
+      "step": 15714
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4608188443866397,
+      "learning_rate": 1.4518952342549189e-05,
+      "loss": 0.9289,
+      "step": 15715
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7688785588596594,
+      "learning_rate": 1.4518271642580663e-05,
+      "loss": 0.7484,
+      "step": 15716
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5361561090621076,
+      "learning_rate": 1.4517590916305335e-05,
+      "loss": 0.8896,
+      "step": 15717
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6591103629029462,
+      "learning_rate": 1.4516910163727158e-05,
+      "loss": 0.9614,
+      "step": 15718
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6533176119285773,
+      "learning_rate": 1.4516229384850105e-05,
+      "loss": 0.9656,
+      "step": 15719
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8204927588636348,
+      "learning_rate": 1.451554857967813e-05,
+      "loss": 0.8856,
+      "step": 15720
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8385575616131333,
+      "learning_rate": 1.4514867748215204e-05,
+      "loss": 0.8724,
+      "step": 15721
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5581954261116995,
+      "learning_rate": 1.4514186890465289e-05,
+      "loss": 0.847,
+      "step": 15722
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.524400318603551,
+      "learning_rate": 1.4513506006432347e-05,
+      "loss": 0.8226,
+      "step": 15723
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5900735982583198,
+      "learning_rate": 1.4512825096120346e-05,
+      "loss": 0.9794,
+      "step": 15724
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5051974985431618,
+      "learning_rate": 1.4512144159533248e-05,
+      "loss": 0.8814,
+      "step": 15725
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.357869055891006,
+      "learning_rate": 1.4511463196675018e-05,
+      "loss": 0.7505,
+      "step": 15726
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.049482206968337,
+      "learning_rate": 1.4510782207549619e-05,
+      "loss": 0.8841,
+      "step": 15727
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.066645482322781,
+      "learning_rate": 1.451010119216102e-05,
+      "loss": 0.9473,
+      "step": 15728
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7479577614536965,
+      "learning_rate": 1.4509420150513183e-05,
+      "loss": 0.8035,
+      "step": 15729
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7424917550438466,
+      "learning_rate": 1.4508739082610075e-05,
+      "loss": 0.8028,
+      "step": 15730
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6422360383081331,
+      "learning_rate": 1.4508057988455659e-05,
+      "loss": 0.809,
+      "step": 15731
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0034244130142738,
+      "learning_rate": 1.4507376868053903e-05,
+      "loss": 0.975,
+      "step": 15732
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7415213630179798,
+      "learning_rate": 1.4506695721408775e-05,
+      "loss": 0.7677,
+      "step": 15733
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.630608693860949,
+      "learning_rate": 1.4506014548524237e-05,
+      "loss": 0.8276,
+      "step": 15734
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4745309278522079,
+      "learning_rate": 1.4505333349404252e-05,
+      "loss": 0.8551,
+      "step": 15735
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6872257325674986,
+      "learning_rate": 1.4504652124052792e-05,
+      "loss": 0.8629,
+      "step": 15736
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6779451640275327,
+      "learning_rate": 1.450397087247382e-05,
+      "loss": 0.7777,
+      "step": 15737
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6356884116435855,
+      "learning_rate": 1.4503289594671306e-05,
+      "loss": 0.9649,
+      "step": 15738
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8181197536612625,
+      "learning_rate": 1.450260829064921e-05,
+      "loss": 0.8158,
+      "step": 15739
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8434444870557107,
+      "learning_rate": 1.4501926960411508e-05,
+      "loss": 0.7725,
+      "step": 15740
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5427420236806297,
+      "learning_rate": 1.450124560396216e-05,
+      "loss": 0.8406,
+      "step": 15741
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.398655701841605,
+      "learning_rate": 1.4500564221305133e-05,
+      "loss": 0.8129,
+      "step": 15742
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7514389729164408,
+      "learning_rate": 1.44998828124444e-05,
+      "loss": 0.8138,
+      "step": 15743
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4924737868710602,
+      "learning_rate": 1.4499201377383923e-05,
+      "loss": 0.8425,
+      "step": 15744
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7287938349035707,
+      "learning_rate": 1.449851991612767e-05,
+      "loss": 0.7769,
+      "step": 15745
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6383598941629962,
+      "learning_rate": 1.449783842867961e-05,
+      "loss": 0.916,
+      "step": 15746
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6087465886941024,
+      "learning_rate": 1.4497156915043712e-05,
+      "loss": 0.8629,
+      "step": 15747
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5717137973222115,
+      "learning_rate": 1.4496475375223941e-05,
+      "loss": 0.9941,
+      "step": 15748
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5123888487825576,
+      "learning_rate": 1.4495793809224268e-05,
+      "loss": 0.8724,
+      "step": 15749
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.913496173005395,
+      "learning_rate": 1.449511221704866e-05,
+      "loss": 0.9492,
+      "step": 15750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0468571774808924,
+      "learning_rate": 1.4494430598701087e-05,
+      "loss": 0.8501,
+      "step": 15751
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6121458041948276,
+      "learning_rate": 1.4493748954185513e-05,
+      "loss": 0.8686,
+      "step": 15752
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7960513893725856,
+      "learning_rate": 1.449306728350591e-05,
+      "loss": 0.8864,
+      "step": 15753
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9861107464970231,
+      "learning_rate": 1.4492385586666249e-05,
+      "loss": 0.9823,
+      "step": 15754
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8600785611431243,
+      "learning_rate": 1.4491703863670496e-05,
+      "loss": 0.8939,
+      "step": 15755
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.038330974586085,
+      "learning_rate": 1.4491022114522621e-05,
+      "loss": 0.8381,
+      "step": 15756
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5034984750469835,
+      "learning_rate": 1.4490340339226596e-05,
+      "loss": 0.8797,
+      "step": 15757
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7736637697415651,
+      "learning_rate": 1.4489658537786385e-05,
+      "loss": 0.8244,
+      "step": 15758
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4769726904213745,
+      "learning_rate": 1.448897671020596e-05,
+      "loss": 0.9541,
+      "step": 15759
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7399404851032436,
+      "learning_rate": 1.4488294856489291e-05,
+      "loss": 0.8893,
+      "step": 15760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.742106158976794,
+      "learning_rate": 1.448761297664035e-05,
+      "loss": 0.983,
+      "step": 15761
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6263358792731015,
+      "learning_rate": 1.4486931070663107e-05,
+      "loss": 0.785,
+      "step": 15762
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7009387566386056,
+      "learning_rate": 1.4486249138561528e-05,
+      "loss": 0.8981,
+      "step": 15763
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.614633086021844,
+      "learning_rate": 1.448556718033959e-05,
+      "loss": 0.8783,
+      "step": 15764
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9200767356738693,
+      "learning_rate": 1.4484885196001255e-05,
+      "loss": 0.8513,
+      "step": 15765
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8024049036586137,
+      "learning_rate": 1.4484203185550503e-05,
+      "loss": 0.8031,
+      "step": 15766
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4948654141512359,
+      "learning_rate": 1.4483521148991295e-05,
+      "loss": 0.8647,
+      "step": 15767
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9803719503966224,
+      "learning_rate": 1.4482839086327614e-05,
+      "loss": 0.8886,
+      "step": 15768
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7473833966878453,
+      "learning_rate": 1.4482156997563417e-05,
+      "loss": 0.6853,
+      "step": 15769
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8182087459080483,
+      "learning_rate": 1.4481474882702688e-05,
+      "loss": 0.8166,
+      "step": 15770
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7479585741833037,
+      "learning_rate": 1.4480792741749395e-05,
+      "loss": 0.7263,
+      "step": 15771
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7610427732238934,
+      "learning_rate": 1.4480110574707504e-05,
+      "loss": 0.9538,
+      "step": 15772
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.929301320675313,
+      "learning_rate": 1.4479428381580993e-05,
+      "loss": 0.9853,
+      "step": 15773
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5349180247761827,
+      "learning_rate": 1.447874616237383e-05,
+      "loss": 0.9118,
+      "step": 15774
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.800665040592605,
+      "learning_rate": 1.447806391708999e-05,
+      "loss": 0.879,
+      "step": 15775
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6700315214991137,
+      "learning_rate": 1.4477381645733446e-05,
+      "loss": 0.7691,
+      "step": 15776
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7284477761390018,
+      "learning_rate": 1.4476699348308165e-05,
+      "loss": 0.9427,
+      "step": 15777
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8245341446604939,
+      "learning_rate": 1.4476017024818126e-05,
+      "loss": 0.7472,
+      "step": 15778
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6402407373214443,
+      "learning_rate": 1.4475334675267296e-05,
+      "loss": 0.8683,
+      "step": 15779
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4689219348516176,
+      "learning_rate": 1.4474652299659653e-05,
+      "loss": 0.795,
+      "step": 15780
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.582914392093331,
+      "learning_rate": 1.4473969897999168e-05,
+      "loss": 0.952,
+      "step": 15781
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7684344432557628,
+      "learning_rate": 1.4473287470289811e-05,
+      "loss": 0.6868,
+      "step": 15782
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.642357694500921,
+      "learning_rate": 1.447260501653556e-05,
+      "loss": 0.8231,
+      "step": 15783
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7285203151888444,
+      "learning_rate": 1.4471922536740385e-05,
+      "loss": 0.9032,
+      "step": 15784
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8710913953817332,
+      "learning_rate": 1.4471240030908265e-05,
+      "loss": 0.9372,
+      "step": 15785
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.638442353070547,
+      "learning_rate": 1.4470557499043167e-05,
+      "loss": 0.8877,
+      "step": 15786
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8329667539823629,
+      "learning_rate": 1.4469874941149069e-05,
+      "loss": 0.7552,
+      "step": 15787
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0756740534515377,
+      "learning_rate": 1.4469192357229942e-05,
+      "loss": 0.9348,
+      "step": 15788
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4877049802441547,
+      "learning_rate": 1.4468509747289764e-05,
+      "loss": 0.8308,
+      "step": 15789
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4790399322438674,
+      "learning_rate": 1.4467827111332506e-05,
+      "loss": 0.8055,
+      "step": 15790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8579409880219109,
+      "learning_rate": 1.4467144449362147e-05,
+      "loss": 0.7341,
+      "step": 15791
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6343235001195004,
+      "learning_rate": 1.4466461761382657e-05,
+      "loss": 0.7736,
+      "step": 15792
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.771712806049089,
+      "learning_rate": 1.4465779047398014e-05,
+      "loss": 0.7093,
+      "step": 15793
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.522451340529338,
+      "learning_rate": 1.446509630741219e-05,
+      "loss": 0.9051,
+      "step": 15794
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6030868319125058,
+      "learning_rate": 1.4464413541429161e-05,
+      "loss": 0.952,
+      "step": 15795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8372401208768918,
+      "learning_rate": 1.4463730749452904e-05,
+      "loss": 0.9373,
+      "step": 15796
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7666372970244709,
+      "learning_rate": 1.4463047931487391e-05,
+      "loss": 0.7249,
+      "step": 15797
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.163218418019009,
+      "learning_rate": 1.4462365087536603e-05,
+      "loss": 0.9277,
+      "step": 15798
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5378152217739887,
+      "learning_rate": 1.446168221760451e-05,
+      "loss": 0.9272,
+      "step": 15799
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8845828119312258,
+      "learning_rate": 1.4460999321695095e-05,
+      "loss": 0.9283,
+      "step": 15800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7587597754956034,
+      "learning_rate": 1.4460316399812326e-05,
+      "loss": 0.6999,
+      "step": 15801
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.25700374015651,
+      "learning_rate": 1.4459633451960184e-05,
+      "loss": 0.8637,
+      "step": 15802
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7178211051026397,
+      "learning_rate": 1.4458950478142643e-05,
+      "loss": 0.9004,
+      "step": 15803
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.579504697058386,
+      "learning_rate": 1.4458267478363683e-05,
+      "loss": 0.8866,
+      "step": 15804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5978342058960038,
+      "learning_rate": 1.4457584452627276e-05,
+      "loss": 0.9167,
+      "step": 15805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7169087848052129,
+      "learning_rate": 1.4456901400937404e-05,
+      "loss": 0.66,
+      "step": 15806
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6073641416321305,
+      "learning_rate": 1.4456218323298037e-05,
+      "loss": 0.7933,
+      "step": 15807
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9724243334020635,
+      "learning_rate": 1.4455535219713157e-05,
+      "loss": 0.8159,
+      "step": 15808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6087096814780961,
+      "learning_rate": 1.4454852090186742e-05,
+      "loss": 0.8865,
+      "step": 15809
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.567640434707544,
+      "learning_rate": 1.4454168934722765e-05,
+      "loss": 0.9976,
+      "step": 15810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5197887915476214,
+      "learning_rate": 1.445348575332521e-05,
+      "loss": 0.9128,
+      "step": 15811
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7646139894261507,
+      "learning_rate": 1.4452802545998048e-05,
+      "loss": 0.7274,
+      "step": 15812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7752646931866074,
+      "learning_rate": 1.4452119312745266e-05,
+      "loss": 0.7873,
+      "step": 15813
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7308331445901959,
+      "learning_rate": 1.4451436053570826e-05,
+      "loss": 0.8289,
+      "step": 15814
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.722406756590779,
+      "learning_rate": 1.4450752768478725e-05,
+      "loss": 0.8911,
+      "step": 15815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7566468726252855,
+      "learning_rate": 1.4450069457472926e-05,
+      "loss": 0.8398,
+      "step": 15816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0154879100552523,
+      "learning_rate": 1.4449386120557418e-05,
+      "loss": 0.8576,
+      "step": 15817
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7000087015649084,
+      "learning_rate": 1.4448702757736172e-05,
+      "loss": 0.8228,
+      "step": 15818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7739362322119876,
+      "learning_rate": 1.4448019369013173e-05,
+      "loss": 0.8086,
+      "step": 15819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7801629774461447,
+      "learning_rate": 1.4447335954392393e-05,
+      "loss": 0.9441,
+      "step": 15820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7586961492987575,
+      "learning_rate": 1.4446652513877818e-05,
+      "loss": 0.7479,
+      "step": 15821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8203092932993223,
+      "learning_rate": 1.4445969047473427e-05,
+      "loss": 0.7692,
+      "step": 15822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4586183590034458,
+      "learning_rate": 1.4445285555183193e-05,
+      "loss": 0.8569,
+      "step": 15823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9175773291788705,
+      "learning_rate": 1.4444602037011099e-05,
+      "loss": 0.8143,
+      "step": 15824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7689651905760438,
+      "learning_rate": 1.4443918492961127e-05,
+      "loss": 0.7597,
+      "step": 15825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6649828883872422,
+      "learning_rate": 1.4443234923037253e-05,
+      "loss": 0.9582,
+      "step": 15826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7194066206313023,
+      "learning_rate": 1.4442551327243456e-05,
+      "loss": 0.8561,
+      "step": 15827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6792373882363736,
+      "learning_rate": 1.4441867705583723e-05,
+      "loss": 0.8896,
+      "step": 15828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8534571574013041,
+      "learning_rate": 1.4441184058062028e-05,
+      "loss": 0.9533,
+      "step": 15829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5266552811946006,
+      "learning_rate": 1.4440500384682353e-05,
+      "loss": 0.8589,
+      "step": 15830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7236126518117332,
+      "learning_rate": 1.443981668544868e-05,
+      "loss": 0.8723,
+      "step": 15831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7259387301526345,
+      "learning_rate": 1.4439132960364988e-05,
+      "loss": 1.0644,
+      "step": 15832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6916774429517847,
+      "learning_rate": 1.4438449209435259e-05,
+      "loss": 0.9163,
+      "step": 15833
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7024363365352506,
+      "learning_rate": 1.4437765432663472e-05,
+      "loss": 0.7956,
+      "step": 15834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8986308289053153,
+      "learning_rate": 1.4437081630053614e-05,
+      "loss": 0.8773,
+      "step": 15835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0598934318273887,
+      "learning_rate": 1.4436397801609658e-05,
+      "loss": 0.7695,
+      "step": 15836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6319966877858867,
+      "learning_rate": 1.443571394733559e-05,
+      "loss": 0.812,
+      "step": 15837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.682635161614237,
+      "learning_rate": 1.4435030067235394e-05,
+      "loss": 0.8338,
+      "step": 15838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7819092324132381,
+      "learning_rate": 1.4434346161313041e-05,
+      "loss": 0.7216,
+      "step": 15839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5688299312348626,
+      "learning_rate": 1.4433662229572529e-05,
+      "loss": 0.8778,
+      "step": 15840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3973100972856696,
+      "learning_rate": 1.4432978272017828e-05,
+      "loss": 0.8022,
+      "step": 15841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7565882610526693,
+      "learning_rate": 1.4432294288652926e-05,
+      "loss": 0.9161,
+      "step": 15842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3669119243843073,
+      "learning_rate": 1.4431610279481802e-05,
+      "loss": 0.8401,
+      "step": 15843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.909947716418847,
+      "learning_rate": 1.4430926244508441e-05,
+      "loss": 0.8018,
+      "step": 15844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8689815731879533,
+      "learning_rate": 1.4430242183736824e-05,
+      "loss": 0.8284,
+      "step": 15845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5664937158692436,
+      "learning_rate": 1.4429558097170932e-05,
+      "loss": 0.8119,
+      "step": 15846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5925085983856728,
+      "learning_rate": 1.4428873984814755e-05,
+      "loss": 0.8524,
+      "step": 15847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5821721260387094,
+      "learning_rate": 1.4428189846672267e-05,
+      "loss": 0.8434,
+      "step": 15848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.615365167178564,
+      "learning_rate": 1.442750568274746e-05,
+      "loss": 0.9731,
+      "step": 15849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5917131144113492,
+      "learning_rate": 1.4426821493044312e-05,
+      "loss": 0.9322,
+      "step": 15850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5515733633481619,
+      "learning_rate": 1.4426137277566805e-05,
+      "loss": 0.8891,
+      "step": 15851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.570802132516546,
+      "learning_rate": 1.4425453036318929e-05,
+      "loss": 0.9423,
+      "step": 15852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6720854329143875,
+      "learning_rate": 1.4424768769304663e-05,
+      "loss": 0.9041,
+      "step": 15853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7757946362624948,
+      "learning_rate": 1.4424084476527991e-05,
+      "loss": 0.9268,
+      "step": 15854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7691916556735883,
+      "learning_rate": 1.44234001579929e-05,
+      "loss": 0.8952,
+      "step": 15855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.544467455186416,
+      "learning_rate": 1.4422715813703372e-05,
+      "loss": 0.7703,
+      "step": 15856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6990452543735737,
+      "learning_rate": 1.4422031443663391e-05,
+      "loss": 0.8642,
+      "step": 15857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.881596706119665,
+      "learning_rate": 1.4421347047876945e-05,
+      "loss": 0.8552,
+      "step": 15858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6732647076690417,
+      "learning_rate": 1.4420662626348017e-05,
+      "loss": 1.0003,
+      "step": 15859
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9246423455893833,
+      "learning_rate": 1.441997817908059e-05,
+      "loss": 0.94,
+      "step": 15860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9173253241391675,
+      "learning_rate": 1.4419293706078654e-05,
+      "loss": 0.8786,
+      "step": 15861
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9407674560127888,
+      "learning_rate": 1.4418609207346189e-05,
+      "loss": 0.7503,
+      "step": 15862
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8641718940621614,
+      "learning_rate": 1.4417924682887181e-05,
+      "loss": 0.977,
+      "step": 15863
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5247898996121212,
+      "learning_rate": 1.4417240132705618e-05,
+      "loss": 0.8907,
+      "step": 15864
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7541143944590247,
+      "learning_rate": 1.4416555556805482e-05,
+      "loss": 0.7978,
+      "step": 15865
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7284425872921413,
+      "learning_rate": 1.4415870955190767e-05,
+      "loss": 0.7504,
+      "step": 15866
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2124825476006755,
+      "learning_rate": 1.441518632786545e-05,
+      "loss": 1.0181,
+      "step": 15867
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.042882755721213,
+      "learning_rate": 1.441450167483352e-05,
+      "loss": 0.8772,
+      "step": 15868
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8285512557822926,
+      "learning_rate": 1.4413816996098962e-05,
+      "loss": 0.7047,
+      "step": 15869
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5790370104561353,
+      "learning_rate": 1.4413132291665768e-05,
+      "loss": 0.8592,
+      "step": 15870
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4560440914665893,
+      "learning_rate": 1.4412447561537919e-05,
+      "loss": 0.9433,
+      "step": 15871
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7701541536371628,
+      "learning_rate": 1.4411762805719402e-05,
+      "loss": 0.9428,
+      "step": 15872
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8104100373524745,
+      "learning_rate": 1.4411078024214207e-05,
+      "loss": 0.8843,
+      "step": 15873
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6295678982386004,
+      "learning_rate": 1.4410393217026317e-05,
+      "loss": 0.8162,
+      "step": 15874
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5136154992041746,
+      "learning_rate": 1.4409708384159724e-05,
+      "loss": 0.9984,
+      "step": 15875
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5309499580561239,
+      "learning_rate": 1.4409023525618411e-05,
+      "loss": 0.9245,
+      "step": 15876
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4008426909636977,
+      "learning_rate": 1.4408338641406368e-05,
+      "loss": 0.8384,
+      "step": 15877
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7615750411727318,
+      "learning_rate": 1.440765373152758e-05,
+      "loss": 0.9574,
+      "step": 15878
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7104457541222948,
+      "learning_rate": 1.4406968795986038e-05,
+      "loss": 0.8005,
+      "step": 15879
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.770910660813783,
+      "learning_rate": 1.4406283834785728e-05,
+      "loss": 0.827,
+      "step": 15880
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6870979678648168,
+      "learning_rate": 1.4405598847930639e-05,
+      "loss": 0.7314,
+      "step": 15881
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6865245722830102,
+      "learning_rate": 1.440491383542476e-05,
+      "loss": 0.8898,
+      "step": 15882
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6857781326686974,
+      "learning_rate": 1.4404228797272076e-05,
+      "loss": 0.9982,
+      "step": 15883
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7345086986031955,
+      "learning_rate": 1.4403543733476578e-05,
+      "loss": 0.8979,
+      "step": 15884
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8164191381549987,
+      "learning_rate": 1.4402858644042254e-05,
+      "loss": 0.753,
+      "step": 15885
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7152183480044465,
+      "learning_rate": 1.4402173528973094e-05,
+      "loss": 0.9445,
+      "step": 15886
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.178951300770026,
+      "learning_rate": 1.4401488388273086e-05,
+      "loss": 0.8963,
+      "step": 15887
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.587017692923561,
+      "learning_rate": 1.4400803221946218e-05,
+      "loss": 0.9401,
+      "step": 15888
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6422025420748685,
+      "learning_rate": 1.4400118029996481e-05,
+      "loss": 0.78,
+      "step": 15889
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5894802728996675,
+      "learning_rate": 1.4399432812427862e-05,
+      "loss": 0.8301,
+      "step": 15890
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6052729522474833,
+      "learning_rate": 1.4398747569244355e-05,
+      "loss": 0.8171,
+      "step": 15891
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.47379778110199,
+      "learning_rate": 1.4398062300449947e-05,
+      "loss": 0.8883,
+      "step": 15892
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6740649487107568,
+      "learning_rate": 1.4397377006048628e-05,
+      "loss": 0.8977,
+      "step": 15893
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.673811276379152,
+      "learning_rate": 1.4396691686044386e-05,
+      "loss": 0.8303,
+      "step": 15894
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.561034425090819,
+      "learning_rate": 1.4396006340441214e-05,
+      "loss": 0.9167,
+      "step": 15895
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.670988296558092,
+      "learning_rate": 1.43953209692431e-05,
+      "loss": 1.0278,
+      "step": 15896
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7257432440558813,
+      "learning_rate": 1.4394635572454036e-05,
+      "loss": 0.9467,
+      "step": 15897
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6738732050654292,
+      "learning_rate": 1.4393950150078016e-05,
+      "loss": 0.8496,
+      "step": 15898
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7559259101265638,
+      "learning_rate": 1.4393264702119023e-05,
+      "loss": 0.7178,
+      "step": 15899
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7327704477772954,
+      "learning_rate": 1.4392579228581054e-05,
+      "loss": 0.8864,
+      "step": 15900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9612562317816442,
+      "learning_rate": 1.4391893729468099e-05,
+      "loss": 0.9438,
+      "step": 15901
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4560299340123,
+      "learning_rate": 1.4391208204784146e-05,
+      "loss": 0.9047,
+      "step": 15902
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7895014753646534,
+      "learning_rate": 1.439052265453319e-05,
+      "loss": 0.8495,
+      "step": 15903
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5165360294809156,
+      "learning_rate": 1.438983707871922e-05,
+      "loss": 0.8441,
+      "step": 15904
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.403311076668676,
+      "learning_rate": 1.4389151477346231e-05,
+      "loss": 0.7938,
+      "step": 15905
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.131884398269451,
+      "learning_rate": 1.438846585041821e-05,
+      "loss": 0.7858,
+      "step": 15906
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6650853874070213,
+      "learning_rate": 1.4387780197939152e-05,
+      "loss": 0.9669,
+      "step": 15907
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.612504716691442,
+      "learning_rate": 1.4387094519913048e-05,
+      "loss": 0.8216,
+      "step": 15908
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7817367278369585,
+      "learning_rate": 1.438640881634389e-05,
+      "loss": 0.9484,
+      "step": 15909
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5373477301489493,
+      "learning_rate": 1.4385723087235674e-05,
+      "loss": 0.8774,
+      "step": 15910
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7394058074146292,
+      "learning_rate": 1.438503733259239e-05,
+      "loss": 0.8879,
+      "step": 15911
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0790090680321476,
+      "learning_rate": 1.4384351552418027e-05,
+      "loss": 0.8679,
+      "step": 15912
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.228719277687429,
+      "learning_rate": 1.4383665746716582e-05,
+      "loss": 0.9071,
+      "step": 15913
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7801966642087436,
+      "learning_rate": 1.4382979915492051e-05,
+      "loss": 0.6967,
+      "step": 15914
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7554475438211874,
+      "learning_rate": 1.4382294058748421e-05,
+      "loss": 0.8825,
+      "step": 15915
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7403832030394455,
+      "learning_rate": 1.4381608176489686e-05,
+      "loss": 0.7967,
+      "step": 15916
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5410119449942614,
+      "learning_rate": 1.438092226871984e-05,
+      "loss": 0.9057,
+      "step": 15917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8721749002148171,
+      "learning_rate": 1.438023633544288e-05,
+      "loss": 0.8595,
+      "step": 15918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8662388934129286,
+      "learning_rate": 1.4379550376662799e-05,
+      "loss": 0.8422,
+      "step": 15919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.925003038271799,
+      "learning_rate": 1.4378864392383585e-05,
+      "loss": 0.7942,
+      "step": 15920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4406349654128416,
+      "learning_rate": 1.437817838260924e-05,
+      "loss": 0.9125,
+      "step": 15921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7739235729417674,
+      "learning_rate": 1.4377492347343753e-05,
+      "loss": 0.7475,
+      "step": 15922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7938898613632842,
+      "learning_rate": 1.4376806286591118e-05,
+      "loss": 0.7624,
+      "step": 15923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.616790030707568,
+      "learning_rate": 1.4376120200355334e-05,
+      "loss": 0.7924,
+      "step": 15924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8158951873079836,
+      "learning_rate": 1.437543408864039e-05,
+      "loss": 0.9913,
+      "step": 15925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.750492101764345,
+      "learning_rate": 1.4374747951450285e-05,
+      "loss": 0.8331,
+      "step": 15926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.755706930432019,
+      "learning_rate": 1.4374061788789014e-05,
+      "loss": 0.7784,
+      "step": 15927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9337671381583388,
+      "learning_rate": 1.4373375600660567e-05,
+      "loss": 0.9778,
+      "step": 15928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.833578684403201,
+      "learning_rate": 1.4372689387068943e-05,
+      "loss": 0.9398,
+      "step": 15929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7643681004835257,
+      "learning_rate": 1.4372003148018139e-05,
+      "loss": 0.7219,
+      "step": 15930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.94792082668785,
+      "learning_rate": 1.4371316883512147e-05,
+      "loss": 0.7703,
+      "step": 15931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6137923542571533,
+      "learning_rate": 1.4370630593554965e-05,
+      "loss": 0.9455,
+      "step": 15932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7697400457784036,
+      "learning_rate": 1.4369944278150587e-05,
+      "loss": 0.908,
+      "step": 15933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8143619293679563,
+      "learning_rate": 1.4369257937303012e-05,
+      "loss": 0.8669,
+      "step": 15934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5757207768201946,
+      "learning_rate": 1.4368571571016232e-05,
+      "loss": 0.944,
+      "step": 15935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.482568825913921,
+      "learning_rate": 1.4367885179294248e-05,
+      "loss": 0.8423,
+      "step": 15936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9701825816114065,
+      "learning_rate": 1.436719876214105e-05,
+      "loss": 0.9469,
+      "step": 15937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.7137545480323535,
+      "learning_rate": 1.4366512319560642e-05,
+      "loss": 0.9797,
+      "step": 15938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6685342336246802,
+      "learning_rate": 1.4365825851557012e-05,
+      "loss": 0.8581,
+      "step": 15939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7650204090187913,
+      "learning_rate": 1.4365139358134166e-05,
+      "loss": 0.7159,
+      "step": 15940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7481256018038488,
+      "learning_rate": 1.4364452839296092e-05,
+      "loss": 0.7963,
+      "step": 15941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6111894028185585,
+      "learning_rate": 1.4363766295046797e-05,
+      "loss": 1.0017,
+      "step": 15942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7984768138754548,
+      "learning_rate": 1.436307972539027e-05,
+      "loss": 0.8916,
+      "step": 15943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5201102281642962,
+      "learning_rate": 1.4362393130330512e-05,
+      "loss": 0.798,
+      "step": 15944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.237045425132667,
+      "learning_rate": 1.4361706509871523e-05,
+      "loss": 0.8581,
+      "step": 15945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7225911147793258,
+      "learning_rate": 1.4361019864017293e-05,
+      "loss": 0.8483,
+      "step": 15946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.9115101065263036,
+      "learning_rate": 1.436033319277183e-05,
+      "loss": 0.9049,
+      "step": 15947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6536243104527457,
+      "learning_rate": 1.4359646496139122e-05,
+      "loss": 0.8586,
+      "step": 15948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4674476384985904,
+      "learning_rate": 1.4358959774123174e-05,
+      "loss": 0.9464,
+      "step": 15949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.605059765920261,
+      "learning_rate": 1.435827302672798e-05,
+      "loss": 0.777,
+      "step": 15950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.729536998674166,
+      "learning_rate": 1.4357586253957545e-05,
+      "loss": 0.953,
+      "step": 15951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7597134284423748,
+      "learning_rate": 1.435689945581586e-05,
+      "loss": 0.7105,
+      "step": 15952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6559295858524465,
+      "learning_rate": 1.4356212632306927e-05,
+      "loss": 0.8332,
+      "step": 15953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.60662922610817,
+      "learning_rate": 1.4355525783434746e-05,
+      "loss": 0.8652,
+      "step": 15954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1073434973429137,
+      "learning_rate": 1.4354838909203315e-05,
+      "loss": 0.8966,
+      "step": 15955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5114079917687853,
+      "learning_rate": 1.4354152009616634e-05,
+      "loss": 0.9479,
+      "step": 15956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7051646550763806,
+      "learning_rate": 1.4353465084678697e-05,
+      "loss": 0.8516,
+      "step": 15957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5833061602914962,
+      "learning_rate": 1.4352778134393513e-05,
+      "loss": 0.8078,
+      "step": 15958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6730000400929366,
+      "learning_rate": 1.4352091158765074e-05,
+      "loss": 0.9337,
+      "step": 15959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6369053350303737,
+      "learning_rate": 1.4351404157797383e-05,
+      "loss": 1.0904,
+      "step": 15960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.810667084479689,
+      "learning_rate": 1.4350717131494438e-05,
+      "loss": 0.9328,
+      "step": 15961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.059989486663376,
+      "learning_rate": 1.4350030079860245e-05,
+      "loss": 0.8665,
+      "step": 15962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7110707464041608,
+      "learning_rate": 1.4349343002898795e-05,
+      "loss": 0.9187,
+      "step": 15963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.534490409554429,
+      "learning_rate": 1.4348655900614094e-05,
+      "loss": 0.9886,
+      "step": 15964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7894456982985452,
+      "learning_rate": 1.4347968773010143e-05,
+      "loss": 0.8788,
+      "step": 15965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6066471865812388,
+      "learning_rate": 1.4347281620090942e-05,
+      "loss": 0.8223,
+      "step": 15966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4765638804165155,
+      "learning_rate": 1.434659444186049e-05,
+      "loss": 0.8715,
+      "step": 15967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.740697925078957,
+      "learning_rate": 1.434590723832279e-05,
+      "loss": 0.9561,
+      "step": 15968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8818386569174224,
+      "learning_rate": 1.4345220009481837e-05,
+      "loss": 0.8863,
+      "step": 15969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7250738409342272,
+      "learning_rate": 1.4344532755341645e-05,
+      "loss": 0.9009,
+      "step": 15970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.771904915302908,
+      "learning_rate": 1.4343845475906201e-05,
+      "loss": 0.8703,
+      "step": 15971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.498033593576951,
+      "learning_rate": 1.434315817117952e-05,
+      "loss": 0.9406,
+      "step": 15972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0303957073714067,
+      "learning_rate": 1.4342470841165592e-05,
+      "loss": 0.846,
+      "step": 15973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4062683879505604,
+      "learning_rate": 1.4341783485868426e-05,
+      "loss": 0.882,
+      "step": 15974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.608566263065127,
+      "learning_rate": 1.4341096105292021e-05,
+      "loss": 0.866,
+      "step": 15975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6265414104931848,
+      "learning_rate": 1.434040869944038e-05,
+      "loss": 0.7228,
+      "step": 15976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5749736197241841,
+      "learning_rate": 1.4339721268317508e-05,
+      "loss": 0.9201,
+      "step": 15977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.575472662583977,
+      "learning_rate": 1.43390338119274e-05,
+      "loss": 0.9186,
+      "step": 15978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5570422760120068,
+      "learning_rate": 1.4338346330274067e-05,
+      "loss": 0.8628,
+      "step": 15979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6358849726816356,
+      "learning_rate": 1.4337658823361509e-05,
+      "loss": 0.8542,
+      "step": 15980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7462345109457302,
+      "learning_rate": 1.4336971291193726e-05,
+      "loss": 0.812,
+      "step": 15981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7876313963935806,
+      "learning_rate": 1.4336283733774722e-05,
+      "loss": 0.8348,
+      "step": 15982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7052057003508856,
+      "learning_rate": 1.4335596151108504e-05,
+      "loss": 0.7997,
+      "step": 15983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6031770509726608,
+      "learning_rate": 1.433490854319907e-05,
+      "loss": 0.9386,
+      "step": 15984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0419398269417752,
+      "learning_rate": 1.4334220910050429e-05,
+      "loss": 0.9241,
+      "step": 15985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6657623421872256,
+      "learning_rate": 1.433353325166658e-05,
+      "loss": 0.8174,
+      "step": 15986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5569383964316152,
+      "learning_rate": 1.4332845568051532e-05,
+      "loss": 0.8642,
+      "step": 15987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8350920446502456,
+      "learning_rate": 1.4332157859209278e-05,
+      "loss": 0.8053,
+      "step": 15988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0255604791600037,
+      "learning_rate": 1.4331470125143835e-05,
+      "loss": 0.8983,
+      "step": 15989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6336523129089375,
+      "learning_rate": 1.43307823658592e-05,
+      "loss": 0.8876,
+      "step": 15990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9799280996958204,
+      "learning_rate": 1.433009458135938e-05,
+      "loss": 0.775,
+      "step": 15991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9228593929723683,
+      "learning_rate": 1.4329406771648375e-05,
+      "loss": 0.918,
+      "step": 15992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.052624524244903,
+      "learning_rate": 1.4328718936730197e-05,
+      "loss": 0.9087,
+      "step": 15993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9443289846102458,
+      "learning_rate": 1.4328031076608847e-05,
+      "loss": 0.8241,
+      "step": 15994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.049317993302208,
+      "learning_rate": 1.4327343191288325e-05,
+      "loss": 0.7115,
+      "step": 15995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.020024246653382,
+      "learning_rate": 1.4326655280772648e-05,
+      "loss": 0.8859,
+      "step": 15996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6431858673816049,
+      "learning_rate": 1.4325967345065808e-05,
+      "loss": 0.9797,
+      "step": 15997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6301084863284592,
+      "learning_rate": 1.432527938417182e-05,
+      "loss": 0.8545,
+      "step": 15998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2537363636711327,
+      "learning_rate": 1.4324591398094683e-05,
+      "loss": 0.8554,
+      "step": 15999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.602153702430731,
+      "learning_rate": 1.4323903386838409e-05,
+      "loss": 0.8561,
+      "step": 16000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.798264155186669,
+      "learning_rate": 1.4323215350406998e-05,
+      "loss": 0.737,
+      "step": 16001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.963101597839504,
+      "learning_rate": 1.4322527288804461e-05,
+      "loss": 0.8531,
+      "step": 16002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.428775689593906,
+      "learning_rate": 1.4321839202034799e-05,
+      "loss": 0.887,
+      "step": 16003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7092966810015404,
+      "learning_rate": 1.4321151090102024e-05,
+      "loss": 0.938,
+      "step": 16004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5174975116425575,
+      "learning_rate": 1.4320462953010137e-05,
+      "loss": 0.8551,
+      "step": 16005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5247581946979676,
+      "learning_rate": 1.4319774790763148e-05,
+      "loss": 0.9004,
+      "step": 16006
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6507089570089444,
+      "learning_rate": 1.4319086603365062e-05,
+      "loss": 0.9597,
+      "step": 16007
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.59846864273824,
+      "learning_rate": 1.4318398390819885e-05,
+      "loss": 0.8154,
+      "step": 16008
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.708956727285678,
+      "learning_rate": 1.4317710153131628e-05,
+      "loss": 0.7223,
+      "step": 16009
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0940484763397995,
+      "learning_rate": 1.4317021890304294e-05,
+      "loss": 0.9761,
+      "step": 16010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5833780287763346,
+      "learning_rate": 1.4316333602341892e-05,
+      "loss": 0.8563,
+      "step": 16011
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0046265925156255,
+      "learning_rate": 1.431564528924843e-05,
+      "loss": 1.0156,
+      "step": 16012
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.643952260830392,
+      "learning_rate": 1.4314956951027914e-05,
+      "loss": 0.8971,
+      "step": 16013
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0219692175153554,
+      "learning_rate": 1.4314268587684353e-05,
+      "loss": 0.8537,
+      "step": 16014
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.348129984424284,
+      "learning_rate": 1.4313580199221754e-05,
+      "loss": 0.8874,
+      "step": 16015
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.575033345807746,
+      "learning_rate": 1.4312891785644124e-05,
+      "loss": 1.0248,
+      "step": 16016
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6863674632932963,
+      "learning_rate": 1.4312203346955478e-05,
+      "loss": 0.8447,
+      "step": 16017
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7920295609720672,
+      "learning_rate": 1.4311514883159812e-05,
+      "loss": 0.7359,
+      "step": 16018
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4578619494840745,
+      "learning_rate": 1.4310826394261148e-05,
+      "loss": 0.8666,
+      "step": 16019
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8236216659945501,
+      "learning_rate": 1.4310137880263481e-05,
+      "loss": 0.9609,
+      "step": 16020
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9479192160119665,
+      "learning_rate": 1.4309449341170832e-05,
+      "loss": 0.8904,
+      "step": 16021
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7613460031691244,
+      "learning_rate": 1.43087607769872e-05,
+      "loss": 0.901,
+      "step": 16022
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7447547856617839,
+      "learning_rate": 1.4308072187716604e-05,
+      "loss": 0.7341,
+      "step": 16023
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0057172641933527,
+      "learning_rate": 1.4307383573363044e-05,
+      "loss": 0.7972,
+      "step": 16024
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0711752666882095,
+      "learning_rate": 1.4306694933930537e-05,
+      "loss": 0.8182,
+      "step": 16025
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.803793865291469,
+      "learning_rate": 1.4306006269423082e-05,
+      "loss": 1.0058,
+      "step": 16026
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.886088563096927,
+      "learning_rate": 1.4305317579844701e-05,
+      "loss": 0.7163,
+      "step": 16027
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0631128191883805,
+      "learning_rate": 1.4304628865199397e-05,
+      "loss": 0.9412,
+      "step": 16028
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4390680246911747,
+      "learning_rate": 1.4303940125491175e-05,
+      "loss": 0.9081,
+      "step": 16029
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5795925016358727,
+      "learning_rate": 1.430325136072406e-05,
+      "loss": 0.9984,
+      "step": 16030
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.768744231530637,
+      "learning_rate": 1.4302562570902048e-05,
+      "loss": 0.7301,
+      "step": 16031
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6473076553822406,
+      "learning_rate": 1.4301873756029156e-05,
+      "loss": 0.7768,
+      "step": 16032
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4121690778061984,
+      "learning_rate": 1.4301184916109395e-05,
+      "loss": 0.8019,
+      "step": 16033
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6088068030904483,
+      "learning_rate": 1.430049605114677e-05,
+      "loss": 0.9804,
+      "step": 16034
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.824001985596462,
+      "learning_rate": 1.4299807161145297e-05,
+      "loss": 0.8453,
+      "step": 16035
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.471334287075639,
+      "learning_rate": 1.4299118246108985e-05,
+      "loss": 0.9289,
+      "step": 16036
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8045008932383217,
+      "learning_rate": 1.4298429306041846e-05,
+      "loss": 0.9816,
+      "step": 16037
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.764518205798001,
+      "learning_rate": 1.4297740340947891e-05,
+      "loss": 0.7572,
+      "step": 16038
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1152260476968197,
+      "learning_rate": 1.429705135083113e-05,
+      "loss": 0.9214,
+      "step": 16039
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5979330312157776,
+      "learning_rate": 1.4296362335695578e-05,
+      "loss": 0.7547,
+      "step": 16040
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9254703022993707,
+      "learning_rate": 1.429567329554524e-05,
+      "loss": 0.8626,
+      "step": 16041
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6038869779605838,
+      "learning_rate": 1.4294984230384135e-05,
+      "loss": 0.9568,
+      "step": 16042
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7268467172267267,
+      "learning_rate": 1.429429514021627e-05,
+      "loss": 0.7168,
+      "step": 16043
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.788737656137684,
+      "learning_rate": 1.4293606025045663e-05,
+      "loss": 0.7997,
+      "step": 16044
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.518856214634286,
+      "learning_rate": 1.429291688487632e-05,
+      "loss": 0.8301,
+      "step": 16045
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9611114665904783,
+      "learning_rate": 1.4292227719712254e-05,
+      "loss": 0.7567,
+      "step": 16046
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8057687024977433,
+      "learning_rate": 1.4291538529557485e-05,
+      "loss": 0.9441,
+      "step": 16047
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.086729575083095,
+      "learning_rate": 1.4290849314416014e-05,
+      "loss": 0.8541,
+      "step": 16048
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.824505879703375,
+      "learning_rate": 1.4290160074291864e-05,
+      "loss": 0.9829,
+      "step": 16049
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7483018544975377,
+      "learning_rate": 1.4289470809189039e-05,
+      "loss": 0.8531,
+      "step": 16050
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.617449365734648,
+      "learning_rate": 1.4288781519111563e-05,
+      "loss": 0.9253,
+      "step": 16051
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7460101884056938,
+      "learning_rate": 1.428809220406344e-05,
+      "loss": 0.9654,
+      "step": 16052
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5497682985310586,
+      "learning_rate": 1.4287402864048686e-05,
+      "loss": 0.8621,
+      "step": 16053
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.49790038744163,
+      "learning_rate": 1.4286713499071313e-05,
+      "loss": 0.984,
+      "step": 16054
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6142954997380097,
+      "learning_rate": 1.428602410913534e-05,
+      "loss": 1.0192,
+      "step": 16055
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5650529804194648,
+      "learning_rate": 1.4285334694244778e-05,
+      "loss": 0.9425,
+      "step": 16056
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6985308120387468,
+      "learning_rate": 1.4284645254403642e-05,
+      "loss": 0.9814,
+      "step": 16057
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5990955980648989,
+      "learning_rate": 1.4283955789615942e-05,
+      "loss": 0.844,
+      "step": 16058
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4537601963319071,
+      "learning_rate": 1.4283266299885696e-05,
+      "loss": 0.8925,
+      "step": 16059
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8838487200575154,
+      "learning_rate": 1.4282576785216917e-05,
+      "loss": 0.9096,
+      "step": 16060
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4543401809447942,
+      "learning_rate": 1.4281887245613621e-05,
+      "loss": 0.9753,
+      "step": 16061
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8271008751141335,
+      "learning_rate": 1.4281197681079821e-05,
+      "loss": 0.9003,
+      "step": 16062
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.515098393573591,
+      "learning_rate": 1.4280508091619534e-05,
+      "loss": 0.7022,
+      "step": 16063
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8290625000054992,
+      "learning_rate": 1.4279818477236773e-05,
+      "loss": 0.9131,
+      "step": 16064
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4313339205980933,
+      "learning_rate": 1.4279128837935554e-05,
+      "loss": 0.8552,
+      "step": 16065
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.609059528207484,
+      "learning_rate": 1.4278439173719892e-05,
+      "loss": 0.8711,
+      "step": 16066
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5080853195367574,
+      "learning_rate": 1.4277749484593801e-05,
+      "loss": 0.9262,
+      "step": 16067
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6223024750318291,
+      "learning_rate": 1.4277059770561304e-05,
+      "loss": 1.0064,
+      "step": 16068
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7491045862098271,
+      "learning_rate": 1.4276370031626403e-05,
+      "loss": 1.0119,
+      "step": 16069
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0825905420767583,
+      "learning_rate": 1.4275680267793125e-05,
+      "loss": 0.7967,
+      "step": 16070
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9843624804917996,
+      "learning_rate": 1.4274990479065483e-05,
+      "loss": 0.9045,
+      "step": 16071
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6952163855067535,
+      "learning_rate": 1.4274300665447496e-05,
+      "loss": 0.8476,
+      "step": 16072
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7526447365882634,
+      "learning_rate": 1.4273610826943171e-05,
+      "loss": 0.9252,
+      "step": 16073
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7387792811767144,
+      "learning_rate": 1.4272920963556537e-05,
+      "loss": 0.727,
+      "step": 16074
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7296060255959798,
+      "learning_rate": 1.4272231075291603e-05,
+      "loss": 1.0458,
+      "step": 16075
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9374933738989353,
+      "learning_rate": 1.4271541162152385e-05,
+      "loss": 0.9337,
+      "step": 16076
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8080076789463957,
+      "learning_rate": 1.4270851224142904e-05,
+      "loss": 0.7652,
+      "step": 16077
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5232507761665317,
+      "learning_rate": 1.427016126126717e-05,
+      "loss": 0.9386,
+      "step": 16078
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7522725738811723,
+      "learning_rate": 1.4269471273529212e-05,
+      "loss": 0.7714,
+      "step": 16079
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6626433839195198,
+      "learning_rate": 1.4268781260933034e-05,
+      "loss": 0.9382,
+      "step": 16080
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.573148024216689,
+      "learning_rate": 1.4268091223482667e-05,
+      "loss": 0.8042,
+      "step": 16081
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.129784857980236,
+      "learning_rate": 1.4267401161182117e-05,
+      "loss": 0.8688,
+      "step": 16082
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6762658831223043,
+      "learning_rate": 1.4266711074035408e-05,
+      "loss": 0.9362,
+      "step": 16083
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7547905932333785,
+      "learning_rate": 1.4266020962046555e-05,
+      "loss": 0.8292,
+      "step": 16084
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3666609929385682,
+      "learning_rate": 1.4265330825219578e-05,
+      "loss": 0.8252,
+      "step": 16085
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6308260107268082,
+      "learning_rate": 1.4264640663558496e-05,
+      "loss": 0.9012,
+      "step": 16086
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.743134362462855,
+      "learning_rate": 1.4263950477067326e-05,
+      "loss": 0.7406,
+      "step": 16087
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4896647648701489,
+      "learning_rate": 1.4263260265750085e-05,
+      "loss": 0.9251,
+      "step": 16088
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.575613443766245,
+      "learning_rate": 1.4262570029610792e-05,
+      "loss": 0.9963,
+      "step": 16089
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5448168532833588,
+      "learning_rate": 1.4261879768653469e-05,
+      "loss": 1.0692,
+      "step": 16090
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7844946403162514,
+      "learning_rate": 1.4261189482882131e-05,
+      "loss": 0.8909,
+      "step": 16091
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5596189319264349,
+      "learning_rate": 1.4260499172300798e-05,
+      "loss": 0.916,
+      "step": 16092
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6427702850880286,
+      "learning_rate": 1.425980883691349e-05,
+      "loss": 0.9252,
+      "step": 16093
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5039194723421188,
+      "learning_rate": 1.425911847672423e-05,
+      "loss": 0.9412,
+      "step": 16094
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8559669335450386,
+      "learning_rate": 1.4258428091737031e-05,
+      "loss": 0.9594,
+      "step": 16095
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3903840812875192,
+      "learning_rate": 1.425773768195592e-05,
+      "loss": 0.8299,
+      "step": 16096
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6868826768492142,
+      "learning_rate": 1.4257047247384904e-05,
+      "loss": 0.9644,
+      "step": 16097
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5669002682562103,
+      "learning_rate": 1.4256356788028018e-05,
+      "loss": 1.0383,
+      "step": 16098
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8343673454955114,
+      "learning_rate": 1.425566630388927e-05,
+      "loss": 0.875,
+      "step": 16099
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8003100698023302,
+      "learning_rate": 1.425497579497269e-05,
+      "loss": 0.801,
+      "step": 16100
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9245687235895999,
+      "learning_rate": 1.425428526128229e-05,
+      "loss": 0.8879,
+      "step": 16101
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5176653451844866,
+      "learning_rate": 1.42535947028221e-05,
+      "loss": 0.9216,
+      "step": 16102
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5754233839707295,
+      "learning_rate": 1.4252904119596129e-05,
+      "loss": 0.8122,
+      "step": 16103
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.45549080864164,
+      "learning_rate": 1.4252213511608406e-05,
+      "loss": 0.8611,
+      "step": 16104
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5212294566959312,
+      "learning_rate": 1.425152287886295e-05,
+      "loss": 0.8725,
+      "step": 16105
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7929512608211216,
+      "learning_rate": 1.4250832221363782e-05,
+      "loss": 0.8908,
+      "step": 16106
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7552748677615169,
+      "learning_rate": 1.425014153911492e-05,
+      "loss": 0.9644,
+      "step": 16107
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.496476225920551,
+      "learning_rate": 1.4249450832120393e-05,
+      "loss": 0.85,
+      "step": 16108
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6837528025638704,
+      "learning_rate": 1.4248760100384216e-05,
+      "loss": 0.8405,
+      "step": 16109
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6505692789657935,
+      "learning_rate": 1.4248069343910412e-05,
+      "loss": 0.8772,
+      "step": 16110
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.644700108934761,
+      "learning_rate": 1.4247378562703004e-05,
+      "loss": 0.9714,
+      "step": 16111
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6757517985513022,
+      "learning_rate": 1.4246687756766013e-05,
+      "loss": 0.9424,
+      "step": 16112
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6959445124719874,
+      "learning_rate": 1.424599692610346e-05,
+      "loss": 1.0756,
+      "step": 16113
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6738424085975205,
+      "learning_rate": 1.424530607071937e-05,
+      "loss": 0.9195,
+      "step": 16114
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.24515353587595,
+      "learning_rate": 1.4244615190617765e-05,
+      "loss": 0.8875,
+      "step": 16115
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6802211881165245,
+      "learning_rate": 1.4243924285802665e-05,
+      "loss": 0.8931,
+      "step": 16116
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6196381144675875,
+      "learning_rate": 1.4243233356278097e-05,
+      "loss": 0.8544,
+      "step": 16117
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9188695826220028,
+      "learning_rate": 1.4242542402048076e-05,
+      "loss": 0.9037,
+      "step": 16118
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2097129690430894,
+      "learning_rate": 1.4241851423116634e-05,
+      "loss": 0.9432,
+      "step": 16119
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5264511109923744,
+      "learning_rate": 1.4241160419487789e-05,
+      "loss": 0.8848,
+      "step": 16120
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.209348498064746,
+      "learning_rate": 1.4240469391165569e-05,
+      "loss": 0.898,
+      "step": 16121
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7897844131167981,
+      "learning_rate": 1.4239778338153984e-05,
+      "loss": 0.8871,
+      "step": 16122
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8042428909398789,
+      "learning_rate": 1.4239087260457076e-05,
+      "loss": 0.8299,
+      "step": 16123
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4901746359652903,
+      "learning_rate": 1.4238396158078859e-05,
+      "loss": 0.9196,
+      "step": 16124
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6793269238908115,
+      "learning_rate": 1.4237705031023354e-05,
+      "loss": 0.8751,
+      "step": 16125
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9612318820446817,
+      "learning_rate": 1.423701387929459e-05,
+      "loss": 1.075,
+      "step": 16126
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7920981515671615,
+      "learning_rate": 1.4236322702896588e-05,
+      "loss": 0.8135,
+      "step": 16127
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7008609637822414,
+      "learning_rate": 1.4235631501833378e-05,
+      "loss": 0.8453,
+      "step": 16128
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6713788045473754,
+      "learning_rate": 1.4234940276108977e-05,
+      "loss": 0.9586,
+      "step": 16129
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7253041641818718,
+      "learning_rate": 1.4234249025727419e-05,
+      "loss": 0.8553,
+      "step": 16130
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7400741492015,
+      "learning_rate": 1.4233557750692715e-05,
+      "loss": 0.7371,
+      "step": 16131
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8030465478820403,
+      "learning_rate": 1.4232866451008902e-05,
+      "loss": 0.8811,
+      "step": 16132
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9774838868291051,
+      "learning_rate": 1.4232175126679999e-05,
+      "loss": 0.8415,
+      "step": 16133
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5578618000674254,
+      "learning_rate": 1.423148377771003e-05,
+      "loss": 0.8844,
+      "step": 16134
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8080753866082352,
+      "learning_rate": 1.4230792404103026e-05,
+      "loss": 0.7149,
+      "step": 16135
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6801321041100057,
+      "learning_rate": 1.4230101005863008e-05,
+      "loss": 0.9609,
+      "step": 16136
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4744915798526992,
+      "learning_rate": 1.4229409582994005e-05,
+      "loss": 0.7317,
+      "step": 16137
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3695278322803035,
+      "learning_rate": 1.422871813550004e-05,
+      "loss": 0.8504,
+      "step": 16138
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7239622156760737,
+      "learning_rate": 1.4228026663385136e-05,
+      "loss": 0.8129,
+      "step": 16139
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1849313118560945,
+      "learning_rate": 1.4227335166653325e-05,
+      "loss": 0.9155,
+      "step": 16140
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7579419561070178,
+      "learning_rate": 1.4226643645308628e-05,
+      "loss": 0.7114,
+      "step": 16141
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.621997027023434,
+      "learning_rate": 1.4225952099355075e-05,
+      "loss": 0.8248,
+      "step": 16142
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7003385938033049,
+      "learning_rate": 1.4225260528796692e-05,
+      "loss": 0.9146,
+      "step": 16143
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7938746516352453,
+      "learning_rate": 1.42245689336375e-05,
+      "loss": 0.9032,
+      "step": 16144
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4098180859454885,
+      "learning_rate": 1.4223877313881535e-05,
+      "loss": 0.8122,
+      "step": 16145
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6668064580890627,
+      "learning_rate": 1.422318566953282e-05,
+      "loss": 0.828,
+      "step": 16146
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.752363733863888,
+      "learning_rate": 1.422249400059538e-05,
+      "loss": 0.8256,
+      "step": 16147
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6089539190962692,
+      "learning_rate": 1.422180230707324e-05,
+      "loss": 0.9187,
+      "step": 16148
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.85099220674368,
+      "learning_rate": 1.4221110588970432e-05,
+      "loss": 1.0494,
+      "step": 16149
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.660103172109404,
+      "learning_rate": 1.4220418846290981e-05,
+      "loss": 0.8821,
+      "step": 16150
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.569980636525123,
+      "learning_rate": 1.4219727079038919e-05,
+      "loss": 0.9059,
+      "step": 16151
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7683228673233472,
+      "learning_rate": 1.4219035287218264e-05,
+      "loss": 0.6995,
+      "step": 16152
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.780568530695834,
+      "learning_rate": 1.4218343470833055e-05,
+      "loss": 0.8537,
+      "step": 16153
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5099945182847705,
+      "learning_rate": 1.4217651629887312e-05,
+      "loss": 0.8395,
+      "step": 16154
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.503006874249419,
+      "learning_rate": 1.4216959764385067e-05,
+      "loss": 0.8275,
+      "step": 16155
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7529384728332291,
+      "learning_rate": 1.4216267874330347e-05,
+      "loss": 0.7713,
+      "step": 16156
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.609053756265527,
+      "learning_rate": 1.421557595972718e-05,
+      "loss": 0.8923,
+      "step": 16157
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5678541351066195,
+      "learning_rate": 1.4214884020579595e-05,
+      "loss": 1.0248,
+      "step": 16158
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8657288040354607,
+      "learning_rate": 1.4214192056891622e-05,
+      "loss": 0.7936,
+      "step": 16159
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.472389785059915,
+      "learning_rate": 1.421350006866729e-05,
+      "loss": 0.8501,
+      "step": 16160
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5508130654544416,
+      "learning_rate": 1.4212808055910623e-05,
+      "loss": 0.8793,
+      "step": 16161
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8568313371367207,
+      "learning_rate": 1.4212116018625658e-05,
+      "loss": 0.9551,
+      "step": 16162
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.777748438858209,
+      "learning_rate": 1.4211423956816415e-05,
+      "loss": 0.7716,
+      "step": 16163
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.553667157477652,
+      "learning_rate": 1.4210731870486931e-05,
+      "loss": 0.9027,
+      "step": 16164
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.79740175013917,
+      "learning_rate": 1.4210039759641232e-05,
+      "loss": 0.852,
+      "step": 16165
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.973217303565555,
+      "learning_rate": 1.4209347624283352e-05,
+      "loss": 0.7962,
+      "step": 16166
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.595925875173632,
+      "learning_rate": 1.4208655464417313e-05,
+      "loss": 0.9325,
+      "step": 16167
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7823764047274144,
+      "learning_rate": 1.4207963280047155e-05,
+      "loss": 0.9698,
+      "step": 16168
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7877241832180093,
+      "learning_rate": 1.4207271071176896e-05,
+      "loss": 0.9621,
+      "step": 16169
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.653002392231291,
+      "learning_rate": 1.4206578837810578e-05,
+      "loss": 0.8874,
+      "step": 16170
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7945627307848646,
+      "learning_rate": 1.4205886579952221e-05,
+      "loss": 0.6935,
+      "step": 16171
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4203765589744162,
+      "learning_rate": 1.4205194297605867e-05,
+      "loss": 0.8909,
+      "step": 16172
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.723538781919491,
+      "learning_rate": 1.4204501990775535e-05,
+      "loss": 0.9432,
+      "step": 16173
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.37430933132267,
+      "learning_rate": 1.4203809659465266e-05,
+      "loss": 0.8353,
+      "step": 16174
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5631640059375935,
+      "learning_rate": 1.4203117303679083e-05,
+      "loss": 0.7753,
+      "step": 16175
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5501947758771197,
+      "learning_rate": 1.4202424923421018e-05,
+      "loss": 0.8491,
+      "step": 16176
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7754768192604076,
+      "learning_rate": 1.420173251869511e-05,
+      "loss": 0.7578,
+      "step": 16177
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7146761405740776,
+      "learning_rate": 1.4201040089505383e-05,
+      "loss": 0.9172,
+      "step": 16178
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8085240725155242,
+      "learning_rate": 1.420034763585587e-05,
+      "loss": 0.8026,
+      "step": 16179
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7460953463570021,
+      "learning_rate": 1.4199655157750602e-05,
+      "loss": 1.011,
+      "step": 16180
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5588641786001716,
+      "learning_rate": 1.4198962655193617e-05,
+      "loss": 1.0333,
+      "step": 16181
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8395291947941719,
+      "learning_rate": 1.4198270128188938e-05,
+      "loss": 0.9017,
+      "step": 16182
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9360587547651111,
+      "learning_rate": 1.4197577576740602e-05,
+      "loss": 0.9172,
+      "step": 16183
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6603425390895021,
+      "learning_rate": 1.419688500085264e-05,
+      "loss": 0.9188,
+      "step": 16184
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7272378610480112,
+      "learning_rate": 1.4196192400529088e-05,
+      "loss": 0.7539,
+      "step": 16185
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8313999852599283,
+      "learning_rate": 1.4195499775773972e-05,
+      "loss": 0.9053,
+      "step": 16186
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8599094525097144,
+      "learning_rate": 1.4194807126591328e-05,
+      "loss": 0.772,
+      "step": 16187
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.549976054092872,
+      "learning_rate": 1.4194114452985189e-05,
+      "loss": 0.8247,
+      "step": 16188
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.81389248008583,
+      "learning_rate": 1.4193421754959586e-05,
+      "loss": 0.8065,
+      "step": 16189
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.106183097032642,
+      "learning_rate": 1.4192729032518557e-05,
+      "loss": 0.8228,
+      "step": 16190
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.110715033007815,
+      "learning_rate": 1.4192036285666133e-05,
+      "loss": 0.9039,
+      "step": 16191
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4723719680496605,
+      "learning_rate": 1.4191343514406345e-05,
+      "loss": 0.9542,
+      "step": 16192
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6543075794353017,
+      "learning_rate": 1.4190650718743229e-05,
+      "loss": 0.9413,
+      "step": 16193
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9450946369229967,
+      "learning_rate": 1.4189957898680815e-05,
+      "loss": 0.9196,
+      "step": 16194
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9416005450632567,
+      "learning_rate": 1.4189265054223142e-05,
+      "loss": 0.9901,
+      "step": 16195
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5753550635453366,
+      "learning_rate": 1.4188572185374241e-05,
+      "loss": 0.8698,
+      "step": 16196
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4448836011397346,
+      "learning_rate": 1.4187879292138144e-05,
+      "loss": 0.8957,
+      "step": 16197
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8349366985829474,
+      "learning_rate": 1.4187186374518893e-05,
+      "loss": 0.8638,
+      "step": 16198
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4693610482587793,
+      "learning_rate": 1.4186493432520513e-05,
+      "loss": 0.8075,
+      "step": 16199
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.745723853798576,
+      "learning_rate": 1.4185800466147048e-05,
+      "loss": 0.9055,
+      "step": 16200
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3875485866725765,
+      "learning_rate": 1.4185107475402522e-05,
+      "loss": 0.7338,
+      "step": 16201
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.57577859658929,
+      "learning_rate": 1.4184414460290979e-05,
+      "loss": 0.858,
+      "step": 16202
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7499418626528247,
+      "learning_rate": 1.4183721420816448e-05,
+      "loss": 0.7852,
+      "step": 16203
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4434138144483106,
+      "learning_rate": 1.4183028356982969e-05,
+      "loss": 0.9274,
+      "step": 16204
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8202298520148832,
+      "learning_rate": 1.4182335268794571e-05,
+      "loss": 0.8288,
+      "step": 16205
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.628993638642422,
+      "learning_rate": 1.4181642156255295e-05,
+      "loss": 0.9619,
+      "step": 16206
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5406394483943529,
+      "learning_rate": 1.4180949019369174e-05,
+      "loss": 0.784,
+      "step": 16207
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.484334539266702,
+      "learning_rate": 1.4180255858140243e-05,
+      "loss": 0.8389,
+      "step": 16208
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9317297939313007,
+      "learning_rate": 1.417956267257254e-05,
+      "loss": 0.8148,
+      "step": 16209
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5016746814568518,
+      "learning_rate": 1.4178869462670102e-05,
+      "loss": 0.8761,
+      "step": 16210
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7400659628577604,
+      "learning_rate": 1.417817622843696e-05,
+      "loss": 0.9185,
+      "step": 16211
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.76521312050275,
+      "learning_rate": 1.4177482969877155e-05,
+      "loss": 0.9155,
+      "step": 16212
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8500437449637933,
+      "learning_rate": 1.4176789686994718e-05,
+      "loss": 0.7025,
+      "step": 16213
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7717187137918733,
+      "learning_rate": 1.4176096379793694e-05,
+      "loss": 0.8908,
+      "step": 16214
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6994170422610644,
+      "learning_rate": 1.4175403048278113e-05,
+      "loss": 0.9359,
+      "step": 16215
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6252202773716788,
+      "learning_rate": 1.4174709692452014e-05,
+      "loss": 0.8033,
+      "step": 16216
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9668205833740395,
+      "learning_rate": 1.4174016312319432e-05,
+      "loss": 0.9183,
+      "step": 16217
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.745710433758525,
+      "learning_rate": 1.4173322907884406e-05,
+      "loss": 0.7795,
+      "step": 16218
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4930668892459797,
+      "learning_rate": 1.4172629479150975e-05,
+      "loss": 0.8666,
+      "step": 16219
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7324215094644906,
+      "learning_rate": 1.417193602612317e-05,
+      "loss": 0.7761,
+      "step": 16220
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6021659703606472,
+      "learning_rate": 1.4171242548805036e-05,
+      "loss": 0.9126,
+      "step": 16221
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6450213772768196,
+      "learning_rate": 1.4170549047200604e-05,
+      "loss": 0.8872,
+      "step": 16222
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.11442654416251,
+      "learning_rate": 1.416985552131392e-05,
+      "loss": 0.8151,
+      "step": 16223
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8092038294437381,
+      "learning_rate": 1.416916197114901e-05,
+      "loss": 0.7307,
+      "step": 16224
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5515299332687198,
+      "learning_rate": 1.4168468396709925e-05,
+      "loss": 0.8812,
+      "step": 16225
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.495199196142013,
+      "learning_rate": 1.4167774798000697e-05,
+      "loss": 0.7613,
+      "step": 16226
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3591018419801455,
+      "learning_rate": 1.416708117502536e-05,
+      "loss": 0.8409,
+      "step": 16227
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8075678127176136,
+      "learning_rate": 1.4166387527787964e-05,
+      "loss": 0.9107,
+      "step": 16228
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6391916556083124,
+      "learning_rate": 1.4165693856292535e-05,
+      "loss": 0.8041,
+      "step": 16229
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5502321395053844,
+      "learning_rate": 1.416500016054312e-05,
+      "loss": 0.851,
+      "step": 16230
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6145365232145137,
+      "learning_rate": 1.4164306440543754e-05,
+      "loss": 0.9479,
+      "step": 16231
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7656234447565498,
+      "learning_rate": 1.4163612696298481e-05,
+      "loss": 0.9001,
+      "step": 16232
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7393470918085183,
+      "learning_rate": 1.4162918927811334e-05,
+      "loss": 0.8598,
+      "step": 16233
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4344434158820867,
+      "learning_rate": 1.4162225135086354e-05,
+      "loss": 0.7398,
+      "step": 16234
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.790838093419211,
+      "learning_rate": 1.4161531318127584e-05,
+      "loss": 0.8674,
+      "step": 16235
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9354150225687774,
+      "learning_rate": 1.416083747693906e-05,
+      "loss": 0.8924,
+      "step": 16236
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5859935364599473,
+      "learning_rate": 1.4160143611524821e-05,
+      "loss": 0.9642,
+      "step": 16237
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6631868142168824,
+      "learning_rate": 1.4159449721888911e-05,
+      "loss": 0.8866,
+      "step": 16238
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.590592660950205,
+      "learning_rate": 1.4158755808035367e-05,
+      "loss": 0.8226,
+      "step": 16239
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7555209534923217,
+      "learning_rate": 1.4158061869968232e-05,
+      "loss": 0.7046,
+      "step": 16240
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6867247852258989,
+      "learning_rate": 1.4157367907691542e-05,
+      "loss": 0.8506,
+      "step": 16241
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6417810323450817,
+      "learning_rate": 1.4156673921209341e-05,
+      "loss": 0.8333,
+      "step": 16242
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.561371295159361,
+      "learning_rate": 1.4155979910525668e-05,
+      "loss": 0.9437,
+      "step": 16243
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7816068336544533,
+      "learning_rate": 1.4155285875644564e-05,
+      "loss": 0.9198,
+      "step": 16244
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5081722489986207,
+      "learning_rate": 1.4154591816570072e-05,
+      "loss": 0.9328,
+      "step": 16245
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0529471082379662,
+      "learning_rate": 1.4153897733306228e-05,
+      "loss": 0.9463,
+      "step": 16246
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.798256088599227,
+      "learning_rate": 1.415320362585708e-05,
+      "loss": 0.8332,
+      "step": 16247
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5817143682326005,
+      "learning_rate": 1.4152509494226662e-05,
+      "loss": 1.0319,
+      "step": 16248
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.433592809440021,
+      "learning_rate": 1.4151815338419022e-05,
+      "loss": 0.7118,
+      "step": 16249
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5804601561795173,
+      "learning_rate": 1.4151121158438195e-05,
+      "loss": 0.8776,
+      "step": 16250
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8950198432951102,
+      "learning_rate": 1.415042695428823e-05,
+      "loss": 0.8896,
+      "step": 16251
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.527258331106619,
+      "learning_rate": 1.4149732725973162e-05,
+      "loss": 0.7887,
+      "step": 16252
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4965219490676474,
+      "learning_rate": 1.414903847349704e-05,
+      "loss": 0.9189,
+      "step": 16253
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.757895527110672,
+      "learning_rate": 1.41483441968639e-05,
+      "loss": 0.9177,
+      "step": 16254
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5621097895152976,
+      "learning_rate": 1.4147649896077789e-05,
+      "loss": 0.7874,
+      "step": 16255
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6651874033928726,
+      "learning_rate": 1.4146955571142745e-05,
+      "loss": 0.9788,
+      "step": 16256
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4736407316805944,
+      "learning_rate": 1.4146261222062813e-05,
+      "loss": 0.9198,
+      "step": 16257
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9543005991704288,
+      "learning_rate": 1.4145566848842036e-05,
+      "loss": 0.8521,
+      "step": 16258
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5817657422488858,
+      "learning_rate": 1.4144872451484455e-05,
+      "loss": 0.9777,
+      "step": 16259
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6449478824474484,
+      "learning_rate": 1.4144178029994118e-05,
+      "loss": 0.8995,
+      "step": 16260
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.681358101678654,
+      "learning_rate": 1.414348358437506e-05,
+      "loss": 1.0014,
+      "step": 16261
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.566208366857859,
+      "learning_rate": 1.414278911463133e-05,
+      "loss": 1.0083,
+      "step": 16262
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5700337925639496,
+      "learning_rate": 1.4142094620766972e-05,
+      "loss": 0.982,
+      "step": 16263
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5494586072583414,
+      "learning_rate": 1.4141400102786027e-05,
+      "loss": 0.8848,
+      "step": 16264
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7376020024693171,
+      "learning_rate": 1.4140705560692538e-05,
+      "loss": 0.7213,
+      "step": 16265
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7537812718246527,
+      "learning_rate": 1.4140010994490552e-05,
+      "loss": 0.7388,
+      "step": 16266
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6504995763313877,
+      "learning_rate": 1.4139316404184111e-05,
+      "loss": 0.7295,
+      "step": 16267
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7720904990414685,
+      "learning_rate": 1.4138621789777258e-05,
+      "loss": 0.926,
+      "step": 16268
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7905715330083342,
+      "learning_rate": 1.413792715127404e-05,
+      "loss": 0.723,
+      "step": 16269
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8228821671008524,
+      "learning_rate": 1.4137232488678503e-05,
+      "loss": 0.9193,
+      "step": 16270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5952737583083465,
+      "learning_rate": 1.4136537801994683e-05,
+      "loss": 0.9768,
+      "step": 16271
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.984655244330877,
+      "learning_rate": 1.4135843091226635e-05,
+      "loss": 0.8833,
+      "step": 16272
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8468212990604661,
+      "learning_rate": 1.4135148356378393e-05,
+      "loss": 0.6776,
+      "step": 16273
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7050468593461334,
+      "learning_rate": 1.4134453597454015e-05,
+      "loss": 0.9515,
+      "step": 16274
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.891559718696087,
+      "learning_rate": 1.4133758814457534e-05,
+      "loss": 0.8782,
+      "step": 16275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6467974138833874,
+      "learning_rate": 1.4133064007393005e-05,
+      "loss": 0.8865,
+      "step": 16276
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7039347533333442,
+      "learning_rate": 1.4132369176264467e-05,
+      "loss": 0.8956,
+      "step": 16277
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4652310157508297,
+      "learning_rate": 1.4131674321075966e-05,
+      "loss": 0.76,
+      "step": 16278
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.66127875042733,
+      "learning_rate": 1.4130979441831552e-05,
+      "loss": 0.848,
+      "step": 16279
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7522700574709615,
+      "learning_rate": 1.4130284538535261e-05,
+      "loss": 0.7381,
+      "step": 16280
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6199648665440212,
+      "learning_rate": 1.4129589611191154e-05,
+      "loss": 0.9196,
+      "step": 16281
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4452020763650002,
+      "learning_rate": 1.4128894659803263e-05,
+      "loss": 0.9461,
+      "step": 16282
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6600206517316949,
+      "learning_rate": 1.4128199684375645e-05,
+      "loss": 0.8058,
+      "step": 16283
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5913190940456072,
+      "learning_rate": 1.412750468491234e-05,
+      "loss": 0.995,
+      "step": 16284
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9898333499358274,
+      "learning_rate": 1.4126809661417396e-05,
+      "loss": 1.0359,
+      "step": 16285
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5049612106209365,
+      "learning_rate": 1.4126114613894857e-05,
+      "loss": 0.9432,
+      "step": 16286
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7157781347409806,
+      "learning_rate": 1.4125419542348775e-05,
+      "loss": 0.857,
+      "step": 16287
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.547704865170717,
+      "learning_rate": 1.4124724446783194e-05,
+      "loss": 0.8916,
+      "step": 16288
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6588476901617883,
+      "learning_rate": 1.412402932720216e-05,
+      "loss": 0.9111,
+      "step": 16289
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.621041612494057,
+      "learning_rate": 1.412333418360972e-05,
+      "loss": 0.9269,
+      "step": 16290
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7475228817726776,
+      "learning_rate": 1.4122639016009927e-05,
+      "loss": 0.7836,
+      "step": 16291
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9604818427060204,
+      "learning_rate": 1.4121943824406822e-05,
+      "loss": 0.9625,
+      "step": 16292
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8736804382824475,
+      "learning_rate": 1.4121248608804456e-05,
+      "loss": 0.9965,
+      "step": 16293
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7679677567664756,
+      "learning_rate": 1.4120553369206876e-05,
+      "loss": 0.6828,
+      "step": 16294
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8246889940010385,
+      "learning_rate": 1.4119858105618129e-05,
+      "loss": 0.8456,
+      "step": 16295
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8690623371283577,
+      "learning_rate": 1.4119162818042265e-05,
+      "loss": 0.9376,
+      "step": 16296
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.79473978156736,
+      "learning_rate": 1.411846750648333e-05,
+      "loss": 0.912,
+      "step": 16297
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1328044639469197,
+      "learning_rate": 1.4117772170945375e-05,
+      "loss": 0.8814,
+      "step": 16298
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.917522050970327,
+      "learning_rate": 1.4117076811432444e-05,
+      "loss": 0.8088,
+      "step": 16299
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6128641215018789,
+      "learning_rate": 1.4116381427948591e-05,
+      "loss": 0.8049,
+      "step": 16300
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7437394662385286,
+      "learning_rate": 1.4115686020497858e-05,
+      "loss": 0.9787,
+      "step": 16301
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7037405121373845,
+      "learning_rate": 1.4114990589084304e-05,
+      "loss": 0.85,
+      "step": 16302
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6796632417012303,
+      "learning_rate": 1.4114295133711966e-05,
+      "loss": 0.843,
+      "step": 16303
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8155805108846934,
+      "learning_rate": 1.4113599654384904e-05,
+      "loss": 0.7826,
+      "step": 16304
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7544996287581849,
+      "learning_rate": 1.411290415110716e-05,
+      "loss": 0.8742,
+      "step": 16305
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5292193756937686,
+      "learning_rate": 1.4112208623882787e-05,
+      "loss": 0.9204,
+      "step": 16306
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6316123067373578,
+      "learning_rate": 1.4111513072715833e-05,
+      "loss": 0.8338,
+      "step": 16307
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.642520045412852,
+      "learning_rate": 1.4110817497610347e-05,
+      "loss": 0.9874,
+      "step": 16308
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0883578863160825,
+      "learning_rate": 1.4110121898570382e-05,
+      "loss": 0.7874,
+      "step": 16309
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6955840019657666,
+      "learning_rate": 1.4109426275599983e-05,
+      "loss": 0.8212,
+      "step": 16310
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.699387439676504,
+      "learning_rate": 1.4108730628703207e-05,
+      "loss": 0.8099,
+      "step": 16311
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.453690855261298,
+      "learning_rate": 1.4108034957884096e-05,
+      "loss": 0.8998,
+      "step": 16312
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5682111036827817,
+      "learning_rate": 1.4107339263146709e-05,
+      "loss": 0.8459,
+      "step": 16313
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0790252947604153,
+      "learning_rate": 1.4106643544495092e-05,
+      "loss": 0.845,
+      "step": 16314
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7741002177308209,
+      "learning_rate": 1.4105947801933295e-05,
+      "loss": 0.7535,
+      "step": 16315
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7205511177478938,
+      "learning_rate": 1.4105252035465368e-05,
+      "loss": 0.8798,
+      "step": 16316
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0585825794852703,
+      "learning_rate": 1.4104556245095368e-05,
+      "loss": 0.8956,
+      "step": 16317
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7703073791066217,
+      "learning_rate": 1.4103860430827342e-05,
+      "loss": 0.9596,
+      "step": 16318
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7004397139990635,
+      "learning_rate": 1.4103164592665337e-05,
+      "loss": 0.9702,
+      "step": 16319
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6464207647443119,
+      "learning_rate": 1.4102468730613412e-05,
+      "loss": 0.8745,
+      "step": 16320
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0426480281027493,
+      "learning_rate": 1.4101772844675617e-05,
+      "loss": 0.8745,
+      "step": 16321
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.637301978749284,
+      "learning_rate": 1.4101076934855997e-05,
+      "loss": 1.0741,
+      "step": 16322
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.536523473127924,
+      "learning_rate": 1.4100381001158612e-05,
+      "loss": 0.9346,
+      "step": 16323
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1375969879933017,
+      "learning_rate": 1.409968504358751e-05,
+      "loss": 0.8934,
+      "step": 16324
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.644278447624677,
+      "learning_rate": 1.4098989062146741e-05,
+      "loss": 0.8771,
+      "step": 16325
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.64442945351321,
+      "learning_rate": 1.4098293056840364e-05,
+      "loss": 0.9114,
+      "step": 16326
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7849301535978043,
+      "learning_rate": 1.4097597027672425e-05,
+      "loss": 0.7216,
+      "step": 16327
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7950241810891158,
+      "learning_rate": 1.409690097464698e-05,
+      "loss": 0.7791,
+      "step": 16328
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6120800629406031,
+      "learning_rate": 1.409620489776808e-05,
+      "loss": 0.8252,
+      "step": 16329
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.825499752307524,
+      "learning_rate": 1.4095508797039781e-05,
+      "loss": 0.836,
+      "step": 16330
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6894971240762542,
+      "learning_rate": 1.4094812672466127e-05,
+      "loss": 0.8162,
+      "step": 16331
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7063182776689823,
+      "learning_rate": 1.4094116524051182e-05,
+      "loss": 0.9012,
+      "step": 16332
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.044287117085008,
+      "learning_rate": 1.4093420351798991e-05,
+      "loss": 0.7896,
+      "step": 16333
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6647161352573046,
+      "learning_rate": 1.4092724155713614e-05,
+      "loss": 0.8941,
+      "step": 16334
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.908058468744354,
+      "learning_rate": 1.4092027935799102e-05,
+      "loss": 0.9317,
+      "step": 16335
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.600901503767181,
+      "learning_rate": 1.4091331692059506e-05,
+      "loss": 0.827,
+      "step": 16336
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7341111285729087,
+      "learning_rate": 1.4090635424498882e-05,
+      "loss": 0.9556,
+      "step": 16337
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5462101714205536,
+      "learning_rate": 1.4089939133121282e-05,
+      "loss": 0.973,
+      "step": 16338
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7735576535764024,
+      "learning_rate": 1.4089242817930764e-05,
+      "loss": 0.9541,
+      "step": 16339
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.603213014755639,
+      "learning_rate": 1.4088546478931378e-05,
+      "loss": 0.8244,
+      "step": 16340
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4858727586195326,
+      "learning_rate": 1.408785011612718e-05,
+      "loss": 0.8892,
+      "step": 16341
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5699381442133467,
+      "learning_rate": 1.4087153729522224e-05,
+      "loss": 0.8371,
+      "step": 16342
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.532042515780512,
+      "learning_rate": 1.4086457319120566e-05,
+      "loss": 0.8263,
+      "step": 16343
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6654225360904267,
+      "learning_rate": 1.408576088492626e-05,
+      "loss": 0.8691,
+      "step": 16344
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4021106800404164,
+      "learning_rate": 1.408506442694336e-05,
+      "loss": 0.8292,
+      "step": 16345
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9638326689197794,
+      "learning_rate": 1.408436794517592e-05,
+      "loss": 0.8425,
+      "step": 16346
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9155529662981339,
+      "learning_rate": 1.4083671439628e-05,
+      "loss": 0.9812,
+      "step": 16347
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4963775550697005,
+      "learning_rate": 1.4082974910303648e-05,
+      "loss": 0.9589,
+      "step": 16348
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7936997522528,
+      "learning_rate": 1.4082278357206928e-05,
+      "loss": 0.7461,
+      "step": 16349
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7734903229107368,
+      "learning_rate": 1.4081581780341887e-05,
+      "loss": 1.0597,
+      "step": 16350
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8541062682559322,
+      "learning_rate": 1.4080885179712588e-05,
+      "loss": 0.9063,
+      "step": 16351
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.354944761467313,
+      "learning_rate": 1.4080188555323079e-05,
+      "loss": 0.9081,
+      "step": 16352
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.546264929501431,
+      "learning_rate": 1.4079491907177428e-05,
+      "loss": 0.8409,
+      "step": 16353
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8096182984430323,
+      "learning_rate": 1.4078795235279677e-05,
+      "loss": 0.7903,
+      "step": 16354
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5368367701523764,
+      "learning_rate": 1.407809853963389e-05,
+      "loss": 0.9083,
+      "step": 16355
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3355362768080496,
+      "learning_rate": 1.4077401820244125e-05,
+      "loss": 0.894,
+      "step": 16356
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6098584990593805,
+      "learning_rate": 1.4076705077114432e-05,
+      "loss": 0.9724,
+      "step": 16357
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7614873642436526,
+      "learning_rate": 1.4076008310248874e-05,
+      "loss": 0.8716,
+      "step": 16358
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8597821481228511,
+      "learning_rate": 1.4075311519651502e-05,
+      "loss": 0.944,
+      "step": 16359
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7016580399331789,
+      "learning_rate": 1.4074614705326379e-05,
+      "loss": 0.8515,
+      "step": 16360
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6014468979801202,
+      "learning_rate": 1.4073917867277557e-05,
+      "loss": 0.8525,
+      "step": 16361
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8382789224783114,
+      "learning_rate": 1.40732210055091e-05,
+      "loss": 0.7909,
+      "step": 16362
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5732155286990215,
+      "learning_rate": 1.4072524120025057e-05,
+      "loss": 0.9016,
+      "step": 16363
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5938316219238917,
+      "learning_rate": 1.4071827210829493e-05,
+      "loss": 0.9079,
+      "step": 16364
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5435599405088172,
+      "learning_rate": 1.4071130277926457e-05,
+      "loss": 0.7284,
+      "step": 16365
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8384467195229062,
+      "learning_rate": 1.4070433321320012e-05,
+      "loss": 0.9636,
+      "step": 16366
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.476131081821491,
+      "learning_rate": 1.4069736341014218e-05,
+      "loss": 0.7952,
+      "step": 16367
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4619791804778894,
+      "learning_rate": 1.406903933701313e-05,
+      "loss": 0.9095,
+      "step": 16368
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5726589561270106,
+      "learning_rate": 1.4068342309320807e-05,
+      "loss": 0.8233,
+      "step": 16369
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7150018368035211,
+      "learning_rate": 1.4067645257941308e-05,
+      "loss": 0.9185,
+      "step": 16370
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7125954276772075,
+      "learning_rate": 1.4066948182878686e-05,
+      "loss": 0.9687,
+      "step": 16371
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6568527508352426,
+      "learning_rate": 1.4066251084137006e-05,
+      "loss": 0.8932,
+      "step": 16372
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5756053753872132,
+      "learning_rate": 1.4065553961720327e-05,
+      "loss": 0.9011,
+      "step": 16373
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.824616282826102,
+      "learning_rate": 1.4064856815632703e-05,
+      "loss": 0.7271,
+      "step": 16374
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7569011904866055,
+      "learning_rate": 1.4064159645878197e-05,
+      "loss": 0.6913,
+      "step": 16375
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5781314720331512,
+      "learning_rate": 1.4063462452460868e-05,
+      "loss": 0.9704,
+      "step": 16376
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7119916827845434,
+      "learning_rate": 1.4062765235384772e-05,
+      "loss": 0.9095,
+      "step": 16377
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.896389990890854,
+      "learning_rate": 1.4062067994653971e-05,
+      "loss": 0.7467,
+      "step": 16378
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9513521571480634,
+      "learning_rate": 1.4061370730272527e-05,
+      "loss": 0.8824,
+      "step": 16379
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.188515726789743,
+      "learning_rate": 1.4060673442244493e-05,
+      "loss": 0.8994,
+      "step": 16380
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8093963146060925,
+      "learning_rate": 1.4059976130573936e-05,
+      "loss": 0.9154,
+      "step": 16381
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6143586865856234,
+      "learning_rate": 1.4059278795264908e-05,
+      "loss": 1.0096,
+      "step": 16382
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.461907347645083,
+      "learning_rate": 1.4058581436321478e-05,
+      "loss": 0.8494,
+      "step": 16383
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9078922784497025,
+      "learning_rate": 1.4057884053747699e-05,
+      "loss": 0.8815,
+      "step": 16384
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.607397068195541,
+      "learning_rate": 1.4057186647547638e-05,
+      "loss": 0.9359,
+      "step": 16385
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.844520610045408,
+      "learning_rate": 1.405648921772535e-05,
+      "loss": 0.8035,
+      "step": 16386
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.61182019340659,
+      "learning_rate": 1.4055791764284897e-05,
+      "loss": 0.7632,
+      "step": 16387
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.181397789662672,
+      "learning_rate": 1.405509428723034e-05,
+      "loss": 0.8269,
+      "step": 16388
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6536230424408038,
+      "learning_rate": 1.405439678656574e-05,
+      "loss": 0.8497,
+      "step": 16389
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9341613690328487,
+      "learning_rate": 1.405369926229516e-05,
+      "loss": 0.8986,
+      "step": 16390
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.679448615701022,
+      "learning_rate": 1.4053001714422658e-05,
+      "loss": 0.808,
+      "step": 16391
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4449656771143606,
+      "learning_rate": 1.4052304142952297e-05,
+      "loss": 0.8314,
+      "step": 16392
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.037709369446287,
+      "learning_rate": 1.405160654788814e-05,
+      "loss": 0.7424,
+      "step": 16393
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6027531230614096,
+      "learning_rate": 1.4050908929234245e-05,
+      "loss": 0.798,
+      "step": 16394
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6527221790036504,
+      "learning_rate": 1.4050211286994678e-05,
+      "loss": 0.9742,
+      "step": 16395
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5415762511107067,
+      "learning_rate": 1.4049513621173496e-05,
+      "loss": 0.915,
+      "step": 16396
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6833962052814062,
+      "learning_rate": 1.4048815931774764e-05,
+      "loss": 0.9226,
+      "step": 16397
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8926732166394793,
+      "learning_rate": 1.4048118218802545e-05,
+      "loss": 0.9414,
+      "step": 16398
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6941573747679186,
+      "learning_rate": 1.40474204822609e-05,
+      "loss": 0.7994,
+      "step": 16399
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5589209856306803,
+      "learning_rate": 1.4046722722153891e-05,
+      "loss": 0.8443,
+      "step": 16400
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6665357785728563,
+      "learning_rate": 1.4046024938485581e-05,
+      "loss": 0.9227,
+      "step": 16401
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7993876180926804,
+      "learning_rate": 1.4045327131260033e-05,
+      "loss": 0.9546,
+      "step": 16402
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6768892451717319,
+      "learning_rate": 1.404462930048131e-05,
+      "loss": 0.7842,
+      "step": 16403
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8068113521587635,
+      "learning_rate": 1.4043931446153476e-05,
+      "loss": 0.7223,
+      "step": 16404
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7873211932201364,
+      "learning_rate": 1.404323356828059e-05,
+      "loss": 0.6901,
+      "step": 16405
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4517476495629629,
+      "learning_rate": 1.404253566686672e-05,
+      "loss": 0.9145,
+      "step": 16406
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7129479692814542,
+      "learning_rate": 1.404183774191593e-05,
+      "loss": 0.865,
+      "step": 16407
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7265899522887902,
+      "learning_rate": 1.4041139793432274e-05,
+      "loss": 0.8541,
+      "step": 16408
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7479596396051621,
+      "learning_rate": 1.4040441821419829e-05,
+      "loss": 0.6953,
+      "step": 16409
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.656448623029043,
+      "learning_rate": 1.4039743825882648e-05,
+      "loss": 0.889,
+      "step": 16410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8505305767900238,
+      "learning_rate": 1.4039045806824802e-05,
+      "loss": 0.8075,
+      "step": 16411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.600467033236167,
+      "learning_rate": 1.4038347764250353e-05,
+      "loss": 0.8782,
+      "step": 16412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5476585254838706,
+      "learning_rate": 1.4037649698163364e-05,
+      "loss": 0.927,
+      "step": 16413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6053508106941217,
+      "learning_rate": 1.4036951608567901e-05,
+      "loss": 0.818,
+      "step": 16414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5472349650917168,
+      "learning_rate": 1.4036253495468026e-05,
+      "loss": 0.8289,
+      "step": 16415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.716296058194572,
+      "learning_rate": 1.4035555358867807e-05,
+      "loss": 0.8728,
+      "step": 16416
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4273356686247443,
+      "learning_rate": 1.4034857198771308e-05,
+      "loss": 0.7799,
+      "step": 16417
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5230190341207017,
+      "learning_rate": 1.4034159015182591e-05,
+      "loss": 0.9128,
+      "step": 16418
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.273045886311453,
+      "learning_rate": 1.4033460808105723e-05,
+      "loss": 0.7771,
+      "step": 16419
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7079486463678506,
+      "learning_rate": 1.403276257754477e-05,
+      "loss": 0.9761,
+      "step": 16420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5551988937009404,
+      "learning_rate": 1.4032064323503796e-05,
+      "loss": 0.8324,
+      "step": 16421
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.440172237628815,
+      "learning_rate": 1.4031366045986867e-05,
+      "loss": 0.9864,
+      "step": 16422
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3497474084599987,
+      "learning_rate": 1.4030667744998048e-05,
+      "loss": 0.8652,
+      "step": 16423
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4658896530412597,
+      "learning_rate": 1.4029969420541406e-05,
+      "loss": 0.8783,
+      "step": 16424
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4822037466701306,
+      "learning_rate": 1.4029271072621007e-05,
+      "loss": 0.8537,
+      "step": 16425
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.571970701267535,
+      "learning_rate": 1.4028572701240915e-05,
+      "loss": 1.0741,
+      "step": 16426
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3927919472545645,
+      "learning_rate": 1.4027874306405196e-05,
+      "loss": 0.9984,
+      "step": 16427
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5782243176693111,
+      "learning_rate": 1.4027175888117921e-05,
+      "loss": 0.8172,
+      "step": 16428
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5250654138160515,
+      "learning_rate": 1.4026477446383149e-05,
+      "loss": 0.8879,
+      "step": 16429
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8999225418813248,
+      "learning_rate": 1.4025778981204954e-05,
+      "loss": 0.8686,
+      "step": 16430
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6527637028410853,
+      "learning_rate": 1.4025080492587397e-05,
+      "loss": 0.7504,
+      "step": 16431
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6066040394727592,
+      "learning_rate": 1.4024381980534549e-05,
+      "loss": 1.0203,
+      "step": 16432
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.593167784067781,
+      "learning_rate": 1.4023683445050473e-05,
+      "loss": 0.7626,
+      "step": 16433
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6352226379844452,
+      "learning_rate": 1.4022984886139241e-05,
+      "loss": 0.9184,
+      "step": 16434
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9136300360507779,
+      "learning_rate": 1.4022286303804912e-05,
+      "loss": 0.7956,
+      "step": 16435
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.788562094370815,
+      "learning_rate": 1.4021587698051567e-05,
+      "loss": 0.714,
+      "step": 16436
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.838667769537779,
+      "learning_rate": 1.4020889068883258e-05,
+      "loss": 0.9219,
+      "step": 16437
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.087121632920503,
+      "learning_rate": 1.4020190416304062e-05,
+      "loss": 0.8862,
+      "step": 16438
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7864507984055416,
+      "learning_rate": 1.4019491740318045e-05,
+      "loss": 0.9707,
+      "step": 16439
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8829699811897964,
+      "learning_rate": 1.4018793040929275e-05,
+      "loss": 0.8358,
+      "step": 16440
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5724260713509537,
+      "learning_rate": 1.401809431814182e-05,
+      "loss": 0.9402,
+      "step": 16441
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7701005361699924,
+      "learning_rate": 1.4017395571959745e-05,
+      "loss": 0.7501,
+      "step": 16442
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7202982497236803,
+      "learning_rate": 1.4016696802387121e-05,
+      "loss": 0.8847,
+      "step": 16443
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7812330513471724,
+      "learning_rate": 1.401599800942802e-05,
+      "loss": 0.7674,
+      "step": 16444
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.424051061753845,
+      "learning_rate": 1.4015299193086507e-05,
+      "loss": 0.8334,
+      "step": 16445
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5622269800568829,
+      "learning_rate": 1.4014600353366648e-05,
+      "loss": 0.9442,
+      "step": 16446
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6886351421909591,
+      "learning_rate": 1.4013901490272517e-05,
+      "loss": 0.9396,
+      "step": 16447
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6298684050937182,
+      "learning_rate": 1.401320260380818e-05,
+      "loss": 0.9389,
+      "step": 16448
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7058673682043577,
+      "learning_rate": 1.4012503693977705e-05,
+      "loss": 0.8607,
+      "step": 16449
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5226965842721887,
+      "learning_rate": 1.4011804760785165e-05,
+      "loss": 0.8924,
+      "step": 16450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5493948170219407,
+      "learning_rate": 1.401110580423463e-05,
+      "loss": 0.8167,
+      "step": 16451
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5838093553817623,
+      "learning_rate": 1.4010406824330162e-05,
+      "loss": 0.856,
+      "step": 16452
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7669042650316356,
+      "learning_rate": 1.4009707821075843e-05,
+      "loss": 0.9792,
+      "step": 16453
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7893535693239998,
+      "learning_rate": 1.4009008794475726e-05,
+      "loss": 0.764,
+      "step": 16454
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5865693342756575,
+      "learning_rate": 1.40083097445339e-05,
+      "loss": 0.7956,
+      "step": 16455
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7956181535459378,
+      "learning_rate": 1.4007610671254417e-05,
+      "loss": 0.7537,
+      "step": 16456
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5963188392486074,
+      "learning_rate": 1.4006911574641363e-05,
+      "loss": 0.9876,
+      "step": 16457
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5589174697655415,
+      "learning_rate": 1.4006212454698798e-05,
+      "loss": 0.8363,
+      "step": 16458
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8495663465389913,
+      "learning_rate": 1.4005513311430795e-05,
+      "loss": 0.853,
+      "step": 16459
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8253501316296912,
+      "learning_rate": 1.4004814144841428e-05,
+      "loss": 0.8783,
+      "step": 16460
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5635801115308259,
+      "learning_rate": 1.4004114954934762e-05,
+      "loss": 1.0237,
+      "step": 16461
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.850866065095635,
+      "learning_rate": 1.4003415741714874e-05,
+      "loss": 0.8772,
+      "step": 16462
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7458414660093525,
+      "learning_rate": 1.4002716505185828e-05,
+      "loss": 0.7269,
+      "step": 16463
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4888105996522287,
+      "learning_rate": 1.4002017245351705e-05,
+      "loss": 0.9789,
+      "step": 16464
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8325959171982691,
+      "learning_rate": 1.4001317962216568e-05,
+      "loss": 0.7943,
+      "step": 16465
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5624119873888969,
+      "learning_rate": 1.400061865578449e-05,
+      "loss": 0.8657,
+      "step": 16466
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0351658022368313,
+      "learning_rate": 1.3999919326059543e-05,
+      "loss": 0.9526,
+      "step": 16467
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4379389607483293,
+      "learning_rate": 1.3999219973045802e-05,
+      "loss": 0.8487,
+      "step": 16468
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8873551523095775,
+      "learning_rate": 1.3998520596747333e-05,
+      "loss": 1.0606,
+      "step": 16469
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7405012960383888,
+      "learning_rate": 1.3997821197168214e-05,
+      "loss": 0.8394,
+      "step": 16470
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7952088312501825,
+      "learning_rate": 1.3997121774312512e-05,
+      "loss": 0.7845,
+      "step": 16471
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.636785128401542,
+      "learning_rate": 1.3996422328184305e-05,
+      "loss": 0.9069,
+      "step": 16472
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7859665631395434,
+      "learning_rate": 1.3995722858787658e-05,
+      "loss": 0.8788,
+      "step": 16473
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5577337496068009,
+      "learning_rate": 1.399502336612665e-05,
+      "loss": 0.8572,
+      "step": 16474
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7125675055244498,
+      "learning_rate": 1.3994323850205349e-05,
+      "loss": 0.8452,
+      "step": 16475
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.415728406757289,
+      "learning_rate": 1.3993624311027833e-05,
+      "loss": 0.8814,
+      "step": 16476
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6171038524071877,
+      "learning_rate": 1.399292474859817e-05,
+      "loss": 0.9535,
+      "step": 16477
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5064978415710462,
+      "learning_rate": 1.3992225162920434e-05,
+      "loss": 0.7785,
+      "step": 16478
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.684273222056987,
+      "learning_rate": 1.3991525553998704e-05,
+      "loss": 0.968,
+      "step": 16479
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6856193993424082,
+      "learning_rate": 1.399082592183704e-05,
+      "loss": 0.8822,
+      "step": 16480
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5319596918099225,
+      "learning_rate": 1.3990126266439532e-05,
+      "loss": 0.9459,
+      "step": 16481
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6105158295470958,
+      "learning_rate": 1.3989426587810241e-05,
+      "loss": 0.7716,
+      "step": 16482
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5704516698445268,
+      "learning_rate": 1.398872688595325e-05,
+      "loss": 0.8061,
+      "step": 16483
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7204002412396469,
+      "learning_rate": 1.3988027160872624e-05,
+      "loss": 0.8203,
+      "step": 16484
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.571309025983938,
+      "learning_rate": 1.3987327412572442e-05,
+      "loss": 0.9489,
+      "step": 16485
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7496229930218923,
+      "learning_rate": 1.3986627641056777e-05,
+      "loss": 0.7446,
+      "step": 16486
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.143809224898404,
+      "learning_rate": 1.3985927846329705e-05,
+      "loss": 0.9437,
+      "step": 16487
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6042638385902839,
+      "learning_rate": 1.3985228028395298e-05,
+      "loss": 0.8334,
+      "step": 16488
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.044688369822954,
+      "learning_rate": 1.3984528187257636e-05,
+      "loss": 0.8479,
+      "step": 16489
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.800834634493447,
+      "learning_rate": 1.3983828322920786e-05,
+      "loss": 0.8922,
+      "step": 16490
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7119947218349065,
+      "learning_rate": 1.3983128435388824e-05,
+      "loss": 0.868,
+      "step": 16491
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5950695022861274,
+      "learning_rate": 1.398242852466583e-05,
+      "loss": 0.9165,
+      "step": 16492
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.610233303807617,
+      "learning_rate": 1.3981728590755877e-05,
+      "loss": 0.8182,
+      "step": 16493
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9770892403423892,
+      "learning_rate": 1.3981028633663036e-05,
+      "loss": 0.9379,
+      "step": 16494
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.578330731496184,
+      "learning_rate": 1.3980328653391389e-05,
+      "loss": 0.8903,
+      "step": 16495
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6012208156702272,
+      "learning_rate": 1.3979628649945007e-05,
+      "loss": 0.715,
+      "step": 16496
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5592942709972182,
+      "learning_rate": 1.3978928623327969e-05,
+      "loss": 0.8402,
+      "step": 16497
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7335970631022575,
+      "learning_rate": 1.3978228573544348e-05,
+      "loss": 0.6721,
+      "step": 16498
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6670432250551634,
+      "learning_rate": 1.3977528500598218e-05,
+      "loss": 0.8947,
+      "step": 16499
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.126098507888115,
+      "learning_rate": 1.3976828404493662e-05,
+      "loss": 0.7952,
+      "step": 16500
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6056015740365606,
+      "learning_rate": 1.3976128285234748e-05,
+      "loss": 0.8981,
+      "step": 16501
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7264184487807395,
+      "learning_rate": 1.3975428142825562e-05,
+      "loss": 0.9697,
+      "step": 16502
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5797350167846873,
+      "learning_rate": 1.3974727977270167e-05,
+      "loss": 0.9897,
+      "step": 16503
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5948358371015465,
+      "learning_rate": 1.3974027788572655e-05,
+      "loss": 0.9048,
+      "step": 16504
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4745074406066971,
+      "learning_rate": 1.397332757673709e-05,
+      "loss": 0.9377,
+      "step": 16505
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.836372649415688,
+      "learning_rate": 1.3972627341767555e-05,
+      "loss": 0.7551,
+      "step": 16506
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7962235324520828,
+      "learning_rate": 1.3971927083668127e-05,
+      "loss": 0.9925,
+      "step": 16507
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7552518409679839,
+      "learning_rate": 1.397122680244288e-05,
+      "loss": 0.7841,
+      "step": 16508
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4972099153298453,
+      "learning_rate": 1.3970526498095897e-05,
+      "loss": 0.8198,
+      "step": 16509
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6652364639338586,
+      "learning_rate": 1.3969826170631247e-05,
+      "loss": 0.9041,
+      "step": 16510
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8853131891226598,
+      "learning_rate": 1.3969125820053018e-05,
+      "loss": 0.8785,
+      "step": 16511
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8110704623451601,
+      "learning_rate": 1.3968425446365275e-05,
+      "loss": 0.9262,
+      "step": 16512
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6208728023378853,
+      "learning_rate": 1.3967725049572109e-05,
+      "loss": 0.7684,
+      "step": 16513
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8669054862915355,
+      "learning_rate": 1.3967024629677585e-05,
+      "loss": 0.8676,
+      "step": 16514
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8127817133500367,
+      "learning_rate": 1.3966324186685795e-05,
+      "loss": 0.7393,
+      "step": 16515
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5801325089002545,
+      "learning_rate": 1.3965623720600805e-05,
+      "loss": 0.914,
+      "step": 16516
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6298701722843174,
+      "learning_rate": 1.3964923231426701e-05,
+      "loss": 0.9515,
+      "step": 16517
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5881473136130522,
+      "learning_rate": 1.3964222719167557e-05,
+      "loss": 0.8426,
+      "step": 16518
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2437156192264758,
+      "learning_rate": 1.3963522183827454e-05,
+      "loss": 0.6891,
+      "step": 16519
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.759361248165486,
+      "learning_rate": 1.3962821625410468e-05,
+      "loss": 0.8836,
+      "step": 16520
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8239612310789746,
+      "learning_rate": 1.3962121043920682e-05,
+      "loss": 0.8962,
+      "step": 16521
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7378430568642484,
+      "learning_rate": 1.3961420439362173e-05,
+      "loss": 1.0126,
+      "step": 16522
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5769663224687969,
+      "learning_rate": 1.396071981173902e-05,
+      "loss": 0.8547,
+      "step": 16523
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5426173131328391,
+      "learning_rate": 1.3960019161055297e-05,
+      "loss": 0.859,
+      "step": 16524
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5523420385320745,
+      "learning_rate": 1.3959318487315096e-05,
+      "loss": 0.8217,
+      "step": 16525
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8539611702180451,
+      "learning_rate": 1.3958617790522486e-05,
+      "loss": 0.7902,
+      "step": 16526
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.705501960824977,
+      "learning_rate": 1.395791707068155e-05,
+      "loss": 0.801,
+      "step": 16527
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4271200501356707,
+      "learning_rate": 1.3957216327796368e-05,
+      "loss": 0.9643,
+      "step": 16528
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6575389938913483,
+      "learning_rate": 1.395651556187102e-05,
+      "loss": 0.9097,
+      "step": 16529
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5400276969635114,
+      "learning_rate": 1.3955814772909588e-05,
+      "loss": 0.8878,
+      "step": 16530
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9230647008020967,
+      "learning_rate": 1.3955113960916146e-05,
+      "loss": 0.8779,
+      "step": 16531
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7717775686733348,
+      "learning_rate": 1.3954413125894783e-05,
+      "loss": 0.8314,
+      "step": 16532
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5685022338497765,
+      "learning_rate": 1.3953712267849568e-05,
+      "loss": 0.8958,
+      "step": 16533
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2435774427194217,
+      "learning_rate": 1.3953011386784597e-05,
+      "loss": 0.8236,
+      "step": 16534
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8468977932314676,
+      "learning_rate": 1.3952310482703935e-05,
+      "loss": 0.9096,
+      "step": 16535
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6735913526058421,
+      "learning_rate": 1.3951609555611673e-05,
+      "loss": 0.7829,
+      "step": 16536
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5267049200858098,
+      "learning_rate": 1.3950908605511888e-05,
+      "loss": 0.7801,
+      "step": 16537
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.398064320802622,
+      "learning_rate": 1.3950207632408662e-05,
+      "loss": 0.8308,
+      "step": 16538
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1288655718747123,
+      "learning_rate": 1.394950663630608e-05,
+      "loss": 0.9915,
+      "step": 16539
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.622215114848383,
+      "learning_rate": 1.3948805617208216e-05,
+      "loss": 0.8514,
+      "step": 16540
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9354862943960744,
+      "learning_rate": 1.394810457511916e-05,
+      "loss": 0.9963,
+      "step": 16541
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.555474669279308,
+      "learning_rate": 1.3947403510042984e-05,
+      "loss": 0.8728,
+      "step": 16542
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6190610384405875,
+      "learning_rate": 1.3946702421983778e-05,
+      "loss": 0.8422,
+      "step": 16543
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.348701263310551,
+      "learning_rate": 1.394600131094562e-05,
+      "loss": 0.832,
+      "step": 16544
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5358160458728358,
+      "learning_rate": 1.3945300176932594e-05,
+      "loss": 0.8956,
+      "step": 16545
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5217191642557746,
+      "learning_rate": 1.3944599019948779e-05,
+      "loss": 0.8787,
+      "step": 16546
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5858442238238846,
+      "learning_rate": 1.3943897839998265e-05,
+      "loss": 0.9606,
+      "step": 16547
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7654763322898677,
+      "learning_rate": 1.3943196637085123e-05,
+      "loss": 0.8574,
+      "step": 16548
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.842713085892872,
+      "learning_rate": 1.3942495411213446e-05,
+      "loss": 0.944,
+      "step": 16549
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.160304252572655,
+      "learning_rate": 1.394179416238731e-05,
+      "loss": 0.8654,
+      "step": 16550
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.490980423851057,
+      "learning_rate": 1.3941092890610803e-05,
+      "loss": 0.7522,
+      "step": 16551
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7647268725996514,
+      "learning_rate": 1.3940391595888e-05,
+      "loss": 0.9868,
+      "step": 16552
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5822990800060968,
+      "learning_rate": 1.3939690278222997e-05,
+      "loss": 0.8851,
+      "step": 16553
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0439815748411276,
+      "learning_rate": 1.3938988937619864e-05,
+      "loss": 1.0015,
+      "step": 16554
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5325555522277352,
+      "learning_rate": 1.3938287574082692e-05,
+      "loss": 0.9693,
+      "step": 16555
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7184694130558971,
+      "learning_rate": 1.3937586187615564e-05,
+      "loss": 0.6547,
+      "step": 16556
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0423223131511277,
+      "learning_rate": 1.3936884778222562e-05,
+      "loss": 0.8464,
+      "step": 16557
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4707431507751123,
+      "learning_rate": 1.3936183345907772e-05,
+      "loss": 0.8572,
+      "step": 16558
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5577714198684194,
+      "learning_rate": 1.3935481890675273e-05,
+      "loss": 0.9588,
+      "step": 16559
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8271784958337018,
+      "learning_rate": 1.3934780412529158e-05,
+      "loss": 0.8581,
+      "step": 16560
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.183996011698615,
+      "learning_rate": 1.3934078911473499e-05,
+      "loss": 1.0059,
+      "step": 16561
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7980786218174996,
+      "learning_rate": 1.3933377387512393e-05,
+      "loss": 0.9266,
+      "step": 16562
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6478992196198186,
+      "learning_rate": 1.3932675840649915e-05,
+      "loss": 0.8226,
+      "step": 16563
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7511557683402357,
+      "learning_rate": 1.3931974270890154e-05,
+      "loss": 1.0021,
+      "step": 16564
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6425486592929648,
+      "learning_rate": 1.393127267823719e-05,
+      "loss": 0.8396,
+      "step": 16565
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6827635920878259,
+      "learning_rate": 1.393057106269512e-05,
+      "loss": 0.9017,
+      "step": 16566
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5179255313728532,
+      "learning_rate": 1.3929869424268016e-05,
+      "loss": 0.8973,
+      "step": 16567
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6757273036631084,
+      "learning_rate": 1.392916776295997e-05,
+      "loss": 0.8084,
+      "step": 16568
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9622727105037956,
+      "learning_rate": 1.3928466078775063e-05,
+      "loss": 0.9186,
+      "step": 16569
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4851222313458114,
+      "learning_rate": 1.3927764371717386e-05,
+      "loss": 1.0243,
+      "step": 16570
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7414189053089943,
+      "learning_rate": 1.3927062641791018e-05,
+      "loss": 0.9958,
+      "step": 16571
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.520474299926393,
+      "learning_rate": 1.3926360889000048e-05,
+      "loss": 0.8828,
+      "step": 16572
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8815953452002712,
+      "learning_rate": 1.3925659113348564e-05,
+      "loss": 0.8822,
+      "step": 16573
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8366173867322462,
+      "learning_rate": 1.3924957314840648e-05,
+      "loss": 0.8431,
+      "step": 16574
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5307335346006175,
+      "learning_rate": 1.3924255493480387e-05,
+      "loss": 0.9158,
+      "step": 16575
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.392695115571882,
+      "learning_rate": 1.392355364927187e-05,
+      "loss": 0.9195,
+      "step": 16576
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7261624516701242,
+      "learning_rate": 1.3922851782219182e-05,
+      "loss": 0.9341,
+      "step": 16577
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.714374680114032,
+      "learning_rate": 1.3922149892326409e-05,
+      "loss": 0.8189,
+      "step": 16578
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7584343618210205,
+      "learning_rate": 1.3921447979597636e-05,
+      "loss": 0.8413,
+      "step": 16579
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4807307511547363,
+      "learning_rate": 1.3920746044036953e-05,
+      "loss": 0.9116,
+      "step": 16580
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7704318374219407,
+      "learning_rate": 1.3920044085648446e-05,
+      "loss": 0.7186,
+      "step": 16581
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7425516287992221,
+      "learning_rate": 1.3919342104436198e-05,
+      "loss": 0.7293,
+      "step": 16582
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9727501203904205,
+      "learning_rate": 1.3918640100404301e-05,
+      "loss": 0.7622,
+      "step": 16583
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.673616074442927,
+      "learning_rate": 1.3917938073556841e-05,
+      "loss": 0.8632,
+      "step": 16584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6741260071508342,
+      "learning_rate": 1.3917236023897906e-05,
+      "loss": 0.8124,
+      "step": 16585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5619940765139682,
+      "learning_rate": 1.3916533951431579e-05,
+      "loss": 0.9804,
+      "step": 16586
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5779650502805318,
+      "learning_rate": 1.3915831856161956e-05,
+      "loss": 0.9039,
+      "step": 16587
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7869156049729449,
+      "learning_rate": 1.3915129738093119e-05,
+      "loss": 0.792,
+      "step": 16588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9424395830417756,
+      "learning_rate": 1.3914427597229156e-05,
+      "loss": 0.9384,
+      "step": 16589
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.766104298640751,
+      "learning_rate": 1.3913725433574157e-05,
+      "loss": 1.0154,
+      "step": 16590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.662798934343351,
+      "learning_rate": 1.391302324713221e-05,
+      "loss": 0.849,
+      "step": 16591
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8892741143055602,
+      "learning_rate": 1.39123210379074e-05,
+      "loss": 0.9694,
+      "step": 16592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4105524969843948,
+      "learning_rate": 1.391161880590382e-05,
+      "loss": 0.9107,
+      "step": 16593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8795494622452114,
+      "learning_rate": 1.3910916551125558e-05,
+      "loss": 1.1014,
+      "step": 16594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7100219132069077,
+      "learning_rate": 1.3910214273576701e-05,
+      "loss": 1.005,
+      "step": 16595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7572380186297698,
+      "learning_rate": 1.3909511973261339e-05,
+      "loss": 0.7577,
+      "step": 16596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7732381319018444,
+      "learning_rate": 1.3908809650183559e-05,
+      "loss": 0.8528,
+      "step": 16597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5426792319434213,
+      "learning_rate": 1.3908107304347453e-05,
+      "loss": 0.8075,
+      "step": 16598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5942128051267412,
+      "learning_rate": 1.3907404935757107e-05,
+      "loss": 0.8998,
+      "step": 16599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6133224749067745,
+      "learning_rate": 1.3906702544416614e-05,
+      "loss": 0.8848,
+      "step": 16600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7405795704644194,
+      "learning_rate": 1.3906000130330061e-05,
+      "loss": 0.8261,
+      "step": 16601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8713864737130903,
+      "learning_rate": 1.3905297693501543e-05,
+      "loss": 0.9029,
+      "step": 16602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.841708393524511,
+      "learning_rate": 1.3904595233935139e-05,
+      "loss": 0.919,
+      "step": 16603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5031528950835722,
+      "learning_rate": 1.3903892751634949e-05,
+      "loss": 0.8634,
+      "step": 16604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6541616976837101,
+      "learning_rate": 1.3903190246605059e-05,
+      "loss": 0.884,
+      "step": 16605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7040395811840658,
+      "learning_rate": 1.390248771884956e-05,
+      "loss": 0.9579,
+      "step": 16606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7983469011656124,
+      "learning_rate": 1.3901785168372542e-05,
+      "loss": 0.8291,
+      "step": 16607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9602450277663295,
+      "learning_rate": 1.3901082595178092e-05,
+      "loss": 1.0282,
+      "step": 16608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.652487549811884,
+      "learning_rate": 1.390037999927031e-05,
+      "loss": 0.9666,
+      "step": 16609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4338777691400688,
+      "learning_rate": 1.3899677380653276e-05,
+      "loss": 0.9065,
+      "step": 16610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0382969843923595,
+      "learning_rate": 1.3898974739331091e-05,
+      "loss": 0.904,
+      "step": 16611
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5551200945808736,
+      "learning_rate": 1.3898272075307833e-05,
+      "loss": 1.0322,
+      "step": 16612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8601685831520691,
+      "learning_rate": 1.3897569388587609e-05,
+      "loss": 0.9163,
+      "step": 16613
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6710631186032048,
+      "learning_rate": 1.3896866679174495e-05,
+      "loss": 0.8586,
+      "step": 16614
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4614909431079044,
+      "learning_rate": 1.3896163947072595e-05,
+      "loss": 0.9238,
+      "step": 16615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5433721666672195,
+      "learning_rate": 1.3895461192285992e-05,
+      "loss": 1.0021,
+      "step": 16616
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5159307130748985,
+      "learning_rate": 1.3894758414818781e-05,
+      "loss": 0.8855,
+      "step": 16617
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7540846418609888,
+      "learning_rate": 1.3894055614675052e-05,
+      "loss": 0.9681,
+      "step": 16618
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6687250376657188,
+      "learning_rate": 1.38933527918589e-05,
+      "loss": 1.0058,
+      "step": 16619
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.669220493770661,
+      "learning_rate": 1.3892649946374417e-05,
+      "loss": 0.9706,
+      "step": 16620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6604359877043007,
+      "learning_rate": 1.3891947078225692e-05,
+      "loss": 0.9754,
+      "step": 16621
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6008719751423472,
+      "learning_rate": 1.389124418741682e-05,
+      "loss": 0.9383,
+      "step": 16622
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.912817977934686,
+      "learning_rate": 1.389054127395189e-05,
+      "loss": 0.916,
+      "step": 16623
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5544353305234242,
+      "learning_rate": 1.3889838337834997e-05,
+      "loss": 0.9946,
+      "step": 16624
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.538773850343323,
+      "learning_rate": 1.3889135379070235e-05,
+      "loss": 0.8046,
+      "step": 16625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4164474206927726,
+      "learning_rate": 1.3888432397661697e-05,
+      "loss": 0.8206,
+      "step": 16626
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2839671308870972,
+      "learning_rate": 1.3887729393613471e-05,
+      "loss": 0.9026,
+      "step": 16627
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6860209314802395,
+      "learning_rate": 1.3887026366929657e-05,
+      "loss": 0.9001,
+      "step": 16628
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5557789237437907,
+      "learning_rate": 1.3886323317614344e-05,
+      "loss": 0.7626,
+      "step": 16629
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4360919042226026,
+      "learning_rate": 1.3885620245671624e-05,
+      "loss": 0.9046,
+      "step": 16630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7564898992292257,
+      "learning_rate": 1.3884917151105595e-05,
+      "loss": 0.7473,
+      "step": 16631
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4967438893240617,
+      "learning_rate": 1.388421403392035e-05,
+      "loss": 0.8642,
+      "step": 16632
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7180080312927388,
+      "learning_rate": 1.3883510894119979e-05,
+      "loss": 0.764,
+      "step": 16633
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4559451460329225,
+      "learning_rate": 1.3882807731708579e-05,
+      "loss": 0.9825,
+      "step": 16634
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7564154747143443,
+      "learning_rate": 1.3882104546690241e-05,
+      "loss": 0.7374,
+      "step": 16635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6400979303070133,
+      "learning_rate": 1.3881401339069064e-05,
+      "loss": 0.9142,
+      "step": 16636
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.33035326539178,
+      "learning_rate": 1.3880698108849138e-05,
+      "loss": 0.9364,
+      "step": 16637
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7976530380467783,
+      "learning_rate": 1.387999485603456e-05,
+      "loss": 0.6873,
+      "step": 16638
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.429205379397053,
+      "learning_rate": 1.3879291580629421e-05,
+      "loss": 0.9181,
+      "step": 16639
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6840301417990426,
+      "learning_rate": 1.3878588282637822e-05,
+      "loss": 0.9942,
+      "step": 16640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1288643804832708,
+      "learning_rate": 1.3877884962063853e-05,
+      "loss": 0.7825,
+      "step": 16641
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.557535616538521,
+      "learning_rate": 1.3877181618911606e-05,
+      "loss": 0.8984,
+      "step": 16642
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6323273281375323,
+      "learning_rate": 1.3876478253185183e-05,
+      "loss": 0.8882,
+      "step": 16643
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7248706466687564,
+      "learning_rate": 1.3875774864888675e-05,
+      "loss": 0.7394,
+      "step": 16644
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.689540150690781,
+      "learning_rate": 1.3875071454026183e-05,
+      "loss": 0.8251,
+      "step": 16645
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7601141205162094,
+      "learning_rate": 1.3874368020601796e-05,
+      "loss": 0.9014,
+      "step": 16646
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5458253245744726,
+      "learning_rate": 1.3873664564619612e-05,
+      "loss": 0.882,
+      "step": 16647
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4186917480118992,
+      "learning_rate": 1.3872961086083725e-05,
+      "loss": 0.7944,
+      "step": 16648
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3461564326784954,
+      "learning_rate": 1.3872257584998235e-05,
+      "loss": 0.9063,
+      "step": 16649
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.671163720947365,
+      "learning_rate": 1.3871554061367232e-05,
+      "loss": 0.961,
+      "step": 16650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3315694800212765,
+      "learning_rate": 1.3870850515194817e-05,
+      "loss": 0.9719,
+      "step": 16651
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.535804636065625,
+      "learning_rate": 1.3870146946485083e-05,
+      "loss": 0.9972,
+      "step": 16652
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.669952141284636,
+      "learning_rate": 1.386944335524213e-05,
+      "loss": 0.8086,
+      "step": 16653
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.620882491736195,
+      "learning_rate": 1.3868739741470055e-05,
+      "loss": 0.8319,
+      "step": 16654
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7521995556702711,
+      "learning_rate": 1.386803610517295e-05,
+      "loss": 0.759,
+      "step": 16655
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5271543558554663,
+      "learning_rate": 1.3867332446354916e-05,
+      "loss": 0.8599,
+      "step": 16656
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6556391576780163,
+      "learning_rate": 1.3866628765020045e-05,
+      "loss": 0.9469,
+      "step": 16657
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7941847600296739,
+      "learning_rate": 1.3865925061172439e-05,
+      "loss": 0.9043,
+      "step": 16658
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8669850478026808,
+      "learning_rate": 1.3865221334816193e-05,
+      "loss": 0.7904,
+      "step": 16659
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7546323320922605,
+      "learning_rate": 1.3864517585955406e-05,
+      "loss": 0.8827,
+      "step": 16660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7255578064904339,
+      "learning_rate": 1.3863813814594172e-05,
+      "loss": 0.7139,
+      "step": 16661
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9597206665766933,
+      "learning_rate": 1.3863110020736592e-05,
+      "loss": 0.8821,
+      "step": 16662
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7801394455319641,
+      "learning_rate": 1.3862406204386761e-05,
+      "loss": 0.806,
+      "step": 16663
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8348421410899458,
+      "learning_rate": 1.3861702365548781e-05,
+      "loss": 0.948,
+      "step": 16664
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6285910167175528,
+      "learning_rate": 1.3860998504226744e-05,
+      "loss": 0.8977,
+      "step": 16665
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.09064408505453,
+      "learning_rate": 1.3860294620424754e-05,
+      "loss": 0.9547,
+      "step": 16666
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7929370148719823,
+      "learning_rate": 1.3859590714146906e-05,
+      "loss": 0.9302,
+      "step": 16667
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4958166812160127,
+      "learning_rate": 1.38588867853973e-05,
+      "loss": 0.7799,
+      "step": 16668
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.592454126945998,
+      "learning_rate": 1.385818283418003e-05,
+      "loss": 0.8167,
+      "step": 16669
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5972414732522149,
+      "learning_rate": 1.3857478860499201e-05,
+      "loss": 0.8641,
+      "step": 16670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3415784582998824,
+      "learning_rate": 1.3856774864358906e-05,
+      "loss": 0.8254,
+      "step": 16671
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9132955805591612,
+      "learning_rate": 1.385607084576325e-05,
+      "loss": 0.8368,
+      "step": 16672
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.086749674183042,
+      "learning_rate": 1.3855366804716326e-05,
+      "loss": 0.7987,
+      "step": 16673
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7813569655803666,
+      "learning_rate": 1.3854662741222238e-05,
+      "loss": 0.7054,
+      "step": 16674
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5485419713918214,
+      "learning_rate": 1.385395865528508e-05,
+      "loss": 0.8799,
+      "step": 16675
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4383469069238073,
+      "learning_rate": 1.3853254546908958e-05,
+      "loss": 0.889,
+      "step": 16676
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7217417378429198,
+      "learning_rate": 1.3852550416097968e-05,
+      "loss": 0.8628,
+      "step": 16677
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.54105854050211,
+      "learning_rate": 1.3851846262856208e-05,
+      "loss": 0.9518,
+      "step": 16678
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7990320556024146,
+      "learning_rate": 1.3851142087187778e-05,
+      "loss": 0.8765,
+      "step": 16679
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6505535666031783,
+      "learning_rate": 1.3850437889096781e-05,
+      "loss": 0.8007,
+      "step": 16680
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6467097427501327,
+      "learning_rate": 1.384973366858732e-05,
+      "loss": 0.8526,
+      "step": 16681
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.141443424377261,
+      "learning_rate": 1.3849029425663485e-05,
+      "loss": 0.8879,
+      "step": 16682
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4876395090982413,
+      "learning_rate": 1.3848325160329386e-05,
+      "loss": 0.8165,
+      "step": 16683
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4643451018577027,
+      "learning_rate": 1.3847620872589116e-05,
+      "loss": 0.824,
+      "step": 16684
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9199385989263331,
+      "learning_rate": 1.3846916562446784e-05,
+      "loss": 0.8672,
+      "step": 16685
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9786174106987846,
+      "learning_rate": 1.384621222990648e-05,
+      "loss": 0.7408,
+      "step": 16686
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.745020900169532,
+      "learning_rate": 1.3845507874972314e-05,
+      "loss": 0.6789,
+      "step": 16687
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5571493009979298,
+      "learning_rate": 1.3844803497648382e-05,
+      "loss": 0.8679,
+      "step": 16688
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8829769971819026,
+      "learning_rate": 1.384409909793879e-05,
+      "loss": 0.8638,
+      "step": 16689
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5471315579994334,
+      "learning_rate": 1.3843394675847635e-05,
+      "loss": 0.9078,
+      "step": 16690
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7809159656274329,
+      "learning_rate": 1.3842690231379014e-05,
+      "loss": 0.7343,
+      "step": 16691
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7302131320726717,
+      "learning_rate": 1.384198576453704e-05,
+      "loss": 0.7724,
+      "step": 16692
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.549923003123743,
+      "learning_rate": 1.3841281275325807e-05,
+      "loss": 0.767,
+      "step": 16693
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6315687335049749,
+      "learning_rate": 1.3840576763749419e-05,
+      "loss": 0.9613,
+      "step": 16694
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6778429451985968,
+      "learning_rate": 1.3839872229811974e-05,
+      "loss": 0.9003,
+      "step": 16695
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.763951862682191,
+      "learning_rate": 1.3839167673517582e-05,
+      "loss": 0.7258,
+      "step": 16696
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4934885736115153,
+      "learning_rate": 1.3838463094870338e-05,
+      "loss": 0.9463,
+      "step": 16697
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6076734013521745,
+      "learning_rate": 1.3837758493874347e-05,
+      "loss": 0.9601,
+      "step": 16698
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5586327026984739,
+      "learning_rate": 1.383705387053371e-05,
+      "loss": 0.7763,
+      "step": 16699
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.724846321173294,
+      "learning_rate": 1.383634922485253e-05,
+      "loss": 0.9132,
+      "step": 16700
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9358535985883036,
+      "learning_rate": 1.3835644556834913e-05,
+      "loss": 0.8309,
+      "step": 16701
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.595194627667604,
+      "learning_rate": 1.3834939866484956e-05,
+      "loss": 0.9605,
+      "step": 16702
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6864316638160752,
+      "learning_rate": 1.3834235153806766e-05,
+      "loss": 0.8115,
+      "step": 16703
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.696552238163376,
+      "learning_rate": 1.3833530418804446e-05,
+      "loss": 0.7974,
+      "step": 16704
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7675960629597558,
+      "learning_rate": 1.3832825661482097e-05,
+      "loss": 0.8566,
+      "step": 16705
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6467499155417182,
+      "learning_rate": 1.3832120881843825e-05,
+      "loss": 0.8861,
+      "step": 16706
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6373609692647595,
+      "learning_rate": 1.383141607989373e-05,
+      "loss": 0.9801,
+      "step": 16707
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1982344843474424,
+      "learning_rate": 1.3830711255635919e-05,
+      "loss": 0.9558,
+      "step": 16708
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7487158981852626,
+      "learning_rate": 1.3830006409074495e-05,
+      "loss": 0.9093,
+      "step": 16709
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8295452879639957,
+      "learning_rate": 1.3829301540213557e-05,
+      "loss": 0.8558,
+      "step": 16710
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.070053159918095,
+      "learning_rate": 1.382859664905722e-05,
+      "loss": 0.7949,
+      "step": 16711
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5597386591404685,
+      "learning_rate": 1.3827891735609575e-05,
+      "loss": 0.7723,
+      "step": 16712
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7225858525593543,
+      "learning_rate": 1.3827186799874737e-05,
+      "loss": 0.69,
+      "step": 16713
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5859985544793027,
+      "learning_rate": 1.3826481841856802e-05,
+      "loss": 0.9411,
+      "step": 16714
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.505061276055601,
+      "learning_rate": 1.382577686155988e-05,
+      "loss": 0.9604,
+      "step": 16715
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8252222628170236,
+      "learning_rate": 1.3825071858988073e-05,
+      "loss": 0.8056,
+      "step": 16716
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6917828529209944,
+      "learning_rate": 1.3824366834145488e-05,
+      "loss": 0.9343,
+      "step": 16717
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5058038020169233,
+      "learning_rate": 1.382366178703623e-05,
+      "loss": 0.8803,
+      "step": 16718
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6007072499423387,
+      "learning_rate": 1.38229567176644e-05,
+      "loss": 0.8307,
+      "step": 16719
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.102044084418658,
+      "learning_rate": 1.3822251626034105e-05,
+      "loss": 0.861,
+      "step": 16720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.576531578269107,
+      "learning_rate": 1.3821546512149453e-05,
+      "loss": 0.9421,
+      "step": 16721
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.187319639659638,
+      "learning_rate": 1.3820841376014546e-05,
+      "loss": 0.9533,
+      "step": 16722
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4873940836401762,
+      "learning_rate": 1.3820136217633491e-05,
+      "loss": 0.8278,
+      "step": 16723
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7123872818942518,
+      "learning_rate": 1.3819431037010393e-05,
+      "loss": 0.8505,
+      "step": 16724
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7919505938404956,
+      "learning_rate": 1.381872583414936e-05,
+      "loss": 0.8603,
+      "step": 16725
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.674094608916325,
+      "learning_rate": 1.3818020609054494e-05,
+      "loss": 0.766,
+      "step": 16726
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0372536174911184,
+      "learning_rate": 1.3817315361729904e-05,
+      "loss": 0.7829,
+      "step": 16727
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7513915445882458,
+      "learning_rate": 1.3816610092179697e-05,
+      "loss": 0.8803,
+      "step": 16728
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.882362402876238,
+      "learning_rate": 1.3815904800407975e-05,
+      "loss": 1.1005,
+      "step": 16729
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3729988949874248,
+      "learning_rate": 1.3815199486418851e-05,
+      "loss": 0.8819,
+      "step": 16730
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.61839866300888,
+      "learning_rate": 1.3814494150216424e-05,
+      "loss": 1.0069,
+      "step": 16731
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2997010492067784,
+      "learning_rate": 1.3813788791804807e-05,
+      "loss": 0.7708,
+      "step": 16732
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7385059477674174,
+      "learning_rate": 1.3813083411188104e-05,
+      "loss": 1.0556,
+      "step": 16733
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4982592479938055,
+      "learning_rate": 1.3812378008370422e-05,
+      "loss": 0.7905,
+      "step": 16734
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5822434163876424,
+      "learning_rate": 1.3811672583355864e-05,
+      "loss": 0.8936,
+      "step": 16735
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7151423536762336,
+      "learning_rate": 1.3810967136148547e-05,
+      "loss": 0.8658,
+      "step": 16736
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.365901781491196,
+      "learning_rate": 1.3810261666752571e-05,
+      "loss": 1.0127,
+      "step": 16737
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6201737888280543,
+      "learning_rate": 1.3809556175172043e-05,
+      "loss": 0.9443,
+      "step": 16738
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.974551686096892,
+      "learning_rate": 1.3808850661411077e-05,
+      "loss": 0.9322,
+      "step": 16739
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5865759359475837,
+      "learning_rate": 1.3808145125473773e-05,
+      "loss": 0.9217,
+      "step": 16740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8057480045842949,
+      "learning_rate": 1.3807439567364246e-05,
+      "loss": 0.9709,
+      "step": 16741
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.368817599007845,
+      "learning_rate": 1.3806733987086597e-05,
+      "loss": 0.9576,
+      "step": 16742
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5502652755088597,
+      "learning_rate": 1.380602838464494e-05,
+      "loss": 0.8448,
+      "step": 16743
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.620829645978544,
+      "learning_rate": 1.3805322760043378e-05,
+      "loss": 0.9396,
+      "step": 16744
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6812562391565475,
+      "learning_rate": 1.3804617113286027e-05,
+      "loss": 0.912,
+      "step": 16745
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8029704350683686,
+      "learning_rate": 1.3803911444376984e-05,
+      "loss": 0.7171,
+      "step": 16746
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6421827481914641,
+      "learning_rate": 1.380320575332037e-05,
+      "loss": 0.8496,
+      "step": 16747
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0722361923291786,
+      "learning_rate": 1.3802500040120285e-05,
+      "loss": 0.8752,
+      "step": 16748
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0826551485596467,
+      "learning_rate": 1.3801794304780843e-05,
+      "loss": 0.8199,
+      "step": 16749
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4683542902297873,
+      "learning_rate": 1.3801088547306149e-05,
+      "loss": 0.9369,
+      "step": 16750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5525132326366884,
+      "learning_rate": 1.3800382767700316e-05,
+      "loss": 0.9408,
+      "step": 16751
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.660026159939657,
+      "learning_rate": 1.3799676965967451e-05,
+      "loss": 0.978,
+      "step": 16752
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5083984244677064,
+      "learning_rate": 1.3798971142111661e-05,
+      "loss": 0.9105,
+      "step": 16753
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.397771737474787,
+      "learning_rate": 1.379826529613706e-05,
+      "loss": 0.7984,
+      "step": 16754
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6455302506305072,
+      "learning_rate": 1.3797559428047755e-05,
+      "loss": 0.9048,
+      "step": 16755
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.536379101542752,
+      "learning_rate": 1.3796853537847859e-05,
+      "loss": 0.8894,
+      "step": 16756
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6257358096781038,
+      "learning_rate": 1.3796147625541479e-05,
+      "loss": 0.8347,
+      "step": 16757
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3870783664936104,
+      "learning_rate": 1.3795441691132726e-05,
+      "loss": 0.907,
+      "step": 16758
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5914581389468414,
+      "learning_rate": 1.3794735734625708e-05,
+      "loss": 0.7909,
+      "step": 16759
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5361529338997253,
+      "learning_rate": 1.3794029756024536e-05,
+      "loss": 1.0031,
+      "step": 16760
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4804192638516855,
+      "learning_rate": 1.3793323755333323e-05,
+      "loss": 0.8828,
+      "step": 16761
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7032020479750605,
+      "learning_rate": 1.3792617732556182e-05,
+      "loss": 0.8987,
+      "step": 16762
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6397620917375262,
+      "learning_rate": 1.3791911687697215e-05,
+      "loss": 0.868,
+      "step": 16763
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7454785728563881,
+      "learning_rate": 1.3791205620760541e-05,
+      "loss": 0.7608,
+      "step": 16764
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8163613804311545,
+      "learning_rate": 1.3790499531750262e-05,
+      "loss": 0.9519,
+      "step": 16765
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6644186502550582,
+      "learning_rate": 1.3789793420670503e-05,
+      "loss": 0.7483,
+      "step": 16766
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0238819264669465,
+      "learning_rate": 1.378908728752536e-05,
+      "loss": 0.9144,
+      "step": 16767
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8056274266766086,
+      "learning_rate": 1.3788381132318955e-05,
+      "loss": 0.9126,
+      "step": 16768
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.822825537315009,
+      "learning_rate": 1.3787674955055395e-05,
+      "loss": 0.7992,
+      "step": 16769
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4736860208309435,
+      "learning_rate": 1.3786968755738792e-05,
+      "loss": 0.8445,
+      "step": 16770
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8246872666446874,
+      "learning_rate": 1.3786262534373256e-05,
+      "loss": 0.8592,
+      "step": 16771
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9281635626747842,
+      "learning_rate": 1.3785556290962902e-05,
+      "loss": 0.8434,
+      "step": 16772
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8052381676297609,
+      "learning_rate": 1.3784850025511842e-05,
+      "loss": 0.9073,
+      "step": 16773
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7012200750228699,
+      "learning_rate": 1.3784143738024188e-05,
+      "loss": 0.8441,
+      "step": 16774
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4405248941865267,
+      "learning_rate": 1.3783437428504048e-05,
+      "loss": 0.9167,
+      "step": 16775
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8041768926638007,
+      "learning_rate": 1.3782731096955537e-05,
+      "loss": 0.8878,
+      "step": 16776
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6759591077685385,
+      "learning_rate": 1.378202474338277e-05,
+      "loss": 0.9986,
+      "step": 16777
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6173449660415742,
+      "learning_rate": 1.3781318367789858e-05,
+      "loss": 0.866,
+      "step": 16778
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6214379927862896,
+      "learning_rate": 1.3780611970180913e-05,
+      "loss": 0.8793,
+      "step": 16779
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.958143399989008,
+      "learning_rate": 1.3779905550560047e-05,
+      "loss": 0.806,
+      "step": 16780
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9715080665383076,
+      "learning_rate": 1.3779199108931374e-05,
+      "loss": 0.786,
+      "step": 16781
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9455706860285071,
+      "learning_rate": 1.3778492645299008e-05,
+      "loss": 0.7269,
+      "step": 16782
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7150984847939617,
+      "learning_rate": 1.3777786159667063e-05,
+      "loss": 1.0032,
+      "step": 16783
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4892223227485588,
+      "learning_rate": 1.3777079652039649e-05,
+      "loss": 0.8524,
+      "step": 16784
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.627990781758099,
+      "learning_rate": 1.3776373122420884e-05,
+      "loss": 0.8603,
+      "step": 16785
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4948164674361437,
+      "learning_rate": 1.3775666570814876e-05,
+      "loss": 0.7869,
+      "step": 16786
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7823090971100862,
+      "learning_rate": 1.3774959997225743e-05,
+      "loss": 0.7017,
+      "step": 16787
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4692911255122922,
+      "learning_rate": 1.3774253401657599e-05,
+      "loss": 0.8882,
+      "step": 16788
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.570472897600453,
+      "learning_rate": 1.3773546784114556e-05,
+      "loss": 0.8566,
+      "step": 16789
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7326896341161027,
+      "learning_rate": 1.377284014460073e-05,
+      "loss": 0.6653,
+      "step": 16790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7173674651396662,
+      "learning_rate": 1.3772133483120229e-05,
+      "loss": 0.8134,
+      "step": 16791
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.607194098128238,
+      "learning_rate": 1.377142679967718e-05,
+      "loss": 0.9433,
+      "step": 16792
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6145379823922836,
+      "learning_rate": 1.3770720094275685e-05,
+      "loss": 0.8801,
+      "step": 16793
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5452366015078876,
+      "learning_rate": 1.3770013366919867e-05,
+      "loss": 0.8569,
+      "step": 16794
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6187486657299155,
+      "learning_rate": 1.3769306617613833e-05,
+      "loss": 0.9185,
+      "step": 16795
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7776151346916467,
+      "learning_rate": 1.3768599846361707e-05,
+      "loss": 0.7594,
+      "step": 16796
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0520733818186288,
+      "learning_rate": 1.37678930531676e-05,
+      "loss": 0.9388,
+      "step": 16797
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5327658514321343,
+      "learning_rate": 1.3767186238035624e-05,
+      "loss": 0.8104,
+      "step": 16798
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4922270378935332,
+      "learning_rate": 1.3766479400969899e-05,
+      "loss": 0.9427,
+      "step": 16799
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6994097315953718,
+      "learning_rate": 1.3765772541974536e-05,
+      "loss": 0.8253,
+      "step": 16800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7956708168239752,
+      "learning_rate": 1.3765065661053654e-05,
+      "loss": 0.9476,
+      "step": 16801
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6230943057994076,
+      "learning_rate": 1.376435875821137e-05,
+      "loss": 0.8434,
+      "step": 16802
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5851117874328984,
+      "learning_rate": 1.3763651833451795e-05,
+      "loss": 0.9033,
+      "step": 16803
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6003975831950248,
+      "learning_rate": 1.3762944886779049e-05,
+      "loss": 0.8319,
+      "step": 16804
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5497869503782602,
+      "learning_rate": 1.3762237918197245e-05,
+      "loss": 0.9773,
+      "step": 16805
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7497758141843893,
+      "learning_rate": 1.37615309277105e-05,
+      "loss": 0.9517,
+      "step": 16806
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6788257487112452,
+      "learning_rate": 1.3760823915322934e-05,
+      "loss": 0.8073,
+      "step": 16807
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8720571299396223,
+      "learning_rate": 1.3760116881038658e-05,
+      "loss": 0.8386,
+      "step": 16808
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.488634969823904,
+      "learning_rate": 1.375940982486179e-05,
+      "loss": 0.9026,
+      "step": 16809
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9670147310512804,
+      "learning_rate": 1.3758702746796454e-05,
+      "loss": 0.8935,
+      "step": 16810
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4552846814764215,
+      "learning_rate": 1.3757995646846755e-05,
+      "loss": 0.8753,
+      "step": 16811
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5985265174814127,
+      "learning_rate": 1.3757288525016816e-05,
+      "loss": 1.0018,
+      "step": 16812
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5327611866386088,
+      "learning_rate": 1.375658138131076e-05,
+      "loss": 0.7937,
+      "step": 16813
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7881294563854275,
+      "learning_rate": 1.3755874215732687e-05,
+      "loss": 0.8678,
+      "step": 16814
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4096385452820575,
+      "learning_rate": 1.3755167028286733e-05,
+      "loss": 0.8225,
+      "step": 16815
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6899992608585976,
+      "learning_rate": 1.3754459818977007e-05,
+      "loss": 0.9093,
+      "step": 16816
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1188191021504714,
+      "learning_rate": 1.3753752587807627e-05,
+      "loss": 0.9347,
+      "step": 16817
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8462792420158624,
+      "learning_rate": 1.3753045334782707e-05,
+      "loss": 0.7773,
+      "step": 16818
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.635231291817316,
+      "learning_rate": 1.3752338059906374e-05,
+      "loss": 0.991,
+      "step": 16819
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5563095138784249,
+      "learning_rate": 1.3751630763182738e-05,
+      "loss": 0.9098,
+      "step": 16820
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4691698577834227,
+      "learning_rate": 1.375092344461592e-05,
+      "loss": 0.8383,
+      "step": 16821
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5887776029118144,
+      "learning_rate": 1.375021610421004e-05,
+      "loss": 0.7417,
+      "step": 16822
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.099455899115646,
+      "learning_rate": 1.3749508741969213e-05,
+      "loss": 0.9893,
+      "step": 16823
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.538300348864719,
+      "learning_rate": 1.3748801357897559e-05,
+      "loss": 0.8713,
+      "step": 16824
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5750615111293094,
+      "learning_rate": 1.3748093951999195e-05,
+      "loss": 0.9623,
+      "step": 16825
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4950794769448588,
+      "learning_rate": 1.3747386524278246e-05,
+      "loss": 0.9471,
+      "step": 16826
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5769748086176174,
+      "learning_rate": 1.3746679074738822e-05,
+      "loss": 0.8618,
+      "step": 16827
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.768307538726499,
+      "learning_rate": 1.3745971603385049e-05,
+      "loss": 0.9482,
+      "step": 16828
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7728481836836787,
+      "learning_rate": 1.3745264110221042e-05,
+      "loss": 0.8294,
+      "step": 16829
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3209352975482638,
+      "learning_rate": 1.3744556595250922e-05,
+      "loss": 1.0096,
+      "step": 16830
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.096180847967985,
+      "learning_rate": 1.3743849058478809e-05,
+      "loss": 0.9251,
+      "step": 16831
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5547212057055688,
+      "learning_rate": 1.374314149990882e-05,
+      "loss": 0.8905,
+      "step": 16832
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1737032624556285,
+      "learning_rate": 1.3742433919545076e-05,
+      "loss": 0.9503,
+      "step": 16833
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6424727790233369,
+      "learning_rate": 1.3741726317391699e-05,
+      "loss": 0.9723,
+      "step": 16834
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.514811228559017,
+      "learning_rate": 1.3741018693452803e-05,
+      "loss": 0.9657,
+      "step": 16835
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.086967617046384,
+      "learning_rate": 1.3740311047732514e-05,
+      "loss": 0.9219,
+      "step": 16836
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6203920950195114,
+      "learning_rate": 1.373960338023495e-05,
+      "loss": 0.9563,
+      "step": 16837
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.678636855422969,
+      "learning_rate": 1.3738895690964231e-05,
+      "loss": 0.9199,
+      "step": 16838
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6637141966522502,
+      "learning_rate": 1.3738187979924478e-05,
+      "loss": 0.8443,
+      "step": 16839
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2449248881600457,
+      "learning_rate": 1.373748024711981e-05,
+      "loss": 0.9195,
+      "step": 16840
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4974151958720239,
+      "learning_rate": 1.3736772492554352e-05,
+      "loss": 0.895,
+      "step": 16841
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6693174748602297,
+      "learning_rate": 1.3736064716232216e-05,
+      "loss": 0.7788,
+      "step": 16842
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5689544672540425,
+      "learning_rate": 1.3735356918157532e-05,
+      "loss": 0.8638,
+      "step": 16843
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7458345894190854,
+      "learning_rate": 1.3734649098334416e-05,
+      "loss": 0.6818,
+      "step": 16844
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.507758758776517,
+      "learning_rate": 1.3733941256766994e-05,
+      "loss": 0.8666,
+      "step": 16845
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9034638871202811,
+      "learning_rate": 1.3733233393459377e-05,
+      "loss": 0.9067,
+      "step": 16846
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7010473143364642,
+      "learning_rate": 1.3732525508415702e-05,
+      "loss": 0.8347,
+      "step": 16847
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4535886842852515,
+      "learning_rate": 1.3731817601640077e-05,
+      "loss": 0.9104,
+      "step": 16848
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.685200894184952,
+      "learning_rate": 1.3731109673136627e-05,
+      "loss": 0.8032,
+      "step": 16849
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.521722342128095,
+      "learning_rate": 1.3730401722909479e-05,
+      "loss": 0.8814,
+      "step": 16850
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.093166255931981,
+      "learning_rate": 1.3729693750962747e-05,
+      "loss": 0.8596,
+      "step": 16851
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5817845877993668,
+      "learning_rate": 1.372898575730056e-05,
+      "loss": 0.9443,
+      "step": 16852
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5418239975371604,
+      "learning_rate": 1.3728277741927037e-05,
+      "loss": 0.9433,
+      "step": 16853
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.653122424027356,
+      "learning_rate": 1.37275697048463e-05,
+      "loss": 0.8264,
+      "step": 16854
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4431054472813243,
+      "learning_rate": 1.3726861646062472e-05,
+      "loss": 0.8929,
+      "step": 16855
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7335017515005495,
+      "learning_rate": 1.3726153565579676e-05,
+      "loss": 0.8352,
+      "step": 16856
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6640573610328016,
+      "learning_rate": 1.3725445463402034e-05,
+      "loss": 0.8721,
+      "step": 16857
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9066826553082385,
+      "learning_rate": 1.3724737339533669e-05,
+      "loss": 0.9302,
+      "step": 16858
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7476220069619672,
+      "learning_rate": 1.3724029193978706e-05,
+      "loss": 0.8289,
+      "step": 16859
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5582095536061789,
+      "learning_rate": 1.3723321026741262e-05,
+      "loss": 0.7271,
+      "step": 16860
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6452241767388864,
+      "learning_rate": 1.3722612837825466e-05,
+      "loss": 0.8777,
+      "step": 16861
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9349054957428975,
+      "learning_rate": 1.3721904627235444e-05,
+      "loss": 0.89,
+      "step": 16862
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6796415177659147,
+      "learning_rate": 1.3721196394975306e-05,
+      "loss": 0.8335,
+      "step": 16863
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6154515468248198,
+      "learning_rate": 1.3720488141049193e-05,
+      "loss": 1.0036,
+      "step": 16864
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.284673880239047,
+      "learning_rate": 1.3719779865461214e-05,
+      "loss": 0.8039,
+      "step": 16865
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.453023474141124,
+      "learning_rate": 1.3719071568215503e-05,
+      "loss": 0.9346,
+      "step": 16866
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7733966637760589,
+      "learning_rate": 1.3718363249316179e-05,
+      "loss": 0.7718,
+      "step": 16867
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.893187671067565,
+      "learning_rate": 1.3717654908767366e-05,
+      "loss": 0.8319,
+      "step": 16868
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5278338017488475,
+      "learning_rate": 1.3716946546573186e-05,
+      "loss": 0.8904,
+      "step": 16869
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5741526558986176,
+      "learning_rate": 1.3716238162737772e-05,
+      "loss": 0.8857,
+      "step": 16870
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5578083355839534,
+      "learning_rate": 1.3715529757265243e-05,
+      "loss": 0.7951,
+      "step": 16871
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.472633149481425,
+      "learning_rate": 1.371482133015972e-05,
+      "loss": 0.8639,
+      "step": 16872
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5828728980170408,
+      "learning_rate": 1.3714112881425332e-05,
+      "loss": 0.8461,
+      "step": 16873
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9925730802532537,
+      "learning_rate": 1.3713404411066203e-05,
+      "loss": 0.7834,
+      "step": 16874
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.033060640987292,
+      "learning_rate": 1.3712695919086458e-05,
+      "loss": 0.9902,
+      "step": 16875
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4790553680718992,
+      "learning_rate": 1.371198740549022e-05,
+      "loss": 0.9561,
+      "step": 16876
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.515732514971738,
+      "learning_rate": 1.3711278870281622e-05,
+      "loss": 0.896,
+      "step": 16877
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3886302185039396,
+      "learning_rate": 1.3710570313464778e-05,
+      "loss": 0.8804,
+      "step": 16878
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7445486430377832,
+      "learning_rate": 1.3709861735043818e-05,
+      "loss": 0.7152,
+      "step": 16879
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7703454018675224,
+      "learning_rate": 1.3709153135022873e-05,
+      "loss": 0.703,
+      "step": 16880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.954027718575454,
+      "learning_rate": 1.3708444513406062e-05,
+      "loss": 0.8553,
+      "step": 16881
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.622103214952932,
+      "learning_rate": 1.3707735870197514e-05,
+      "loss": 0.9094,
+      "step": 16882
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.44924632678647,
+      "learning_rate": 1.3707027205401353e-05,
+      "loss": 0.9546,
+      "step": 16883
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6473085604180453,
+      "learning_rate": 1.3706318519021704e-05,
+      "loss": 0.8729,
+      "step": 16884
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8190613202843671,
+      "learning_rate": 1.37056098110627e-05,
+      "loss": 0.7573,
+      "step": 16885
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0580368705814203,
+      "learning_rate": 1.3704901081528457e-05,
+      "loss": 0.8885,
+      "step": 16886
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.2928461822851425,
+      "learning_rate": 1.3704192330423109e-05,
+      "loss": 0.8389,
+      "step": 16887
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6533871761172678,
+      "learning_rate": 1.3703483557750783e-05,
+      "loss": 0.8651,
+      "step": 16888
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.612208534037318,
+      "learning_rate": 1.3702774763515601e-05,
+      "loss": 0.8893,
+      "step": 16889
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5761924578192528,
+      "learning_rate": 1.3702065947721691e-05,
+      "loss": 0.8099,
+      "step": 16890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.456025172467882,
+      "learning_rate": 1.3701357110373184e-05,
+      "loss": 0.9533,
+      "step": 16891
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4618266891749763,
+      "learning_rate": 1.3700648251474204e-05,
+      "loss": 1.004,
+      "step": 16892
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.44068584916304,
+      "learning_rate": 1.3699939371028873e-05,
+      "loss": 0.8128,
+      "step": 16893
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7095428030252056,
+      "learning_rate": 1.3699230469041331e-05,
+      "loss": 0.8709,
+      "step": 16894
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4680183052162377,
+      "learning_rate": 1.3698521545515693e-05,
+      "loss": 0.9812,
+      "step": 16895
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1643541772713597,
+      "learning_rate": 1.3697812600456094e-05,
+      "loss": 0.8048,
+      "step": 16896
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5662798171658636,
+      "learning_rate": 1.3697103633866657e-05,
+      "loss": 0.8093,
+      "step": 16897
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.760156706681401,
+      "learning_rate": 1.3696394645751514e-05,
+      "loss": 0.7439,
+      "step": 16898
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7919167331711943,
+      "learning_rate": 1.369568563611479e-05,
+      "loss": 1.0114,
+      "step": 16899
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7068937087870293,
+      "learning_rate": 1.3694976604960615e-05,
+      "loss": 0.8155,
+      "step": 16900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5310730440201716,
+      "learning_rate": 1.3694267552293117e-05,
+      "loss": 0.7785,
+      "step": 16901
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0635555716689753,
+      "learning_rate": 1.3693558478116425e-05,
+      "loss": 0.9986,
+      "step": 16902
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7654397460424313,
+      "learning_rate": 1.369284938243466e-05,
+      "loss": 0.7214,
+      "step": 16903
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7727699291564463,
+      "learning_rate": 1.3692140265251962e-05,
+      "loss": 0.8892,
+      "step": 16904
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9039732133622496,
+      "learning_rate": 1.3691431126572454e-05,
+      "loss": 0.8749,
+      "step": 16905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.542636615581762,
+      "learning_rate": 1.3690721966400264e-05,
+      "loss": 0.8857,
+      "step": 16906
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4202727082554838,
+      "learning_rate": 1.3690012784739524e-05,
+      "loss": 0.9437,
+      "step": 16907
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.520016598059137,
+      "learning_rate": 1.368930358159436e-05,
+      "loss": 0.7842,
+      "step": 16908
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6381739446445382,
+      "learning_rate": 1.3688594356968903e-05,
+      "loss": 0.8569,
+      "step": 16909
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1529909061309325,
+      "learning_rate": 1.368788511086728e-05,
+      "loss": 0.7668,
+      "step": 16910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6201108028563138,
+      "learning_rate": 1.3687175843293625e-05,
+      "loss": 0.8171,
+      "step": 16911
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.259656824778582,
+      "learning_rate": 1.3686466554252062e-05,
+      "loss": 0.9471,
+      "step": 16912
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5637450356269107,
+      "learning_rate": 1.3685757243746728e-05,
+      "loss": 0.8804,
+      "step": 16913
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.573278001473684,
+      "learning_rate": 1.3685047911781744e-05,
+      "loss": 0.9099,
+      "step": 16914
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.188842506125727,
+      "learning_rate": 1.368433855836125e-05,
+      "loss": 0.9577,
+      "step": 16915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7395444035585661,
+      "learning_rate": 1.3683629183489364e-05,
+      "loss": 0.7409,
+      "step": 16916
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.697723129334704,
+      "learning_rate": 1.3682919787170228e-05,
+      "loss": 0.7646,
+      "step": 16917
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5504256548381814,
+      "learning_rate": 1.3682210369407961e-05,
+      "loss": 0.9068,
+      "step": 16918
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1835953805702837,
+      "learning_rate": 1.3681500930206707e-05,
+      "loss": 0.9381,
+      "step": 16919
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.546877636769515,
+      "learning_rate": 1.3680791469570585e-05,
+      "loss": 0.9085,
+      "step": 16920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8757257406036805,
+      "learning_rate": 1.3680081987503728e-05,
+      "loss": 0.8511,
+      "step": 16921
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.805372790033491,
+      "learning_rate": 1.3679372484010269e-05,
+      "loss": 0.9902,
+      "step": 16922
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4876517313931261,
+      "learning_rate": 1.3678662959094338e-05,
+      "loss": 0.968,
+      "step": 16923
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8513156753168694,
+      "learning_rate": 1.3677953412760072e-05,
+      "loss": 0.9372,
+      "step": 16924
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6458630274290953,
+      "learning_rate": 1.3677243845011589e-05,
+      "loss": 0.8683,
+      "step": 16925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4688680616333751,
+      "learning_rate": 1.3676534255853036e-05,
+      "loss": 0.8382,
+      "step": 16926
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4907595470468487,
+      "learning_rate": 1.3675824645288533e-05,
+      "loss": 0.8067,
+      "step": 16927
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7891608863811139,
+      "learning_rate": 1.3675115013322215e-05,
+      "loss": 0.7735,
+      "step": 16928
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.805264313751026,
+      "learning_rate": 1.3674405359958214e-05,
+      "loss": 0.755,
+      "step": 16929
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9456918023156828,
+      "learning_rate": 1.367369568520066e-05,
+      "loss": 0.7988,
+      "step": 16930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.467627142691281,
+      "learning_rate": 1.367298598905369e-05,
+      "loss": 0.9433,
+      "step": 16931
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7326427298792053,
+      "learning_rate": 1.367227627152143e-05,
+      "loss": 0.7741,
+      "step": 16932
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.708130012031855,
+      "learning_rate": 1.3671566532608015e-05,
+      "loss": 0.8038,
+      "step": 16933
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8330002935812018,
+      "learning_rate": 1.367085677231758e-05,
+      "loss": 0.7647,
+      "step": 16934
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4949544890313615,
+      "learning_rate": 1.3670146990654252e-05,
+      "loss": 0.8925,
+      "step": 16935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9351972901414478,
+      "learning_rate": 1.3669437187622168e-05,
+      "loss": 0.9478,
+      "step": 16936
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6144662242652648,
+      "learning_rate": 1.3668727363225459e-05,
+      "loss": 0.9669,
+      "step": 16937
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8244982520518744,
+      "learning_rate": 1.3668017517468255e-05,
+      "loss": 0.8591,
+      "step": 16938
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6109831927575704,
+      "learning_rate": 1.3667307650354693e-05,
+      "loss": 1.0163,
+      "step": 16939
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4500928433377451,
+      "learning_rate": 1.3666597761888906e-05,
+      "loss": 0.9273,
+      "step": 16940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.525874480756064,
+      "learning_rate": 1.3665887852075026e-05,
+      "loss": 0.9042,
+      "step": 16941
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7362459816043334,
+      "learning_rate": 1.3665177920917185e-05,
+      "loss": 0.8547,
+      "step": 16942
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9320239390167369,
+      "learning_rate": 1.3664467968419522e-05,
+      "loss": 0.7491,
+      "step": 16943
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.413704136596176,
+      "learning_rate": 1.366375799458616e-05,
+      "loss": 0.8545,
+      "step": 16944
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6533813133311561,
+      "learning_rate": 1.3663047999421243e-05,
+      "loss": 0.9557,
+      "step": 16945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9156502717493318,
+      "learning_rate": 1.36623379829289e-05,
+      "loss": 0.8579,
+      "step": 16946
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.56775552755636,
+      "learning_rate": 1.3661627945113268e-05,
+      "loss": 1.0426,
+      "step": 16947
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6409419465489732,
+      "learning_rate": 1.3660917885978473e-05,
+      "loss": 0.933,
+      "step": 16948
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6579287780527463,
+      "learning_rate": 1.366020780552866e-05,
+      "loss": 0.8557,
+      "step": 16949
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.882955556038465,
+      "learning_rate": 1.3659497703767957e-05,
+      "loss": 0.8733,
+      "step": 16950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6686625045727765,
+      "learning_rate": 1.36587875807005e-05,
+      "loss": 0.9033,
+      "step": 16951
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5373498922981998,
+      "learning_rate": 1.3658077436330421e-05,
+      "loss": 0.8227,
+      "step": 16952
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4099766133155394,
+      "learning_rate": 1.365736727066186e-05,
+      "loss": 0.7571,
+      "step": 16953
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3138680665986615,
+      "learning_rate": 1.3656657083698947e-05,
+      "loss": 0.8335,
+      "step": 16954
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7982996010692565,
+      "learning_rate": 1.365594687544582e-05,
+      "loss": 0.871,
+      "step": 16955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7526097735775497,
+      "learning_rate": 1.3655236645906613e-05,
+      "loss": 0.9526,
+      "step": 16956
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7201047648807188,
+      "learning_rate": 1.3654526395085461e-05,
+      "loss": 0.808,
+      "step": 16957
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7061627585229828,
+      "learning_rate": 1.3653816122986499e-05,
+      "loss": 0.996,
+      "step": 16958
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5551173081252077,
+      "learning_rate": 1.3653105829613862e-05,
+      "loss": 0.8541,
+      "step": 16959
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.657177786742928,
+      "learning_rate": 1.3652395514971686e-05,
+      "loss": 0.9228,
+      "step": 16960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.390811473771705,
+      "learning_rate": 1.365168517906411e-05,
+      "loss": 0.8738,
+      "step": 16961
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6251577742710552,
+      "learning_rate": 1.3650974821895262e-05,
+      "loss": 0.9773,
+      "step": 16962
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4345758475860027,
+      "learning_rate": 1.3650264443469285e-05,
+      "loss": 0.8346,
+      "step": 16963
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8977104406337482,
+      "learning_rate": 1.3649554043790316e-05,
+      "loss": 0.9425,
+      "step": 16964
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6393816071869527,
+      "learning_rate": 1.3648843622862483e-05,
+      "loss": 0.915,
+      "step": 16965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.913246506222284,
+      "learning_rate": 1.3648133180689932e-05,
+      "loss": 0.9325,
+      "step": 16966
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6686120971187424,
+      "learning_rate": 1.3647422717276788e-05,
+      "loss": 0.827,
+      "step": 16967
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8050992959716465,
+      "learning_rate": 1.3646712232627201e-05,
+      "loss": 0.9315,
+      "step": 16968
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4242000793675706,
+      "learning_rate": 1.3646001726745298e-05,
+      "loss": 0.9331,
+      "step": 16969
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.713545071678637,
+      "learning_rate": 1.3645291199635218e-05,
+      "loss": 0.8276,
+      "step": 16970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7406226805731904,
+      "learning_rate": 1.36445806513011e-05,
+      "loss": 0.9314,
+      "step": 16971
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5999137381768893,
+      "learning_rate": 1.3643870081747076e-05,
+      "loss": 0.776,
+      "step": 16972
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1746273159272813,
+      "learning_rate": 1.364315949097729e-05,
+      "loss": 0.798,
+      "step": 16973
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8467924940242648,
+      "learning_rate": 1.3642448878995876e-05,
+      "loss": 0.8669,
+      "step": 16974
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7928477524051942,
+      "learning_rate": 1.3641738245806973e-05,
+      "loss": 0.7839,
+      "step": 16975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5237693052755479,
+      "learning_rate": 1.364102759141471e-05,
+      "loss": 0.8648,
+      "step": 16976
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.693190476867517,
+      "learning_rate": 1.3640316915823242e-05,
+      "loss": 0.8873,
+      "step": 16977
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.800084946398855,
+      "learning_rate": 1.363960621903669e-05,
+      "loss": 0.8809,
+      "step": 16978
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.658784734268649,
+      "learning_rate": 1.36388955010592e-05,
+      "loss": 0.8748,
+      "step": 16979
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3859435708289711,
+      "learning_rate": 1.3638184761894907e-05,
+      "loss": 0.8148,
+      "step": 16980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5247182191156883,
+      "learning_rate": 1.363747400154795e-05,
+      "loss": 0.9165,
+      "step": 16981
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7897948527815138,
+      "learning_rate": 1.3636763220022468e-05,
+      "loss": 0.7566,
+      "step": 16982
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5263334695846746,
+      "learning_rate": 1.3636052417322601e-05,
+      "loss": 0.9593,
+      "step": 16983
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1239172648196596,
+      "learning_rate": 1.3635341593452484e-05,
+      "loss": 0.8777,
+      "step": 16984
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.066247192889507,
+      "learning_rate": 1.3634630748416257e-05,
+      "loss": 0.861,
+      "step": 16985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7775006543157097,
+      "learning_rate": 1.3633919882218063e-05,
+      "loss": 0.9435,
+      "step": 16986
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6586947760408774,
+      "learning_rate": 1.3633208994862033e-05,
+      "loss": 0.8263,
+      "step": 16987
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8591559995597053,
+      "learning_rate": 1.363249808635231e-05,
+      "loss": 0.8459,
+      "step": 16988
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7958932869090836,
+      "learning_rate": 1.3631787156693035e-05,
+      "loss": 1.0085,
+      "step": 16989
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6208772801664448,
+      "learning_rate": 1.3631076205888344e-05,
+      "loss": 0.8445,
+      "step": 16990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6076458355289518,
+      "learning_rate": 1.363036523394238e-05,
+      "loss": 1.0585,
+      "step": 16991
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6642981961021943,
+      "learning_rate": 1.3629654240859279e-05,
+      "loss": 0.9117,
+      "step": 16992
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7931631741517887,
+      "learning_rate": 1.362894322664318e-05,
+      "loss": 0.6957,
+      "step": 16993
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7871506881791672,
+      "learning_rate": 1.3628232191298227e-05,
+      "loss": 0.9181,
+      "step": 16994
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0121879727300613,
+      "learning_rate": 1.3627521134828553e-05,
+      "loss": 0.9184,
+      "step": 16995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5663958462050964,
+      "learning_rate": 1.3626810057238311e-05,
+      "loss": 0.8755,
+      "step": 16996
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5656630405858614,
+      "learning_rate": 1.3626098958531624e-05,
+      "loss": 0.8427,
+      "step": 16997
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5183623827283617,
+      "learning_rate": 1.3625387838712647e-05,
+      "loss": 0.8637,
+      "step": 16998
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7502942770657695,
+      "learning_rate": 1.3624676697785511e-05,
+      "loss": 0.788,
+      "step": 16999
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0260424156206605,
+      "learning_rate": 1.3623965535754363e-05,
+      "loss": 0.877,
+      "step": 17000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5349513558967836,
+      "learning_rate": 1.3623254352623336e-05,
+      "loss": 0.8552,
+      "step": 17001
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7661539925339853,
+      "learning_rate": 1.3622543148396577e-05,
+      "loss": 0.7555,
+      "step": 17002
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9002091805344792,
+      "learning_rate": 1.3621831923078224e-05,
+      "loss": 0.8733,
+      "step": 17003
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8205099327167447,
+      "learning_rate": 1.362112067667242e-05,
+      "loss": 0.8605,
+      "step": 17004
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6100977048154101,
+      "learning_rate": 1.3620409409183304e-05,
+      "loss": 0.984,
+      "step": 17005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.901016554358025,
+      "learning_rate": 1.3619698120615019e-05,
+      "loss": 0.8905,
+      "step": 17006
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5404643318692286,
+      "learning_rate": 1.3618986810971707e-05,
+      "loss": 0.8689,
+      "step": 17007
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5114640186197872,
+      "learning_rate": 1.3618275480257506e-05,
+      "loss": 0.7428,
+      "step": 17008
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.102232154125245,
+      "learning_rate": 1.361756412847656e-05,
+      "loss": 0.7688,
+      "step": 17009
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7452382044351241,
+      "learning_rate": 1.361685275563301e-05,
+      "loss": 0.7506,
+      "step": 17010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6778241469596982,
+      "learning_rate": 1.3616141361730999e-05,
+      "loss": 0.9679,
+      "step": 17011
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7920027814565499,
+      "learning_rate": 1.3615429946774668e-05,
+      "loss": 0.7594,
+      "step": 17012
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.015381688814437,
+      "learning_rate": 1.3614718510768156e-05,
+      "loss": 0.9101,
+      "step": 17013
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9510891262469963,
+      "learning_rate": 1.3614007053715612e-05,
+      "loss": 0.9068,
+      "step": 17014
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.155152624141226,
+      "learning_rate": 1.3613295575621177e-05,
+      "loss": 0.7799,
+      "step": 17015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.902397336161669,
+      "learning_rate": 1.3612584076488986e-05,
+      "loss": 0.9211,
+      "step": 17016
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.137684294502645,
+      "learning_rate": 1.361187255632319e-05,
+      "loss": 0.7407,
+      "step": 17017
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1611855768205293,
+      "learning_rate": 1.3611161015127926e-05,
+      "loss": 0.9729,
+      "step": 17018
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5189594075604298,
+      "learning_rate": 1.3610449452907345e-05,
+      "loss": 0.8221,
+      "step": 17019
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6705561584943822,
+      "learning_rate": 1.3609737869665579e-05,
+      "loss": 1.0338,
+      "step": 17020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9149112626097815,
+      "learning_rate": 1.3609026265406775e-05,
+      "loss": 0.8287,
+      "step": 17021
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6441468434237163,
+      "learning_rate": 1.3608314640135084e-05,
+      "loss": 0.89,
+      "step": 17022
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9020309010161682,
+      "learning_rate": 1.3607602993854637e-05,
+      "loss": 0.8622,
+      "step": 17023
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7787250978905406,
+      "learning_rate": 1.3606891326569584e-05,
+      "loss": 0.7867,
+      "step": 17024
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6028148728798666,
+      "learning_rate": 1.3606179638284067e-05,
+      "loss": 0.802,
+      "step": 17025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6017762134049112,
+      "learning_rate": 1.3605467929002233e-05,
+      "loss": 0.956,
+      "step": 17026
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.951472917469815,
+      "learning_rate": 1.3604756198728222e-05,
+      "loss": 0.8214,
+      "step": 17027
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6163194633621079,
+      "learning_rate": 1.3604044447466182e-05,
+      "loss": 0.7765,
+      "step": 17028
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3772731723881924,
+      "learning_rate": 1.360333267522025e-05,
+      "loss": 0.884,
+      "step": 17029
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.033914601173184,
+      "learning_rate": 1.3602620881994577e-05,
+      "loss": 0.876,
+      "step": 17030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7862236443794546,
+      "learning_rate": 1.3601909067793304e-05,
+      "loss": 0.7451,
+      "step": 17031
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9533725279180438,
+      "learning_rate": 1.3601197232620575e-05,
+      "loss": 0.7826,
+      "step": 17032
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7244785032525082,
+      "learning_rate": 1.3600485376480536e-05,
+      "loss": 0.7,
+      "step": 17033
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9404147106728014,
+      "learning_rate": 1.3599773499377333e-05,
+      "loss": 0.9095,
+      "step": 17034
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5934430637723225,
+      "learning_rate": 1.359906160131511e-05,
+      "loss": 0.7515,
+      "step": 17035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.492570085116686,
+      "learning_rate": 1.3598349682298006e-05,
+      "loss": 0.8514,
+      "step": 17036
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5125437845143277,
+      "learning_rate": 1.3597637742330174e-05,
+      "loss": 0.81,
+      "step": 17037
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7992266509328911,
+      "learning_rate": 1.3596925781415755e-05,
+      "loss": 0.879,
+      "step": 17038
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8349936976999035,
+      "learning_rate": 1.3596213799558898e-05,
+      "loss": 0.9738,
+      "step": 17039
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7292572984469367,
+      "learning_rate": 1.359550179676374e-05,
+      "loss": 0.7389,
+      "step": 17040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.275384076563372,
+      "learning_rate": 1.3594789773034439e-05,
+      "loss": 0.9197,
+      "step": 17041
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7530218143712167,
+      "learning_rate": 1.3594077728375129e-05,
+      "loss": 0.7383,
+      "step": 17042
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6697808175959932,
+      "learning_rate": 1.3593365662789966e-05,
+      "loss": 0.8742,
+      "step": 17043
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.013459148690383,
+      "learning_rate": 1.3592653576283083e-05,
+      "loss": 0.8996,
+      "step": 17044
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7470186126185554,
+      "learning_rate": 1.359194146885864e-05,
+      "loss": 0.9782,
+      "step": 17045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6023310374325723,
+      "learning_rate": 1.359122934052077e-05,
+      "loss": 0.9453,
+      "step": 17046
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6114514630067582,
+      "learning_rate": 1.3590517191273634e-05,
+      "loss": 0.8642,
+      "step": 17047
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.807577445384451,
+      "learning_rate": 1.3589805021121363e-05,
+      "loss": 0.9106,
+      "step": 17048
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.469663867413236,
+      "learning_rate": 1.3589092830068116e-05,
+      "loss": 0.9019,
+      "step": 17049
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7613468802604025,
+      "learning_rate": 1.3588380618118031e-05,
+      "loss": 0.6721,
+      "step": 17050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.702293693207459,
+      "learning_rate": 1.3587668385275258e-05,
+      "loss": 0.8928,
+      "step": 17051
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.676038679950981,
+      "learning_rate": 1.3586956131543944e-05,
+      "loss": 0.9732,
+      "step": 17052
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9701681790533447,
+      "learning_rate": 1.3586243856928238e-05,
+      "loss": 0.8624,
+      "step": 17053
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7832658206313017,
+      "learning_rate": 1.3585531561432284e-05,
+      "loss": 0.7742,
+      "step": 17054
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6328025073505428,
+      "learning_rate": 1.3584819245060228e-05,
+      "loss": 0.9467,
+      "step": 17055
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7870253510929495,
+      "learning_rate": 1.3584106907816222e-05,
+      "loss": 0.9789,
+      "step": 17056
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6770922282251117,
+      "learning_rate": 1.3583394549704404e-05,
+      "loss": 0.9066,
+      "step": 17057
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8178808214005224,
+      "learning_rate": 1.3582682170728939e-05,
+      "loss": 0.8855,
+      "step": 17058
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6043954647116523,
+      "learning_rate": 1.3581969770893959e-05,
+      "loss": 0.7608,
+      "step": 17059
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5673429422788956,
+      "learning_rate": 1.3581257350203616e-05,
+      "loss": 0.9201,
+      "step": 17060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6958738869248984,
+      "learning_rate": 1.3580544908662062e-05,
+      "loss": 0.8427,
+      "step": 17061
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1402084421032104,
+      "learning_rate": 1.3579832446273438e-05,
+      "loss": 0.9168,
+      "step": 17062
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8325633865777209,
+      "learning_rate": 1.35791199630419e-05,
+      "loss": 0.9896,
+      "step": 17063
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7801063884926894,
+      "learning_rate": 1.357840745897159e-05,
+      "loss": 0.7673,
+      "step": 17064
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.855398459828516,
+      "learning_rate": 1.3577694934066659e-05,
+      "loss": 0.876,
+      "step": 17065
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4478279673383296,
+      "learning_rate": 1.3576982388331258e-05,
+      "loss": 0.8504,
+      "step": 17066
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.775787847684351,
+      "learning_rate": 1.357626982176953e-05,
+      "loss": 1.0067,
+      "step": 17067
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.812332162329284,
+      "learning_rate": 1.3575557234385628e-05,
+      "loss": 0.8821,
+      "step": 17068
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1812850083929582,
+      "learning_rate": 1.3574844626183698e-05,
+      "loss": 0.7488,
+      "step": 17069
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8470386183131793,
+      "learning_rate": 1.3574131997167891e-05,
+      "loss": 0.9296,
+      "step": 17070
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7753078552729115,
+      "learning_rate": 1.3573419347342358e-05,
+      "loss": 0.7505,
+      "step": 17071
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6275215412823845,
+      "learning_rate": 1.3572706676711247e-05,
+      "loss": 0.8066,
+      "step": 17072
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5091214442910565,
+      "learning_rate": 1.3571993985278706e-05,
+      "loss": 0.9456,
+      "step": 17073
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7688674499299268,
+      "learning_rate": 1.3571281273048885e-05,
+      "loss": 0.9342,
+      "step": 17074
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6613411639560927,
+      "learning_rate": 1.3570568540025933e-05,
+      "loss": 0.8542,
+      "step": 17075
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6965328268840403,
+      "learning_rate": 1.3569855786213998e-05,
+      "loss": 0.9708,
+      "step": 17076
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4992655431214625,
+      "learning_rate": 1.3569143011617238e-05,
+      "loss": 0.9518,
+      "step": 17077
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7689987477202784,
+      "learning_rate": 1.3568430216239792e-05,
+      "loss": 0.8785,
+      "step": 17078
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7118883256295492,
+      "learning_rate": 1.356771740008582e-05,
+      "loss": 1.0271,
+      "step": 17079
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4634928586946128,
+      "learning_rate": 1.3567004563159467e-05,
+      "loss": 0.8286,
+      "step": 17080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.112457701176178,
+      "learning_rate": 1.3566291705464883e-05,
+      "loss": 0.9616,
+      "step": 17081
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6827979576393768,
+      "learning_rate": 1.3565578827006218e-05,
+      "loss": 0.8378,
+      "step": 17082
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8502817188847485,
+      "learning_rate": 1.3564865927787627e-05,
+      "loss": 0.9576,
+      "step": 17083
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.551076991589521,
+      "learning_rate": 1.3564153007813257e-05,
+      "loss": 1.0312,
+      "step": 17084
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7559042508232596,
+      "learning_rate": 1.356344006708726e-05,
+      "loss": 0.8655,
+      "step": 17085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7256657466785332,
+      "learning_rate": 1.3562727105613784e-05,
+      "loss": 0.8446,
+      "step": 17086
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.038072080269816,
+      "learning_rate": 1.3562014123396988e-05,
+      "loss": 0.8214,
+      "step": 17087
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.474610590685122,
+      "learning_rate": 1.3561301120441015e-05,
+      "loss": 0.8195,
+      "step": 17088
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7337115577343833,
+      "learning_rate": 1.3560588096750018e-05,
+      "loss": 0.8981,
+      "step": 17089
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9528590114791107,
+      "learning_rate": 1.355987505232815e-05,
+      "loss": 0.9552,
+      "step": 17090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5736588024361318,
+      "learning_rate": 1.3559161987179565e-05,
+      "loss": 1.0784,
+      "step": 17091
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7950362517773415,
+      "learning_rate": 1.3558448901308411e-05,
+      "loss": 0.9661,
+      "step": 17092
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5618453424072023,
+      "learning_rate": 1.355773579471884e-05,
+      "loss": 0.9105,
+      "step": 17093
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.580930821140351,
+      "learning_rate": 1.3557022667415009e-05,
+      "loss": 0.8655,
+      "step": 17094
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6467011456692866,
+      "learning_rate": 1.3556309519401061e-05,
+      "loss": 0.8643,
+      "step": 17095
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.182810263126561,
+      "learning_rate": 1.3555596350681158e-05,
+      "loss": 0.9924,
+      "step": 17096
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5068393621989724,
+      "learning_rate": 1.355488316125944e-05,
+      "loss": 0.9577,
+      "step": 17097
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7643098799976362,
+      "learning_rate": 1.3554169951140073e-05,
+      "loss": 0.9956,
+      "step": 17098
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7434490854723474,
+      "learning_rate": 1.35534567203272e-05,
+      "loss": 0.8878,
+      "step": 17099
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7014737081455535,
+      "learning_rate": 1.355274346882498e-05,
+      "loss": 0.834,
+      "step": 17100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5859555313669091,
+      "learning_rate": 1.355203019663756e-05,
+      "loss": 0.8077,
+      "step": 17101
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.519689450553425,
+      "learning_rate": 1.3551316903769097e-05,
+      "loss": 0.7964,
+      "step": 17102
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4578086889080504,
+      "learning_rate": 1.3550603590223744e-05,
+      "loss": 0.9882,
+      "step": 17103
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.490750234730187,
+      "learning_rate": 1.3549890256005649e-05,
+      "loss": 0.8636,
+      "step": 17104
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2142860538933236,
+      "learning_rate": 1.354917690111897e-05,
+      "loss": 0.8963,
+      "step": 17105
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5169454524249901,
+      "learning_rate": 1.3548463525567858e-05,
+      "loss": 0.9294,
+      "step": 17106
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6601542052263403,
+      "learning_rate": 1.3547750129356474e-05,
+      "loss": 0.8566,
+      "step": 17107
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6989718180698052,
+      "learning_rate": 1.3547036712488961e-05,
+      "loss": 0.9177,
+      "step": 17108
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7974192431931229,
+      "learning_rate": 1.3546323274969479e-05,
+      "loss": 0.7937,
+      "step": 17109
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.667300040511061,
+      "learning_rate": 1.354560981680218e-05,
+      "loss": 0.897,
+      "step": 17110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5539618266296984,
+      "learning_rate": 1.3544896337991219e-05,
+      "loss": 1.0235,
+      "step": 17111
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9436342300203866,
+      "learning_rate": 1.3544182838540745e-05,
+      "loss": 0.9555,
+      "step": 17112
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8720277997894947,
+      "learning_rate": 1.354346931845492e-05,
+      "loss": 0.9008,
+      "step": 17113
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6867334988315528,
+      "learning_rate": 1.3542755777737892e-05,
+      "loss": 0.9164,
+      "step": 17114
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9232804808143744,
+      "learning_rate": 1.3542042216393825e-05,
+      "loss": 0.9435,
+      "step": 17115
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8867667410936457,
+      "learning_rate": 1.3541328634426858e-05,
+      "loss": 0.9382,
+      "step": 17116
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9471300795743343,
+      "learning_rate": 1.3540615031841163e-05,
+      "loss": 0.9809,
+      "step": 17117
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4865091318159926,
+      "learning_rate": 1.353990140864088e-05,
+      "loss": 0.8766,
+      "step": 17118
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.634437638369363,
+      "learning_rate": 1.353918776483017e-05,
+      "loss": 0.8053,
+      "step": 17119
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6038066908069823,
+      "learning_rate": 1.353847410041319e-05,
+      "loss": 0.8386,
+      "step": 17120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.636142225776054,
+      "learning_rate": 1.3537760415394095e-05,
+      "loss": 0.7167,
+      "step": 17121
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6516185329319581,
+      "learning_rate": 1.3537046709777037e-05,
+      "loss": 0.8104,
+      "step": 17122
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8197094135679673,
+      "learning_rate": 1.3536332983566177e-05,
+      "loss": 0.7227,
+      "step": 17123
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5945503765120337,
+      "learning_rate": 1.3535619236765665e-05,
+      "loss": 0.9498,
+      "step": 17124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0573033926996565,
+      "learning_rate": 1.3534905469379654e-05,
+      "loss": 0.8972,
+      "step": 17125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9784159284536245,
+      "learning_rate": 1.3534191681412309e-05,
+      "loss": 0.8913,
+      "step": 17126
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4089445073669693,
+      "learning_rate": 1.353347787286778e-05,
+      "loss": 0.8508,
+      "step": 17127
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6909759971359224,
+      "learning_rate": 1.3532764043750224e-05,
+      "loss": 0.9719,
+      "step": 17128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8949564159909122,
+      "learning_rate": 1.3532050194063795e-05,
+      "loss": 0.9022,
+      "step": 17129
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9105165073492614,
+      "learning_rate": 1.3531336323812656e-05,
+      "loss": 0.8989,
+      "step": 17130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9290094036328993,
+      "learning_rate": 1.3530622433000958e-05,
+      "loss": 0.8899,
+      "step": 17131
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5511090955300384,
+      "learning_rate": 1.3529908521632854e-05,
+      "loss": 1.0122,
+      "step": 17132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7676971681792975,
+      "learning_rate": 1.352919458971251e-05,
+      "loss": 0.9263,
+      "step": 17133
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7156101932955012,
+      "learning_rate": 1.3528480637244075e-05,
+      "loss": 0.8515,
+      "step": 17134
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.438347486593679,
+      "learning_rate": 1.3527766664231709e-05,
+      "loss": 0.8518,
+      "step": 17135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6680634958190623,
+      "learning_rate": 1.352705267067957e-05,
+      "loss": 0.9524,
+      "step": 17136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7487885822377139,
+      "learning_rate": 1.3526338656591813e-05,
+      "loss": 0.7349,
+      "step": 17137
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4778101104629209,
+      "learning_rate": 1.3525624621972595e-05,
+      "loss": 0.9854,
+      "step": 17138
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6222410371090439,
+      "learning_rate": 1.3524910566826075e-05,
+      "loss": 0.8484,
+      "step": 17139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2174348769208305,
+      "learning_rate": 1.3524196491156409e-05,
+      "loss": 0.9352,
+      "step": 17140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5213535332000285,
+      "learning_rate": 1.3523482394967756e-05,
+      "loss": 0.7598,
+      "step": 17141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4789450782473517,
+      "learning_rate": 1.3522768278264273e-05,
+      "loss": 0.9303,
+      "step": 17142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.65206387769962,
+      "learning_rate": 1.3522054141050118e-05,
+      "loss": 0.7871,
+      "step": 17143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5595381822423389,
+      "learning_rate": 1.352133998332945e-05,
+      "loss": 0.9359,
+      "step": 17144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0027192111326655,
+      "learning_rate": 1.3520625805106425e-05,
+      "loss": 0.887,
+      "step": 17145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.988845078738925,
+      "learning_rate": 1.3519911606385199e-05,
+      "loss": 0.883,
+      "step": 17146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.487201554112785,
+      "learning_rate": 1.3519197387169939e-05,
+      "loss": 0.8257,
+      "step": 17147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6918232741426067,
+      "learning_rate": 1.3518483147464792e-05,
+      "loss": 0.7795,
+      "step": 17148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5310821076614942,
+      "learning_rate": 1.3517768887273923e-05,
+      "loss": 0.9683,
+      "step": 17149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5851801145521764,
+      "learning_rate": 1.3517054606601492e-05,
+      "loss": 0.9035,
+      "step": 17150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6099863534480712,
+      "learning_rate": 1.3516340305451654e-05,
+      "loss": 0.8609,
+      "step": 17151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.576128157156495,
+      "learning_rate": 1.3515625983828572e-05,
+      "loss": 0.7376,
+      "step": 17152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6542835277280146,
+      "learning_rate": 1.35149116417364e-05,
+      "loss": 0.8292,
+      "step": 17153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5407887314505262,
+      "learning_rate": 1.35141972791793e-05,
+      "loss": 0.9706,
+      "step": 17154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7905738828159201,
+      "learning_rate": 1.351348289616143e-05,
+      "loss": 0.7443,
+      "step": 17155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4319453119321264,
+      "learning_rate": 1.3512768492686954e-05,
+      "loss": 1.0048,
+      "step": 17156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.506143793874407,
+      "learning_rate": 1.351205406876002e-05,
+      "loss": 0.8065,
+      "step": 17157
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8020018506318185,
+      "learning_rate": 1.3511339624384805e-05,
+      "loss": 0.9206,
+      "step": 17158
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5731505424488976,
+      "learning_rate": 1.3510625159565453e-05,
+      "loss": 0.9679,
+      "step": 17159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7570564902629018,
+      "learning_rate": 1.3509910674306133e-05,
+      "loss": 0.8562,
+      "step": 17160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.538655945991603,
+      "learning_rate": 1.3509196168610999e-05,
+      "loss": 0.8946,
+      "step": 17161
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4921579685124053,
+      "learning_rate": 1.3508481642484215e-05,
+      "loss": 1.0651,
+      "step": 17162
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7714343997908886,
+      "learning_rate": 1.350776709592994e-05,
+      "loss": 0.7386,
+      "step": 17163
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5229425865101087,
+      "learning_rate": 1.3507052528952333e-05,
+      "loss": 0.956,
+      "step": 17164
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2845052828377224,
+      "learning_rate": 1.3506337941555559e-05,
+      "loss": 0.8079,
+      "step": 17165
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.906047154128926,
+      "learning_rate": 1.3505623333743774e-05,
+      "loss": 1.0461,
+      "step": 17166
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5103710301733757,
+      "learning_rate": 1.350490870552114e-05,
+      "loss": 1.0337,
+      "step": 17167
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5303449039303783,
+      "learning_rate": 1.3504194056891817e-05,
+      "loss": 0.8936,
+      "step": 17168
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.300151243099542,
+      "learning_rate": 1.3503479387859967e-05,
+      "loss": 0.755,
+      "step": 17169
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4803772633026895,
+      "learning_rate": 1.3502764698429753e-05,
+      "loss": 0.908,
+      "step": 17170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.666880655901407,
+      "learning_rate": 1.3502049988605332e-05,
+      "loss": 0.7055,
+      "step": 17171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5131649978328394,
+      "learning_rate": 1.3501335258390868e-05,
+      "loss": 0.8874,
+      "step": 17172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.863703944093084,
+      "learning_rate": 1.3500620507790523e-05,
+      "loss": 0.852,
+      "step": 17173
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7543589263950552,
+      "learning_rate": 1.3499905736808455e-05,
+      "loss": 1.0198,
+      "step": 17174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1027074929706027,
+      "learning_rate": 1.3499190945448828e-05,
+      "loss": 0.8872,
+      "step": 17175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5548546953107931,
+      "learning_rate": 1.3498476133715803e-05,
+      "loss": 0.8267,
+      "step": 17176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5423010090082834,
+      "learning_rate": 1.3497761301613543e-05,
+      "loss": 0.9879,
+      "step": 17177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5810926248490111,
+      "learning_rate": 1.3497046449146209e-05,
+      "loss": 0.9032,
+      "step": 17178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5928394209235317,
+      "learning_rate": 1.3496331576317964e-05,
+      "loss": 0.8395,
+      "step": 17179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.770610615646385,
+      "learning_rate": 1.3495616683132969e-05,
+      "loss": 0.7366,
+      "step": 17180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7427954440403073,
+      "learning_rate": 1.349490176959539e-05,
+      "loss": 0.7489,
+      "step": 17181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5083248129991274,
+      "learning_rate": 1.3494186835709383e-05,
+      "loss": 0.886,
+      "step": 17182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8072400066657326,
+      "learning_rate": 1.3493471881479114e-05,
+      "loss": 0.8328,
+      "step": 17183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7121689940389533,
+      "learning_rate": 1.3492756906908748e-05,
+      "loss": 0.8209,
+      "step": 17184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5001688640936166,
+      "learning_rate": 1.3492041912002442e-05,
+      "loss": 0.8443,
+      "step": 17185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5751932170239682,
+      "learning_rate": 1.3491326896764366e-05,
+      "loss": 0.8051,
+      "step": 17186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6918487836451968,
+      "learning_rate": 1.3490611861198676e-05,
+      "loss": 0.9308,
+      "step": 17187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8090725262549225,
+      "learning_rate": 1.3489896805309541e-05,
+      "loss": 0.9313,
+      "step": 17188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7379190052264804,
+      "learning_rate": 1.3489181729101122e-05,
+      "loss": 0.9045,
+      "step": 17189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.299514607529934,
+      "learning_rate": 1.3488466632577581e-05,
+      "loss": 0.7852,
+      "step": 17190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.650750507404524,
+      "learning_rate": 1.3487751515743084e-05,
+      "loss": 0.9407,
+      "step": 17191
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5691133024581336,
+      "learning_rate": 1.3487036378601794e-05,
+      "loss": 0.8084,
+      "step": 17192
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2415059033726785,
+      "learning_rate": 1.3486321221157872e-05,
+      "loss": 0.7724,
+      "step": 17193
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5657113185608706,
+      "learning_rate": 1.3485606043415487e-05,
+      "loss": 0.8778,
+      "step": 17194
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2187819310435577,
+      "learning_rate": 1.3484890845378798e-05,
+      "loss": 0.788,
+      "step": 17195
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7499780884869347,
+      "learning_rate": 1.3484175627051974e-05,
+      "loss": 0.863,
+      "step": 17196
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9107896333176889,
+      "learning_rate": 1.3483460388439172e-05,
+      "loss": 0.7617,
+      "step": 17197
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8016972043007599,
+      "learning_rate": 1.3482745129544564e-05,
+      "loss": 0.7472,
+      "step": 17198
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8718934837518921,
+      "learning_rate": 1.3482029850372309e-05,
+      "loss": 0.9747,
+      "step": 17199
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.64241864359552,
+      "learning_rate": 1.3481314550926576e-05,
+      "loss": 0.8894,
+      "step": 17200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9092985626563708,
+      "learning_rate": 1.3480599231211524e-05,
+      "loss": 0.9005,
+      "step": 17201
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.697949059943226,
+      "learning_rate": 1.3479883891231326e-05,
+      "loss": 0.8372,
+      "step": 17202
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6667742990787406,
+      "learning_rate": 1.347916853099014e-05,
+      "loss": 0.8295,
+      "step": 17203
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6575657090821212,
+      "learning_rate": 1.3478453150492132e-05,
+      "loss": 0.8724,
+      "step": 17204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.487157041814631,
+      "learning_rate": 1.3477737749741473e-05,
+      "loss": 0.8109,
+      "step": 17205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8907578607658038,
+      "learning_rate": 1.3477022328742318e-05,
+      "loss": 0.8795,
+      "step": 17206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9145484154260246,
+      "learning_rate": 1.3476306887498842e-05,
+      "loss": 0.9308,
+      "step": 17207
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3711792319220009,
+      "learning_rate": 1.3475591426015204e-05,
+      "loss": 0.9252,
+      "step": 17208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5742155215288536,
+      "learning_rate": 1.3474875944295576e-05,
+      "loss": 0.7571,
+      "step": 17209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5370188691199942,
+      "learning_rate": 1.3474160442344118e-05,
+      "loss": 0.8048,
+      "step": 17210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7216917316310246,
+      "learning_rate": 1.3473444920164997e-05,
+      "loss": 0.6837,
+      "step": 17211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6415825298571562,
+      "learning_rate": 1.3472729377762381e-05,
+      "loss": 0.9286,
+      "step": 17212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5990576704384478,
+      "learning_rate": 1.3472013815140437e-05,
+      "loss": 0.8318,
+      "step": 17213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6309400643086347,
+      "learning_rate": 1.3471298232303326e-05,
+      "loss": 0.9254,
+      "step": 17214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7213834857121761,
+      "learning_rate": 1.347058262925522e-05,
+      "loss": 0.8299,
+      "step": 17215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5849728199517208,
+      "learning_rate": 1.346986700600028e-05,
+      "loss": 0.8785,
+      "step": 17216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4962372988382104,
+      "learning_rate": 1.346915136254268e-05,
+      "loss": 0.9819,
+      "step": 17217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.03924373142978,
+      "learning_rate": 1.3468435698886582e-05,
+      "loss": 1.0306,
+      "step": 17218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8222338166527788,
+      "learning_rate": 1.3467720015036152e-05,
+      "loss": 0.7095,
+      "step": 17219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8933115763617915,
+      "learning_rate": 1.3467004310995557e-05,
+      "loss": 0.8052,
+      "step": 17220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.06663618585636,
+      "learning_rate": 1.3466288586768968e-05,
+      "loss": 0.7989,
+      "step": 17221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6337195337522028,
+      "learning_rate": 1.3465572842360546e-05,
+      "loss": 0.9958,
+      "step": 17222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9373717738785579,
+      "learning_rate": 1.3464857077774465e-05,
+      "loss": 0.994,
+      "step": 17223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3544536318296103,
+      "learning_rate": 1.346414129301489e-05,
+      "loss": 0.7956,
+      "step": 17224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8378689901443437,
+      "learning_rate": 1.3463425488085982e-05,
+      "loss": 0.9387,
+      "step": 17225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5892150012765995,
+      "learning_rate": 1.346270966299192e-05,
+      "loss": 0.8399,
+      "step": 17226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.209639145495442,
+      "learning_rate": 1.3461993817736861e-05,
+      "loss": 0.9528,
+      "step": 17227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.190162469881313,
+      "learning_rate": 1.3461277952324982e-05,
+      "loss": 0.8953,
+      "step": 17228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3210555155529722,
+      "learning_rate": 1.3460562066760444e-05,
+      "loss": 0.8572,
+      "step": 17229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.867435767375178,
+      "learning_rate": 1.3459846161047421e-05,
+      "loss": 0.7747,
+      "step": 17230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7267857701259108,
+      "learning_rate": 1.3459130235190077e-05,
+      "loss": 1.083,
+      "step": 17231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7109332027956972,
+      "learning_rate": 1.3458414289192582e-05,
+      "loss": 0.8873,
+      "step": 17232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7517640866462596,
+      "learning_rate": 1.3457698323059102e-05,
+      "loss": 0.7311,
+      "step": 17233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7484876968930572,
+      "learning_rate": 1.3456982336793811e-05,
+      "loss": 0.9341,
+      "step": 17234
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9404537064333325,
+      "learning_rate": 1.3456266330400873e-05,
+      "loss": 0.7202,
+      "step": 17235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4933127304844878,
+      "learning_rate": 1.3455550303884458e-05,
+      "loss": 0.974,
+      "step": 17236
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5661562998672096,
+      "learning_rate": 1.3454834257248734e-05,
+      "loss": 0.7667,
+      "step": 17237
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0084912878316246,
+      "learning_rate": 1.3454118190497874e-05,
+      "loss": 0.9588,
+      "step": 17238
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5139376024836613,
+      "learning_rate": 1.3453402103636042e-05,
+      "loss": 0.9464,
+      "step": 17239
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5029611617148808,
+      "learning_rate": 1.3452685996667409e-05,
+      "loss": 0.8596,
+      "step": 17240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8121567843524173,
+      "learning_rate": 1.3451969869596146e-05,
+      "loss": 0.8399,
+      "step": 17241
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5968410089639444,
+      "learning_rate": 1.3451253722426423e-05,
+      "loss": 0.9086,
+      "step": 17242
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8934295489618556,
+      "learning_rate": 1.3450537555162406e-05,
+      "loss": 0.8876,
+      "step": 17243
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6261134336867689,
+      "learning_rate": 1.3449821367808269e-05,
+      "loss": 0.8884,
+      "step": 17244
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.806642087026286,
+      "learning_rate": 1.3449105160368179e-05,
+      "loss": 0.7328,
+      "step": 17245
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.495574888919458,
+      "learning_rate": 1.3448388932846306e-05,
+      "loss": 0.9037,
+      "step": 17246
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9639531762692846,
+      "learning_rate": 1.3447672685246822e-05,
+      "loss": 0.8822,
+      "step": 17247
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6628046873844375,
+      "learning_rate": 1.3446956417573894e-05,
+      "loss": 0.9856,
+      "step": 17248
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.644154641260656,
+      "learning_rate": 1.3446240129831699e-05,
+      "loss": 0.8298,
+      "step": 17249
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4567609076087882,
+      "learning_rate": 1.3445523822024398e-05,
+      "loss": 0.8675,
+      "step": 17250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.459946787846933,
+      "learning_rate": 1.344480749415617e-05,
+      "loss": 0.7694,
+      "step": 17251
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6432072678100724,
+      "learning_rate": 1.3444091146231178e-05,
+      "loss": 0.8531,
+      "step": 17252
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7983921025565497,
+      "learning_rate": 1.3443374778253597e-05,
+      "loss": 0.8916,
+      "step": 17253
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4660120227309779,
+      "learning_rate": 1.3442658390227604e-05,
+      "loss": 0.95,
+      "step": 17254
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4816326971611042,
+      "learning_rate": 1.3441941982157358e-05,
+      "loss": 0.8883,
+      "step": 17255
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7860462264587098,
+      "learning_rate": 1.3441225554047039e-05,
+      "loss": 0.8325,
+      "step": 17256
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.588655361698687,
+      "learning_rate": 1.3440509105900813e-05,
+      "loss": 0.8327,
+      "step": 17257
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6478187484309514,
+      "learning_rate": 1.3439792637722859e-05,
+      "loss": 0.9264,
+      "step": 17258
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6052282938896798,
+      "learning_rate": 1.3439076149517337e-05,
+      "loss": 0.8522,
+      "step": 17259
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5138283381812863,
+      "learning_rate": 1.343835964128843e-05,
+      "loss": 0.8884,
+      "step": 17260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.882638752002436,
+      "learning_rate": 1.3437643113040302e-05,
+      "loss": 0.7821,
+      "step": 17261
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.355809121702026,
+      "learning_rate": 1.3436926564777127e-05,
+      "loss": 0.9086,
+      "step": 17262
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8642567043965574,
+      "learning_rate": 1.3436209996503078e-05,
+      "loss": 0.8488,
+      "step": 17263
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.431114757881154,
+      "learning_rate": 1.3435493408222326e-05,
+      "loss": 0.91,
+      "step": 17264
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7013069802941512,
+      "learning_rate": 1.3434776799939045e-05,
+      "loss": 0.8861,
+      "step": 17265
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9951032280900383,
+      "learning_rate": 1.3434060171657403e-05,
+      "loss": 0.9503,
+      "step": 17266
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.51416640071264,
+      "learning_rate": 1.3433343523381578e-05,
+      "loss": 0.8516,
+      "step": 17267
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.345690700959556,
+      "learning_rate": 1.343262685511574e-05,
+      "loss": 0.925,
+      "step": 17268
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8985735544695186,
+      "learning_rate": 1.3431910166864062e-05,
+      "loss": 0.8267,
+      "step": 17269
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4609915589384632,
+      "learning_rate": 1.3431193458630716e-05,
+      "loss": 0.797,
+      "step": 17270
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6410071647571984,
+      "learning_rate": 1.3430476730419874e-05,
+      "loss": 0.9303,
+      "step": 17271
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6173426712506114,
+      "learning_rate": 1.3429759982235712e-05,
+      "loss": 0.9701,
+      "step": 17272
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9850223446439161,
+      "learning_rate": 1.3429043214082403e-05,
+      "loss": 0.8,
+      "step": 17273
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4676482523940861,
+      "learning_rate": 1.3428326425964115e-05,
+      "loss": 0.699,
+      "step": 17274
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4898929054350032,
+      "learning_rate": 1.3427609617885029e-05,
+      "loss": 0.8467,
+      "step": 17275
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7509337775438534,
+      "learning_rate": 1.3426892789849312e-05,
+      "loss": 0.8263,
+      "step": 17276
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5522933864746402,
+      "learning_rate": 1.3426175941861144e-05,
+      "loss": 0.8566,
+      "step": 17277
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9574617452625123,
+      "learning_rate": 1.3425459073924689e-05,
+      "loss": 0.8001,
+      "step": 17278
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3894330129670596,
+      "learning_rate": 1.3424742186044131e-05,
+      "loss": 0.8458,
+      "step": 17279
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7605385730208105,
+      "learning_rate": 1.3424025278223637e-05,
+      "loss": 0.8157,
+      "step": 17280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4633979954427527,
+      "learning_rate": 1.3423308350467388e-05,
+      "loss": 0.9153,
+      "step": 17281
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.616634674758321,
+      "learning_rate": 1.342259140277955e-05,
+      "loss": 0.8427,
+      "step": 17282
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5665781710736673,
+      "learning_rate": 1.3421874435164303e-05,
+      "loss": 0.9486,
+      "step": 17283
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.934815857447527,
+      "learning_rate": 1.3421157447625822e-05,
+      "loss": 0.8747,
+      "step": 17284
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.536876813160923,
+      "learning_rate": 1.3420440440168275e-05,
+      "loss": 0.7618,
+      "step": 17285
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6240667738174643,
+      "learning_rate": 1.3419723412795843e-05,
+      "loss": 0.7793,
+      "step": 17286
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6054774194431707,
+      "learning_rate": 1.3419006365512697e-05,
+      "loss": 0.9266,
+      "step": 17287
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9292570607585768,
+      "learning_rate": 1.3418289298323015e-05,
+      "loss": 0.8825,
+      "step": 17288
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.506077735298858,
+      "learning_rate": 1.3417572211230969e-05,
+      "loss": 0.8196,
+      "step": 17289
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.477199805850461,
+      "learning_rate": 1.3416855104240739e-05,
+      "loss": 0.9428,
+      "step": 17290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4536034120928467,
+      "learning_rate": 1.3416137977356491e-05,
+      "loss": 1.0443,
+      "step": 17291
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.822289646791211,
+      "learning_rate": 1.341542083058241e-05,
+      "loss": 0.819,
+      "step": 17292
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7604751999853682,
+      "learning_rate": 1.3414703663922668e-05,
+      "loss": 0.7442,
+      "step": 17293
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8368515223370708,
+      "learning_rate": 1.3413986477381439e-05,
+      "loss": 0.7584,
+      "step": 17294
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.778232531643465,
+      "learning_rate": 1.3413269270962902e-05,
+      "loss": 0.7341,
+      "step": 17295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4705520142578632,
+      "learning_rate": 1.341255204467123e-05,
+      "loss": 0.7754,
+      "step": 17296
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6618321792559347,
+      "learning_rate": 1.3411834798510596e-05,
+      "loss": 0.8568,
+      "step": 17297
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9804618552618611,
+      "learning_rate": 1.3411117532485185e-05,
+      "loss": 0.8652,
+      "step": 17298
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8647617845481947,
+      "learning_rate": 1.3410400246599163e-05,
+      "loss": 1.0067,
+      "step": 17299
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.49160573503705,
+      "learning_rate": 1.3409682940856714e-05,
+      "loss": 0.9881,
+      "step": 17300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.720562453404529,
+      "learning_rate": 1.3408965615262008e-05,
+      "loss": 0.9439,
+      "step": 17301
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5124229400038043,
+      "learning_rate": 1.3408248269819228e-05,
+      "loss": 0.7954,
+      "step": 17302
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0254389308272525,
+      "learning_rate": 1.3407530904532546e-05,
+      "loss": 0.7603,
+      "step": 17303
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8094871901124248,
+      "learning_rate": 1.3406813519406142e-05,
+      "loss": 0.9788,
+      "step": 17304
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4856344257940837,
+      "learning_rate": 1.3406096114444193e-05,
+      "loss": 1.0417,
+      "step": 17305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8104130953833257,
+      "learning_rate": 1.340537868965087e-05,
+      "loss": 1.0049,
+      "step": 17306
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7447067664120686,
+      "learning_rate": 1.3404661245030357e-05,
+      "loss": 0.9377,
+      "step": 17307
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9803727518359524,
+      "learning_rate": 1.3403943780586826e-05,
+      "loss": 0.8765,
+      "step": 17308
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.77276952109171,
+      "learning_rate": 1.3403226296324458e-05,
+      "loss": 0.751,
+      "step": 17309
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3957352473541365,
+      "learning_rate": 1.3402508792247427e-05,
+      "loss": 0.9712,
+      "step": 17310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7869073717702929,
+      "learning_rate": 1.3401791268359916e-05,
+      "loss": 0.7073,
+      "step": 17311
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3053922700395155,
+      "learning_rate": 1.3401073724666098e-05,
+      "loss": 0.895,
+      "step": 17312
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7863834694530497,
+      "learning_rate": 1.3400356161170152e-05,
+      "loss": 0.7608,
+      "step": 17313
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7442366626337039,
+      "learning_rate": 1.3399638577876255e-05,
+      "loss": 0.7159,
+      "step": 17314
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.011848601746936,
+      "learning_rate": 1.3398920974788585e-05,
+      "loss": 0.94,
+      "step": 17315
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7405746597513068,
+      "learning_rate": 1.3398203351911323e-05,
+      "loss": 0.8503,
+      "step": 17316
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5233628277061437,
+      "learning_rate": 1.3397485709248643e-05,
+      "loss": 0.9315,
+      "step": 17317
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.649716125596699,
+      "learning_rate": 1.3396768046804728e-05,
+      "loss": 0.7481,
+      "step": 17318
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3662736007331393,
+      "learning_rate": 1.3396050364583752e-05,
+      "loss": 0.893,
+      "step": 17319
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.533606711195375,
+      "learning_rate": 1.3395332662589898e-05,
+      "loss": 0.9803,
+      "step": 17320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.001762209037638,
+      "learning_rate": 1.339461494082734e-05,
+      "loss": 0.8513,
+      "step": 17321
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8402664448086319,
+      "learning_rate": 1.3393897199300262e-05,
+      "loss": 0.9048,
+      "step": 17322
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8895153251358334,
+      "learning_rate": 1.3393179438012836e-05,
+      "loss": 0.8366,
+      "step": 17323
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5458284221449627,
+      "learning_rate": 1.3392461656969246e-05,
+      "loss": 0.8396,
+      "step": 17324
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6584881069652766,
+      "learning_rate": 1.3391743856173673e-05,
+      "loss": 0.8799,
+      "step": 17325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0983565322357522,
+      "learning_rate": 1.3391026035630294e-05,
+      "loss": 0.9464,
+      "step": 17326
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7952474272063792,
+      "learning_rate": 1.3390308195343282e-05,
+      "loss": 0.8765,
+      "step": 17327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6056109477137233,
+      "learning_rate": 1.338959033531683e-05,
+      "loss": 0.8009,
+      "step": 17328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8895434349788074,
+      "learning_rate": 1.3388872455555102e-05,
+      "loss": 0.9486,
+      "step": 17329
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4533973809772582,
+      "learning_rate": 1.3388154556062292e-05,
+      "loss": 0.8027,
+      "step": 17330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2019898855104487,
+      "learning_rate": 1.338743663684257e-05,
+      "loss": 1.0973,
+      "step": 17331
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4467168567260364,
+      "learning_rate": 1.3386718697900123e-05,
+      "loss": 0.9851,
+      "step": 17332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6731002683144451,
+      "learning_rate": 1.3386000739239125e-05,
+      "loss": 0.8691,
+      "step": 17333
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8069212802814008,
+      "learning_rate": 1.3385282760863758e-05,
+      "loss": 0.716,
+      "step": 17334
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.545233605116997,
+      "learning_rate": 1.3384564762778203e-05,
+      "loss": 0.8339,
+      "step": 17335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8021926808047721,
+      "learning_rate": 1.3383846744986642e-05,
+      "loss": 0.754,
+      "step": 17336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5446502648021754,
+      "learning_rate": 1.3383128707493253e-05,
+      "loss": 0.9152,
+      "step": 17337
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8488233083868364,
+      "learning_rate": 1.3382410650302218e-05,
+      "loss": 0.9405,
+      "step": 17338
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5318215862135347,
+      "learning_rate": 1.3381692573417718e-05,
+      "loss": 0.8915,
+      "step": 17339
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7662451607568013,
+      "learning_rate": 1.3380974476843932e-05,
+      "loss": 0.874,
+      "step": 17340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9370476816230735,
+      "learning_rate": 1.338025636058504e-05,
+      "loss": 0.9527,
+      "step": 17341
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7839299332294287,
+      "learning_rate": 1.3379538224645228e-05,
+      "loss": 1.0709,
+      "step": 17342
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7283543817852421,
+      "learning_rate": 1.3378820069028675e-05,
+      "loss": 0.7336,
+      "step": 17343
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3917561518400974,
+      "learning_rate": 1.3378101893739561e-05,
+      "loss": 0.8031,
+      "step": 17344
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.738599292050644,
+      "learning_rate": 1.3377383698782066e-05,
+      "loss": 0.8993,
+      "step": 17345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6890918758909703,
+      "learning_rate": 1.3376665484160378e-05,
+      "loss": 0.7284,
+      "step": 17346
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.729904256636404,
+      "learning_rate": 1.3375947249878675e-05,
+      "loss": 0.8815,
+      "step": 17347
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.758558724197751,
+      "learning_rate": 1.3375228995941135e-05,
+      "loss": 0.7436,
+      "step": 17348
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6865873771620108,
+      "learning_rate": 1.3374510722351945e-05,
+      "loss": 1.0447,
+      "step": 17349
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0787338896888565,
+      "learning_rate": 1.3373792429115282e-05,
+      "loss": 0.8658,
+      "step": 17350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.654271884983362,
+      "learning_rate": 1.3373074116235334e-05,
+      "loss": 0.7789,
+      "step": 17351
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5803325603900589,
+      "learning_rate": 1.3372355783716282e-05,
+      "loss": 0.7255,
+      "step": 17352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0737237861842024,
+      "learning_rate": 1.3371637431562303e-05,
+      "loss": 0.9417,
+      "step": 17353
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6326674931777874,
+      "learning_rate": 1.337091905977759e-05,
+      "loss": 0.8823,
+      "step": 17354
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5561515695440105,
+      "learning_rate": 1.337020066836631e-05,
+      "loss": 0.8857,
+      "step": 17355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.762444562112901,
+      "learning_rate": 1.3369482257332664e-05,
+      "loss": 0.7621,
+      "step": 17356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.027487394596067,
+      "learning_rate": 1.336876382668082e-05,
+      "loss": 0.931,
+      "step": 17357
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.611775261926004,
+      "learning_rate": 1.3368045376414968e-05,
+      "loss": 0.9795,
+      "step": 17358
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6041747671367639,
+      "learning_rate": 1.3367326906539286e-05,
+      "loss": 0.896,
+      "step": 17359
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.986594222193457,
+      "learning_rate": 1.3366608417057966e-05,
+      "loss": 0.8574,
+      "step": 17360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4904561960422165,
+      "learning_rate": 1.3365889907975184e-05,
+      "loss": 0.8678,
+      "step": 17361
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7595318034991456,
+      "learning_rate": 1.3365171379295126e-05,
+      "loss": 0.8273,
+      "step": 17362
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5341307745644879,
+      "learning_rate": 1.3364452831021972e-05,
+      "loss": 0.9003,
+      "step": 17363
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.707688264872562,
+      "learning_rate": 1.3363734263159911e-05,
+      "loss": 0.8793,
+      "step": 17364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7642424783575552,
+      "learning_rate": 1.3363015675713124e-05,
+      "loss": 0.9487,
+      "step": 17365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7320004561331908,
+      "learning_rate": 1.3362297068685794e-05,
+      "loss": 0.9621,
+      "step": 17366
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4579711902051407,
+      "learning_rate": 1.3361578442082107e-05,
+      "loss": 0.9346,
+      "step": 17367
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.597373041591722,
+      "learning_rate": 1.3360859795906244e-05,
+      "loss": 0.7793,
+      "step": 17368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7968145619583376,
+      "learning_rate": 1.3360141130162394e-05,
+      "loss": 0.7791,
+      "step": 17369
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.649570285474899,
+      "learning_rate": 1.3359422444854738e-05,
+      "loss": 0.7838,
+      "step": 17370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.52605336704452,
+      "learning_rate": 1.3358703739987459e-05,
+      "loss": 0.9817,
+      "step": 17371
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5324124217402733,
+      "learning_rate": 1.3357985015564744e-05,
+      "loss": 1.0174,
+      "step": 17372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4979078646126234,
+      "learning_rate": 1.335726627159078e-05,
+      "loss": 0.9033,
+      "step": 17373
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.944881394581637,
+      "learning_rate": 1.3356547508069745e-05,
+      "loss": 1.0472,
+      "step": 17374
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.910108742050126,
+      "learning_rate": 1.335582872500583e-05,
+      "loss": 0.7797,
+      "step": 17375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5922130863340913,
+      "learning_rate": 1.3355109922403218e-05,
+      "loss": 0.8274,
+      "step": 17376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5028132292997731,
+      "learning_rate": 1.3354391100266096e-05,
+      "loss": 0.8019,
+      "step": 17377
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.613630627386991,
+      "learning_rate": 1.335367225859864e-05,
+      "loss": 0.8744,
+      "step": 17378
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5715037848230493,
+      "learning_rate": 1.3352953397405049e-05,
+      "loss": 0.8668,
+      "step": 17379
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.899390215619828,
+      "learning_rate": 1.3352234516689497e-05,
+      "loss": 0.9544,
+      "step": 17380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5890120686306795,
+      "learning_rate": 1.3351515616456178e-05,
+      "loss": 0.8789,
+      "step": 17381
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8837115606229686,
+      "learning_rate": 1.3350796696709272e-05,
+      "loss": 0.9368,
+      "step": 17382
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6268173329757945,
+      "learning_rate": 1.3350077757452969e-05,
+      "loss": 0.9137,
+      "step": 17383
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8464416623818136,
+      "learning_rate": 1.334935879869145e-05,
+      "loss": 0.8582,
+      "step": 17384
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.982636761793019,
+      "learning_rate": 1.3348639820428906e-05,
+      "loss": 0.8317,
+      "step": 17385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5344960617916785,
+      "learning_rate": 1.3347920822669518e-05,
+      "loss": 0.8249,
+      "step": 17386
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.702498419178236,
+      "learning_rate": 1.3347201805417476e-05,
+      "loss": 0.9291,
+      "step": 17387
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5224592060204056,
+      "learning_rate": 1.3346482768676968e-05,
+      "loss": 0.9726,
+      "step": 17388
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5416355215260893,
+      "learning_rate": 1.3345763712452174e-05,
+      "loss": 0.8516,
+      "step": 17389
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5181352574282938,
+      "learning_rate": 1.3345044636747288e-05,
+      "loss": 0.9154,
+      "step": 17390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5753444087907496,
+      "learning_rate": 1.3344325541566492e-05,
+      "loss": 0.9098,
+      "step": 17391
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6515471946080762,
+      "learning_rate": 1.334360642691397e-05,
+      "loss": 0.9952,
+      "step": 17392
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6333542063803674,
+      "learning_rate": 1.3342887292793918e-05,
+      "loss": 0.8919,
+      "step": 17393
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1218740317732387,
+      "learning_rate": 1.3342168139210517e-05,
+      "loss": 0.8776,
+      "step": 17394
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1933087315916304,
+      "learning_rate": 1.3341448966167954e-05,
+      "loss": 0.8084,
+      "step": 17395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5152126622499007,
+      "learning_rate": 1.3340729773670416e-05,
+      "loss": 0.8256,
+      "step": 17396
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7773894906701502,
+      "learning_rate": 1.3340010561722093e-05,
+      "loss": 0.7701,
+      "step": 17397
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8716679721458362,
+      "learning_rate": 1.3339291330327172e-05,
+      "loss": 0.87,
+      "step": 17398
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5342046043914919,
+      "learning_rate": 1.3338572079489835e-05,
+      "loss": 0.9235,
+      "step": 17399
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5145971153980864,
+      "learning_rate": 1.3337852809214278e-05,
+      "loss": 0.8258,
+      "step": 17400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7690755735228179,
+      "learning_rate": 1.3337133519504685e-05,
+      "loss": 0.9057,
+      "step": 17401
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6691683477514205,
+      "learning_rate": 1.3336414210365245e-05,
+      "loss": 0.7388,
+      "step": 17402
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5506813021387127,
+      "learning_rate": 1.3335694881800145e-05,
+      "loss": 0.9126,
+      "step": 17403
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8210405879513145,
+      "learning_rate": 1.3334975533813572e-05,
+      "loss": 0.9241,
+      "step": 17404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5009383756024628,
+      "learning_rate": 1.3334256166409717e-05,
+      "loss": 0.6694,
+      "step": 17405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5465401036115762,
+      "learning_rate": 1.3333536779592766e-05,
+      "loss": 0.8381,
+      "step": 17406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8182891247514443,
+      "learning_rate": 1.333281737336691e-05,
+      "loss": 0.888,
+      "step": 17407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.826483835778875,
+      "learning_rate": 1.3332097947736333e-05,
+      "loss": 0.6919,
+      "step": 17408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6352264405941415,
+      "learning_rate": 1.333137850270523e-05,
+      "loss": 0.8975,
+      "step": 17409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4641658131128734,
+      "learning_rate": 1.3330659038277785e-05,
+      "loss": 0.8929,
+      "step": 17410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7888904177487686,
+      "learning_rate": 1.3329939554458188e-05,
+      "loss": 0.6752,
+      "step": 17411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4740872036283053,
+      "learning_rate": 1.332922005125063e-05,
+      "loss": 0.8023,
+      "step": 17412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.790234208460739,
+      "learning_rate": 1.3328500528659301e-05,
+      "loss": 0.7858,
+      "step": 17413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6374733068660168,
+      "learning_rate": 1.3327780986688384e-05,
+      "loss": 0.8056,
+      "step": 17414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5257135096364367,
+      "learning_rate": 1.3327061425342073e-05,
+      "loss": 0.8904,
+      "step": 17415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7814093533403912,
+      "learning_rate": 1.332634184462456e-05,
+      "loss": 0.8729,
+      "step": 17416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.591492083943541,
+      "learning_rate": 1.3325622244540028e-05,
+      "loss": 0.8821,
+      "step": 17417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6841779047955352,
+      "learning_rate": 1.3324902625092672e-05,
+      "loss": 0.845,
+      "step": 17418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.251627608042936,
+      "learning_rate": 1.3324182986286683e-05,
+      "loss": 0.8363,
+      "step": 17419
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5316564538383908,
+      "learning_rate": 1.3323463328126245e-05,
+      "loss": 0.846,
+      "step": 17420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4267509549303812,
+      "learning_rate": 1.3322743650615552e-05,
+      "loss": 0.7506,
+      "step": 17421
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7444762625032852,
+      "learning_rate": 1.3322023953758794e-05,
+      "loss": 0.9905,
+      "step": 17422
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.110565120571417,
+      "learning_rate": 1.332130423756016e-05,
+      "loss": 0.9606,
+      "step": 17423
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7177634021697403,
+      "learning_rate": 1.3320584502023841e-05,
+      "loss": 0.9977,
+      "step": 17424
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.851844297238669,
+      "learning_rate": 1.3319864747154028e-05,
+      "loss": 0.86,
+      "step": 17425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7706958757796387,
+      "learning_rate": 1.3319144972954915e-05,
+      "loss": 0.8572,
+      "step": 17426
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7176007317129607,
+      "learning_rate": 1.3318425179430683e-05,
+      "loss": 0.8533,
+      "step": 17427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6143243273891161,
+      "learning_rate": 1.3317705366585534e-05,
+      "loss": 0.8282,
+      "step": 17428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7256310175872285,
+      "learning_rate": 1.331698553442365e-05,
+      "loss": 0.7375,
+      "step": 17429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0460095052939353,
+      "learning_rate": 1.331626568294923e-05,
+      "loss": 0.7024,
+      "step": 17430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6159115720974038,
+      "learning_rate": 1.3315545812166457e-05,
+      "loss": 0.8573,
+      "step": 17431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6311379315066108,
+      "learning_rate": 1.3314825922079531e-05,
+      "loss": 0.9417,
+      "step": 17432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6894871279879213,
+      "learning_rate": 1.3314106012692633e-05,
+      "loss": 0.915,
+      "step": 17433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5589998935311746,
+      "learning_rate": 1.331338608400997e-05,
+      "loss": 0.8177,
+      "step": 17434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4085357808346994,
+      "learning_rate": 1.3312666136035716e-05,
+      "loss": 0.8743,
+      "step": 17435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5935154231932769,
+      "learning_rate": 1.3311946168774072e-05,
+      "loss": 0.8814,
+      "step": 17436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5747393442505457,
+      "learning_rate": 1.3311226182229233e-05,
+      "loss": 0.9719,
+      "step": 17437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4612445513806973,
+      "learning_rate": 1.3310506176405383e-05,
+      "loss": 1.0076,
+      "step": 17438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6497510211325597,
+      "learning_rate": 1.3309786151306721e-05,
+      "loss": 0.9926,
+      "step": 17439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9799882272016824,
+      "learning_rate": 1.3309066106937432e-05,
+      "loss": 0.9238,
+      "step": 17440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.660939592070548,
+      "learning_rate": 1.3308346043301717e-05,
+      "loss": 0.9249,
+      "step": 17441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4232100150179536,
+      "learning_rate": 1.3307625960403763e-05,
+      "loss": 0.7619,
+      "step": 17442
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8898302821519313,
+      "learning_rate": 1.330690585824776e-05,
+      "loss": 0.854,
+      "step": 17443
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6856191400249945,
+      "learning_rate": 1.3306185736837908e-05,
+      "loss": 1.0827,
+      "step": 17444
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7950699164228139,
+      "learning_rate": 1.3305465596178395e-05,
+      "loss": 0.8529,
+      "step": 17445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3829851680428933,
+      "learning_rate": 1.3304745436273415e-05,
+      "loss": 0.8426,
+      "step": 17446
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6559408352218106,
+      "learning_rate": 1.330402525712716e-05,
+      "loss": 0.8349,
+      "step": 17447
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8115535351214276,
+      "learning_rate": 1.3303305058743823e-05,
+      "loss": 0.9146,
+      "step": 17448
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6934807402821659,
+      "learning_rate": 1.33025848411276e-05,
+      "loss": 0.8935,
+      "step": 17449
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8774858934199803,
+      "learning_rate": 1.330186460428268e-05,
+      "loss": 0.8869,
+      "step": 17450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8033609355112367,
+      "learning_rate": 1.3301144348213262e-05,
+      "loss": 0.7073,
+      "step": 17451
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5610571456226396,
+      "learning_rate": 1.3300424072923534e-05,
+      "loss": 0.8875,
+      "step": 17452
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0317804249065685,
+      "learning_rate": 1.3299703778417694e-05,
+      "loss": 0.8285,
+      "step": 17453
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7740453797377442,
+      "learning_rate": 1.3298983464699933e-05,
+      "loss": 0.8954,
+      "step": 17454
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6233236761137981,
+      "learning_rate": 1.3298263131774445e-05,
+      "loss": 0.8424,
+      "step": 17455
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4855966580136317,
+      "learning_rate": 1.3297542779645428e-05,
+      "loss": 0.8608,
+      "step": 17456
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7252732378405284,
+      "learning_rate": 1.3296822408317067e-05,
+      "loss": 0.8765,
+      "step": 17457
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7959837299157504,
+      "learning_rate": 1.3296102017793569e-05,
+      "loss": 0.8982,
+      "step": 17458
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.639273493662068,
+      "learning_rate": 1.3295381608079116e-05,
+      "loss": 0.8898,
+      "step": 17459
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0791519241753336,
+      "learning_rate": 1.3294661179177912e-05,
+      "loss": 0.8757,
+      "step": 17460
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8009845782538267,
+      "learning_rate": 1.3293940731094143e-05,
+      "loss": 0.7569,
+      "step": 17461
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9870383480742897,
+      "learning_rate": 1.3293220263832015e-05,
+      "loss": 0.9675,
+      "step": 17462
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6123671575352805,
+      "learning_rate": 1.3292499777395711e-05,
+      "loss": 0.8286,
+      "step": 17463
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7564592001185513,
+      "learning_rate": 1.3291779271789433e-05,
+      "loss": 0.7352,
+      "step": 17464
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8263421587740487,
+      "learning_rate": 1.3291058747017373e-05,
+      "loss": 0.8267,
+      "step": 17465
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4876369867789485,
+      "learning_rate": 1.3290338203083728e-05,
+      "loss": 0.8246,
+      "step": 17466
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5916638113204362,
+      "learning_rate": 1.328961763999269e-05,
+      "loss": 0.8911,
+      "step": 17467
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7743012497099093,
+      "learning_rate": 1.3288897057748458e-05,
+      "loss": 0.8803,
+      "step": 17468
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5438728546433338,
+      "learning_rate": 1.3288176456355226e-05,
+      "loss": 0.9985,
+      "step": 17469
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6871799897103859,
+      "learning_rate": 1.3287455835817189e-05,
+      "loss": 0.8819,
+      "step": 17470
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4489926350331215,
+      "learning_rate": 1.3286735196138544e-05,
+      "loss": 0.8001,
+      "step": 17471
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6009538698859782,
+      "learning_rate": 1.3286014537323486e-05,
+      "loss": 0.8018,
+      "step": 17472
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5312963249567055,
+      "learning_rate": 1.3285293859376212e-05,
+      "loss": 0.8647,
+      "step": 17473
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5270392493265512,
+      "learning_rate": 1.3284573162300916e-05,
+      "loss": 0.9542,
+      "step": 17474
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7351097936803244,
+      "learning_rate": 1.3283852446101794e-05,
+      "loss": 0.8809,
+      "step": 17475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4608138208044918,
+      "learning_rate": 1.3283131710783046e-05,
+      "loss": 0.954,
+      "step": 17476
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7999084468490956,
+      "learning_rate": 1.3282410956348865e-05,
+      "loss": 0.8676,
+      "step": 17477
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6102343981334615,
+      "learning_rate": 1.3281690182803448e-05,
+      "loss": 0.8405,
+      "step": 17478
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4787676222776556,
+      "learning_rate": 1.3280969390150993e-05,
+      "loss": 0.995,
+      "step": 17479
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7300735824943165,
+      "learning_rate": 1.328024857839569e-05,
+      "loss": 0.7978,
+      "step": 17480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5934144467702678,
+      "learning_rate": 1.327952774754175e-05,
+      "loss": 0.8365,
+      "step": 17481
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6639444971234278,
+      "learning_rate": 1.3278806897593352e-05,
+      "loss": 0.8567,
+      "step": 17482
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6997696609822448,
+      "learning_rate": 1.3278086028554711e-05,
+      "loss": 0.9606,
+      "step": 17483
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6195357290212102,
+      "learning_rate": 1.327736514043001e-05,
+      "loss": 1.0061,
+      "step": 17484
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.62392814037138,
+      "learning_rate": 1.327664423322345e-05,
+      "loss": 0.8782,
+      "step": 17485
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7293045971845624,
+      "learning_rate": 1.3275923306939235e-05,
+      "loss": 0.9538,
+      "step": 17486
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7040422470240952,
+      "learning_rate": 1.3275202361581552e-05,
+      "loss": 0.7325,
+      "step": 17487
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.671995287838064,
+      "learning_rate": 1.327448139715461e-05,
+      "loss": 0.8552,
+      "step": 17488
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5770511104644045,
+      "learning_rate": 1.3273760413662596e-05,
+      "loss": 0.8325,
+      "step": 17489
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8392653950180171,
+      "learning_rate": 1.3273039411109714e-05,
+      "loss": 0.808,
+      "step": 17490
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7065007165322426,
+      "learning_rate": 1.3272318389500158e-05,
+      "loss": 0.8649,
+      "step": 17491
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8806856917653247,
+      "learning_rate": 1.3271597348838133e-05,
+      "loss": 0.9617,
+      "step": 17492
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6004542493302076,
+      "learning_rate": 1.327087628912783e-05,
+      "loss": 0.9855,
+      "step": 17493
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4622825477567125,
+      "learning_rate": 1.327015521037345e-05,
+      "loss": 0.8522,
+      "step": 17494
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.671079490698955,
+      "learning_rate": 1.326943411257919e-05,
+      "loss": 0.8286,
+      "step": 17495
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5191647549909033,
+      "learning_rate": 1.3268712995749251e-05,
+      "loss": 0.9095,
+      "step": 17496
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5503592907806536,
+      "learning_rate": 1.3267991859887827e-05,
+      "loss": 0.8301,
+      "step": 17497
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6253379398885446,
+      "learning_rate": 1.3267270704999122e-05,
+      "loss": 0.9095,
+      "step": 17498
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.526251136658383,
+      "learning_rate": 1.3266549531087333e-05,
+      "loss": 0.8683,
+      "step": 17499
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8001414261871371,
+      "learning_rate": 1.3265828338156658e-05,
+      "loss": 0.8983,
+      "step": 17500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.633157789132308,
+      "learning_rate": 1.3265107126211296e-05,
+      "loss": 0.9254,
+      "step": 17501
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8646807980985853,
+      "learning_rate": 1.326438589525545e-05,
+      "loss": 0.855,
+      "step": 17502
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3613697921861623,
+      "learning_rate": 1.326366464529331e-05,
+      "loss": 0.7683,
+      "step": 17503
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7817322967308865,
+      "learning_rate": 1.3262943376329085e-05,
+      "loss": 0.6916,
+      "step": 17504
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.747813796985393,
+      "learning_rate": 1.326222208836697e-05,
+      "loss": 0.8228,
+      "step": 17505
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7783999276138978,
+      "learning_rate": 1.3261500781411165e-05,
+      "loss": 0.8464,
+      "step": 17506
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8272770634754723,
+      "learning_rate": 1.3260779455465871e-05,
+      "loss": 0.8281,
+      "step": 17507
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3854730005983276,
+      "learning_rate": 1.3260058110535283e-05,
+      "loss": 0.8228,
+      "step": 17508
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8304489641118084,
+      "learning_rate": 1.325933674662361e-05,
+      "loss": 0.7478,
+      "step": 17509
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7037228477715065,
+      "learning_rate": 1.3258615363735041e-05,
+      "loss": 0.819,
+      "step": 17510
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6993755761828666,
+      "learning_rate": 1.3257893961873787e-05,
+      "loss": 0.8666,
+      "step": 17511
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6529396901081008,
+      "learning_rate": 1.3257172541044039e-05,
+      "loss": 0.7738,
+      "step": 17512
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7655095427368894,
+      "learning_rate": 1.3256451101250002e-05,
+      "loss": 0.7484,
+      "step": 17513
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.618520986270851,
+      "learning_rate": 1.3255729642495878e-05,
+      "loss": 0.9396,
+      "step": 17514
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.373291464148554,
+      "learning_rate": 1.3255008164785863e-05,
+      "loss": 0.9284,
+      "step": 17515
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9121623327804327,
+      "learning_rate": 1.325428666812416e-05,
+      "loss": 0.8368,
+      "step": 17516
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8009050632738064,
+      "learning_rate": 1.3253565152514971e-05,
+      "loss": 0.9561,
+      "step": 17517
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5178077709371987,
+      "learning_rate": 1.3252843617962494e-05,
+      "loss": 0.9957,
+      "step": 17518
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5576003518809078,
+      "learning_rate": 1.3252122064470934e-05,
+      "loss": 0.9445,
+      "step": 17519
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7841909620075487,
+      "learning_rate": 1.3251400492044487e-05,
+      "loss": 0.7889,
+      "step": 17520
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9275818071301491,
+      "learning_rate": 1.3250678900687356e-05,
+      "loss": 0.8964,
+      "step": 17521
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6746045051537102,
+      "learning_rate": 1.3249957290403748e-05,
+      "loss": 0.8425,
+      "step": 17522
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4202445212667607,
+      "learning_rate": 1.3249235661197857e-05,
+      "loss": 0.9606,
+      "step": 17523
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7275623036738277,
+      "learning_rate": 1.3248514013073886e-05,
+      "loss": 0.9084,
+      "step": 17524
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.887006299396486,
+      "learning_rate": 1.3247792346036039e-05,
+      "loss": 0.9046,
+      "step": 17525
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.107752381752136,
+      "learning_rate": 1.3247070660088519e-05,
+      "loss": 0.8883,
+      "step": 17526
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6138825994462251,
+      "learning_rate": 1.3246348955235521e-05,
+      "loss": 0.8541,
+      "step": 17527
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5090973237857577,
+      "learning_rate": 1.3245627231481258e-05,
+      "loss": 0.7948,
+      "step": 17528
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7468124405936336,
+      "learning_rate": 1.324490548882992e-05,
+      "loss": 0.7855,
+      "step": 17529
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7138195163327075,
+      "learning_rate": 1.3244183727285717e-05,
+      "loss": 0.8186,
+      "step": 17530
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4600068474053072,
+      "learning_rate": 1.3243461946852849e-05,
+      "loss": 0.9551,
+      "step": 17531
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6014383448729606,
+      "learning_rate": 1.324274014753552e-05,
+      "loss": 0.8959,
+      "step": 17532
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.02323321746498,
+      "learning_rate": 1.3242018329337928e-05,
+      "loss": 0.7187,
+      "step": 17533
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7425560801568436,
+      "learning_rate": 1.3241296492264282e-05,
+      "loss": 0.8541,
+      "step": 17534
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7885587309860612,
+      "learning_rate": 1.324057463631878e-05,
+      "loss": 0.7249,
+      "step": 17535
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8134787091004707,
+      "learning_rate": 1.3239852761505627e-05,
+      "loss": 0.9321,
+      "step": 17536
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5269333625046069,
+      "learning_rate": 1.3239130867829028e-05,
+      "loss": 0.9187,
+      "step": 17537
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8026355979530686,
+      "learning_rate": 1.3238408955293179e-05,
+      "loss": 0.8912,
+      "step": 17538
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7779932155457951,
+      "learning_rate": 1.3237687023902292e-05,
+      "loss": 0.821,
+      "step": 17539
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7401649211246264,
+      "learning_rate": 1.3236965073660562e-05,
+      "loss": 0.9286,
+      "step": 17540
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5845688871482815,
+      "learning_rate": 1.3236243104572202e-05,
+      "loss": 0.9301,
+      "step": 17541
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6702764152698177,
+      "learning_rate": 1.3235521116641405e-05,
+      "loss": 0.8126,
+      "step": 17542
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6092042852322948,
+      "learning_rate": 1.3234799109872382e-05,
+      "loss": 0.9054,
+      "step": 17543
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7330861409190992,
+      "learning_rate": 1.3234077084269334e-05,
+      "loss": 0.9215,
+      "step": 17544
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9216949471650422,
+      "learning_rate": 1.3233355039836466e-05,
+      "loss": 0.8682,
+      "step": 17545
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7144329676363343,
+      "learning_rate": 1.323263297657798e-05,
+      "loss": 0.9964,
+      "step": 17546
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5821291891183653,
+      "learning_rate": 1.3231910894498082e-05,
+      "loss": 0.7789,
+      "step": 17547
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6309807928152689,
+      "learning_rate": 1.3231188793600976e-05,
+      "loss": 0.9762,
+      "step": 17548
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7717120366598337,
+      "learning_rate": 1.3230466673890865e-05,
+      "loss": 0.732,
+      "step": 17549
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6858801357915407,
+      "learning_rate": 1.3229744535371953e-05,
+      "loss": 0.8505,
+      "step": 17550
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.956260242391665,
+      "learning_rate": 1.3229022378048448e-05,
+      "loss": 0.8735,
+      "step": 17551
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9561714482644519,
+      "learning_rate": 1.322830020192455e-05,
+      "loss": 1.021,
+      "step": 17552
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.692203977003499,
+      "learning_rate": 1.3227578007004471e-05,
+      "loss": 0.8548,
+      "step": 17553
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6892736501193328,
+      "learning_rate": 1.3226855793292406e-05,
+      "loss": 0.7572,
+      "step": 17554
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7222633421595186,
+      "learning_rate": 1.322613356079257e-05,
+      "loss": 0.931,
+      "step": 17555
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.539640276922755,
+      "learning_rate": 1.3225411309509158e-05,
+      "loss": 0.8458,
+      "step": 17556
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5309287262082258,
+      "learning_rate": 1.322468903944638e-05,
+      "loss": 1.0398,
+      "step": 17557
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.586749851020259,
+      "learning_rate": 1.3223966750608447e-05,
+      "loss": 0.8195,
+      "step": 17558
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5162822659928317,
+      "learning_rate": 1.3223244442999554e-05,
+      "loss": 0.7657,
+      "step": 17559
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5226705186405851,
+      "learning_rate": 1.3222522116623915e-05,
+      "loss": 0.8109,
+      "step": 17560
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.805681200831019,
+      "learning_rate": 1.3221799771485728e-05,
+      "loss": 0.7865,
+      "step": 17561
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7010265129063316,
+      "learning_rate": 1.3221077407589207e-05,
+      "loss": 0.8818,
+      "step": 17562
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.564199633329986,
+      "learning_rate": 1.3220355024938549e-05,
+      "loss": 0.9136,
+      "step": 17563
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9092478345912474,
+      "learning_rate": 1.3219632623537968e-05,
+      "loss": 0.9045,
+      "step": 17564
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9855200398565798,
+      "learning_rate": 1.3218910203391666e-05,
+      "loss": 0.872,
+      "step": 17565
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4490990792356555,
+      "learning_rate": 1.321818776450385e-05,
+      "loss": 0.9643,
+      "step": 17566
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3353331976604552,
+      "learning_rate": 1.3217465306878725e-05,
+      "loss": 0.82,
+      "step": 17567
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7680918872866218,
+      "learning_rate": 1.3216742830520497e-05,
+      "loss": 0.7201,
+      "step": 17568
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.890705360916382,
+      "learning_rate": 1.3216020335433375e-05,
+      "loss": 0.7582,
+      "step": 17569
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.326819129703437,
+      "learning_rate": 1.3215297821621565e-05,
+      "loss": 0.88,
+      "step": 17570
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5278815748408652,
+      "learning_rate": 1.3214575289089272e-05,
+      "loss": 0.9309,
+      "step": 17571
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.642663541763193,
+      "learning_rate": 1.3213852737840705e-05,
+      "loss": 0.9736,
+      "step": 17572
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6792542828677048,
+      "learning_rate": 1.3213130167880071e-05,
+      "loss": 0.8995,
+      "step": 17573
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7948893440011067,
+      "learning_rate": 1.3212407579211575e-05,
+      "loss": 0.7593,
+      "step": 17574
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5058888871466622,
+      "learning_rate": 1.3211684971839422e-05,
+      "loss": 0.8815,
+      "step": 17575
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.780832800699716,
+      "learning_rate": 1.3210962345767826e-05,
+      "loss": 0.9075,
+      "step": 17576
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6826516634004485,
+      "learning_rate": 1.3210239701000989e-05,
+      "loss": 0.9194,
+      "step": 17577
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4437798378349034,
+      "learning_rate": 1.320951703754312e-05,
+      "loss": 0.9041,
+      "step": 17578
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7356590796773376,
+      "learning_rate": 1.3208794355398431e-05,
+      "loss": 0.6996,
+      "step": 17579
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7115700888265517,
+      "learning_rate": 1.320807165457112e-05,
+      "loss": 1.0355,
+      "step": 17580
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5303261984926773,
+      "learning_rate": 1.3207348935065403e-05,
+      "loss": 0.882,
+      "step": 17581
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8414277353274513,
+      "learning_rate": 1.3206626196885481e-05,
+      "loss": 0.7779,
+      "step": 17582
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9327326996419343,
+      "learning_rate": 1.320590344003557e-05,
+      "loss": 0.9544,
+      "step": 17583
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4867593226569582,
+      "learning_rate": 1.3205180664519872e-05,
+      "loss": 0.8838,
+      "step": 17584
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7813643202362317,
+      "learning_rate": 1.3204457870342599e-05,
+      "loss": 0.751,
+      "step": 17585
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2431935024213474,
+      "learning_rate": 1.320373505750796e-05,
+      "loss": 0.8347,
+      "step": 17586
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3913852581713506,
+      "learning_rate": 1.3203012226020156e-05,
+      "loss": 0.926,
+      "step": 17587
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.797606326664616,
+      "learning_rate": 1.3202289375883405e-05,
+      "loss": 0.945,
+      "step": 17588
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7833789448116177,
+      "learning_rate": 1.3201566507101908e-05,
+      "loss": 0.8755,
+      "step": 17589
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7689887292711204,
+      "learning_rate": 1.3200843619679882e-05,
+      "loss": 0.9996,
+      "step": 17590
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.715576157233122,
+      "learning_rate": 1.3200120713621525e-05,
+      "loss": 0.7439,
+      "step": 17591
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.581979126314403,
+      "learning_rate": 1.3199397788931057e-05,
+      "loss": 0.7857,
+      "step": 17592
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9017528316791148,
+      "learning_rate": 1.319867484561268e-05,
+      "loss": 1.0002,
+      "step": 17593
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5039425277553207,
+      "learning_rate": 1.3197951883670606e-05,
+      "loss": 0.9193,
+      "step": 17594
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7417464821690747,
+      "learning_rate": 1.3197228903109044e-05,
+      "loss": 0.9247,
+      "step": 17595
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.522086345903614,
+      "learning_rate": 1.3196505903932203e-05,
+      "loss": 0.8676,
+      "step": 17596
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.715419462622194,
+      "learning_rate": 1.3195782886144291e-05,
+      "loss": 0.8682,
+      "step": 17597
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8368258960676629,
+      "learning_rate": 1.3195059849749519e-05,
+      "loss": 1.015,
+      "step": 17598
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.802716708608865,
+      "learning_rate": 1.3194336794752099e-05,
+      "loss": 0.7601,
+      "step": 17599
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9857810500422353,
+      "learning_rate": 1.3193613721156237e-05,
+      "loss": 0.7918,
+      "step": 17600
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7221038517678493,
+      "learning_rate": 1.3192890628966146e-05,
+      "loss": 0.9583,
+      "step": 17601
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.541008795531506,
+      "learning_rate": 1.3192167518186036e-05,
+      "loss": 0.9528,
+      "step": 17602
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.747706120743197,
+      "learning_rate": 1.3191444388820114e-05,
+      "loss": 0.7988,
+      "step": 17603
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7993052008454646,
+      "learning_rate": 1.3190721240872595e-05,
+      "loss": 0.9927,
+      "step": 17604
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6112122780886102,
+      "learning_rate": 1.3189998074347684e-05,
+      "loss": 0.8512,
+      "step": 17605
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5707802165124238,
+      "learning_rate": 1.3189274889249598e-05,
+      "loss": 0.9028,
+      "step": 17606
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.657640948987155,
+      "learning_rate": 1.3188551685582542e-05,
+      "loss": 0.8396,
+      "step": 17607
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6063060825894449,
+      "learning_rate": 1.3187828463350725e-05,
+      "loss": 0.8477,
+      "step": 17608
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7610326380333429,
+      "learning_rate": 1.3187105222558368e-05,
+      "loss": 0.7662,
+      "step": 17609
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6749609005654147,
+      "learning_rate": 1.3186381963209666e-05,
+      "loss": 0.8628,
+      "step": 17610
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5591393868505037,
+      "learning_rate": 1.3185658685308848e-05,
+      "loss": 0.8999,
+      "step": 17611
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7581709295308663,
+      "learning_rate": 1.3184935388860111e-05,
+      "loss": 0.7511,
+      "step": 17612
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4556291122847607,
+      "learning_rate": 1.3184212073867678e-05,
+      "loss": 0.8064,
+      "step": 17613
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5696751345030495,
+      "learning_rate": 1.3183488740335748e-05,
+      "loss": 0.8749,
+      "step": 17614
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.068924788631991,
+      "learning_rate": 1.3182765388268544e-05,
+      "loss": 0.9167,
+      "step": 17615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.847686206998054,
+      "learning_rate": 1.318204201767027e-05,
+      "loss": 0.9456,
+      "step": 17616
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9945804678633756,
+      "learning_rate": 1.3181318628545138e-05,
+      "loss": 0.8381,
+      "step": 17617
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5842739033285453,
+      "learning_rate": 1.3180595220897365e-05,
+      "loss": 0.8851,
+      "step": 17618
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7241023112973908,
+      "learning_rate": 1.3179871794731158e-05,
+      "loss": 0.9011,
+      "step": 17619
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9421521961105719,
+      "learning_rate": 1.3179148350050733e-05,
+      "loss": 0.9566,
+      "step": 17620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6389964460574102,
+      "learning_rate": 1.3178424886860292e-05,
+      "loss": 0.8063,
+      "step": 17621
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4265597156230994,
+      "learning_rate": 1.3177701405164065e-05,
+      "loss": 0.7933,
+      "step": 17622
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.417290711776006,
+      "learning_rate": 1.3176977904966248e-05,
+      "loss": 0.7657,
+      "step": 17623
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8481378974694707,
+      "learning_rate": 1.3176254386271063e-05,
+      "loss": 0.9804,
+      "step": 17624
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8552851403829895,
+      "learning_rate": 1.3175530849082718e-05,
+      "loss": 0.9138,
+      "step": 17625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6201081867721894,
+      "learning_rate": 1.3174807293405427e-05,
+      "loss": 0.8558,
+      "step": 17626
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6488242647255857,
+      "learning_rate": 1.3174083719243402e-05,
+      "loss": 0.8619,
+      "step": 17627
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7302277094476438,
+      "learning_rate": 1.3173360126600858e-05,
+      "loss": 0.7551,
+      "step": 17628
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5671308105049904,
+      "learning_rate": 1.3172636515482007e-05,
+      "loss": 0.9308,
+      "step": 17629
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6381216298634442,
+      "learning_rate": 1.3171912885891063e-05,
+      "loss": 0.8469,
+      "step": 17630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6323093606155379,
+      "learning_rate": 1.3171189237832234e-05,
+      "loss": 0.9164,
+      "step": 17631
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.81560820559938,
+      "learning_rate": 1.3170465571309738e-05,
+      "loss": 1.0111,
+      "step": 17632
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.484538490066388,
+      "learning_rate": 1.316974188632779e-05,
+      "loss": 0.8438,
+      "step": 17633
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8698266112683888,
+      "learning_rate": 1.31690181828906e-05,
+      "loss": 0.9044,
+      "step": 17634
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5985510815922976,
+      "learning_rate": 1.3168294461002385e-05,
+      "loss": 0.8508,
+      "step": 17635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.117465937160862,
+      "learning_rate": 1.3167570720667353e-05,
+      "loss": 0.6865,
+      "step": 17636
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6079960196882856,
+      "learning_rate": 1.3166846961889725e-05,
+      "loss": 0.9296,
+      "step": 17637
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6448633631709015,
+      "learning_rate": 1.316612318467371e-05,
+      "loss": 0.9046,
+      "step": 17638
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6433691921765758,
+      "learning_rate": 1.3165399389023523e-05,
+      "loss": 0.9391,
+      "step": 17639
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7766163909004378,
+      "learning_rate": 1.3164675574943377e-05,
+      "loss": 0.9024,
+      "step": 17640
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.798855056846456,
+      "learning_rate": 1.3163951742437493e-05,
+      "loss": 0.7652,
+      "step": 17641
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.763961199689911,
+      "learning_rate": 1.3163227891510073e-05,
+      "loss": 0.7866,
+      "step": 17642
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6164118695364509,
+      "learning_rate": 1.3162504022165345e-05,
+      "loss": 0.9123,
+      "step": 17643
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.724805961160548,
+      "learning_rate": 1.3161780134407515e-05,
+      "loss": 0.986,
+      "step": 17644
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7398381491577024,
+      "learning_rate": 1.3161056228240798e-05,
+      "loss": 0.7763,
+      "step": 17645
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9105747290973347,
+      "learning_rate": 1.3160332303669414e-05,
+      "loss": 0.9322,
+      "step": 17646
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8701459891918848,
+      "learning_rate": 1.3159608360697573e-05,
+      "loss": 0.8997,
+      "step": 17647
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4564836525883929,
+      "learning_rate": 1.3158884399329492e-05,
+      "loss": 0.8686,
+      "step": 17648
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6810336873678635,
+      "learning_rate": 1.3158160419569388e-05,
+      "loss": 0.8138,
+      "step": 17649
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5386155512447612,
+      "learning_rate": 1.3157436421421472e-05,
+      "loss": 0.9143,
+      "step": 17650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.139122682627176,
+      "learning_rate": 1.3156712404889962e-05,
+      "loss": 0.9392,
+      "step": 17651
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5126978802305164,
+      "learning_rate": 1.3155988369979071e-05,
+      "loss": 0.8697,
+      "step": 17652
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5652962251147662,
+      "learning_rate": 1.315526431669302e-05,
+      "loss": 0.9146,
+      "step": 17653
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.651052487727938,
+      "learning_rate": 1.315454024503602e-05,
+      "loss": 0.936,
+      "step": 17654
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6003553803476303,
+      "learning_rate": 1.3153816155012287e-05,
+      "loss": 0.8044,
+      "step": 17655
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5187141565668878,
+      "learning_rate": 1.3153092046626039e-05,
+      "loss": 0.7876,
+      "step": 17656
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5443981539065348,
+      "learning_rate": 1.315236791988149e-05,
+      "loss": 0.8365,
+      "step": 17657
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1277668770518456,
+      "learning_rate": 1.315164377478286e-05,
+      "loss": 0.8965,
+      "step": 17658
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8080625411297684,
+      "learning_rate": 1.3150919611334358e-05,
+      "loss": 0.926,
+      "step": 17659
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7224402059727724,
+      "learning_rate": 1.3150195429540207e-05,
+      "loss": 0.8363,
+      "step": 17660
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.644521971151708,
+      "learning_rate": 1.3149471229404617e-05,
+      "loss": 0.9227,
+      "step": 17661
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7767217214908932,
+      "learning_rate": 1.3148747010931814e-05,
+      "loss": 0.7097,
+      "step": 17662
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9256434168525638,
+      "learning_rate": 1.3148022774126003e-05,
+      "loss": 0.8731,
+      "step": 17663
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4515187425449765,
+      "learning_rate": 1.3147298518991413e-05,
+      "loss": 0.8173,
+      "step": 17664
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5034925752464814,
+      "learning_rate": 1.3146574245532254e-05,
+      "loss": 0.907,
+      "step": 17665
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.623252789494366,
+      "learning_rate": 1.3145849953752739e-05,
+      "loss": 0.8091,
+      "step": 17666
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.738048187933362,
+      "learning_rate": 1.3145125643657093e-05,
+      "loss": 0.8649,
+      "step": 17667
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4600078558080867,
+      "learning_rate": 1.3144401315249529e-05,
+      "loss": 0.9261,
+      "step": 17668
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7685338174114205,
+      "learning_rate": 1.3143676968534264e-05,
+      "loss": 0.7263,
+      "step": 17669
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7612032859659458,
+      "learning_rate": 1.3142952603515515e-05,
+      "loss": 0.6958,
+      "step": 17670
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6722905192069653,
+      "learning_rate": 1.3142228220197503e-05,
+      "loss": 0.8834,
+      "step": 17671
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6645165701927405,
+      "learning_rate": 1.3141503818584444e-05,
+      "loss": 0.8148,
+      "step": 17672
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5783461537534882,
+      "learning_rate": 1.3140779398680557e-05,
+      "loss": 0.898,
+      "step": 17673
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7236467023693725,
+      "learning_rate": 1.3140054960490053e-05,
+      "loss": 0.952,
+      "step": 17674
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.473259383461351,
+      "learning_rate": 1.3139330504017159e-05,
+      "loss": 0.8218,
+      "step": 17675
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.628189581299897,
+      "learning_rate": 1.3138606029266087e-05,
+      "loss": 0.8872,
+      "step": 17676
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7853409684011152,
+      "learning_rate": 1.3137881536241058e-05,
+      "loss": 0.6744,
+      "step": 17677
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3387403742221813,
+      "learning_rate": 1.3137157024946288e-05,
+      "loss": 0.9002,
+      "step": 17678
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4556207935025172,
+      "learning_rate": 1.3136432495385997e-05,
+      "loss": 0.7926,
+      "step": 17679
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0201947691424316,
+      "learning_rate": 1.3135707947564402e-05,
+      "loss": 0.9085,
+      "step": 17680
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5609427943400325,
+      "learning_rate": 1.3134983381485724e-05,
+      "loss": 0.8886,
+      "step": 17681
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.54131110598082,
+      "learning_rate": 1.3134258797154183e-05,
+      "loss": 0.9219,
+      "step": 17682
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1645931579090085,
+      "learning_rate": 1.3133534194573992e-05,
+      "loss": 0.9072,
+      "step": 17683
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.354533769679122,
+      "learning_rate": 1.3132809573749371e-05,
+      "loss": 0.9185,
+      "step": 17684
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8692774390605318,
+      "learning_rate": 1.3132084934684544e-05,
+      "loss": 0.7906,
+      "step": 17685
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7037684264255497,
+      "learning_rate": 1.3131360277383725e-05,
+      "loss": 0.9475,
+      "step": 17686
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8952857731146346,
+      "learning_rate": 1.3130635601851136e-05,
+      "loss": 0.9078,
+      "step": 17687
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.571680636136486,
+      "learning_rate": 1.3129910908090997e-05,
+      "loss": 0.9295,
+      "step": 17688
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1025891172232023,
+      "learning_rate": 1.3129186196107522e-05,
+      "loss": 0.7209,
+      "step": 17689
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.534708895260463,
+      "learning_rate": 1.3128461465904938e-05,
+      "loss": 0.9586,
+      "step": 17690
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4735474997865605,
+      "learning_rate": 1.3127736717487459e-05,
+      "loss": 0.9417,
+      "step": 17691
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.650336394198448,
+      "learning_rate": 1.3127011950859307e-05,
+      "loss": 0.8035,
+      "step": 17692
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8182429587416266,
+      "learning_rate": 1.31262871660247e-05,
+      "loss": 0.7594,
+      "step": 17693
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8455717803363316,
+      "learning_rate": 1.3125562362987863e-05,
+      "loss": 0.7181,
+      "step": 17694
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.724650362004695,
+      "learning_rate": 1.3124837541753008e-05,
+      "loss": 0.7637,
+      "step": 17695
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.628486636517688,
+      "learning_rate": 1.3124112702324363e-05,
+      "loss": 0.9181,
+      "step": 17696
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5311394607549509,
+      "learning_rate": 1.3123387844706144e-05,
+      "loss": 0.8436,
+      "step": 17697
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7858196504057449,
+      "learning_rate": 1.312266296890257e-05,
+      "loss": 0.7535,
+      "step": 17698
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7657787416825161,
+      "learning_rate": 1.3121938074917866e-05,
+      "loss": 0.7447,
+      "step": 17699
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.579519025745858,
+      "learning_rate": 1.3121213162756247e-05,
+      "loss": 0.9704,
+      "step": 17700
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.098182666827929,
+      "learning_rate": 1.3120488232421939e-05,
+      "loss": 0.8947,
+      "step": 17701
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8199871821533078,
+      "learning_rate": 1.311976328391916e-05,
+      "loss": 0.7873,
+      "step": 17702
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4093405308104474,
+      "learning_rate": 1.3119038317252133e-05,
+      "loss": 0.851,
+      "step": 17703
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4700631947247227,
+      "learning_rate": 1.3118313332425077e-05,
+      "loss": 0.835,
+      "step": 17704
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0263159503261527,
+      "learning_rate": 1.3117588329442213e-05,
+      "loss": 0.8225,
+      "step": 17705
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5342508930239946,
+      "learning_rate": 1.3116863308307762e-05,
+      "loss": 1.0199,
+      "step": 17706
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8205141835264418,
+      "learning_rate": 1.3116138269025947e-05,
+      "loss": 0.8845,
+      "step": 17707
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5954854249769883,
+      "learning_rate": 1.3115413211600987e-05,
+      "loss": 0.8911,
+      "step": 17708
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7779946000547519,
+      "learning_rate": 1.3114688136037109e-05,
+      "loss": 0.7858,
+      "step": 17709
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4625248149224614,
+      "learning_rate": 1.3113963042338527e-05,
+      "loss": 0.9274,
+      "step": 17710
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6008833322485365,
+      "learning_rate": 1.3113237930509469e-05,
+      "loss": 0.8606,
+      "step": 17711
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4453278506718714,
+      "learning_rate": 1.3112512800554149e-05,
+      "loss": 0.8942,
+      "step": 17712
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7621453656729962,
+      "learning_rate": 1.31117876524768e-05,
+      "loss": 0.8677,
+      "step": 17713
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7335664873027106,
+      "learning_rate": 1.3111062486281632e-05,
+      "loss": 0.883,
+      "step": 17714
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9021834726155709,
+      "learning_rate": 1.3110337301972877e-05,
+      "loss": 0.8225,
+      "step": 17715
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.626531335496666,
+      "learning_rate": 1.3109612099554754e-05,
+      "loss": 0.932,
+      "step": 17716
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.732011214616894,
+      "learning_rate": 1.3108886879031485e-05,
+      "loss": 0.9357,
+      "step": 17717
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7656509141252584,
+      "learning_rate": 1.310816164040729e-05,
+      "loss": 0.7543,
+      "step": 17718
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8631415128698183,
+      "learning_rate": 1.3107436383686394e-05,
+      "loss": 0.8929,
+      "step": 17719
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5568434855010505,
+      "learning_rate": 1.3106711108873022e-05,
+      "loss": 1.022,
+      "step": 17720
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6208373385286805,
+      "learning_rate": 1.3105985815971392e-05,
+      "loss": 0.8069,
+      "step": 17721
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5943368779935483,
+      "learning_rate": 1.3105260504985732e-05,
+      "loss": 0.8837,
+      "step": 17722
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7990284087697931,
+      "learning_rate": 1.3104535175920261e-05,
+      "loss": 0.7181,
+      "step": 17723
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0698458593974722,
+      "learning_rate": 1.3103809828779202e-05,
+      "loss": 0.9089,
+      "step": 17724
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.597778765200159,
+      "learning_rate": 1.310308446356678e-05,
+      "loss": 0.8498,
+      "step": 17725
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5772205680630205,
+      "learning_rate": 1.3102359080287219e-05,
+      "loss": 0.7912,
+      "step": 17726
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.236513921565685,
+      "learning_rate": 1.3101633678944741e-05,
+      "loss": 0.9562,
+      "step": 17727
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7824498346174578,
+      "learning_rate": 1.3100908259543567e-05,
+      "loss": 0.9927,
+      "step": 17728
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.598998896010327,
+      "learning_rate": 1.3100182822087928e-05,
+      "loss": 0.8126,
+      "step": 17729
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7148560089276528,
+      "learning_rate": 1.309945736658204e-05,
+      "loss": 0.8807,
+      "step": 17730
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7589898881197247,
+      "learning_rate": 1.309873189303013e-05,
+      "loss": 0.7116,
+      "step": 17731
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.686400443024025,
+      "learning_rate": 1.3098006401436422e-05,
+      "loss": 0.9482,
+      "step": 17732
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.75975472613126,
+      "learning_rate": 1.3097280891805141e-05,
+      "loss": 0.8347,
+      "step": 17733
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4733271758228423,
+      "learning_rate": 1.309655536414051e-05,
+      "loss": 0.7312,
+      "step": 17734
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8958871081987987,
+      "learning_rate": 1.3095829818446753e-05,
+      "loss": 0.905,
+      "step": 17735
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9719566189229936,
+      "learning_rate": 1.3095104254728095e-05,
+      "loss": 0.8843,
+      "step": 17736
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5406644591251324,
+      "learning_rate": 1.3094378672988759e-05,
+      "loss": 0.8591,
+      "step": 17737
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.190301971710499,
+      "learning_rate": 1.3093653073232972e-05,
+      "loss": 0.8315,
+      "step": 17738
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7782509088523906,
+      "learning_rate": 1.3092927455464959e-05,
+      "loss": 0.9172,
+      "step": 17739
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5849186942292126,
+      "learning_rate": 1.3092201819688937e-05,
+      "loss": 0.9947,
+      "step": 17740
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8670493434800386,
+      "learning_rate": 1.3091476165909142e-05,
+      "loss": 0.8096,
+      "step": 17741
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8098300920683252,
+      "learning_rate": 1.3090750494129792e-05,
+      "loss": 0.8527,
+      "step": 17742
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.198391120547303,
+      "learning_rate": 1.3090024804355117e-05,
+      "loss": 0.8806,
+      "step": 17743
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7796836757868704,
+      "learning_rate": 1.3089299096589336e-05,
+      "loss": 0.9516,
+      "step": 17744
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.453635942891441,
+      "learning_rate": 1.308857337083668e-05,
+      "loss": 0.9773,
+      "step": 17745
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7718637004219782,
+      "learning_rate": 1.3087847627101372e-05,
+      "loss": 0.8431,
+      "step": 17746
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6320467172828061,
+      "learning_rate": 1.3087121865387636e-05,
+      "loss": 0.7955,
+      "step": 17747
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.8142906063969395,
+      "learning_rate": 1.30863960856997e-05,
+      "loss": 0.8823,
+      "step": 17748
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.755484842950872,
+      "learning_rate": 1.3085670288041787e-05,
+      "loss": 0.9395,
+      "step": 17749
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.343593191101696,
+      "learning_rate": 1.3084944472418126e-05,
+      "loss": 0.937,
+      "step": 17750
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0635568673782365,
+      "learning_rate": 1.3084218638832943e-05,
+      "loss": 0.9344,
+      "step": 17751
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4846235417760456,
+      "learning_rate": 1.3083492787290462e-05,
+      "loss": 0.9596,
+      "step": 17752
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5910667605922604,
+      "learning_rate": 1.308276691779491e-05,
+      "loss": 0.8959,
+      "step": 17753
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.970241587190273,
+      "learning_rate": 1.3082041030350512e-05,
+      "loss": 0.9354,
+      "step": 17754
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7354180982550378,
+      "learning_rate": 1.3081315124961496e-05,
+      "loss": 0.8352,
+      "step": 17755
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7697922178172037,
+      "learning_rate": 1.3080589201632086e-05,
+      "loss": 0.7759,
+      "step": 17756
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.402180599847784,
+      "learning_rate": 1.3079863260366512e-05,
+      "loss": 0.9518,
+      "step": 17757
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.84685418458741,
+      "learning_rate": 1.3079137301168998e-05,
+      "loss": 0.8046,
+      "step": 17758
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.593961050528826,
+      "learning_rate": 1.3078411324043772e-05,
+      "loss": 0.9118,
+      "step": 17759
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6169152814466807,
+      "learning_rate": 1.3077685328995065e-05,
+      "loss": 0.8238,
+      "step": 17760
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7365373450300627,
+      "learning_rate": 1.3076959316027093e-05,
+      "loss": 0.8767,
+      "step": 17761
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6654827831862802,
+      "learning_rate": 1.3076233285144096e-05,
+      "loss": 0.9196,
+      "step": 17762
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.654908542093687,
+      "learning_rate": 1.307550723635029e-05,
+      "loss": 0.9503,
+      "step": 17763
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0091265455388507,
+      "learning_rate": 1.3074781169649908e-05,
+      "loss": 0.8175,
+      "step": 17764
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8292198777756892,
+      "learning_rate": 1.3074055085047174e-05,
+      "loss": 0.9314,
+      "step": 17765
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6859031044367137,
+      "learning_rate": 1.3073328982546325e-05,
+      "loss": 0.8145,
+      "step": 17766
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4985818561984934,
+      "learning_rate": 1.3072602862151577e-05,
+      "loss": 0.8192,
+      "step": 17767
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8971435031458619,
+      "learning_rate": 1.307187672386716e-05,
+      "loss": 0.8722,
+      "step": 17768
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6397572990363418,
+      "learning_rate": 1.3071150567697313e-05,
+      "loss": 0.776,
+      "step": 17769
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.465877106267277,
+      "learning_rate": 1.3070424393646246e-05,
+      "loss": 0.9346,
+      "step": 17770
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8000380407918415,
+      "learning_rate": 1.3069698201718202e-05,
+      "loss": 0.9441,
+      "step": 17771
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9008485876532157,
+      "learning_rate": 1.3068971991917398e-05,
+      "loss": 0.75,
+      "step": 17772
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5712240685512104,
+      "learning_rate": 1.3068245764248073e-05,
+      "loss": 0.7919,
+      "step": 17773
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6391973960349013,
+      "learning_rate": 1.3067519518714447e-05,
+      "loss": 0.8979,
+      "step": 17774
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.494437967855948,
+      "learning_rate": 1.306679325532075e-05,
+      "loss": 1.0126,
+      "step": 17775
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6238154601364758,
+      "learning_rate": 1.3066066974071213e-05,
+      "loss": 0.9281,
+      "step": 17776
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7244812710684083,
+      "learning_rate": 1.3065340674970064e-05,
+      "loss": 0.7922,
+      "step": 17777
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7463098567681463,
+      "learning_rate": 1.3064614358021532e-05,
+      "loss": 0.8073,
+      "step": 17778
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5518523370548143,
+      "learning_rate": 1.3063888023229842e-05,
+      "loss": 0.7448,
+      "step": 17779
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9499326878549568,
+      "learning_rate": 1.3063161670599227e-05,
+      "loss": 1.0121,
+      "step": 17780
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6561391974444237,
+      "learning_rate": 1.3062435300133917e-05,
+      "loss": 1.0221,
+      "step": 17781
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7929334643295874,
+      "learning_rate": 1.3061708911838138e-05,
+      "loss": 0.7229,
+      "step": 17782
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6147503015404534,
+      "learning_rate": 1.306098250571612e-05,
+      "loss": 0.8136,
+      "step": 17783
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.687597688202173,
+      "learning_rate": 1.3060256081772092e-05,
+      "loss": 0.8694,
+      "step": 17784
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.876924973002267,
+      "learning_rate": 1.3059529640010284e-05,
+      "loss": 0.76,
+      "step": 17785
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8975048095084959,
+      "learning_rate": 1.3058803180434927e-05,
+      "loss": 0.8283,
+      "step": 17786
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5349283624810617,
+      "learning_rate": 1.3058076703050249e-05,
+      "loss": 0.8378,
+      "step": 17787
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6083670297435682,
+      "learning_rate": 1.3057350207860484e-05,
+      "loss": 0.9628,
+      "step": 17788
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4730605341953387,
+      "learning_rate": 1.305662369486985e-05,
+      "loss": 0.8886,
+      "step": 17789
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8314832145240731,
+      "learning_rate": 1.3055897164082594e-05,
+      "loss": 0.7615,
+      "step": 17790
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5299678832279011,
+      "learning_rate": 1.305517061550293e-05,
+      "loss": 0.9859,
+      "step": 17791
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.560025184923582,
+      "learning_rate": 1.3054444049135099e-05,
+      "loss": 0.9342,
+      "step": 17792
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7804071160220923,
+      "learning_rate": 1.3053717464983327e-05,
+      "loss": 0.7724,
+      "step": 17793
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6047829615483875,
+      "learning_rate": 1.3052990863051846e-05,
+      "loss": 0.9444,
+      "step": 17794
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5153069882619983,
+      "learning_rate": 1.3052264243344885e-05,
+      "loss": 0.9005,
+      "step": 17795
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.683275475833677,
+      "learning_rate": 1.3051537605866675e-05,
+      "loss": 0.8544,
+      "step": 17796
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5657550397405746,
+      "learning_rate": 1.305081095062145e-05,
+      "loss": 0.8983,
+      "step": 17797
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.828906064697272,
+      "learning_rate": 1.3050084277613433e-05,
+      "loss": 0.9298,
+      "step": 17798
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6182320040017797,
+      "learning_rate": 1.3049357586846862e-05,
+      "loss": 0.9193,
+      "step": 17799
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6919807243349899,
+      "learning_rate": 1.3048630878325966e-05,
+      "loss": 0.7296,
+      "step": 17800
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5877226138769474,
+      "learning_rate": 1.3047904152054974e-05,
+      "loss": 0.8135,
+      "step": 17801
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.417643796221635,
+      "learning_rate": 1.3047177408038119e-05,
+      "loss": 0.8975,
+      "step": 17802
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8776812584821472,
+      "learning_rate": 1.3046450646279633e-05,
+      "loss": 0.8169,
+      "step": 17803
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7785032327584396,
+      "learning_rate": 1.3045723866783749e-05,
+      "loss": 0.7421,
+      "step": 17804
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7970986819531123,
+      "learning_rate": 1.3044997069554695e-05,
+      "loss": 0.8365,
+      "step": 17805
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6917777770149025,
+      "learning_rate": 1.3044270254596706e-05,
+      "loss": 0.8195,
+      "step": 17806
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7993416129371892,
+      "learning_rate": 1.3043543421914007e-05,
+      "loss": 0.7272,
+      "step": 17807
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9715886322320815,
+      "learning_rate": 1.3042816571510836e-05,
+      "loss": 0.9156,
+      "step": 17808
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.621181711843115,
+      "learning_rate": 1.3042089703391425e-05,
+      "loss": 0.9673,
+      "step": 17809
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.596583148151477,
+      "learning_rate": 1.3041362817560007e-05,
+      "loss": 0.774,
+      "step": 17810
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6270419948333599,
+      "learning_rate": 1.304063591402081e-05,
+      "loss": 0.8151,
+      "step": 17811
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5410883275424774,
+      "learning_rate": 1.3039908992778065e-05,
+      "loss": 0.9029,
+      "step": 17812
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5422317019220038,
+      "learning_rate": 1.3039182053836013e-05,
+      "loss": 0.745,
+      "step": 17813
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.042010697789202,
+      "learning_rate": 1.3038455097198875e-05,
+      "loss": 0.9323,
+      "step": 17814
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7408883830800184,
+      "learning_rate": 1.3037728122870891e-05,
+      "loss": 0.8096,
+      "step": 17815
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.825196260912023,
+      "learning_rate": 1.3037001130856296e-05,
+      "loss": 0.9095,
+      "step": 17816
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0598755772893296,
+      "learning_rate": 1.3036274121159313e-05,
+      "loss": 0.9301,
+      "step": 17817
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7989242144411803,
+      "learning_rate": 1.3035547093784187e-05,
+      "loss": 0.7406,
+      "step": 17818
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0769712830636813,
+      "learning_rate": 1.3034820048735139e-05,
+      "loss": 0.7706,
+      "step": 17819
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.596463109305129,
+      "learning_rate": 1.3034092986016413e-05,
+      "loss": 0.8754,
+      "step": 17820
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.033143031679587,
+      "learning_rate": 1.3033365905632234e-05,
+      "loss": 0.9543,
+      "step": 17821
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5023348179934848,
+      "learning_rate": 1.3032638807586842e-05,
+      "loss": 0.8784,
+      "step": 17822
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.607451911295979,
+      "learning_rate": 1.303191169188446e-05,
+      "loss": 0.8601,
+      "step": 17823
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.67837208229207,
+      "learning_rate": 1.3031184558529335e-05,
+      "loss": 0.8252,
+      "step": 17824
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6749608367440019,
+      "learning_rate": 1.303045740752569e-05,
+      "loss": 0.8678,
+      "step": 17825
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5458548612389897,
+      "learning_rate": 1.3029730238877764e-05,
+      "loss": 0.902,
+      "step": 17826
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.797271165797005,
+      "learning_rate": 1.3029003052589789e-05,
+      "loss": 0.9476,
+      "step": 17827
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5780484631060467,
+      "learning_rate": 1.3028275848665998e-05,
+      "loss": 0.9513,
+      "step": 17828
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6069440437450615,
+      "learning_rate": 1.3027548627110629e-05,
+      "loss": 0.9914,
+      "step": 17829
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5933435659176354,
+      "learning_rate": 1.3026821387927911e-05,
+      "loss": 0.9674,
+      "step": 17830
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6195219346801342,
+      "learning_rate": 1.302609413112208e-05,
+      "loss": 0.8328,
+      "step": 17831
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6140241551899066,
+      "learning_rate": 1.3025366856697372e-05,
+      "loss": 0.8957,
+      "step": 17832
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.838873326214219,
+      "learning_rate": 1.3024639564658022e-05,
+      "loss": 0.9657,
+      "step": 17833
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7519029251426472,
+      "learning_rate": 1.3023912255008261e-05,
+      "loss": 0.9193,
+      "step": 17834
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.770240858446838,
+      "learning_rate": 1.3023184927752326e-05,
+      "loss": 1.0086,
+      "step": 17835
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.595948279929593,
+      "learning_rate": 1.3022457582894452e-05,
+      "loss": 0.9198,
+      "step": 17836
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6719272638570657,
+      "learning_rate": 1.302173022043887e-05,
+      "loss": 0.8837,
+      "step": 17837
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5682310301853457,
+      "learning_rate": 1.302100284038982e-05,
+      "loss": 0.7965,
+      "step": 17838
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7037387670756186,
+      "learning_rate": 1.3020275442751538e-05,
+      "loss": 0.8624,
+      "step": 17839
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4610662219346877,
+      "learning_rate": 1.3019548027528253e-05,
+      "loss": 0.8609,
+      "step": 17840
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4942823779892542,
+      "learning_rate": 1.3018820594724204e-05,
+      "loss": 0.8711,
+      "step": 17841
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5433523844983268,
+      "learning_rate": 1.3018093144343624e-05,
+      "loss": 0.8123,
+      "step": 17842
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6443629383990397,
+      "learning_rate": 1.3017365676390755e-05,
+      "loss": 0.7317,
+      "step": 17843
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7296798424317683,
+      "learning_rate": 1.3016638190869822e-05,
+      "loss": 0.7203,
+      "step": 17844
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9550822161311103,
+      "learning_rate": 1.3015910687785072e-05,
+      "loss": 0.9025,
+      "step": 17845
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7470728334348187,
+      "learning_rate": 1.3015183167140731e-05,
+      "loss": 0.8668,
+      "step": 17846
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4987705442195076,
+      "learning_rate": 1.3014455628941041e-05,
+      "loss": 0.9274,
+      "step": 17847
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6839975738783783,
+      "learning_rate": 1.3013728073190235e-05,
+      "loss": 0.9618,
+      "step": 17848
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7058497176266056,
+      "learning_rate": 1.3013000499892549e-05,
+      "loss": 0.6924,
+      "step": 17849
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.726251424379372,
+      "learning_rate": 1.3012272909052221e-05,
+      "loss": 0.9278,
+      "step": 17850
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6861267692071695,
+      "learning_rate": 1.3011545300673486e-05,
+      "loss": 0.7907,
+      "step": 17851
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8477256272438625,
+      "learning_rate": 1.3010817674760583e-05,
+      "loss": 0.8495,
+      "step": 17852
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1730507409363566,
+      "learning_rate": 1.3010090031317746e-05,
+      "loss": 0.9716,
+      "step": 17853
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.944564605850094,
+      "learning_rate": 1.300936237034921e-05,
+      "loss": 0.9518,
+      "step": 17854
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1751928825558373,
+      "learning_rate": 1.3008634691859214e-05,
+      "loss": 0.7601,
+      "step": 17855
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.649795310586642,
+      "learning_rate": 1.3007906995851994e-05,
+      "loss": 0.8716,
+      "step": 17856
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8452282841147012,
+      "learning_rate": 1.3007179282331788e-05,
+      "loss": 0.8326,
+      "step": 17857
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1741739032179974,
+      "learning_rate": 1.3006451551302833e-05,
+      "loss": 0.79,
+      "step": 17858
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7269594559926096,
+      "learning_rate": 1.3005723802769363e-05,
+      "loss": 0.8766,
+      "step": 17859
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6024433199244674,
+      "learning_rate": 1.300499603673562e-05,
+      "loss": 0.939,
+      "step": 17860
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.831402766935018,
+      "learning_rate": 1.3004268253205837e-05,
+      "loss": 0.8365,
+      "step": 17861
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6773424486665827,
+      "learning_rate": 1.3003540452184254e-05,
+      "loss": 0.8636,
+      "step": 17862
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5118075494951577,
+      "learning_rate": 1.3002812633675105e-05,
+      "loss": 0.9989,
+      "step": 17863
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.608908110796196,
+      "learning_rate": 1.3002084797682633e-05,
+      "loss": 0.9534,
+      "step": 17864
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6316861606693334,
+      "learning_rate": 1.300135694421107e-05,
+      "loss": 0.7804,
+      "step": 17865
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.037485785613404,
+      "learning_rate": 1.3000629073264658e-05,
+      "loss": 1.0072,
+      "step": 17866
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.202935561530704,
+      "learning_rate": 1.2999901184847636e-05,
+      "loss": 0.9536,
+      "step": 17867
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.02766354720404,
+      "learning_rate": 1.2999173278964237e-05,
+      "loss": 0.8553,
+      "step": 17868
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8491658444135652,
+      "learning_rate": 1.2998445355618704e-05,
+      "loss": 0.8498,
+      "step": 17869
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4952063623435325,
+      "learning_rate": 1.2997717414815269e-05,
+      "loss": 0.9909,
+      "step": 17870
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8715899033384624,
+      "learning_rate": 1.299698945655818e-05,
+      "loss": 0.7735,
+      "step": 17871
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6813576125233316,
+      "learning_rate": 1.2996261480851664e-05,
+      "loss": 0.9077,
+      "step": 17872
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4821589790476186,
+      "learning_rate": 1.299553348769997e-05,
+      "loss": 0.8374,
+      "step": 17873
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5605502027856868,
+      "learning_rate": 1.2994805477107323e-05,
+      "loss": 0.7874,
+      "step": 17874
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.867284852608871,
+      "learning_rate": 1.299407744907798e-05,
+      "loss": 0.9321,
+      "step": 17875
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7076748350447386,
+      "learning_rate": 1.2993349403616167e-05,
+      "loss": 0.7175,
+      "step": 17876
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7423139880914191,
+      "learning_rate": 1.2992621340726124e-05,
+      "loss": 0.8465,
+      "step": 17877
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7617285904340407,
+      "learning_rate": 1.2991893260412093e-05,
+      "loss": 0.7379,
+      "step": 17878
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.516337053724594,
+      "learning_rate": 1.2991165162678315e-05,
+      "loss": 0.8754,
+      "step": 17879
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7510540093748881,
+      "learning_rate": 1.2990437047529022e-05,
+      "loss": 0.7164,
+      "step": 17880
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5772893201274292,
+      "learning_rate": 1.298970891496846e-05,
+      "loss": 0.8852,
+      "step": 17881
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8643387581590534,
+      "learning_rate": 1.2988980765000867e-05,
+      "loss": 0.8622,
+      "step": 17882
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6301060859560916,
+      "learning_rate": 1.298825259763048e-05,
+      "loss": 0.9853,
+      "step": 17883
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6056848637816126,
+      "learning_rate": 1.298752441286154e-05,
+      "loss": 0.8717,
+      "step": 17884
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5881058567518163,
+      "learning_rate": 1.2986796210698288e-05,
+      "loss": 0.9783,
+      "step": 17885
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6851226393213241,
+      "learning_rate": 1.2986067991144963e-05,
+      "loss": 0.8469,
+      "step": 17886
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1095774105872818,
+      "learning_rate": 1.2985339754205806e-05,
+      "loss": 0.8719,
+      "step": 17887
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8482886868501827,
+      "learning_rate": 1.2984611499885053e-05,
+      "loss": 0.7772,
+      "step": 17888
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.721850155748886,
+      "learning_rate": 1.2983883228186949e-05,
+      "loss": 0.8224,
+      "step": 17889
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7047762540413947,
+      "learning_rate": 1.2983154939115734e-05,
+      "loss": 0.9331,
+      "step": 17890
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.747452172936446,
+      "learning_rate": 1.2982426632675641e-05,
+      "loss": 0.8917,
+      "step": 17891
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6982333736844932,
+      "learning_rate": 1.2981698308870923e-05,
+      "loss": 0.9183,
+      "step": 17892
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5264422005618903,
+      "learning_rate": 1.2980969967705806e-05,
+      "loss": 0.9047,
+      "step": 17893
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8818965101397696,
+      "learning_rate": 1.2980241609184546e-05,
+      "loss": 0.8894,
+      "step": 17894
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6904372666732814,
+      "learning_rate": 1.2979513233311367e-05,
+      "loss": 0.7904,
+      "step": 17895
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4383519820016835,
+      "learning_rate": 1.2978784840090526e-05,
+      "loss": 0.8991,
+      "step": 17896
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8744179757666057,
+      "learning_rate": 1.2978056429526254e-05,
+      "loss": 0.9064,
+      "step": 17897
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9234836146003358,
+      "learning_rate": 1.2977328001622793e-05,
+      "loss": 0.8117,
+      "step": 17898
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7540821460192246,
+      "learning_rate": 1.2976599556384385e-05,
+      "loss": 0.8946,
+      "step": 17899
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6286946917123746,
+      "learning_rate": 1.2975871093815275e-05,
+      "loss": 0.851,
+      "step": 17900
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2545777853240367,
+      "learning_rate": 1.2975142613919702e-05,
+      "loss": 0.9317,
+      "step": 17901
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.975356868843869,
+      "learning_rate": 1.2974414116701903e-05,
+      "loss": 0.8912,
+      "step": 17902
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9024494587461183,
+      "learning_rate": 1.2973685602166129e-05,
+      "loss": 0.9532,
+      "step": 17903
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6405255569075812,
+      "learning_rate": 1.297295707031661e-05,
+      "loss": 0.7534,
+      "step": 17904
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7320909091414403,
+      "learning_rate": 1.2972228521157598e-05,
+      "loss": 0.8232,
+      "step": 17905
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7200929377637457,
+      "learning_rate": 1.2971499954693327e-05,
+      "loss": 0.8693,
+      "step": 17906
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.821474609993956,
+      "learning_rate": 1.2970771370928045e-05,
+      "loss": 0.9356,
+      "step": 17907
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7265094358483971,
+      "learning_rate": 1.2970042769865992e-05,
+      "loss": 0.7637,
+      "step": 17908
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8189869059152135,
+      "learning_rate": 1.2969314151511407e-05,
+      "loss": 0.8003,
+      "step": 17909
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5404984494520795,
+      "learning_rate": 1.2968585515868537e-05,
+      "loss": 0.8947,
+      "step": 17910
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5979389294580568,
+      "learning_rate": 1.2967856862941623e-05,
+      "loss": 0.8846,
+      "step": 17911
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8098485825208879,
+      "learning_rate": 1.2967128192734903e-05,
+      "loss": 0.8187,
+      "step": 17912
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6611271366984501,
+      "learning_rate": 1.2966399505252628e-05,
+      "loss": 0.9841,
+      "step": 17913
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.906513849593737,
+      "learning_rate": 1.2965670800499032e-05,
+      "loss": 0.8954,
+      "step": 17914
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.888036956175318,
+      "learning_rate": 1.2964942078478365e-05,
+      "loss": 0.867,
+      "step": 17915
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7539560763360025,
+      "learning_rate": 1.2964213339194865e-05,
+      "loss": 0.8939,
+      "step": 17916
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6169481170660043,
+      "learning_rate": 1.2963484582652775e-05,
+      "loss": 0.9791,
+      "step": 17917
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.423464573377525,
+      "learning_rate": 1.2962755808856341e-05,
+      "loss": 0.897,
+      "step": 17918
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.107668751692523,
+      "learning_rate": 1.2962027017809804e-05,
+      "loss": 0.9063,
+      "step": 17919
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7700443580629224,
+      "learning_rate": 1.296129820951741e-05,
+      "loss": 0.9115,
+      "step": 17920
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.684391534136726,
+      "learning_rate": 1.2960569383983397e-05,
+      "loss": 0.9386,
+      "step": 17921
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.428835122963612,
+      "learning_rate": 1.2959840541212017e-05,
+      "loss": 0.8834,
+      "step": 17922
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5781639280842543,
+      "learning_rate": 1.29591116812075e-05,
+      "loss": 0.8672,
+      "step": 17923
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4758680784611948,
+      "learning_rate": 1.2958382803974106e-05,
+      "loss": 0.9257,
+      "step": 17924
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5093593344320646,
+      "learning_rate": 1.2957653909516065e-05,
+      "loss": 0.9836,
+      "step": 17925
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4766967442141488,
+      "learning_rate": 1.2956924997837633e-05,
+      "loss": 0.7814,
+      "step": 17926
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5914339490307816,
+      "learning_rate": 1.2956196068943043e-05,
+      "loss": 0.6896,
+      "step": 17927
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.251438866768747,
+      "learning_rate": 1.2955467122836545e-05,
+      "loss": 0.7572,
+      "step": 17928
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.698350869818794,
+      "learning_rate": 1.295473815952238e-05,
+      "loss": 1.0196,
+      "step": 17929
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.002690967220908,
+      "learning_rate": 1.2954009179004794e-05,
+      "loss": 0.8824,
+      "step": 17930
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5472150703604726,
+      "learning_rate": 1.2953280181288033e-05,
+      "loss": 0.8118,
+      "step": 17931
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6323196456427478,
+      "learning_rate": 1.2952551166376339e-05,
+      "loss": 0.9923,
+      "step": 17932
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.708276253210386,
+      "learning_rate": 1.2951822134273957e-05,
+      "loss": 0.9481,
+      "step": 17933
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8120406669160841,
+      "learning_rate": 1.2951093084985131e-05,
+      "loss": 0.7278,
+      "step": 17934
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.628931012647213,
+      "learning_rate": 1.295036401851411e-05,
+      "loss": 0.8606,
+      "step": 17935
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2704850307726545,
+      "learning_rate": 1.2949634934865134e-05,
+      "loss": 1.008,
+      "step": 17936
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.411013116944502,
+      "learning_rate": 1.294890583404245e-05,
+      "loss": 0.929,
+      "step": 17937
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5995450637342203,
+      "learning_rate": 1.29481767160503e-05,
+      "loss": 0.9556,
+      "step": 17938
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4877353064865613,
+      "learning_rate": 1.2947447580892933e-05,
+      "loss": 0.921,
+      "step": 17939
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5367102099776115,
+      "learning_rate": 1.2946718428574595e-05,
+      "loss": 0.904,
+      "step": 17940
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6024209550128685,
+      "learning_rate": 1.294598925909953e-05,
+      "loss": 0.8864,
+      "step": 17941
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5780024090470437,
+      "learning_rate": 1.294526007247198e-05,
+      "loss": 0.8474,
+      "step": 17942
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5700231491339507,
+      "learning_rate": 1.2944530868696194e-05,
+      "loss": 0.9492,
+      "step": 17943
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4056139667225467,
+      "learning_rate": 1.2943801647776418e-05,
+      "loss": 0.8159,
+      "step": 17944
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.556024363967329,
+      "learning_rate": 1.2943072409716897e-05,
+      "loss": 0.8476,
+      "step": 17945
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.755001367258458,
+      "learning_rate": 1.2942343154521875e-05,
+      "loss": 0.8106,
+      "step": 17946
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6985866955087887,
+      "learning_rate": 1.2941613882195601e-05,
+      "loss": 0.9465,
+      "step": 17947
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8460207877711903,
+      "learning_rate": 1.294088459274232e-05,
+      "loss": 0.8621,
+      "step": 17948
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4755879114932737,
+      "learning_rate": 1.2940155286166279e-05,
+      "loss": 0.9477,
+      "step": 17949
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5019862840968994,
+      "learning_rate": 1.2939425962471719e-05,
+      "loss": 0.928,
+      "step": 17950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8620456535942378,
+      "learning_rate": 1.293869662166289e-05,
+      "loss": 1.0075,
+      "step": 17951
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5840089053184243,
+      "learning_rate": 1.2937967263744044e-05,
+      "loss": 0.9258,
+      "step": 17952
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6252905265414086,
+      "learning_rate": 1.2937237888719418e-05,
+      "loss": 0.8816,
+      "step": 17953
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.072329699592405,
+      "learning_rate": 1.2936508496593268e-05,
+      "loss": 0.9151,
+      "step": 17954
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7665340264776759,
+      "learning_rate": 1.2935779087369832e-05,
+      "loss": 0.8868,
+      "step": 17955
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4896940500153173,
+      "learning_rate": 1.2935049661053363e-05,
+      "loss": 0.8636,
+      "step": 17956
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6937105058050261,
+      "learning_rate": 1.2934320217648101e-05,
+      "loss": 1.0249,
+      "step": 17957
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.492863792212677,
+      "learning_rate": 1.2933590757158303e-05,
+      "loss": 0.8785,
+      "step": 17958
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.117967409884998,
+      "learning_rate": 1.293286127958821e-05,
+      "loss": 0.9508,
+      "step": 17959
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8765171442124802,
+      "learning_rate": 1.2932131784942069e-05,
+      "loss": 0.9435,
+      "step": 17960
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6876315395475865,
+      "learning_rate": 1.2931402273224128e-05,
+      "loss": 0.8968,
+      "step": 17961
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0655450721902153,
+      "learning_rate": 1.2930672744438634e-05,
+      "loss": 0.9096,
+      "step": 17962
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5809070773490035,
+      "learning_rate": 1.2929943198589836e-05,
+      "loss": 0.8812,
+      "step": 17963
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.805852285431323,
+      "learning_rate": 1.2929213635681983e-05,
+      "loss": 0.8609,
+      "step": 17964
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8843184988636108,
+      "learning_rate": 1.292848405571932e-05,
+      "loss": 0.9417,
+      "step": 17965
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4511021135069824,
+      "learning_rate": 1.2927754458706094e-05,
+      "loss": 0.8148,
+      "step": 17966
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5783171892651549,
+      "learning_rate": 1.2927024844646556e-05,
+      "loss": 0.9336,
+      "step": 17967
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6949340719765587,
+      "learning_rate": 1.2926295213544953e-05,
+      "loss": 0.8383,
+      "step": 17968
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6989048227139936,
+      "learning_rate": 1.2925565565405536e-05,
+      "loss": 0.9337,
+      "step": 17969
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8705308651025152,
+      "learning_rate": 1.2924835900232542e-05,
+      "loss": 0.8115,
+      "step": 17970
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.21911986398871,
+      "learning_rate": 1.2924106218030235e-05,
+      "loss": 0.8031,
+      "step": 17971
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.668044554445292,
+      "learning_rate": 1.292337651880285e-05,
+      "loss": 0.7955,
+      "step": 17972
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.111057620001355,
+      "learning_rate": 1.2922646802554649e-05,
+      "loss": 0.9522,
+      "step": 17973
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4225863510525223,
+      "learning_rate": 1.2921917069289864e-05,
+      "loss": 0.8119,
+      "step": 17974
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6027936740425717,
+      "learning_rate": 1.2921187319012761e-05,
+      "loss": 0.8676,
+      "step": 17975
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.695840941742233,
+      "learning_rate": 1.2920457551727575e-05,
+      "loss": 0.868,
+      "step": 17976
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6918907174085318,
+      "learning_rate": 1.2919727767438562e-05,
+      "loss": 0.8526,
+      "step": 17977
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.63079988574912,
+      "learning_rate": 1.2918997966149969e-05,
+      "loss": 1.0477,
+      "step": 17978
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6082712706859699,
+      "learning_rate": 1.2918268147866047e-05,
+      "loss": 0.7567,
+      "step": 17979
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9671751627668226,
+      "learning_rate": 1.2917538312591043e-05,
+      "loss": 0.853,
+      "step": 17980
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8883327300414092,
+      "learning_rate": 1.2916808460329206e-05,
+      "loss": 0.9773,
+      "step": 17981
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4894702623826055,
+      "learning_rate": 1.2916078591084788e-05,
+      "loss": 0.7936,
+      "step": 17982
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8502530661990233,
+      "learning_rate": 1.291534870486204e-05,
+      "loss": 0.9419,
+      "step": 17983
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.589594997373097,
+      "learning_rate": 1.2914618801665205e-05,
+      "loss": 0.9428,
+      "step": 17984
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6637693506385054,
+      "learning_rate": 1.2913888881498537e-05,
+      "loss": 0.9411,
+      "step": 17985
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7634616198070707,
+      "learning_rate": 1.2913158944366286e-05,
+      "loss": 0.757,
+      "step": 17986
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8052726314548393,
+      "learning_rate": 1.2912428990272704e-05,
+      "loss": 0.7273,
+      "step": 17987
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4716164083733876,
+      "learning_rate": 1.2911699019222035e-05,
+      "loss": 0.8894,
+      "step": 17988
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6479629921160959,
+      "learning_rate": 1.2910969031218532e-05,
+      "loss": 0.8706,
+      "step": 17989
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.558532268725037,
+      "learning_rate": 1.2910239026266446e-05,
+      "loss": 0.8666,
+      "step": 17990
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5890197161580146,
+      "learning_rate": 1.290950900437003e-05,
+      "loss": 0.8543,
+      "step": 17991
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.795167480246627,
+      "learning_rate": 1.2908778965533533e-05,
+      "loss": 0.9479,
+      "step": 17992
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7108570081925987,
+      "learning_rate": 1.2908048909761197e-05,
+      "loss": 0.7324,
+      "step": 17993
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7755562487466754,
+      "learning_rate": 1.2907318837057286e-05,
+      "loss": 0.8593,
+      "step": 17994
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8967792564901043,
+      "learning_rate": 1.2906588747426038e-05,
+      "loss": 0.8846,
+      "step": 17995
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7138322835306494,
+      "learning_rate": 1.2905858640871717e-05,
+      "loss": 0.8436,
+      "step": 17996
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7859339749691073,
+      "learning_rate": 1.2905128517398562e-05,
+      "loss": 0.7959,
+      "step": 17997
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8694706655509734,
+      "learning_rate": 1.2904398377010832e-05,
+      "loss": 0.8295,
+      "step": 17998
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5420075315994872,
+      "learning_rate": 1.2903668219712773e-05,
+      "loss": 0.8433,
+      "step": 17999
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4215775105692465,
+      "learning_rate": 1.2902938045508638e-05,
+      "loss": 0.8835,
+      "step": 18000
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8055259465506643,
+      "learning_rate": 1.2902207854402682e-05,
+      "loss": 0.7439,
+      "step": 18001
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4852246232829673,
+      "learning_rate": 1.2901477646399147e-05,
+      "loss": 0.8147,
+      "step": 18002
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3349772966396953,
+      "learning_rate": 1.2900747421502297e-05,
+      "loss": 0.8823,
+      "step": 18003
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4425551429016863,
+      "learning_rate": 1.2900017179716373e-05,
+      "loss": 0.8247,
+      "step": 18004
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.43857242945152,
+      "learning_rate": 1.2899286921045633e-05,
+      "loss": 0.8966,
+      "step": 18005
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.684041556738538,
+      "learning_rate": 1.2898556645494327e-05,
+      "loss": 0.8448,
+      "step": 18006
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7952601971716959,
+      "learning_rate": 1.2897826353066705e-05,
+      "loss": 0.7152,
+      "step": 18007
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7443155135748692,
+      "learning_rate": 1.289709604376702e-05,
+      "loss": 0.6838,
+      "step": 18008
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.552172892270324,
+      "learning_rate": 1.2896365717599526e-05,
+      "loss": 0.9384,
+      "step": 18009
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7280598157488485,
+      "learning_rate": 1.2895635374568474e-05,
+      "loss": 0.8703,
+      "step": 18010
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.839257921809358,
+      "learning_rate": 1.2894905014678117e-05,
+      "loss": 0.9789,
+      "step": 18011
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6047719248549774,
+      "learning_rate": 1.2894174637932703e-05,
+      "loss": 0.8265,
+      "step": 18012
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0409948131193785,
+      "learning_rate": 1.289344424433649e-05,
+      "loss": 0.8781,
+      "step": 18013
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5603665927627899,
+      "learning_rate": 1.2892713833893731e-05,
+      "loss": 0.9099,
+      "step": 18014
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9683047155569626,
+      "learning_rate": 1.2891983406608673e-05,
+      "loss": 0.7426,
+      "step": 18015
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.781639296828722,
+      "learning_rate": 1.2891252962485573e-05,
+      "loss": 1.0386,
+      "step": 18016
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5610170627702233,
+      "learning_rate": 1.2890522501528682e-05,
+      "loss": 0.8461,
+      "step": 18017
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7649616763405693,
+      "learning_rate": 1.2889792023742253e-05,
+      "loss": 0.9249,
+      "step": 18018
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4881274633029415,
+      "learning_rate": 1.2889061529130542e-05,
+      "loss": 0.9806,
+      "step": 18019
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6212294289458882,
+      "learning_rate": 1.2888331017697804e-05,
+      "loss": 0.8443,
+      "step": 18020
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.560233195478341,
+      "learning_rate": 1.2887600489448282e-05,
+      "loss": 0.9073,
+      "step": 18021
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6936105543544084,
+      "learning_rate": 1.2886869944386237e-05,
+      "loss": 0.8448,
+      "step": 18022
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4695713764263163,
+      "learning_rate": 1.2886139382515924e-05,
+      "loss": 0.7615,
+      "step": 18023
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.741043712686792,
+      "learning_rate": 1.2885408803841593e-05,
+      "loss": 0.7431,
+      "step": 18024
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6053659784414316,
+      "learning_rate": 1.2884678208367493e-05,
+      "loss": 0.8494,
+      "step": 18025
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7266793730450956,
+      "learning_rate": 1.288394759609789e-05,
+      "loss": 0.9802,
+      "step": 18026
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7623747506068665,
+      "learning_rate": 1.2883216967037029e-05,
+      "loss": 0.7887,
+      "step": 18027
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5233127805622761,
+      "learning_rate": 1.2882486321189165e-05,
+      "loss": 0.9162,
+      "step": 18028
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9147989482659775,
+      "learning_rate": 1.2881755658558555e-05,
+      "loss": 0.7957,
+      "step": 18029
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7297252049791776,
+      "learning_rate": 1.2881024979149448e-05,
+      "loss": 0.8999,
+      "step": 18030
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6870598035900282,
+      "learning_rate": 1.2880294282966104e-05,
+      "loss": 0.8482,
+      "step": 18031
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7315902415378194,
+      "learning_rate": 1.2879563570012773e-05,
+      "loss": 0.7299,
+      "step": 18032
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6201947482263783,
+      "learning_rate": 1.2878832840293713e-05,
+      "loss": 0.8913,
+      "step": 18033
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.468535555525804,
+      "learning_rate": 1.2878102093813176e-05,
+      "loss": 0.8842,
+      "step": 18034
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5323511334538154,
+      "learning_rate": 1.287737133057542e-05,
+      "loss": 0.9082,
+      "step": 18035
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8175439301937605,
+      "learning_rate": 1.2876640550584693e-05,
+      "loss": 0.7533,
+      "step": 18036
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5981899247415652,
+      "learning_rate": 1.2875909753845255e-05,
+      "loss": 0.7946,
+      "step": 18037
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8439638056438599,
+      "learning_rate": 1.2875178940361361e-05,
+      "loss": 0.926,
+      "step": 18038
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6312436622940352,
+      "learning_rate": 1.2874448110137263e-05,
+      "loss": 0.8463,
+      "step": 18039
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8958796441882346,
+      "learning_rate": 1.2873717263177221e-05,
+      "loss": 0.949,
+      "step": 18040
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9376244575609791,
+      "learning_rate": 1.2872986399485487e-05,
+      "loss": 0.9082,
+      "step": 18041
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6647793148076735,
+      "learning_rate": 1.2872255519066314e-05,
+      "loss": 0.7812,
+      "step": 18042
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6289484803082244,
+      "learning_rate": 1.2871524621923963e-05,
+      "loss": 0.7916,
+      "step": 18043
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.833820656775926,
+      "learning_rate": 1.2870793708062684e-05,
+      "loss": 0.9316,
+      "step": 18044
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.056852603187017,
+      "learning_rate": 1.2870062777486738e-05,
+      "loss": 1.0016,
+      "step": 18045
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7440982678265124,
+      "learning_rate": 1.2869331830200374e-05,
+      "loss": 0.8696,
+      "step": 18046
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6503742984487564,
+      "learning_rate": 1.2868600866207856e-05,
+      "loss": 0.9288,
+      "step": 18047
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.824840986982623,
+      "learning_rate": 1.2867869885513432e-05,
+      "loss": 0.817,
+      "step": 18048
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.902199290088178,
+      "learning_rate": 1.2867138888121362e-05,
+      "loss": 0.9692,
+      "step": 18049
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6449126198659303,
+      "learning_rate": 1.2866407874035904e-05,
+      "loss": 0.9453,
+      "step": 18050
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5435068776907277,
+      "learning_rate": 1.2865676843261308e-05,
+      "loss": 0.9702,
+      "step": 18051
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7073966959592066,
+      "learning_rate": 1.2864945795801837e-05,
+      "loss": 0.9467,
+      "step": 18052
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.772418445282735,
+      "learning_rate": 1.2864214731661743e-05,
+      "loss": 0.9556,
+      "step": 18053
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7695153382514344,
+      "learning_rate": 1.2863483650845288e-05,
+      "loss": 0.802,
+      "step": 18054
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.814701071641801,
+      "learning_rate": 1.2862752553356721e-05,
+      "loss": 0.7682,
+      "step": 18055
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6016042022284012,
+      "learning_rate": 1.2862021439200306e-05,
+      "loss": 0.8493,
+      "step": 18056
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7794176620705839,
+      "learning_rate": 1.2861290308380289e-05,
+      "loss": 0.7641,
+      "step": 18057
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4734055564951123,
+      "learning_rate": 1.286055916090094e-05,
+      "loss": 0.8404,
+      "step": 18058
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.761878885534444,
+      "learning_rate": 1.2859827996766508e-05,
+      "loss": 0.8201,
+      "step": 18059
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6379481629671562,
+      "learning_rate": 1.2859096815981254e-05,
+      "loss": 0.7794,
+      "step": 18060
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8792236439174106,
+      "learning_rate": 1.285836561854943e-05,
+      "loss": 0.9333,
+      "step": 18061
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8875119258630615,
+      "learning_rate": 1.2857634404475297e-05,
+      "loss": 0.865,
+      "step": 18062
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2977994291734136,
+      "learning_rate": 1.2856903173763114e-05,
+      "loss": 0.9012,
+      "step": 18063
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7698990779373114,
+      "learning_rate": 1.2856171926417134e-05,
+      "loss": 0.8592,
+      "step": 18064
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6179715569676494,
+      "learning_rate": 1.2855440662441617e-05,
+      "loss": 0.792,
+      "step": 18065
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5852305576711834,
+      "learning_rate": 1.285470938184082e-05,
+      "loss": 0.9301,
+      "step": 18066
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6497299504728822,
+      "learning_rate": 1.2853978084619004e-05,
+      "loss": 0.9902,
+      "step": 18067
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.15182444152932,
+      "learning_rate": 1.285324677078042e-05,
+      "loss": 0.8372,
+      "step": 18068
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2200435557698635,
+      "learning_rate": 1.2852515440329331e-05,
+      "loss": 0.8968,
+      "step": 18069
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8179612441339644,
+      "learning_rate": 1.2851784093269997e-05,
+      "loss": 0.7886,
+      "step": 18070
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5910344767274611,
+      "learning_rate": 1.2851052729606673e-05,
+      "loss": 0.7616,
+      "step": 18071
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4795297725706422,
+      "learning_rate": 1.2850321349343613e-05,
+      "loss": 0.9107,
+      "step": 18072
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9305953589897884,
+      "learning_rate": 1.2849589952485084e-05,
+      "loss": 0.9581,
+      "step": 18073
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.646489855277369,
+      "learning_rate": 1.2848858539035339e-05,
+      "loss": 0.7148,
+      "step": 18074
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8537410263260832,
+      "learning_rate": 1.2848127108998638e-05,
+      "loss": 0.8907,
+      "step": 18075
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.624396500229369,
+      "learning_rate": 1.2847395662379236e-05,
+      "loss": 0.8012,
+      "step": 18076
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6202391528925146,
+      "learning_rate": 1.28466641991814e-05,
+      "loss": 0.9972,
+      "step": 18077
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4399903879844338,
+      "learning_rate": 1.284593271940938e-05,
+      "loss": 0.8562,
+      "step": 18078
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7908613819868229,
+      "learning_rate": 1.2845201223067441e-05,
+      "loss": 0.7379,
+      "step": 18079
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7402257494138715,
+      "learning_rate": 1.2844469710159837e-05,
+      "loss": 0.8783,
+      "step": 18080
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.645591489191759,
+      "learning_rate": 1.2843738180690833e-05,
+      "loss": 0.8438,
+      "step": 18081
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5287000981006655,
+      "learning_rate": 1.2843006634664683e-05,
+      "loss": 0.8442,
+      "step": 18082
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.600636865385744,
+      "learning_rate": 1.2842275072085649e-05,
+      "loss": 0.8447,
+      "step": 18083
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.669992317803949,
+      "learning_rate": 1.2841543492957988e-05,
+      "loss": 0.8837,
+      "step": 18084
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2850819425260025,
+      "learning_rate": 1.2840811897285963e-05,
+      "loss": 1.0461,
+      "step": 18085
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6889676833342575,
+      "learning_rate": 1.284008028507383e-05,
+      "loss": 0.8742,
+      "step": 18086
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6744596985842233,
+      "learning_rate": 1.2839348656325851e-05,
+      "loss": 0.9221,
+      "step": 18087
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5630904095132445,
+      "learning_rate": 1.2838617011046288e-05,
+      "loss": 0.8073,
+      "step": 18088
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6608402435822691,
+      "learning_rate": 1.2837885349239396e-05,
+      "loss": 0.8641,
+      "step": 18089
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7618190303535289,
+      "learning_rate": 1.2837153670909437e-05,
+      "loss": 0.748,
+      "step": 18090
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6888517354190442,
+      "learning_rate": 1.2836421976060672e-05,
+      "loss": 0.8741,
+      "step": 18091
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8726294309840037,
+      "learning_rate": 1.283569026469736e-05,
+      "loss": 0.9172,
+      "step": 18092
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6208194770122293,
+      "learning_rate": 1.2834958536823758e-05,
+      "loss": 0.9113,
+      "step": 18093
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4751734087502377,
+      "learning_rate": 1.2834226792444136e-05,
+      "loss": 0.9139,
+      "step": 18094
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6962227163945105,
+      "learning_rate": 1.2833495031562744e-05,
+      "loss": 0.8553,
+      "step": 18095
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.793418223224445,
+      "learning_rate": 1.283276325418385e-05,
+      "loss": 0.9302,
+      "step": 18096
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5316995576726837,
+      "learning_rate": 1.2832031460311707e-05,
+      "loss": 0.7749,
+      "step": 18097
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6020560490130082,
+      "learning_rate": 1.2831299649950582e-05,
+      "loss": 0.933,
+      "step": 18098
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6896373099534576,
+      "learning_rate": 1.2830567823104736e-05,
+      "loss": 0.9507,
+      "step": 18099
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5634456145388407,
+      "learning_rate": 1.2829835979778424e-05,
+      "loss": 0.8822,
+      "step": 18100
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4856092610089244,
+      "learning_rate": 1.2829104119975917e-05,
+      "loss": 0.7532,
+      "step": 18101
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5343353235527946,
+      "learning_rate": 1.2828372243701463e-05,
+      "loss": 0.8034,
+      "step": 18102
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7805368523304437,
+      "learning_rate": 1.2827640350959338e-05,
+      "loss": 0.9087,
+      "step": 18103
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7850815719860258,
+      "learning_rate": 1.2826908441753789e-05,
+      "loss": 0.8643,
+      "step": 18104
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8071622731406827,
+      "learning_rate": 1.2826176516089088e-05,
+      "loss": 0.9204,
+      "step": 18105
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8776577774616248,
+      "learning_rate": 1.2825444573969494e-05,
+      "loss": 0.9249,
+      "step": 18106
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.629974884540499,
+      "learning_rate": 1.2824712615399264e-05,
+      "loss": 0.8727,
+      "step": 18107
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6103552856172385,
+      "learning_rate": 1.2823980640382663e-05,
+      "loss": 0.8855,
+      "step": 18108
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.795379288040788,
+      "learning_rate": 1.2823248648923954e-05,
+      "loss": 0.8854,
+      "step": 18109
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7091865490127347,
+      "learning_rate": 1.2822516641027396e-05,
+      "loss": 0.9556,
+      "step": 18110
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5619472835516137,
+      "learning_rate": 1.2821784616697253e-05,
+      "loss": 0.8634,
+      "step": 18111
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5789962467591983,
+      "learning_rate": 1.2821052575937786e-05,
+      "loss": 0.8538,
+      "step": 18112
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6314075669343238,
+      "learning_rate": 1.282032051875326e-05,
+      "loss": 0.8478,
+      "step": 18113
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4550691935965991,
+      "learning_rate": 1.2819588445147932e-05,
+      "loss": 0.907,
+      "step": 18114
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7706454850002213,
+      "learning_rate": 1.2818856355126071e-05,
+      "loss": 0.8121,
+      "step": 18115
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4726050167031846,
+      "learning_rate": 1.2818124248691932e-05,
+      "loss": 0.7651,
+      "step": 18116
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7273056473078865,
+      "learning_rate": 1.2817392125849785e-05,
+      "loss": 0.9351,
+      "step": 18117
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.649228843530059,
+      "learning_rate": 1.2816659986603887e-05,
+      "loss": 0.9909,
+      "step": 18118
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8648459447449823,
+      "learning_rate": 1.2815927830958504e-05,
+      "loss": 0.8644,
+      "step": 18119
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.558075905050804,
+      "learning_rate": 1.2815195658917896e-05,
+      "loss": 0.819,
+      "step": 18120
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7672087591352212,
+      "learning_rate": 1.2814463470486329e-05,
+      "loss": 0.7288,
+      "step": 18121
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8246558890622417,
+      "learning_rate": 1.2813731265668066e-05,
+      "loss": 0.8812,
+      "step": 18122
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4841736822873384,
+      "learning_rate": 1.2812999044467364e-05,
+      "loss": 0.8636,
+      "step": 18123
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.761500800791249,
+      "learning_rate": 1.2812266806888496e-05,
+      "loss": 0.9021,
+      "step": 18124
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6399684133212866,
+      "learning_rate": 1.2811534552935716e-05,
+      "loss": 0.9082,
+      "step": 18125
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4876677809357306,
+      "learning_rate": 1.2810802282613297e-05,
+      "loss": 0.9597,
+      "step": 18126
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6838257375596344,
+      "learning_rate": 1.2810069995925491e-05,
+      "loss": 0.9058,
+      "step": 18127
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4988971332372507,
+      "learning_rate": 1.2809337692876573e-05,
+      "loss": 0.8123,
+      "step": 18128
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6713728469172124,
+      "learning_rate": 1.2808605373470799e-05,
+      "loss": 0.8805,
+      "step": 18129
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.402145735984714,
+      "learning_rate": 1.2807873037712437e-05,
+      "loss": 0.8281,
+      "step": 18130
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.659300387672471,
+      "learning_rate": 1.2807140685605748e-05,
+      "loss": 0.8867,
+      "step": 18131
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.777069073503282,
+      "learning_rate": 1.2806408317154997e-05,
+      "loss": 0.9128,
+      "step": 18132
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.152864503545277,
+      "learning_rate": 1.2805675932364449e-05,
+      "loss": 0.8068,
+      "step": 18133
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.941175655527147,
+      "learning_rate": 1.2804943531238366e-05,
+      "loss": 0.8752,
+      "step": 18134
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5208154072944957,
+      "learning_rate": 1.2804211113781016e-05,
+      "loss": 0.8835,
+      "step": 18135
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.699514220619898,
+      "learning_rate": 1.280347867999666e-05,
+      "loss": 0.953,
+      "step": 18136
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5970270417728745,
+      "learning_rate": 1.2802746229889563e-05,
+      "loss": 0.9447,
+      "step": 18137
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7115162567536049,
+      "learning_rate": 1.280201376346399e-05,
+      "loss": 0.7913,
+      "step": 18138
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6212257372815142,
+      "learning_rate": 1.2801281280724207e-05,
+      "loss": 0.956,
+      "step": 18139
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4346824909532296,
+      "learning_rate": 1.2800548781674476e-05,
+      "loss": 0.8476,
+      "step": 18140
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5197894002084873,
+      "learning_rate": 1.2799816266319062e-05,
+      "loss": 1.0197,
+      "step": 18141
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7989725380630084,
+      "learning_rate": 1.2799083734662237e-05,
+      "loss": 0.7641,
+      "step": 18142
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5707448313356347,
+      "learning_rate": 1.2798351186708258e-05,
+      "loss": 0.8534,
+      "step": 18143
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2517544645926586,
+      "learning_rate": 1.2797618622461387e-05,
+      "loss": 0.9249,
+      "step": 18144
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.54735431300082,
+      "learning_rate": 1.2796886041925903e-05,
+      "loss": 1.0007,
+      "step": 18145
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5023096608428228,
+      "learning_rate": 1.2796153445106057e-05,
+      "loss": 0.898,
+      "step": 18146
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7501039862937346,
+      "learning_rate": 1.279542083200612e-05,
+      "loss": 0.9025,
+      "step": 18147
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5480145568648276,
+      "learning_rate": 1.2794688202630361e-05,
+      "loss": 0.8714,
+      "step": 18148
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0782452633859676,
+      "learning_rate": 1.2793955556983042e-05,
+      "loss": 0.7495,
+      "step": 18149
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6147797905517896,
+      "learning_rate": 1.2793222895068431e-05,
+      "loss": 0.7818,
+      "step": 18150
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6753264283718075,
+      "learning_rate": 1.2792490216890786e-05,
+      "loss": 0.9327,
+      "step": 18151
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.654996847538567,
+      "learning_rate": 1.2791757522454387e-05,
+      "loss": 1.0614,
+      "step": 18152
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.477233808483804,
+      "learning_rate": 1.2791024811763484e-05,
+      "loss": 0.7607,
+      "step": 18153
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1181199360952943,
+      "learning_rate": 1.2790292084822355e-05,
+      "loss": 0.9754,
+      "step": 18154
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.742249339372961,
+      "learning_rate": 1.278955934163526e-05,
+      "loss": 0.7594,
+      "step": 18155
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4562011965011135,
+      "learning_rate": 1.2788826582206473e-05,
+      "loss": 0.8174,
+      "step": 18156
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7748224140565039,
+      "learning_rate": 1.2788093806540249e-05,
+      "loss": 0.8695,
+      "step": 18157
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6927762375295463,
+      "learning_rate": 1.2787361014640862e-05,
+      "loss": 0.9104,
+      "step": 18158
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7228835216255594,
+      "learning_rate": 1.2786628206512577e-05,
+      "loss": 0.8516,
+      "step": 18159
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9957317732372257,
+      "learning_rate": 1.2785895382159658e-05,
+      "loss": 0.8908,
+      "step": 18160
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6041813419492792,
+      "learning_rate": 1.2785162541586378e-05,
+      "loss": 0.9625,
+      "step": 18161
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7135663046744865,
+      "learning_rate": 1.2784429684796997e-05,
+      "loss": 0.7601,
+      "step": 18162
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.525751682051214,
+      "learning_rate": 1.2783696811795787e-05,
+      "loss": 0.7542,
+      "step": 18163
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7087073652184472,
+      "learning_rate": 1.278296392258701e-05,
+      "loss": 0.8144,
+      "step": 18164
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4822119953984223,
+      "learning_rate": 1.2782231017174937e-05,
+      "loss": 0.9857,
+      "step": 18165
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8321725605860426,
+      "learning_rate": 1.2781498095563835e-05,
+      "loss": 0.8747,
+      "step": 18166
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6061229405179307,
+      "learning_rate": 1.2780765157757971e-05,
+      "loss": 0.7994,
+      "step": 18167
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6369429822563504,
+      "learning_rate": 1.278003220376161e-05,
+      "loss": 0.8494,
+      "step": 18168
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4057120237049985,
+      "learning_rate": 1.2779299233579024e-05,
+      "loss": 0.8346,
+      "step": 18169
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.706315554117808,
+      "learning_rate": 1.2778566247214474e-05,
+      "loss": 0.7249,
+      "step": 18170
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5026931239184218,
+      "learning_rate": 1.2777833244672235e-05,
+      "loss": 0.8524,
+      "step": 18171
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5665515274665887,
+      "learning_rate": 1.277710022595657e-05,
+      "loss": 0.8668,
+      "step": 18172
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6143431652306777,
+      "learning_rate": 1.2776367191071752e-05,
+      "loss": 0.8658,
+      "step": 18173
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6928468364112867,
+      "learning_rate": 1.2775634140022039e-05,
+      "loss": 0.9622,
+      "step": 18174
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7309920094410165,
+      "learning_rate": 1.2774901072811709e-05,
+      "loss": 0.9464,
+      "step": 18175
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7228296723039712,
+      "learning_rate": 1.2774167989445025e-05,
+      "loss": 0.7143,
+      "step": 18176
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7914600883978439,
+      "learning_rate": 1.2773434889926259e-05,
+      "loss": 0.7434,
+      "step": 18177
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.469101411616545,
+      "learning_rate": 1.2772701774259674e-05,
+      "loss": 0.8824,
+      "step": 18178
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6815491327389012,
+      "learning_rate": 1.2771968642449544e-05,
+      "loss": 0.9542,
+      "step": 18179
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.840430868800679,
+      "learning_rate": 1.2771235494500133e-05,
+      "loss": 0.9231,
+      "step": 18180
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.797659417809228,
+      "learning_rate": 1.2770502330415712e-05,
+      "loss": 0.9257,
+      "step": 18181
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8861152611026761,
+      "learning_rate": 1.276976915020055e-05,
+      "loss": 0.8665,
+      "step": 18182
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7419351317927154,
+      "learning_rate": 1.2769035953858913e-05,
+      "loss": 0.88,
+      "step": 18183
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8119727257781328,
+      "learning_rate": 1.2768302741395076e-05,
+      "loss": 0.8428,
+      "step": 18184
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8600617855144976,
+      "learning_rate": 1.27675695128133e-05,
+      "loss": 0.9539,
+      "step": 18185
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7448619686510697,
+      "learning_rate": 1.2766836268117862e-05,
+      "loss": 0.7203,
+      "step": 18186
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1318748643155745,
+      "learning_rate": 1.2766103007313024e-05,
+      "loss": 1.0075,
+      "step": 18187
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6848807938182495,
+      "learning_rate": 1.2765369730403061e-05,
+      "loss": 0.8453,
+      "step": 18188
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4677364834835585,
+      "learning_rate": 1.2764636437392239e-05,
+      "loss": 0.9466,
+      "step": 18189
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7777073585828067,
+      "learning_rate": 1.2763903128284827e-05,
+      "loss": 0.8818,
+      "step": 18190
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6225557079893642,
+      "learning_rate": 1.2763169803085095e-05,
+      "loss": 0.7226,
+      "step": 18191
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9143728118418935,
+      "learning_rate": 1.2762436461797318e-05,
+      "loss": 0.8431,
+      "step": 18192
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6078722729703274,
+      "learning_rate": 1.2761703104425757e-05,
+      "loss": 0.8125,
+      "step": 18193
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5553524028388597,
+      "learning_rate": 1.2760969730974692e-05,
+      "loss": 0.8071,
+      "step": 18194
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5356359222872786,
+      "learning_rate": 1.276023634144838e-05,
+      "loss": 0.946,
+      "step": 18195
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6336616114331317,
+      "learning_rate": 1.2759502935851102e-05,
+      "loss": 0.7751,
+      "step": 18196
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5771715862609637,
+      "learning_rate": 1.2758769514187123e-05,
+      "loss": 0.8322,
+      "step": 18197
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5450159755103106,
+      "learning_rate": 1.2758036076460716e-05,
+      "loss": 0.9468,
+      "step": 18198
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.098676687272786,
+      "learning_rate": 1.2757302622676149e-05,
+      "loss": 0.9592,
+      "step": 18199
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6998545345307672,
+      "learning_rate": 1.2756569152837693e-05,
+      "loss": 0.837,
+      "step": 18200
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5214280555258104,
+      "learning_rate": 1.275583566694962e-05,
+      "loss": 0.9497,
+      "step": 18201
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7524770041272966,
+      "learning_rate": 1.2755102165016195e-05,
+      "loss": 0.7623,
+      "step": 18202
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4049855206983655,
+      "learning_rate": 1.2754368647041699e-05,
+      "loss": 0.8312,
+      "step": 18203
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6781483124716687,
+      "learning_rate": 1.275363511303039e-05,
+      "loss": 0.9851,
+      "step": 18204
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6140064610763196,
+      "learning_rate": 1.2752901562986553e-05,
+      "loss": 0.8759,
+      "step": 18205
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7789784459807861,
+      "learning_rate": 1.2752167996914446e-05,
+      "loss": 0.7198,
+      "step": 18206
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.757623282454152,
+      "learning_rate": 1.2751434414818349e-05,
+      "loss": 0.8289,
+      "step": 18207
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9836682060493531,
+      "learning_rate": 1.2750700816702528e-05,
+      "loss": 0.7998,
+      "step": 18208
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.717206125112564,
+      "learning_rate": 1.2749967202571256e-05,
+      "loss": 0.847,
+      "step": 18209
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4760794546430058,
+      "learning_rate": 1.2749233572428805e-05,
+      "loss": 0.9429,
+      "step": 18210
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7536627131638243,
+      "learning_rate": 1.2748499926279445e-05,
+      "loss": 0.7674,
+      "step": 18211
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6138329543052796,
+      "learning_rate": 1.2747766264127449e-05,
+      "loss": 0.805,
+      "step": 18212
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5541463420888835,
+      "learning_rate": 1.2747032585977089e-05,
+      "loss": 0.7353,
+      "step": 18213
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6527580652492166,
+      "learning_rate": 1.2746298891832631e-05,
+      "loss": 0.8911,
+      "step": 18214
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0227542746609437,
+      "learning_rate": 1.2745565181698357e-05,
+      "loss": 0.9875,
+      "step": 18215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6556528596182907,
+      "learning_rate": 1.2744831455578528e-05,
+      "loss": 1.0201,
+      "step": 18216
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6093655979749477,
+      "learning_rate": 1.2744097713477426e-05,
+      "loss": 0.8604,
+      "step": 18217
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4975112722393562,
+      "learning_rate": 1.2743363955399315e-05,
+      "loss": 0.8867,
+      "step": 18218
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7199141493251946,
+      "learning_rate": 1.2742630181348473e-05,
+      "loss": 0.8602,
+      "step": 18219
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6247443822294851,
+      "learning_rate": 1.2741896391329167e-05,
+      "loss": 0.8475,
+      "step": 18220
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8029652192446164,
+      "learning_rate": 1.2741162585345674e-05,
+      "loss": 0.9346,
+      "step": 18221
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3949222846805474,
+      "learning_rate": 1.2740428763402266e-05,
+      "loss": 0.8944,
+      "step": 18222
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.401799770850324,
+      "learning_rate": 1.2739694925503209e-05,
+      "loss": 0.8456,
+      "step": 18223
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7340571066669088,
+      "learning_rate": 1.2738961071652787e-05,
+      "loss": 0.7452,
+      "step": 18224
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.582955304276001,
+      "learning_rate": 1.2738227201855263e-05,
+      "loss": 0.916,
+      "step": 18225
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8653079293325432,
+      "learning_rate": 1.2737493316114916e-05,
+      "loss": 0.8795,
+      "step": 18226
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.670197938142245,
+      "learning_rate": 1.273675941443601e-05,
+      "loss": 0.8547,
+      "step": 18227
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9428713733630232,
+      "learning_rate": 1.2736025496822833e-05,
+      "loss": 0.871,
+      "step": 18228
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5464558859373176,
+      "learning_rate": 1.2735291563279644e-05,
+      "loss": 0.9136,
+      "step": 18229
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.636665171624559,
+      "learning_rate": 1.2734557613810723e-05,
+      "loss": 0.8406,
+      "step": 18230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.589224262249379,
+      "learning_rate": 1.2733823648420341e-05,
+      "loss": 0.8339,
+      "step": 18231
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3072549440506114,
+      "learning_rate": 1.273308966711277e-05,
+      "loss": 0.8412,
+      "step": 18232
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.467522247963152,
+      "learning_rate": 1.2732355669892289e-05,
+      "loss": 0.8286,
+      "step": 18233
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0972225184890014,
+      "learning_rate": 1.2731621656763165e-05,
+      "loss": 0.815,
+      "step": 18234
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5999794689523383,
+      "learning_rate": 1.273088762772968e-05,
+      "loss": 0.9498,
+      "step": 18235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8817226535296265,
+      "learning_rate": 1.2730153582796097e-05,
+      "loss": 0.8538,
+      "step": 18236
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5713764975806979,
+      "learning_rate": 1.2729419521966702e-05,
+      "loss": 0.8838,
+      "step": 18237
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8081538747172774,
+      "learning_rate": 1.2728685445245757e-05,
+      "loss": 0.6582,
+      "step": 18238
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7967898516401418,
+      "learning_rate": 1.2727951352637544e-05,
+      "loss": 0.7633,
+      "step": 18239
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.437750313401959,
+      "learning_rate": 1.2727217244146334e-05,
+      "loss": 0.8343,
+      "step": 18240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.892509224862396,
+      "learning_rate": 1.2726483119776401e-05,
+      "loss": 0.8533,
+      "step": 18241
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5445846682497715,
+      "learning_rate": 1.272574897953202e-05,
+      "loss": 0.9288,
+      "step": 18242
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6439086247653976,
+      "learning_rate": 1.2725014823417468e-05,
+      "loss": 0.8729,
+      "step": 18243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5339204096301482,
+      "learning_rate": 1.2724280651437014e-05,
+      "loss": 0.8754,
+      "step": 18244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.593113968077865,
+      "learning_rate": 1.2723546463594936e-05,
+      "loss": 0.8095,
+      "step": 18245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.514302990035699,
+      "learning_rate": 1.272281225989551e-05,
+      "loss": 0.9097,
+      "step": 18246
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.082129783190385,
+      "learning_rate": 1.2722078040343007e-05,
+      "loss": 0.9407,
+      "step": 18247
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.027370497206923,
+      "learning_rate": 1.2721343804941704e-05,
+      "loss": 0.8414,
+      "step": 18248
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.586016693070012,
+      "learning_rate": 1.2720609553695877e-05,
+      "loss": 0.888,
+      "step": 18249
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.461488308017603,
+      "learning_rate": 1.27198752866098e-05,
+      "loss": 0.9228,
+      "step": 18250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8251658340497834,
+      "learning_rate": 1.2719141003687746e-05,
+      "loss": 0.9134,
+      "step": 18251
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5830222059267092,
+      "learning_rate": 1.2718406704933997e-05,
+      "loss": 0.9205,
+      "step": 18252
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7565076070867445,
+      "learning_rate": 1.2717672390352818e-05,
+      "loss": 0.8154,
+      "step": 18253
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9990718409812056,
+      "learning_rate": 1.2716938059948493e-05,
+      "loss": 0.7767,
+      "step": 18254
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5544356291712393,
+      "learning_rate": 1.2716203713725293e-05,
+      "loss": 0.8662,
+      "step": 18255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.730074475859568,
+      "learning_rate": 1.2715469351687498e-05,
+      "loss": 0.868,
+      "step": 18256
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5871430218107996,
+      "learning_rate": 1.2714734973839375e-05,
+      "loss": 0.8968,
+      "step": 18257
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.633934373355433,
+      "learning_rate": 1.2714000580185211e-05,
+      "loss": 0.8021,
+      "step": 18258
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6893162067903937,
+      "learning_rate": 1.2713266170729275e-05,
+      "loss": 0.9449,
+      "step": 18259
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6283138152475944,
+      "learning_rate": 1.2712531745475842e-05,
+      "loss": 0.8196,
+      "step": 18260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7791880025362963,
+      "learning_rate": 1.2711797304429192e-05,
+      "loss": 0.7448,
+      "step": 18261
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6221236118847422,
+      "learning_rate": 1.27110628475936e-05,
+      "loss": 0.867,
+      "step": 18262
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.83734008063034,
+      "learning_rate": 1.2710328374973342e-05,
+      "loss": 0.8547,
+      "step": 18263
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6886716535376387,
+      "learning_rate": 1.2709593886572693e-05,
+      "loss": 0.8571,
+      "step": 18264
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.522990587595425,
+      "learning_rate": 1.270885938239593e-05,
+      "loss": 0.9048,
+      "step": 18265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.550111495213649,
+      "learning_rate": 1.2708124862447331e-05,
+      "loss": 0.8717,
+      "step": 18266
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5552162385413584,
+      "learning_rate": 1.2707390326731173e-05,
+      "loss": 0.886,
+      "step": 18267
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5959318394635378,
+      "learning_rate": 1.2706655775251728e-05,
+      "loss": 0.7029,
+      "step": 18268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6909775364072028,
+      "learning_rate": 1.270592120801328e-05,
+      "loss": 0.89,
+      "step": 18269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7715311857140368,
+      "learning_rate": 1.2705186625020098e-05,
+      "loss": 0.9383,
+      "step": 18270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7515975916306777,
+      "learning_rate": 1.2704452026276468e-05,
+      "loss": 0.7186,
+      "step": 18271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7964272144247562,
+      "learning_rate": 1.2703717411786658e-05,
+      "loss": 0.8956,
+      "step": 18272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5056749087210046,
+      "learning_rate": 1.2702982781554953e-05,
+      "loss": 0.815,
+      "step": 18273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6647294462968187,
+      "learning_rate": 1.2702248135585622e-05,
+      "loss": 0.9697,
+      "step": 18274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.130095038735106,
+      "learning_rate": 1.2701513473882951e-05,
+      "loss": 0.8221,
+      "step": 18275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7354833892317605,
+      "learning_rate": 1.270077879645121e-05,
+      "loss": 0.7258,
+      "step": 18276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.810752910746951,
+      "learning_rate": 1.2700044103294683e-05,
+      "loss": 0.8213,
+      "step": 18277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7686647099316647,
+      "learning_rate": 1.2699309394417641e-05,
+      "loss": 0.7478,
+      "step": 18278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6013680455049604,
+      "learning_rate": 1.2698574669824366e-05,
+      "loss": 0.837,
+      "step": 18279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.561659267521168,
+      "learning_rate": 1.2697839929519136e-05,
+      "loss": 0.8003,
+      "step": 18280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6253780111549905,
+      "learning_rate": 1.2697105173506225e-05,
+      "loss": 0.8482,
+      "step": 18281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5128138546828265,
+      "learning_rate": 1.269637040178992e-05,
+      "loss": 0.9045,
+      "step": 18282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5675624649754847,
+      "learning_rate": 1.2695635614374484e-05,
+      "loss": 0.8586,
+      "step": 18283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2074808937862653,
+      "learning_rate": 1.2694900811264212e-05,
+      "loss": 0.8402,
+      "step": 18284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6683643292529846,
+      "learning_rate": 1.269416599246337e-05,
+      "loss": 0.887,
+      "step": 18285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5829124578054368,
+      "learning_rate": 1.2693431157976243e-05,
+      "loss": 0.8647,
+      "step": 18286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6643428247014922,
+      "learning_rate": 1.2692696307807106e-05,
+      "loss": 0.9618,
+      "step": 18287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9542482938115338,
+      "learning_rate": 1.2691961441960238e-05,
+      "loss": 0.9009,
+      "step": 18288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3971596422534325,
+      "learning_rate": 1.269122656043992e-05,
+      "loss": 0.8628,
+      "step": 18289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7616178231962583,
+      "learning_rate": 1.2690491663250428e-05,
+      "loss": 0.9354,
+      "step": 18290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4261030983108427,
+      "learning_rate": 1.268975675039604e-05,
+      "loss": 0.9099,
+      "step": 18291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2152663001328516,
+      "learning_rate": 1.268902182188104e-05,
+      "loss": 0.9796,
+      "step": 18292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6406699901752875,
+      "learning_rate": 1.26882868777097e-05,
+      "loss": 0.8854,
+      "step": 18293
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4792411590843308,
+      "learning_rate": 1.2687551917886304e-05,
+      "loss": 0.8645,
+      "step": 18294
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8467547614755437,
+      "learning_rate": 1.2686816942415133e-05,
+      "loss": 0.7932,
+      "step": 18295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7831929916266784,
+      "learning_rate": 1.268608195130046e-05,
+      "loss": 0.8761,
+      "step": 18296
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7925334160861747,
+      "learning_rate": 1.2685346944546566e-05,
+      "loss": 0.7094,
+      "step": 18297
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7529957837920076,
+      "learning_rate": 1.2684611922157735e-05,
+      "loss": 0.7194,
+      "step": 18298
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5901789398466364,
+      "learning_rate": 1.2683876884138242e-05,
+      "loss": 1.0604,
+      "step": 18299
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.50532564805071,
+      "learning_rate": 1.2683141830492368e-05,
+      "loss": 0.8358,
+      "step": 18300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.749218778901198,
+      "learning_rate": 1.2682406761224394e-05,
+      "loss": 0.8873,
+      "step": 18301
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6093030759129285,
+      "learning_rate": 1.2681671676338597e-05,
+      "loss": 1.0071,
+      "step": 18302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5843044608140133,
+      "learning_rate": 1.268093657583926e-05,
+      "loss": 0.8145,
+      "step": 18303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6093570766605407,
+      "learning_rate": 1.2680201459730661e-05,
+      "loss": 0.8502,
+      "step": 18304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6928200197588688,
+      "learning_rate": 1.267946632801708e-05,
+      "loss": 0.8591,
+      "step": 18305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.734824388020016,
+      "learning_rate": 1.2678731180702796e-05,
+      "loss": 0.8035,
+      "step": 18306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.630623600243373,
+      "learning_rate": 1.2677996017792095e-05,
+      "loss": 0.9171,
+      "step": 18307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5829709454947958,
+      "learning_rate": 1.2677260839289249e-05,
+      "loss": 0.8263,
+      "step": 18308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3922661197968618,
+      "learning_rate": 1.267652564519855e-05,
+      "loss": 0.8402,
+      "step": 18309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.578979259685592,
+      "learning_rate": 1.2675790435524264e-05,
+      "loss": 0.9364,
+      "step": 18310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5245336074710443,
+      "learning_rate": 1.2675055210270681e-05,
+      "loss": 1.0457,
+      "step": 18311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8226800891124676,
+      "learning_rate": 1.2674319969442082e-05,
+      "loss": 0.6931,
+      "step": 18312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.736890418782748,
+      "learning_rate": 1.2673584713042743e-05,
+      "loss": 0.7413,
+      "step": 18313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7373319762118843,
+      "learning_rate": 1.2672849441076948e-05,
+      "loss": 0.7139,
+      "step": 18314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5235825711610727,
+      "learning_rate": 1.2672114153548977e-05,
+      "loss": 0.8263,
+      "step": 18315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.990503193520299,
+      "learning_rate": 1.2671378850463112e-05,
+      "loss": 0.8414,
+      "step": 18316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5429132507776004,
+      "learning_rate": 1.2670643531823634e-05,
+      "loss": 0.7793,
+      "step": 18317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.50334475757701,
+      "learning_rate": 1.2669908197634822e-05,
+      "loss": 0.7748,
+      "step": 18318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2038235967619606,
+      "learning_rate": 1.2669172847900958e-05,
+      "loss": 0.8315,
+      "step": 18319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5442917152137328,
+      "learning_rate": 1.266843748262633e-05,
+      "loss": 0.9895,
+      "step": 18320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6768810939791388,
+      "learning_rate": 1.266770210181521e-05,
+      "loss": 0.8662,
+      "step": 18321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4471275635000282,
+      "learning_rate": 1.2666966705471886e-05,
+      "loss": 0.7985,
+      "step": 18322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6545651102732042,
+      "learning_rate": 1.2666231293600634e-05,
+      "loss": 0.7716,
+      "step": 18323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.860031091081099,
+      "learning_rate": 1.2665495866205745e-05,
+      "loss": 0.8728,
+      "step": 18324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6167787723596065,
+      "learning_rate": 1.2664760423291492e-05,
+      "loss": 0.9054,
+      "step": 18325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7524906441764396,
+      "learning_rate": 1.266402496486216e-05,
+      "loss": 0.7894,
+      "step": 18326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6983381998139613,
+      "learning_rate": 1.2663289490922029e-05,
+      "loss": 0.9742,
+      "step": 18327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7680410875756916,
+      "learning_rate": 1.266255400147539e-05,
+      "loss": 0.7581,
+      "step": 18328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8029960918796606,
+      "learning_rate": 1.266181849652651e-05,
+      "loss": 0.9451,
+      "step": 18329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7447495169257496,
+      "learning_rate": 1.2661082976079684e-05,
+      "loss": 0.9029,
+      "step": 18330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6174771419818672,
+      "learning_rate": 1.2660347440139194e-05,
+      "loss": 0.7471,
+      "step": 18331
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5014468450165281,
+      "learning_rate": 1.2659611888709313e-05,
+      "loss": 0.8481,
+      "step": 18332
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5705227836543882,
+      "learning_rate": 1.2658876321794333e-05,
+      "loss": 0.9058,
+      "step": 18333
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7928308638800514,
+      "learning_rate": 1.2658140739398532e-05,
+      "loss": 0.7451,
+      "step": 18334
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8037173052462647,
+      "learning_rate": 1.2657405141526196e-05,
+      "loss": 0.894,
+      "step": 18335
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6233131272288361,
+      "learning_rate": 1.2656669528181603e-05,
+      "loss": 0.9728,
+      "step": 18336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7264120761872406,
+      "learning_rate": 1.2655933899369042e-05,
+      "loss": 0.9544,
+      "step": 18337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4123814401396944,
+      "learning_rate": 1.2655198255092791e-05,
+      "loss": 0.9729,
+      "step": 18338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3868207786108482,
+      "learning_rate": 1.2654462595357136e-05,
+      "loss": 0.9497,
+      "step": 18339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8367815306730597,
+      "learning_rate": 1.2653726920166358e-05,
+      "loss": 0.9351,
+      "step": 18340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.708905723005073,
+      "learning_rate": 1.2652991229524742e-05,
+      "loss": 0.9233,
+      "step": 18341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.986120208663105,
+      "learning_rate": 1.2652255523436572e-05,
+      "loss": 0.9399,
+      "step": 18342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.778590818342607,
+      "learning_rate": 1.2651519801906131e-05,
+      "loss": 0.8126,
+      "step": 18343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.228713521995806,
+      "learning_rate": 1.2650784064937702e-05,
+      "loss": 0.9472,
+      "step": 18344
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6258083112007804,
+      "learning_rate": 1.265004831253557e-05,
+      "loss": 1.0165,
+      "step": 18345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6174776863165647,
+      "learning_rate": 1.2649312544704016e-05,
+      "loss": 0.7637,
+      "step": 18346
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6168933556236438,
+      "learning_rate": 1.2648576761447324e-05,
+      "loss": 0.8597,
+      "step": 18347
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.830665902951621,
+      "learning_rate": 1.2647840962769781e-05,
+      "loss": 0.792,
+      "step": 18348
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.677477106311701,
+      "learning_rate": 1.2647105148675671e-05,
+      "loss": 0.8523,
+      "step": 18349
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.72754656451501,
+      "learning_rate": 1.2646369319169275e-05,
+      "loss": 0.8516,
+      "step": 18350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2765414290653316,
+      "learning_rate": 1.2645633474254882e-05,
+      "loss": 0.9768,
+      "step": 18351
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8044352445444467,
+      "learning_rate": 1.2644897613936773e-05,
+      "loss": 0.845,
+      "step": 18352
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.573829452640754,
+      "learning_rate": 1.2644161738219229e-05,
+      "loss": 0.8738,
+      "step": 18353
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.850483585125042,
+      "learning_rate": 1.2643425847106543e-05,
+      "loss": 0.8886,
+      "step": 18354
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.696717961905162,
+      "learning_rate": 1.264268994060299e-05,
+      "loss": 0.8722,
+      "step": 18355
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6575329942955195,
+      "learning_rate": 1.2641954018712863e-05,
+      "loss": 0.9613,
+      "step": 18356
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8382926838375755,
+      "learning_rate": 1.264121808144044e-05,
+      "loss": 0.9427,
+      "step": 18357
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0641980035840586,
+      "learning_rate": 1.2640482128790012e-05,
+      "loss": 0.8542,
+      "step": 18358
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6415560235297169,
+      "learning_rate": 1.2639746160765858e-05,
+      "loss": 0.9524,
+      "step": 18359
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9568756124105253,
+      "learning_rate": 1.2639010177372271e-05,
+      "loss": 0.893,
+      "step": 18360
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.768080229551764,
+      "learning_rate": 1.2638274178613527e-05,
+      "loss": 0.7889,
+      "step": 18361
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.648969357988285,
+      "learning_rate": 1.2637538164493916e-05,
+      "loss": 0.9198,
+      "step": 18362
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8170998066921881,
+      "learning_rate": 1.2636802135017726e-05,
+      "loss": 0.696,
+      "step": 18363
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.514261466437139,
+      "learning_rate": 1.2636066090189237e-05,
+      "loss": 0.9503,
+      "step": 18364
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6255149412274994,
+      "learning_rate": 1.2635330030012735e-05,
+      "loss": 0.9153,
+      "step": 18365
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6466995507648687,
+      "learning_rate": 1.2634593954492509e-05,
+      "loss": 0.6973,
+      "step": 18366
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6724087460108648,
+      "learning_rate": 1.2633857863632843e-05,
+      "loss": 0.9353,
+      "step": 18367
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8371360601788982,
+      "learning_rate": 1.2633121757438022e-05,
+      "loss": 0.9409,
+      "step": 18368
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.180170577871771,
+      "learning_rate": 1.2632385635912332e-05,
+      "loss": 0.9082,
+      "step": 18369
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7526343366954316,
+      "learning_rate": 1.2631649499060061e-05,
+      "loss": 0.8251,
+      "step": 18370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4814577521571672,
+      "learning_rate": 1.2630913346885491e-05,
+      "loss": 0.8152,
+      "step": 18371
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.900425136311973,
+      "learning_rate": 1.2630177179392913e-05,
+      "loss": 0.8309,
+      "step": 18372
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7851890297103643,
+      "learning_rate": 1.262944099658661e-05,
+      "loss": 1.0006,
+      "step": 18373
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0825422699414715,
+      "learning_rate": 1.262870479847087e-05,
+      "loss": 0.9704,
+      "step": 18374
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5587926063815762,
+      "learning_rate": 1.2627968585049979e-05,
+      "loss": 0.9821,
+      "step": 18375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7821864511918716,
+      "learning_rate": 1.262723235632822e-05,
+      "loss": 0.7067,
+      "step": 18376
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.817433156236562,
+      "learning_rate": 1.2626496112309885e-05,
+      "loss": 0.8483,
+      "step": 18377
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8586542004698283,
+      "learning_rate": 1.2625759852999256e-05,
+      "loss": 0.9791,
+      "step": 18378
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6487065868998478,
+      "learning_rate": 1.2625023578400622e-05,
+      "loss": 0.7198,
+      "step": 18379
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8293391199698592,
+      "learning_rate": 1.2624287288518272e-05,
+      "loss": 0.9063,
+      "step": 18380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7308470106267058,
+      "learning_rate": 1.2623550983356487e-05,
+      "loss": 0.9488,
+      "step": 18381
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7145238872794069,
+      "learning_rate": 1.2622814662919562e-05,
+      "loss": 0.7541,
+      "step": 18382
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7675675880035076,
+      "learning_rate": 1.2622078327211776e-05,
+      "loss": 0.8789,
+      "step": 18383
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.653259422928094,
+      "learning_rate": 1.2621341976237423e-05,
+      "loss": 0.9619,
+      "step": 18384
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5853998792508799,
+      "learning_rate": 1.2620605610000782e-05,
+      "loss": 0.9488,
+      "step": 18385
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7465483094592076,
+      "learning_rate": 1.2619869228506151e-05,
+      "loss": 0.882,
+      "step": 18386
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6597473417910427,
+      "learning_rate": 1.2619132831757808e-05,
+      "loss": 0.8819,
+      "step": 18387
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5755855270199663,
+      "learning_rate": 1.2618396419760049e-05,
+      "loss": 0.8881,
+      "step": 18388
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.699374494766327,
+      "learning_rate": 1.2617659992517153e-05,
+      "loss": 0.9706,
+      "step": 18389
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5666577233074102,
+      "learning_rate": 1.2616923550033413e-05,
+      "loss": 0.9304,
+      "step": 18390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.145659209304529,
+      "learning_rate": 1.2616187092313114e-05,
+      "loss": 0.7902,
+      "step": 18391
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6841177400503968,
+      "learning_rate": 1.2615450619360547e-05,
+      "loss": 0.859,
+      "step": 18392
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4992185250196268,
+      "learning_rate": 1.2614714131179999e-05,
+      "loss": 0.9297,
+      "step": 18393
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.501543559399095,
+      "learning_rate": 1.2613977627775757e-05,
+      "loss": 0.862,
+      "step": 18394
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6380004027998762,
+      "learning_rate": 1.261324110915211e-05,
+      "loss": 0.8277,
+      "step": 18395
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8211908610564698,
+      "learning_rate": 1.2612504575313344e-05,
+      "loss": 0.8125,
+      "step": 18396
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.820459868187642,
+      "learning_rate": 1.2611768026263752e-05,
+      "loss": 0.7541,
+      "step": 18397
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7243964578795703,
+      "learning_rate": 1.2611031462007619e-05,
+      "loss": 0.8226,
+      "step": 18398
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.676642165902507,
+      "learning_rate": 1.2610294882549232e-05,
+      "loss": 0.716,
+      "step": 18399
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.947489804970619,
+      "learning_rate": 1.2609558287892884e-05,
+      "loss": 0.9171,
+      "step": 18400
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6251798224934435,
+      "learning_rate": 1.2608821678042862e-05,
+      "loss": 0.9111,
+      "step": 18401
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0889218605591084,
+      "learning_rate": 1.2608085053003452e-05,
+      "loss": 0.8952,
+      "step": 18402
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1690913498858717,
+      "learning_rate": 1.2607348412778949e-05,
+      "loss": 0.811,
+      "step": 18403
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.814688803829184,
+      "learning_rate": 1.2606611757373633e-05,
+      "loss": 0.7487,
+      "step": 18404
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4269687169840728,
+      "learning_rate": 1.2605875086791804e-05,
+      "loss": 0.9821,
+      "step": 18405
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7695201790383598,
+      "learning_rate": 1.2605138401037738e-05,
+      "loss": 0.697,
+      "step": 18406
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9699077110664938,
+      "learning_rate": 1.2604401700115738e-05,
+      "loss": 0.8377,
+      "step": 18407
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6279602366497592,
+      "learning_rate": 1.2603664984030082e-05,
+      "loss": 0.9628,
+      "step": 18408
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4862584164375563,
+      "learning_rate": 1.2602928252785068e-05,
+      "loss": 0.8505,
+      "step": 18409
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6903329474547366,
+      "learning_rate": 1.260219150638498e-05,
+      "loss": 0.8883,
+      "step": 18410
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4804346635158998,
+      "learning_rate": 1.2601454744834108e-05,
+      "loss": 0.9242,
+      "step": 18411
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3899587221486847,
+      "learning_rate": 1.2600717968136742e-05,
+      "loss": 0.7728,
+      "step": 18412
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2985254956156385,
+      "learning_rate": 1.2599981176297176e-05,
+      "loss": 0.8668,
+      "step": 18413
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9658247998689051,
+      "learning_rate": 1.2599244369319694e-05,
+      "loss": 0.8717,
+      "step": 18414
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8077940828209313,
+      "learning_rate": 1.2598507547208589e-05,
+      "loss": 0.7311,
+      "step": 18415
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9387279741274344,
+      "learning_rate": 1.259777070996815e-05,
+      "loss": 0.7497,
+      "step": 18416
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.105091444474579,
+      "learning_rate": 1.259703385760267e-05,
+      "loss": 0.8629,
+      "step": 18417
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2286465826257995,
+      "learning_rate": 1.2596296990116435e-05,
+      "loss": 0.7491,
+      "step": 18418
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7879050594891945,
+      "learning_rate": 1.2595560107513735e-05,
+      "loss": 0.8979,
+      "step": 18419
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8120964247720378,
+      "learning_rate": 1.2594823209798865e-05,
+      "loss": 0.725,
+      "step": 18420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8082216656958454,
+      "learning_rate": 1.2594086296976111e-05,
+      "loss": 0.8109,
+      "step": 18421
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6456290087581906,
+      "learning_rate": 1.2593349369049767e-05,
+      "loss": 0.8816,
+      "step": 18422
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6260153435401017,
+      "learning_rate": 1.259261242602412e-05,
+      "loss": 0.8464,
+      "step": 18423
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.543843673960858,
+      "learning_rate": 1.2591875467903465e-05,
+      "loss": 0.8794,
+      "step": 18424
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6051341361124833,
+      "learning_rate": 1.2591138494692087e-05,
+      "loss": 0.7412,
+      "step": 18425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.01443276262704,
+      "learning_rate": 1.2590401506394285e-05,
+      "loss": 0.8799,
+      "step": 18426
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7734715324309952,
+      "learning_rate": 1.258966450301434e-05,
+      "loss": 0.7099,
+      "step": 18427
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1518905966970783,
+      "learning_rate": 1.2588927484556552e-05,
+      "loss": 0.911,
+      "step": 18428
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4815110347107734,
+      "learning_rate": 1.2588190451025209e-05,
+      "loss": 0.8369,
+      "step": 18429
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5264199930031959,
+      "learning_rate": 1.2587453402424598e-05,
+      "loss": 0.9812,
+      "step": 18430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.927616156417253,
+      "learning_rate": 1.2586716338759018e-05,
+      "loss": 1.0247,
+      "step": 18431
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6297590457913718,
+      "learning_rate": 1.2585979260032754e-05,
+      "loss": 1.0304,
+      "step": 18432
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.844072726991944,
+      "learning_rate": 1.2585242166250105e-05,
+      "loss": 0.8285,
+      "step": 18433
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7679748541411686,
+      "learning_rate": 1.2584505057415352e-05,
+      "loss": 0.8841,
+      "step": 18434
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6912973312388342,
+      "learning_rate": 1.2583767933532795e-05,
+      "loss": 0.6746,
+      "step": 18435
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8236846275740997,
+      "learning_rate": 1.258303079460672e-05,
+      "loss": 0.9088,
+      "step": 18436
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7856118647372774,
+      "learning_rate": 1.2582293640641427e-05,
+      "loss": 0.9389,
+      "step": 18437
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6808526147669451,
+      "learning_rate": 1.2581556471641199e-05,
+      "loss": 0.9059,
+      "step": 18438
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.620756189245074,
+      "learning_rate": 1.2580819287610332e-05,
+      "loss": 0.8638,
+      "step": 18439
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7729060471924702,
+      "learning_rate": 1.2580082088553121e-05,
+      "loss": 0.7833,
+      "step": 18440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9488235212967908,
+      "learning_rate": 1.2579344874473854e-05,
+      "loss": 0.8353,
+      "step": 18441
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9093253179377832,
+      "learning_rate": 1.2578607645376823e-05,
+      "loss": 0.807,
+      "step": 18442
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.681675527016743,
+      "learning_rate": 1.2577870401266324e-05,
+      "loss": 0.8459,
+      "step": 18443
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5698731475979664,
+      "learning_rate": 1.2577133142146646e-05,
+      "loss": 0.8585,
+      "step": 18444
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7252952206191328,
+      "learning_rate": 1.2576395868022082e-05,
+      "loss": 0.6701,
+      "step": 18445
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.619487418445573,
+      "learning_rate": 1.2575658578896928e-05,
+      "loss": 0.9004,
+      "step": 18446
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5370834926579489,
+      "learning_rate": 1.2574921274775472e-05,
+      "loss": 0.9146,
+      "step": 18447
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6451005964878802,
+      "learning_rate": 1.257418395566201e-05,
+      "loss": 0.9219,
+      "step": 18448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5140737025592743,
+      "learning_rate": 1.2573446621560833e-05,
+      "loss": 0.9203,
+      "step": 18449
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4865533770094046,
+      "learning_rate": 1.2572709272476237e-05,
+      "loss": 0.8928,
+      "step": 18450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.549271564442596,
+      "learning_rate": 1.2571971908412512e-05,
+      "loss": 0.8004,
+      "step": 18451
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8159498550959134,
+      "learning_rate": 1.2571234529373953e-05,
+      "loss": 0.7243,
+      "step": 18452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7211502869746764,
+      "learning_rate": 1.2570497135364851e-05,
+      "loss": 0.8143,
+      "step": 18453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5726454608812332,
+      "learning_rate": 1.2569759726389505e-05,
+      "loss": 0.7869,
+      "step": 18454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8609241566288195,
+      "learning_rate": 1.2569022302452199e-05,
+      "loss": 0.7147,
+      "step": 18455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7356180621828035,
+      "learning_rate": 1.2568284863557236e-05,
+      "loss": 0.7094,
+      "step": 18456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.193449711568744,
+      "learning_rate": 1.25675474097089e-05,
+      "loss": 0.8351,
+      "step": 18457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8163650595846562,
+      "learning_rate": 1.2566809940911497e-05,
+      "loss": 0.7386,
+      "step": 18458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.939080657099378,
+      "learning_rate": 1.2566072457169308e-05,
+      "loss": 0.9641,
+      "step": 18459
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.555239926988521,
+      "learning_rate": 1.2565334958486636e-05,
+      "loss": 0.8119,
+      "step": 18460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8032569122111516,
+      "learning_rate": 1.2564597444867772e-05,
+      "loss": 0.6926,
+      "step": 18461
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8346155803359971,
+      "learning_rate": 1.256385991631701e-05,
+      "loss": 0.8968,
+      "step": 18462
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6938915650601531,
+      "learning_rate": 1.2563122372838642e-05,
+      "loss": 0.855,
+      "step": 18463
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8317340625226362,
+      "learning_rate": 1.2562384814436961e-05,
+      "loss": 0.7553,
+      "step": 18464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.568180994968676,
+      "learning_rate": 1.256164724111627e-05,
+      "loss": 0.6592,
+      "step": 18465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7911598005902045,
+      "learning_rate": 1.2560909652880854e-05,
+      "loss": 0.8626,
+      "step": 18466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.628527698964872,
+      "learning_rate": 1.2560172049735016e-05,
+      "loss": 0.7996,
+      "step": 18467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7336461006816265,
+      "learning_rate": 1.2559434431683042e-05,
+      "loss": 0.7107,
+      "step": 18468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.604512300948614,
+      "learning_rate": 1.2558696798729231e-05,
+      "loss": 0.8518,
+      "step": 18469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5886730967672438,
+      "learning_rate": 1.255795915087788e-05,
+      "loss": 0.8734,
+      "step": 18470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6497868550120274,
+      "learning_rate": 1.2557221488133276e-05,
+      "loss": 0.837,
+      "step": 18471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.569382581762408,
+      "learning_rate": 1.255648381049972e-05,
+      "loss": 0.8062,
+      "step": 18472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6506696126662068,
+      "learning_rate": 1.2555746117981506e-05,
+      "loss": 0.7283,
+      "step": 18473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.833253128167209,
+      "learning_rate": 1.2555008410582932e-05,
+      "loss": 0.9211,
+      "step": 18474
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6506896988058968,
+      "learning_rate": 1.255427068830829e-05,
+      "loss": 0.833,
+      "step": 18475
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.137256162506627,
+      "learning_rate": 1.255353295116187e-05,
+      "loss": 0.9072,
+      "step": 18476
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5744600054211237,
+      "learning_rate": 1.2552795199147979e-05,
+      "loss": 0.8725,
+      "step": 18477
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8939738217205584,
+      "learning_rate": 1.2552057432270903e-05,
+      "loss": 0.8667,
+      "step": 18478
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7224217883986421,
+      "learning_rate": 1.255131965053494e-05,
+      "loss": 0.9328,
+      "step": 18479
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.483010756508878,
+      "learning_rate": 1.2550581853944388e-05,
+      "loss": 0.7509,
+      "step": 18480
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7866450136811267,
+      "learning_rate": 1.254984404250354e-05,
+      "loss": 0.7314,
+      "step": 18481
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1309870514400155,
+      "learning_rate": 1.2549106216216693e-05,
+      "loss": 0.9758,
+      "step": 18482
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5328726575976108,
+      "learning_rate": 1.2548368375088143e-05,
+      "loss": 0.955,
+      "step": 18483
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.674295632068562,
+      "learning_rate": 1.2547630519122187e-05,
+      "loss": 0.8988,
+      "step": 18484
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.56146950267762,
+      "learning_rate": 1.2546892648323114e-05,
+      "loss": 0.9401,
+      "step": 18485
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.491143239939247,
+      "learning_rate": 1.2546154762695233e-05,
+      "loss": 0.9225,
+      "step": 18486
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6926881697623712,
+      "learning_rate": 1.2545416862242827e-05,
+      "loss": 0.9529,
+      "step": 18487
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.071480398651526,
+      "learning_rate": 1.2544678946970201e-05,
+      "loss": 0.7655,
+      "step": 18488
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6105752652791168,
+      "learning_rate": 1.2543941016881647e-05,
+      "loss": 0.972,
+      "step": 18489
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6140856297657666,
+      "learning_rate": 1.2543203071981467e-05,
+      "loss": 0.8153,
+      "step": 18490
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5179472696239926,
+      "learning_rate": 1.254246511227395e-05,
+      "loss": 0.7919,
+      "step": 18491
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7806859548461356,
+      "learning_rate": 1.2541727137763395e-05,
+      "loss": 0.7982,
+      "step": 18492
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5269023805445359,
+      "learning_rate": 1.2540989148454102e-05,
+      "loss": 0.846,
+      "step": 18493
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5251293314605345,
+      "learning_rate": 1.2540251144350366e-05,
+      "loss": 0.7628,
+      "step": 18494
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5139149834771306,
+      "learning_rate": 1.2539513125456484e-05,
+      "loss": 1.006,
+      "step": 18495
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6114983480662641,
+      "learning_rate": 1.253877509177675e-05,
+      "loss": 0.8441,
+      "step": 18496
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5810769517765577,
+      "learning_rate": 1.2538037043315465e-05,
+      "loss": 0.8334,
+      "step": 18497
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7911912949571818,
+      "learning_rate": 1.2537298980076926e-05,
+      "loss": 0.8977,
+      "step": 18498
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5655316389103584,
+      "learning_rate": 1.2536560902065427e-05,
+      "loss": 0.8833,
+      "step": 18499
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.435536872129883,
+      "learning_rate": 1.253582280928527e-05,
+      "loss": 0.8286,
+      "step": 18500
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5119486046918005,
+      "learning_rate": 1.2535084701740748e-05,
+      "loss": 0.962,
+      "step": 18501
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.894720348372885,
+      "learning_rate": 1.2534346579436158e-05,
+      "loss": 0.7338,
+      "step": 18502
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5792129059487923,
+      "learning_rate": 1.2533608442375803e-05,
+      "loss": 0.8167,
+      "step": 18503
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9213888370994223,
+      "learning_rate": 1.2532870290563976e-05,
+      "loss": 0.9326,
+      "step": 18504
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4540202957207435,
+      "learning_rate": 1.253213212400498e-05,
+      "loss": 0.9357,
+      "step": 18505
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5001471986325972,
+      "learning_rate": 1.2531393942703104e-05,
+      "loss": 0.8724,
+      "step": 18506
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9816758446410039,
+      "learning_rate": 1.2530655746662657e-05,
+      "loss": 0.8683,
+      "step": 18507
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.087853227183185,
+      "learning_rate": 1.2529917535887923e-05,
+      "loss": 1.0032,
+      "step": 18508
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.445930967996537,
+      "learning_rate": 1.2529179310383215e-05,
+      "loss": 0.9524,
+      "step": 18509
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.586329650189806,
+      "learning_rate": 1.252844107015282e-05,
+      "loss": 0.874,
+      "step": 18510
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8806991125075323,
+      "learning_rate": 1.2527702815201043e-05,
+      "loss": 0.7357,
+      "step": 18511
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8450220075217754,
+      "learning_rate": 1.2526964545532182e-05,
+      "loss": 0.9579,
+      "step": 18512
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7826576128324804,
+      "learning_rate": 1.2526226261150527e-05,
+      "loss": 0.7418,
+      "step": 18513
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4834007166205534,
+      "learning_rate": 1.2525487962060389e-05,
+      "loss": 0.9117,
+      "step": 18514
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8643890325358029,
+      "learning_rate": 1.2524749648266054e-05,
+      "loss": 0.7998,
+      "step": 18515
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7851956474039918,
+      "learning_rate": 1.2524011319771833e-05,
+      "loss": 0.779,
+      "step": 18516
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7095704765057167,
+      "learning_rate": 1.2523272976582018e-05,
+      "loss": 0.9048,
+      "step": 18517
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5637234070983068,
+      "learning_rate": 1.252253461870091e-05,
+      "loss": 0.9782,
+      "step": 18518
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7895310851323376,
+      "learning_rate": 1.2521796246132804e-05,
+      "loss": 0.7614,
+      "step": 18519
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4985738070233963,
+      "learning_rate": 1.2521057858882002e-05,
+      "loss": 0.8645,
+      "step": 18520
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7619593767502519,
+      "learning_rate": 1.2520319456952806e-05,
+      "loss": 0.7283,
+      "step": 18521
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9459643381815117,
+      "learning_rate": 1.2519581040349508e-05,
+      "loss": 0.8842,
+      "step": 18522
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6547094037486785,
+      "learning_rate": 1.2518842609076414e-05,
+      "loss": 0.7547,
+      "step": 18523
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6379867716174878,
+      "learning_rate": 1.2518104163137821e-05,
+      "loss": 0.911,
+      "step": 18524
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7294119032621071,
+      "learning_rate": 1.2517365702538026e-05,
+      "loss": 0.9097,
+      "step": 18525
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.76296941274947,
+      "learning_rate": 1.2516627227281336e-05,
+      "loss": 0.7144,
+      "step": 18526
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6537843454806018,
+      "learning_rate": 1.2515888737372042e-05,
+      "loss": 0.8865,
+      "step": 18527
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.639791107506259,
+      "learning_rate": 1.2515150232814446e-05,
+      "loss": 0.9655,
+      "step": 18528
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9461485513060979,
+      "learning_rate": 1.251441171361285e-05,
+      "loss": 0.9286,
+      "step": 18529
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8627531412036629,
+      "learning_rate": 1.2513673179771555e-05,
+      "loss": 0.7758,
+      "step": 18530
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5795428281041237,
+      "learning_rate": 1.2512934631294858e-05,
+      "loss": 0.8383,
+      "step": 18531
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0840415106315286,
+      "learning_rate": 1.251219606818706e-05,
+      "loss": 0.7247,
+      "step": 18532
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9872804892728055,
+      "learning_rate": 1.251145749045246e-05,
+      "loss": 0.8695,
+      "step": 18533
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.458454700640228,
+      "learning_rate": 1.2510718898095359e-05,
+      "loss": 0.7663,
+      "step": 18534
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4767169428726241,
+      "learning_rate": 1.2509980291120061e-05,
+      "loss": 0.8556,
+      "step": 18535
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.941048133881919,
+      "learning_rate": 1.2509241669530858e-05,
+      "loss": 0.8505,
+      "step": 18536
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5888541976641557,
+      "learning_rate": 1.250850303333206e-05,
+      "loss": 0.8392,
+      "step": 18537
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7911179089349049,
+      "learning_rate": 1.250776438252796e-05,
+      "loss": 0.7956,
+      "step": 18538
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5930258375443358,
+      "learning_rate": 1.2507025717122865e-05,
+      "loss": 0.8546,
+      "step": 18539
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4349992842804022,
+      "learning_rate": 1.2506287037121072e-05,
+      "loss": 0.7167,
+      "step": 18540
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9204535574431536,
+      "learning_rate": 1.250554834252688e-05,
+      "loss": 0.8127,
+      "step": 18541
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6574548219448695,
+      "learning_rate": 1.2504809633344594e-05,
+      "loss": 0.8776,
+      "step": 18542
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8891461157529665,
+      "learning_rate": 1.2504070909578515e-05,
+      "loss": 0.8347,
+      "step": 18543
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0674190839732014,
+      "learning_rate": 1.2503332171232938e-05,
+      "loss": 0.9403,
+      "step": 18544
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9323645799394715,
+      "learning_rate": 1.250259341831217e-05,
+      "loss": 0.776,
+      "step": 18545
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.700043066416679,
+      "learning_rate": 1.2501854650820513e-05,
+      "loss": 0.9498,
+      "step": 18546
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4401888560758092,
+      "learning_rate": 1.2501115868762264e-05,
+      "loss": 0.8815,
+      "step": 18547
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4038467987293557,
+      "learning_rate": 1.2500377072141727e-05,
+      "loss": 0.995,
+      "step": 18548
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.760414272764136,
+      "learning_rate": 1.2499638260963204e-05,
+      "loss": 0.7848,
+      "step": 18549
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8045040173516476,
+      "learning_rate": 1.2498899435230996e-05,
+      "loss": 0.8165,
+      "step": 18550
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5613595131219713,
+      "learning_rate": 1.2498160594949402e-05,
+      "loss": 0.8793,
+      "step": 18551
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.89056745461118,
+      "learning_rate": 1.2497421740122728e-05,
+      "loss": 0.9002,
+      "step": 18552
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8053841191740316,
+      "learning_rate": 1.2496682870755275e-05,
+      "loss": 0.9349,
+      "step": 18553
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5350543181044463,
+      "learning_rate": 1.2495943986851342e-05,
+      "loss": 0.8555,
+      "step": 18554
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3768646937027846,
+      "learning_rate": 1.2495205088415233e-05,
+      "loss": 0.878,
+      "step": 18555
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6651088775473764,
+      "learning_rate": 1.2494466175451253e-05,
+      "loss": 0.9598,
+      "step": 18556
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6684123508041824,
+      "learning_rate": 1.2493727247963695e-05,
+      "loss": 0.7703,
+      "step": 18557
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6454115447160715,
+      "learning_rate": 1.2492988305956875e-05,
+      "loss": 0.8399,
+      "step": 18558
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6791853225775828,
+      "learning_rate": 1.249224934943508e-05,
+      "loss": 0.91,
+      "step": 18559
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.74945701969555,
+      "learning_rate": 1.2491510378402628e-05,
+      "loss": 0.8823,
+      "step": 18560
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8008822225930126,
+      "learning_rate": 1.2490771392863805e-05,
+      "loss": 0.8469,
+      "step": 18561
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5160541095374338,
+      "learning_rate": 1.2490032392822929e-05,
+      "loss": 0.9736,
+      "step": 18562
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5310462009513413,
+      "learning_rate": 1.2489293378284298e-05,
+      "loss": 0.8051,
+      "step": 18563
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.512569994411533,
+      "learning_rate": 1.2488554349252205e-05,
+      "loss": 0.8541,
+      "step": 18564
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.527011644298313,
+      "learning_rate": 1.2487815305730967e-05,
+      "loss": 0.8989,
+      "step": 18565
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7616959014394012,
+      "learning_rate": 1.2487076247724876e-05,
+      "loss": 0.7731,
+      "step": 18566
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6851570928404156,
+      "learning_rate": 1.2486337175238246e-05,
+      "loss": 0.8349,
+      "step": 18567
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7916037231518003,
+      "learning_rate": 1.2485598088275367e-05,
+      "loss": 0.9319,
+      "step": 18568
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5082369102115452,
+      "learning_rate": 1.2484858986840553e-05,
+      "loss": 0.8252,
+      "step": 18569
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8239478261810695,
+      "learning_rate": 1.2484119870938102e-05,
+      "loss": 0.9417,
+      "step": 18570
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7768441369028714,
+      "learning_rate": 1.2483380740572319e-05,
+      "loss": 0.8244,
+      "step": 18571
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6706477429468913,
+      "learning_rate": 1.2482641595747508e-05,
+      "loss": 0.8242,
+      "step": 18572
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7874292438890546,
+      "learning_rate": 1.2481902436467972e-05,
+      "loss": 0.742,
+      "step": 18573
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6048676582714014,
+      "learning_rate": 1.2481163262738011e-05,
+      "loss": 0.81,
+      "step": 18574
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5858854519039918,
+      "learning_rate": 1.2480424074561934e-05,
+      "loss": 0.8589,
+      "step": 18575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3651474238833896,
+      "learning_rate": 1.2479684871944044e-05,
+      "loss": 0.915,
+      "step": 18576
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5928170801465793,
+      "learning_rate": 1.247894565488864e-05,
+      "loss": 0.9937,
+      "step": 18577
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5098689562360772,
+      "learning_rate": 1.2478206423400034e-05,
+      "loss": 0.9492,
+      "step": 18578
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.723021242696674,
+      "learning_rate": 1.2477467177482524e-05,
+      "loss": 0.8399,
+      "step": 18579
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.44398601135251,
+      "learning_rate": 1.2476727917140415e-05,
+      "loss": 0.7768,
+      "step": 18580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.499264651644108,
+      "learning_rate": 1.2475988642378011e-05,
+      "loss": 0.8561,
+      "step": 18581
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.473181992223792,
+      "learning_rate": 1.247524935319962e-05,
+      "loss": 0.8577,
+      "step": 18582
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.943238652226026,
+      "learning_rate": 1.247451004960954e-05,
+      "loss": 0.9073,
+      "step": 18583
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5420097305495493,
+      "learning_rate": 1.2473770731612084e-05,
+      "loss": 0.8565,
+      "step": 18584
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.564033255899291,
+      "learning_rate": 1.2473031399211548e-05,
+      "loss": 0.7917,
+      "step": 18585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.692953913641639,
+      "learning_rate": 1.2472292052412243e-05,
+      "loss": 0.889,
+      "step": 18586
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5201452073783752,
+      "learning_rate": 1.2471552691218465e-05,
+      "loss": 0.8381,
+      "step": 18587
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5584306782862434,
+      "learning_rate": 1.247081331563453e-05,
+      "loss": 0.9328,
+      "step": 18588
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8937844104004105,
+      "learning_rate": 1.2470073925664736e-05,
+      "loss": 0.8355,
+      "step": 18589
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7363536330528169,
+      "learning_rate": 1.2469334521313392e-05,
+      "loss": 0.7582,
+      "step": 18590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8014600089308617,
+      "learning_rate": 1.2468595102584798e-05,
+      "loss": 0.9095,
+      "step": 18591
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8417853743020346,
+      "learning_rate": 1.2467855669483262e-05,
+      "loss": 0.7749,
+      "step": 18592
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6458395603368392,
+      "learning_rate": 1.2467116222013089e-05,
+      "loss": 0.8098,
+      "step": 18593
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7499788469919815,
+      "learning_rate": 1.2466376760178584e-05,
+      "loss": 0.9619,
+      "step": 18594
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3658041969335515,
+      "learning_rate": 1.2465637283984056e-05,
+      "loss": 0.7659,
+      "step": 18595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7036507085411485,
+      "learning_rate": 1.2464897793433803e-05,
+      "loss": 0.807,
+      "step": 18596
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.836940403729908,
+      "learning_rate": 1.2464158288532136e-05,
+      "loss": 0.8448,
+      "step": 18597
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9591780076097198,
+      "learning_rate": 1.246341876928336e-05,
+      "loss": 0.8669,
+      "step": 18598
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5662694395859873,
+      "learning_rate": 1.2462679235691779e-05,
+      "loss": 0.7954,
+      "step": 18599
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.894747372225828,
+      "learning_rate": 1.24619396877617e-05,
+      "loss": 0.7534,
+      "step": 18600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5910189936400794,
+      "learning_rate": 1.2461200125497428e-05,
+      "loss": 0.8526,
+      "step": 18601
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4094343737609631,
+      "learning_rate": 1.2460460548903273e-05,
+      "loss": 0.8542,
+      "step": 18602
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7959828345270996,
+      "learning_rate": 1.2459720957983535e-05,
+      "loss": 0.7938,
+      "step": 18603
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4679231536982116,
+      "learning_rate": 1.2458981352742522e-05,
+      "loss": 0.9101,
+      "step": 18604
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6610288427145619,
+      "learning_rate": 1.2458241733184542e-05,
+      "loss": 0.8703,
+      "step": 18605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8544494327429744,
+      "learning_rate": 1.2457502099313901e-05,
+      "loss": 0.8577,
+      "step": 18606
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.716002110571162,
+      "learning_rate": 1.2456762451134905e-05,
+      "loss": 0.8621,
+      "step": 18607
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6297244202274543,
+      "learning_rate": 1.2456022788651858e-05,
+      "loss": 0.7661,
+      "step": 18608
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5616288011650663,
+      "learning_rate": 1.2455283111869072e-05,
+      "loss": 0.7602,
+      "step": 18609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.134741314798565,
+      "learning_rate": 1.2454543420790847e-05,
+      "loss": 0.8964,
+      "step": 18610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6809452332605797,
+      "learning_rate": 1.2453803715421496e-05,
+      "loss": 0.7572,
+      "step": 18611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.641287635365525,
+      "learning_rate": 1.245306399576532e-05,
+      "loss": 0.8737,
+      "step": 18612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6426238471327235,
+      "learning_rate": 1.2452324261826632e-05,
+      "loss": 0.9688,
+      "step": 18613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7635936891679327,
+      "learning_rate": 1.2451584513609736e-05,
+      "loss": 0.8523,
+      "step": 18614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5638014750200886,
+      "learning_rate": 1.2450844751118934e-05,
+      "loss": 0.9353,
+      "step": 18615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8310332302512824,
+      "learning_rate": 1.2450104974358544e-05,
+      "loss": 0.7372,
+      "step": 18616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5125079840118587,
+      "learning_rate": 1.2449365183332862e-05,
+      "loss": 0.9116,
+      "step": 18617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5009334178896252,
+      "learning_rate": 1.2448625378046205e-05,
+      "loss": 0.9065,
+      "step": 18618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6717523286183988,
+      "learning_rate": 1.244788555850287e-05,
+      "loss": 0.9664,
+      "step": 18619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.566306415913165,
+      "learning_rate": 1.2447145724707178e-05,
+      "loss": 0.7103,
+      "step": 18620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.552200203588767,
+      "learning_rate": 1.2446405876663424e-05,
+      "loss": 0.9291,
+      "step": 18621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.823010969550428,
+      "learning_rate": 1.2445666014375918e-05,
+      "loss": 0.8683,
+      "step": 18622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5092774785208973,
+      "learning_rate": 1.2444926137848974e-05,
+      "loss": 0.839,
+      "step": 18623
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.733610819915312,
+      "learning_rate": 1.2444186247086895e-05,
+      "loss": 0.7225,
+      "step": 18624
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3570977350771676,
+      "learning_rate": 1.244344634209399e-05,
+      "loss": 0.8545,
+      "step": 18625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5666637167624657,
+      "learning_rate": 1.2442706422874567e-05,
+      "loss": 0.8995,
+      "step": 18626
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4441305841144776,
+      "learning_rate": 1.2441966489432936e-05,
+      "loss": 0.9431,
+      "step": 18627
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7318440461353428,
+      "learning_rate": 1.24412265417734e-05,
+      "loss": 0.8763,
+      "step": 18628
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.464315332391821,
+      "learning_rate": 1.2440486579900268e-05,
+      "loss": 0.8963,
+      "step": 18629
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7839645100207328,
+      "learning_rate": 1.2439746603817855e-05,
+      "loss": 0.7497,
+      "step": 18630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7294042623390606,
+      "learning_rate": 1.2439006613530462e-05,
+      "loss": 0.8309,
+      "step": 18631
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.505046489789278,
+      "learning_rate": 1.2438266609042403e-05,
+      "loss": 0.8952,
+      "step": 18632
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.566020313067043,
+      "learning_rate": 1.2437526590357982e-05,
+      "loss": 0.8923,
+      "step": 18633
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5447803965208087,
+      "learning_rate": 1.243678655748151e-05,
+      "loss": 0.9006,
+      "step": 18634
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8754563227083025,
+      "learning_rate": 1.2436046510417297e-05,
+      "loss": 0.9314,
+      "step": 18635
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8154982356887486,
+      "learning_rate": 1.2435306449169646e-05,
+      "loss": 0.7444,
+      "step": 18636
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7402565883921663,
+      "learning_rate": 1.2434566373742873e-05,
+      "loss": 0.8833,
+      "step": 18637
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8324919837708926,
+      "learning_rate": 1.2433826284141282e-05,
+      "loss": 0.9143,
+      "step": 18638
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6044830713554388,
+      "learning_rate": 1.2433086180369184e-05,
+      "loss": 0.9228,
+      "step": 18639
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3910870701063516,
+      "learning_rate": 1.2432346062430889e-05,
+      "loss": 0.7759,
+      "step": 18640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4829404784340945,
+      "learning_rate": 1.2431605930330706e-05,
+      "loss": 0.7804,
+      "step": 18641
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6329428763547988,
+      "learning_rate": 1.243086578407294e-05,
+      "loss": 0.8629,
+      "step": 18642
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.575302984666261,
+      "learning_rate": 1.2430125623661909e-05,
+      "loss": 0.9684,
+      "step": 18643
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8077564987309949,
+      "learning_rate": 1.2429385449101912e-05,
+      "loss": 0.8613,
+      "step": 18644
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6127827602087064,
+      "learning_rate": 1.2428645260397267e-05,
+      "loss": 0.8749,
+      "step": 18645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.807572199263825,
+      "learning_rate": 1.2427905057552279e-05,
+      "loss": 0.8824,
+      "step": 18646
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5239618156417738,
+      "learning_rate": 1.242716484057126e-05,
+      "loss": 0.8921,
+      "step": 18647
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6698593443392535,
+      "learning_rate": 1.2426424609458518e-05,
+      "loss": 0.968,
+      "step": 18648
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6656841463162972,
+      "learning_rate": 1.2425684364218365e-05,
+      "loss": 0.9049,
+      "step": 18649
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1168515367960783,
+      "learning_rate": 1.2424944104855107e-05,
+      "loss": 0.8622,
+      "step": 18650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6320764366649956,
+      "learning_rate": 1.2424203831373058e-05,
+      "loss": 0.9699,
+      "step": 18651
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8044736703435609,
+      "learning_rate": 1.2423463543776529e-05,
+      "loss": 0.7406,
+      "step": 18652
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8075842264749513,
+      "learning_rate": 1.2422723242069826e-05,
+      "loss": 0.8121,
+      "step": 18653
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5301396865155434,
+      "learning_rate": 1.2421982926257262e-05,
+      "loss": 0.8102,
+      "step": 18654
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5026374048606255,
+      "learning_rate": 1.2421242596343144e-05,
+      "loss": 0.8763,
+      "step": 18655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.730072557679017,
+      "learning_rate": 1.242050225233179e-05,
+      "loss": 0.7783,
+      "step": 18656
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5793455067954516,
+      "learning_rate": 1.24197618942275e-05,
+      "loss": 0.803,
+      "step": 18657
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.352060376186119,
+      "learning_rate": 1.2419021522034596e-05,
+      "loss": 0.8339,
+      "step": 18658
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.582393294446416,
+      "learning_rate": 1.2418281135757378e-05,
+      "loss": 0.9127,
+      "step": 18659
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.061410399698006,
+      "learning_rate": 1.2417540735400161e-05,
+      "loss": 0.9088,
+      "step": 18660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8461358036903808,
+      "learning_rate": 1.241680032096726e-05,
+      "loss": 0.8407,
+      "step": 18661
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.489036342479071,
+      "learning_rate": 1.2416059892462979e-05,
+      "loss": 0.9246,
+      "step": 18662
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.636611624482529,
+      "learning_rate": 1.2415319449891638e-05,
+      "loss": 0.9206,
+      "step": 18663
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7145957973554087,
+      "learning_rate": 1.2414578993257535e-05,
+      "loss": 0.8785,
+      "step": 18664
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4278789753745402,
+      "learning_rate": 1.2413838522564997e-05,
+      "loss": 0.8466,
+      "step": 18665
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8929448576032195,
+      "learning_rate": 1.2413098037818321e-05,
+      "loss": 0.8872,
+      "step": 18666
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7550836967009166,
+      "learning_rate": 1.2412357539021826e-05,
+      "loss": 0.7477,
+      "step": 18667
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4383592510816436,
+      "learning_rate": 1.2411617026179821e-05,
+      "loss": 0.9978,
+      "step": 18668
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8085245486017179,
+      "learning_rate": 1.241087649929662e-05,
+      "loss": 0.7497,
+      "step": 18669
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4508829217817198,
+      "learning_rate": 1.241013595837653e-05,
+      "loss": 0.8649,
+      "step": 18670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.595891234378773,
+      "learning_rate": 1.240939540342387e-05,
+      "loss": 0.9757,
+      "step": 18671
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6467574291292482,
+      "learning_rate": 1.2408654834442942e-05,
+      "loss": 1.0508,
+      "step": 18672
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6274751577588575,
+      "learning_rate": 1.2407914251438065e-05,
+      "loss": 0.7746,
+      "step": 18673
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7849950997268758,
+      "learning_rate": 1.2407173654413549e-05,
+      "loss": 0.7504,
+      "step": 18674
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.791125001155868,
+      "learning_rate": 1.2406433043373707e-05,
+      "loss": 0.9678,
+      "step": 18675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8939404954366625,
+      "learning_rate": 1.2405692418322847e-05,
+      "loss": 0.8801,
+      "step": 18676
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6372963545418382,
+      "learning_rate": 1.2404951779265287e-05,
+      "loss": 0.8098,
+      "step": 18677
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.725601784200772,
+      "learning_rate": 1.2404211126205335e-05,
+      "loss": 0.8702,
+      "step": 18678
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.580412359294269,
+      "learning_rate": 1.2403470459147305e-05,
+      "loss": 0.9397,
+      "step": 18679
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.852432353098998,
+      "learning_rate": 1.2402729778095507e-05,
+      "loss": 0.8253,
+      "step": 18680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6025711808183853,
+      "learning_rate": 1.2401989083054258e-05,
+      "loss": 0.933,
+      "step": 18681
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6130913480583517,
+      "learning_rate": 1.2401248374027868e-05,
+      "loss": 0.8928,
+      "step": 18682
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7456292448391977,
+      "learning_rate": 1.240050765102065e-05,
+      "loss": 0.736,
+      "step": 18683
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8780386747159292,
+      "learning_rate": 1.2399766914036914e-05,
+      "loss": 0.9348,
+      "step": 18684
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6685006698615217,
+      "learning_rate": 1.2399026163080978e-05,
+      "loss": 0.9653,
+      "step": 18685
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.578054908729407,
+      "learning_rate": 1.2398285398157153e-05,
+      "loss": 0.724,
+      "step": 18686
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7918585298349232,
+      "learning_rate": 1.2397544619269748e-05,
+      "loss": 0.8389,
+      "step": 18687
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6015475427078858,
+      "learning_rate": 1.2396803826423083e-05,
+      "loss": 0.8866,
+      "step": 18688
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7315751804691224,
+      "learning_rate": 1.2396063019621463e-05,
+      "loss": 0.9258,
+      "step": 18689
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7435955603796052,
+      "learning_rate": 1.2395322198869209e-05,
+      "loss": 0.9488,
+      "step": 18690
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1412431379585186,
+      "learning_rate": 1.2394581364170626e-05,
+      "loss": 0.8378,
+      "step": 18691
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2335745380967773,
+      "learning_rate": 1.2393840515530038e-05,
+      "loss": 0.7598,
+      "step": 18692
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8296668806940577,
+      "learning_rate": 1.239309965295175e-05,
+      "loss": 0.8221,
+      "step": 18693
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7899345782081604,
+      "learning_rate": 1.2392358776440076e-05,
+      "loss": 0.7391,
+      "step": 18694
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6221455912580618,
+      "learning_rate": 1.2391617885999336e-05,
+      "loss": 0.9191,
+      "step": 18695
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7680785527328293,
+      "learning_rate": 1.2390876981633838e-05,
+      "loss": 0.9139,
+      "step": 18696
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4897223308628513,
+      "learning_rate": 1.2390136063347895e-05,
+      "loss": 0.828,
+      "step": 18697
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2076761219957954,
+      "learning_rate": 1.2389395131145823e-05,
+      "loss": 0.8408,
+      "step": 18698
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7826555355135969,
+      "learning_rate": 1.2388654185031941e-05,
+      "loss": 0.7677,
+      "step": 18699
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7512558739042291,
+      "learning_rate": 1.2387913225010554e-05,
+      "loss": 0.748,
+      "step": 18700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.614926486728685,
+      "learning_rate": 1.2387172251085982e-05,
+      "loss": 0.8753,
+      "step": 18701
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6522486419753628,
+      "learning_rate": 1.2386431263262538e-05,
+      "loss": 0.8148,
+      "step": 18702
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7480026230199581,
+      "learning_rate": 1.2385690261544531e-05,
+      "loss": 0.7489,
+      "step": 18703
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.588767563016141,
+      "learning_rate": 1.2384949245936284e-05,
+      "loss": 0.7892,
+      "step": 18704
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.674708508768864,
+      "learning_rate": 1.2384208216442106e-05,
+      "loss": 0.8586,
+      "step": 18705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3480086260903226,
+      "learning_rate": 1.2383467173066315e-05,
+      "loss": 0.7903,
+      "step": 18706
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6190588580529972,
+      "learning_rate": 1.2382726115813224e-05,
+      "loss": 0.8394,
+      "step": 18707
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7577312780011937,
+      "learning_rate": 1.2381985044687144e-05,
+      "loss": 0.6912,
+      "step": 18708
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.50886600822469,
+      "learning_rate": 1.2381243959692394e-05,
+      "loss": 0.8613,
+      "step": 18709
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7603593994774995,
+      "learning_rate": 1.2380502860833287e-05,
+      "loss": 0.7496,
+      "step": 18710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8710139121369376,
+      "learning_rate": 1.237976174811414e-05,
+      "loss": 0.9611,
+      "step": 18711
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6259148782193025,
+      "learning_rate": 1.2379020621539266e-05,
+      "loss": 0.8701,
+      "step": 18712
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5441809475705959,
+      "learning_rate": 1.2378279481112983e-05,
+      "loss": 0.9432,
+      "step": 18713
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.855906393708446,
+      "learning_rate": 1.2377538326839603e-05,
+      "loss": 0.797,
+      "step": 18714
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.101243091300917,
+      "learning_rate": 1.2376797158723439e-05,
+      "loss": 0.854,
+      "step": 18715
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4581990365523332,
+      "learning_rate": 1.2376055976768815e-05,
+      "loss": 0.7138,
+      "step": 18716
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4844401000987408,
+      "learning_rate": 1.2375314780980034e-05,
+      "loss": 0.7695,
+      "step": 18717
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7504085408904981,
+      "learning_rate": 1.2374573571361426e-05,
+      "loss": 0.739,
+      "step": 18718
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.427981354694639,
+      "learning_rate": 1.2373832347917292e-05,
+      "loss": 0.8682,
+      "step": 18719
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.724568834798001,
+      "learning_rate": 1.2373091110651958e-05,
+      "loss": 0.8696,
+      "step": 18720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4557781196679114,
+      "learning_rate": 1.2372349859569737e-05,
+      "loss": 0.7804,
+      "step": 18721
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5037742375282768,
+      "learning_rate": 1.2371608594674943e-05,
+      "loss": 0.8169,
+      "step": 18722
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7810848631710562,
+      "learning_rate": 1.237086731597189e-05,
+      "loss": 0.764,
+      "step": 18723
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6851295321989177,
+      "learning_rate": 1.2370126023464901e-05,
+      "loss": 0.9367,
+      "step": 18724
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6615335221654088,
+      "learning_rate": 1.2369384717158285e-05,
+      "loss": 1.0448,
+      "step": 18725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8456824941402974,
+      "learning_rate": 1.2368643397056364e-05,
+      "loss": 0.7957,
+      "step": 18726
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8629558827591057,
+      "learning_rate": 1.2367902063163448e-05,
+      "loss": 0.7761,
+      "step": 18727
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5905383941946392,
+      "learning_rate": 1.236716071548386e-05,
+      "loss": 0.8163,
+      "step": 18728
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6109947303465508,
+      "learning_rate": 1.236641935402191e-05,
+      "loss": 0.8496,
+      "step": 18729
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8017818376397845,
+      "learning_rate": 1.2365677978781919e-05,
+      "loss": 0.6704,
+      "step": 18730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7630951655743177,
+      "learning_rate": 1.2364936589768202e-05,
+      "loss": 0.7478,
+      "step": 18731
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3858004713924927,
+      "learning_rate": 1.2364195186985073e-05,
+      "loss": 0.98,
+      "step": 18732
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7724057805067543,
+      "learning_rate": 1.2363453770436853e-05,
+      "loss": 0.9605,
+      "step": 18733
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.879205307623305,
+      "learning_rate": 1.2362712340127856e-05,
+      "loss": 0.9165,
+      "step": 18734
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7946913652430418,
+      "learning_rate": 1.23619708960624e-05,
+      "loss": 0.9477,
+      "step": 18735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8616669547122406,
+      "learning_rate": 1.2361229438244803e-05,
+      "loss": 0.861,
+      "step": 18736
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8278232539973984,
+      "learning_rate": 1.2360487966679383e-05,
+      "loss": 0.9231,
+      "step": 18737
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.875395148705261,
+      "learning_rate": 1.2359746481370447e-05,
+      "loss": 1.0135,
+      "step": 18738
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7730872065212102,
+      "learning_rate": 1.2359004982322328e-05,
+      "loss": 0.9452,
+      "step": 18739
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.455217665103522,
+      "learning_rate": 1.2358263469539328e-05,
+      "loss": 0.9073,
+      "step": 18740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.588233869342729,
+      "learning_rate": 1.2357521943025779e-05,
+      "loss": 0.9722,
+      "step": 18741
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8745129595322911,
+      "learning_rate": 1.2356780402785985e-05,
+      "loss": 0.8384,
+      "step": 18742
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5790002060642236,
+      "learning_rate": 1.2356038848824272e-05,
+      "loss": 0.8908,
+      "step": 18743
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.643045064566485,
+      "learning_rate": 1.2355297281144953e-05,
+      "loss": 0.8979,
+      "step": 18744
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.145961022361672,
+      "learning_rate": 1.2354555699752351e-05,
+      "loss": 0.9391,
+      "step": 18745
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.73588809552938,
+      "learning_rate": 1.2353814104650777e-05,
+      "loss": 0.73,
+      "step": 18746
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5698897092248845,
+      "learning_rate": 1.2353072495844551e-05,
+      "loss": 0.7979,
+      "step": 18747
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9061128925233608,
+      "learning_rate": 1.2352330873337995e-05,
+      "loss": 0.9209,
+      "step": 18748
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6292758878772022,
+      "learning_rate": 1.2351589237135422e-05,
+      "loss": 0.8608,
+      "step": 18749
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3618432605713964,
+      "learning_rate": 1.2350847587241155e-05,
+      "loss": 0.8899,
+      "step": 18750
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6118941620191998,
+      "learning_rate": 1.2350105923659506e-05,
+      "loss": 0.9042,
+      "step": 18751
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0339406934301465,
+      "learning_rate": 1.2349364246394797e-05,
+      "loss": 0.9299,
+      "step": 18752
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7607038850659311,
+      "learning_rate": 1.2348622555451346e-05,
+      "loss": 0.9266,
+      "step": 18753
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.846098405811811,
+      "learning_rate": 1.2347880850833472e-05,
+      "loss": 0.8387,
+      "step": 18754
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8087989277375491,
+      "learning_rate": 1.234713913254549e-05,
+      "loss": 0.907,
+      "step": 18755
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8312257441913444,
+      "learning_rate": 1.2346397400591721e-05,
+      "loss": 0.7984,
+      "step": 18756
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0213424134768845,
+      "learning_rate": 1.2345655654976485e-05,
+      "loss": 0.8581,
+      "step": 18757
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8605375267369506,
+      "learning_rate": 1.2344913895704099e-05,
+      "loss": 0.8332,
+      "step": 18758
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.822752962283348,
+      "learning_rate": 1.2344172122778882e-05,
+      "loss": 0.875,
+      "step": 18759
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4644743792229589,
+      "learning_rate": 1.234343033620515e-05,
+      "loss": 0.8098,
+      "step": 18760
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.148036966024165,
+      "learning_rate": 1.2342688535987228e-05,
+      "loss": 0.9448,
+      "step": 18761
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5266768678444596,
+      "learning_rate": 1.234194672212943e-05,
+      "loss": 0.9978,
+      "step": 18762
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.678766296767539,
+      "learning_rate": 1.234120489463608e-05,
+      "loss": 0.9476,
+      "step": 18763
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5791779768099115,
+      "learning_rate": 1.234046305351149e-05,
+      "loss": 0.8872,
+      "step": 18764
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7257271294432652,
+      "learning_rate": 1.2339721198759987e-05,
+      "loss": 0.8376,
+      "step": 18765
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.203534763932395,
+      "learning_rate": 1.2338979330385881e-05,
+      "loss": 0.8422,
+      "step": 18766
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.005168673228064,
+      "learning_rate": 1.23382374483935e-05,
+      "loss": 0.8885,
+      "step": 18767
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7092577872400916,
+      "learning_rate": 1.2337495552787157e-05,
+      "loss": 0.9929,
+      "step": 18768
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7546641692159987,
+      "learning_rate": 1.2336753643571181e-05,
+      "loss": 0.6915,
+      "step": 18769
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8314464090954636,
+      "learning_rate": 1.2336011720749881e-05,
+      "loss": 1.0079,
+      "step": 18770
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7666603105541452,
+      "learning_rate": 1.2335269784327583e-05,
+      "loss": 0.7691,
+      "step": 18771
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0921806583844185,
+      "learning_rate": 1.2334527834308606e-05,
+      "loss": 0.8195,
+      "step": 18772
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5995464934474029,
+      "learning_rate": 1.2333785870697265e-05,
+      "loss": 0.9284,
+      "step": 18773
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.226998219317238,
+      "learning_rate": 1.2333043893497886e-05,
+      "loss": 0.8442,
+      "step": 18774
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.689293684526434,
+      "learning_rate": 1.2332301902714789e-05,
+      "loss": 0.7963,
+      "step": 18775
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5928653605219782,
+      "learning_rate": 1.233155989835229e-05,
+      "loss": 0.6714,
+      "step": 18776
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.528532021723323,
+      "learning_rate": 1.233081788041471e-05,
+      "loss": 0.9418,
+      "step": 18777
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5200708418647844,
+      "learning_rate": 1.2330075848906371e-05,
+      "loss": 0.753,
+      "step": 18778
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.077304694599245,
+      "learning_rate": 1.2329333803831593e-05,
+      "loss": 0.9055,
+      "step": 18779
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5644575455785594,
+      "learning_rate": 1.2328591745194696e-05,
+      "loss": 0.908,
+      "step": 18780
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4485817242218446,
+      "learning_rate": 1.2327849673000001e-05,
+      "loss": 0.9063,
+      "step": 18781
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7927776899855739,
+      "learning_rate": 1.2327107587251827e-05,
+      "loss": 0.7805,
+      "step": 18782
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7264084475019599,
+      "learning_rate": 1.2326365487954497e-05,
+      "loss": 0.86,
+      "step": 18783
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8977253410440179,
+      "learning_rate": 1.232562337511233e-05,
+      "loss": 0.7323,
+      "step": 18784
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7389359414399823,
+      "learning_rate": 1.232488124872965e-05,
+      "loss": 0.6837,
+      "step": 18785
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8145290174167594,
+      "learning_rate": 1.2324139108810775e-05,
+      "loss": 0.9206,
+      "step": 18786
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7437538320874731,
+      "learning_rate": 1.2323396955360023e-05,
+      "loss": 0.7801,
+      "step": 18787
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.538508937907924,
+      "learning_rate": 1.232265478838172e-05,
+      "loss": 0.9032,
+      "step": 18788
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4006231670554359,
+      "learning_rate": 1.2321912607880184e-05,
+      "loss": 0.8252,
+      "step": 18789
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7926907313917528,
+      "learning_rate": 1.2321170413859742e-05,
+      "loss": 0.9567,
+      "step": 18790
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1783538593844702,
+      "learning_rate": 1.2320428206324703e-05,
+      "loss": 1.0053,
+      "step": 18791
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.576591990542889,
+      "learning_rate": 1.2319685985279406e-05,
+      "loss": 0.966,
+      "step": 18792
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.611749442823432,
+      "learning_rate": 1.2318943750728157e-05,
+      "loss": 1.072,
+      "step": 18793
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7162434032426104,
+      "learning_rate": 1.2318201502675285e-05,
+      "loss": 0.8402,
+      "step": 18794
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7590631450835388,
+      "learning_rate": 1.231745924112511e-05,
+      "loss": 0.9605,
+      "step": 18795
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5214791441860862,
+      "learning_rate": 1.231671696608195e-05,
+      "loss": 0.8964,
+      "step": 18796
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7556854385396479,
+      "learning_rate": 1.2315974677550136e-05,
+      "loss": 0.9933,
+      "step": 18797
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5508875512199185,
+      "learning_rate": 1.231523237553398e-05,
+      "loss": 0.776,
+      "step": 18798
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.591897575620431,
+      "learning_rate": 1.2314490060037812e-05,
+      "loss": 0.9182,
+      "step": 18799
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7705912904041259,
+      "learning_rate": 1.2313747731065947e-05,
+      "loss": 0.9102,
+      "step": 18800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8317213125301408,
+      "learning_rate": 1.2313005388622712e-05,
+      "loss": 0.8042,
+      "step": 18801
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5985146549784346,
+      "learning_rate": 1.2312263032712425e-05,
+      "loss": 0.9429,
+      "step": 18802
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6542260544752139,
+      "learning_rate": 1.2311520663339411e-05,
+      "loss": 0.9628,
+      "step": 18803
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.584687532312805,
+      "learning_rate": 1.2310778280507994e-05,
+      "loss": 0.7635,
+      "step": 18804
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8500577015934423,
+      "learning_rate": 1.2310035884222493e-05,
+      "loss": 0.9332,
+      "step": 18805
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.564583026088539,
+      "learning_rate": 1.2309293474487232e-05,
+      "loss": 0.8556,
+      "step": 18806
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6008375550043548,
+      "learning_rate": 1.2308551051306533e-05,
+      "loss": 0.9155,
+      "step": 18807
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7311357734422639,
+      "learning_rate": 1.230780861468472e-05,
+      "loss": 0.9023,
+      "step": 18808
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.752970143494864,
+      "learning_rate": 1.2307066164626113e-05,
+      "loss": 0.824,
+      "step": 18809
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7595045383038022,
+      "learning_rate": 1.2306323701135036e-05,
+      "loss": 0.7083,
+      "step": 18810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6216228119623575,
+      "learning_rate": 1.2305581224215812e-05,
+      "loss": 0.9249,
+      "step": 18811
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.775689525886216,
+      "learning_rate": 1.2304838733872768e-05,
+      "loss": 0.8123,
+      "step": 18812
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6641205096116747,
+      "learning_rate": 1.2304096230110219e-05,
+      "loss": 0.8516,
+      "step": 18813
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.002427371319954,
+      "learning_rate": 1.2303353712932495e-05,
+      "loss": 0.9748,
+      "step": 18814
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.479753566830966,
+      "learning_rate": 1.2302611182343915e-05,
+      "loss": 0.9163,
+      "step": 18815
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7561461918000876,
+      "learning_rate": 1.2301868638348805e-05,
+      "loss": 1.0358,
+      "step": 18816
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9604159585713294,
+      "learning_rate": 1.2301126080951484e-05,
+      "loss": 0.8071,
+      "step": 18817
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7663551401473169,
+      "learning_rate": 1.2300383510156284e-05,
+      "loss": 0.7178,
+      "step": 18818
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5467777795114848,
+      "learning_rate": 1.2299640925967518e-05,
+      "loss": 0.9408,
+      "step": 18819
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0945810309815283,
+      "learning_rate": 1.2298898328389517e-05,
+      "loss": 0.8256,
+      "step": 18820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7615644689773086,
+      "learning_rate": 1.2298155717426598e-05,
+      "loss": 0.7185,
+      "step": 18821
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7152740771482438,
+      "learning_rate": 1.2297413093083094e-05,
+      "loss": 0.7863,
+      "step": 18822
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9813763288373687,
+      "learning_rate": 1.2296670455363322e-05,
+      "loss": 0.9726,
+      "step": 18823
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8266601833995093,
+      "learning_rate": 1.2295927804271606e-05,
+      "loss": 0.938,
+      "step": 18824
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8895373618731865,
+      "learning_rate": 1.2295185139812274e-05,
+      "loss": 0.9513,
+      "step": 18825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.459341460629806,
+      "learning_rate": 1.2294442461989644e-05,
+      "loss": 0.9584,
+      "step": 18826
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8608245522727376,
+      "learning_rate": 1.2293699770808047e-05,
+      "loss": 0.8655,
+      "step": 18827
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.788943297833864,
+      "learning_rate": 1.2292957066271802e-05,
+      "loss": 0.8511,
+      "step": 18828
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6203186897581032,
+      "learning_rate": 1.2292214348385235e-05,
+      "loss": 0.8504,
+      "step": 18829
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7408899050013416,
+      "learning_rate": 1.229147161715267e-05,
+      "loss": 0.9063,
+      "step": 18830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.753664958923447,
+      "learning_rate": 1.2290728872578432e-05,
+      "loss": 0.6938,
+      "step": 18831
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6746110686747877,
+      "learning_rate": 1.2289986114666847e-05,
+      "loss": 0.7952,
+      "step": 18832
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8493443619459837,
+      "learning_rate": 1.2289243343422234e-05,
+      "loss": 0.9502,
+      "step": 18833
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6773916574888208,
+      "learning_rate": 1.2288500558848926e-05,
+      "loss": 0.9369,
+      "step": 18834
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.896697073358042,
+      "learning_rate": 1.228775776095124e-05,
+      "loss": 0.8815,
+      "step": 18835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8759252051194888,
+      "learning_rate": 1.2287014949733506e-05,
+      "loss": 0.8389,
+      "step": 18836
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.751305535755409,
+      "learning_rate": 1.2286272125200048e-05,
+      "loss": 0.8934,
+      "step": 18837
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6357272379996264,
+      "learning_rate": 1.2285529287355188e-05,
+      "loss": 0.8247,
+      "step": 18838
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.733224380965276,
+      "learning_rate": 1.2284786436203253e-05,
+      "loss": 0.8194,
+      "step": 18839
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7093850610559969,
+      "learning_rate": 1.2284043571748565e-05,
+      "loss": 0.8798,
+      "step": 18840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.183326472879425,
+      "learning_rate": 1.2283300693995459e-05,
+      "loss": 0.9764,
+      "step": 18841
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9586873686283306,
+      "learning_rate": 1.2282557802948247e-05,
+      "loss": 1.0051,
+      "step": 18842
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8319852766402478,
+      "learning_rate": 1.2281814898611267e-05,
+      "loss": 0.8702,
+      "step": 18843
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3107954628010403,
+      "learning_rate": 1.2281071980988832e-05,
+      "loss": 1.0663,
+      "step": 18844
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7227081100922335,
+      "learning_rate": 1.2280329050085276e-05,
+      "loss": 0.9352,
+      "step": 18845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7330414488368723,
+      "learning_rate": 1.2279586105904924e-05,
+      "loss": 0.6768,
+      "step": 18846
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9156287188594994,
+      "learning_rate": 1.2278843148452095e-05,
+      "loss": 1.0522,
+      "step": 18847
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.847709634098031,
+      "learning_rate": 1.2278100177731128e-05,
+      "loss": 0.8352,
+      "step": 18848
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.948398273831439,
+      "learning_rate": 1.2277357193746333e-05,
+      "loss": 0.6789,
+      "step": 18849
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6007558882670936,
+      "learning_rate": 1.2276614196502047e-05,
+      "loss": 0.9558,
+      "step": 18850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5217444530566218,
+      "learning_rate": 1.2275871186002592e-05,
+      "loss": 0.8499,
+      "step": 18851
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6316477770822129,
+      "learning_rate": 1.2275128162252297e-05,
+      "loss": 0.8228,
+      "step": 18852
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5568415193262657,
+      "learning_rate": 1.2274385125255481e-05,
+      "loss": 0.9426,
+      "step": 18853
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.514841524142983,
+      "learning_rate": 1.2273642075016476e-05,
+      "loss": 0.8539,
+      "step": 18854
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7726132977303182,
+      "learning_rate": 1.2272899011539607e-05,
+      "loss": 0.8947,
+      "step": 18855
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.558920583010346,
+      "learning_rate": 1.2272155934829203e-05,
+      "loss": 1.0026,
+      "step": 18856
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.650368340277607,
+      "learning_rate": 1.2271412844889585e-05,
+      "loss": 0.9117,
+      "step": 18857
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5976510067500467,
+      "learning_rate": 1.2270669741725083e-05,
+      "loss": 0.9547,
+      "step": 18858
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6882194196685814,
+      "learning_rate": 1.2269926625340022e-05,
+      "loss": 0.912,
+      "step": 18859
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.355704207382237,
+      "learning_rate": 1.226918349573873e-05,
+      "loss": 0.7959,
+      "step": 18860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7995795835485113,
+      "learning_rate": 1.2268440352925534e-05,
+      "loss": 0.7129,
+      "step": 18861
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.554960243609933,
+      "learning_rate": 1.226769719690476e-05,
+      "loss": 0.8589,
+      "step": 18862
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8290192203687623,
+      "learning_rate": 1.2266954027680736e-05,
+      "loss": 0.6744,
+      "step": 18863
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7470287695343942,
+      "learning_rate": 1.2266210845257787e-05,
+      "loss": 0.9847,
+      "step": 18864
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5212913245408846,
+      "learning_rate": 1.226546764964024e-05,
+      "loss": 0.9147,
+      "step": 18865
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8617849554129926,
+      "learning_rate": 1.2264724440832423e-05,
+      "loss": 0.7986,
+      "step": 18866
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7702773243685823,
+      "learning_rate": 1.2263981218838668e-05,
+      "loss": 0.9756,
+      "step": 18867
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3709462163637673,
+      "learning_rate": 1.226323798366329e-05,
+      "loss": 0.9345,
+      "step": 18868
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.614017919694362,
+      "learning_rate": 1.226249473531063e-05,
+      "loss": 0.8981,
+      "step": 18869
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6389092344083815,
+      "learning_rate": 1.2261751473785004e-05,
+      "loss": 0.9503,
+      "step": 18870
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5287986527823032,
+      "learning_rate": 1.2261008199090751e-05,
+      "loss": 0.9882,
+      "step": 18871
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7591180145117696,
+      "learning_rate": 1.2260264911232187e-05,
+      "loss": 0.7218,
+      "step": 18872
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7793277672434467,
+      "learning_rate": 1.225952161021365e-05,
+      "loss": 0.8289,
+      "step": 18873
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7988521421462111,
+      "learning_rate": 1.225877829603946e-05,
+      "loss": 0.8896,
+      "step": 18874
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8322642902917805,
+      "learning_rate": 1.2258034968713948e-05,
+      "loss": 0.754,
+      "step": 18875
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8309195898760039,
+      "learning_rate": 1.2257291628241441e-05,
+      "loss": 0.7122,
+      "step": 18876
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7394808115594629,
+      "learning_rate": 1.2256548274626267e-05,
+      "loss": 0.8702,
+      "step": 18877
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5025383823762162,
+      "learning_rate": 1.2255804907872755e-05,
+      "loss": 0.8569,
+      "step": 18878
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.980510028284554,
+      "learning_rate": 1.2255061527985235e-05,
+      "loss": 0.8748,
+      "step": 18879
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.710338498015874,
+      "learning_rate": 1.2254318134968029e-05,
+      "loss": 0.7967,
+      "step": 18880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.759475440867496,
+      "learning_rate": 1.2253574728825472e-05,
+      "loss": 0.8582,
+      "step": 18881
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.632255592427676,
+      "learning_rate": 1.2252831309561888e-05,
+      "loss": 0.9076,
+      "step": 18882
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7183644008531969,
+      "learning_rate": 1.2252087877181606e-05,
+      "loss": 0.6592,
+      "step": 18883
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3520355460579467,
+      "learning_rate": 1.2251344431688959e-05,
+      "loss": 0.9521,
+      "step": 18884
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5114100559447696,
+      "learning_rate": 1.225060097308827e-05,
+      "loss": 0.9679,
+      "step": 18885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2853125713216267,
+      "learning_rate": 1.2249857501383869e-05,
+      "loss": 0.9902,
+      "step": 18886
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7885453322784315,
+      "learning_rate": 1.2249114016580083e-05,
+      "loss": 0.727,
+      "step": 18887
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8004842112416264,
+      "learning_rate": 1.224837051868125e-05,
+      "loss": 0.8709,
+      "step": 18888
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8426766684767222,
+      "learning_rate": 1.2247627007691684e-05,
+      "loss": 0.8567,
+      "step": 18889
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.58407880931381,
+      "learning_rate": 1.2246883483615731e-05,
+      "loss": 0.9725,
+      "step": 18890
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5116463221166714,
+      "learning_rate": 1.2246139946457703e-05,
+      "loss": 0.9301,
+      "step": 18891
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4891990471212895,
+      "learning_rate": 1.2245396396221938e-05,
+      "loss": 0.9649,
+      "step": 18892
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6138148435616853,
+      "learning_rate": 1.2244652832912766e-05,
+      "loss": 0.8561,
+      "step": 18893
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.927681416013689,
+      "learning_rate": 1.2243909256534517e-05,
+      "loss": 0.8428,
+      "step": 18894
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.276981945245245,
+      "learning_rate": 1.2243165667091517e-05,
+      "loss": 0.8354,
+      "step": 18895
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3730885247459719,
+      "learning_rate": 1.224242206458809e-05,
+      "loss": 0.8468,
+      "step": 18896
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7304485343029723,
+      "learning_rate": 1.224167844902858e-05,
+      "loss": 0.7299,
+      "step": 18897
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0073923701035756,
+      "learning_rate": 1.2240934820417303e-05,
+      "loss": 0.8206,
+      "step": 18898
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.343640874157528,
+      "learning_rate": 1.2240191178758598e-05,
+      "loss": 0.9282,
+      "step": 18899
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6778152024156607,
+      "learning_rate": 1.2239447524056786e-05,
+      "loss": 0.8501,
+      "step": 18900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4816341000692548,
+      "learning_rate": 1.2238703856316208e-05,
+      "loss": 0.8401,
+      "step": 18901
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.543796679486251,
+      "learning_rate": 1.2237960175541184e-05,
+      "loss": 0.8393,
+      "step": 18902
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5522168571620332,
+      "learning_rate": 1.2237216481736047e-05,
+      "loss": 1.0454,
+      "step": 18903
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.509382773722956,
+      "learning_rate": 1.2236472774905128e-05,
+      "loss": 0.8991,
+      "step": 18904
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5349551799763026,
+      "learning_rate": 1.2235729055052757e-05,
+      "loss": 0.9168,
+      "step": 18905
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.73844012593542,
+      "learning_rate": 1.2234985322183262e-05,
+      "loss": 0.9383,
+      "step": 18906
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7402060397179455,
+      "learning_rate": 1.2234241576300974e-05,
+      "loss": 0.9457,
+      "step": 18907
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6314789305640325,
+      "learning_rate": 1.2233497817410226e-05,
+      "loss": 0.9292,
+      "step": 18908
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.166096461578911,
+      "learning_rate": 1.2232754045515348e-05,
+      "loss": 0.7934,
+      "step": 18909
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5025447816326483,
+      "learning_rate": 1.2232010260620665e-05,
+      "loss": 0.7939,
+      "step": 18910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.638252688804098,
+      "learning_rate": 1.2231266462730514e-05,
+      "loss": 0.91,
+      "step": 18911
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.575284148741023,
+      "learning_rate": 1.2230522651849224e-05,
+      "loss": 0.9058,
+      "step": 18912
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4664320565073155,
+      "learning_rate": 1.2229778827981122e-05,
+      "loss": 0.814,
+      "step": 18913
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.84730550809867,
+      "learning_rate": 1.2229034991130544e-05,
+      "loss": 0.8546,
+      "step": 18914
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5752327301739237,
+      "learning_rate": 1.2228291141301818e-05,
+      "loss": 0.9334,
+      "step": 18915
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8666003611351614,
+      "learning_rate": 1.2227547278499276e-05,
+      "loss": 0.8425,
+      "step": 18916
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5380701252147233,
+      "learning_rate": 1.222680340272725e-05,
+      "loss": 0.9159,
+      "step": 18917
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.745045685127159,
+      "learning_rate": 1.222605951399007e-05,
+      "loss": 0.8107,
+      "step": 18918
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6112556550678556,
+      "learning_rate": 1.2225315612292061e-05,
+      "loss": 0.9517,
+      "step": 18919
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4999114422683946,
+      "learning_rate": 1.2224571697637566e-05,
+      "loss": 0.8891,
+      "step": 18920
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.621040191577132,
+      "learning_rate": 1.2223827770030907e-05,
+      "loss": 0.9317,
+      "step": 18921
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7040796203177533,
+      "learning_rate": 1.2223083829476421e-05,
+      "loss": 0.9028,
+      "step": 18922
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8407935868480967,
+      "learning_rate": 1.2222339875978435e-05,
+      "loss": 0.8262,
+      "step": 18923
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8698149999218288,
+      "learning_rate": 1.2221595909541283e-05,
+      "loss": 0.9652,
+      "step": 18924
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5909333376801855,
+      "learning_rate": 1.22208519301693e-05,
+      "loss": 0.9087,
+      "step": 18925
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7295450051176063,
+      "learning_rate": 1.222010793786681e-05,
+      "loss": 0.7184,
+      "step": 18926
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6437169657243254,
+      "learning_rate": 1.2219363932638148e-05,
+      "loss": 0.8207,
+      "step": 18927
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7298469903243836,
+      "learning_rate": 1.2218619914487648e-05,
+      "loss": 0.715,
+      "step": 18928
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1582726821084974,
+      "learning_rate": 1.2217875883419642e-05,
+      "loss": 0.8707,
+      "step": 18929
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.758750391102662,
+      "learning_rate": 1.2217131839438457e-05,
+      "loss": 0.7452,
+      "step": 18930
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5290403026399473,
+      "learning_rate": 1.221638778254843e-05,
+      "loss": 0.8923,
+      "step": 18931
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5034255262089218,
+      "learning_rate": 1.2215643712753893e-05,
+      "loss": 0.8779,
+      "step": 18932
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5523205535025302,
+      "learning_rate": 1.2214899630059176e-05,
+      "loss": 0.7868,
+      "step": 18933
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7142540271825517,
+      "learning_rate": 1.2214155534468612e-05,
+      "loss": 0.9487,
+      "step": 18934
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.542415678579918,
+      "learning_rate": 1.2213411425986533e-05,
+      "loss": 1.057,
+      "step": 18935
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.794375414921963,
+      "learning_rate": 1.2212667304617272e-05,
+      "loss": 0.7678,
+      "step": 18936
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7709019998871187,
+      "learning_rate": 1.2211923170365163e-05,
+      "loss": 0.8122,
+      "step": 18937
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7078118801811113,
+      "learning_rate": 1.2211179023234534e-05,
+      "loss": 0.7452,
+      "step": 18938
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8316045837569221,
+      "learning_rate": 1.2210434863229725e-05,
+      "loss": 1.041,
+      "step": 18939
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4949383857855734,
+      "learning_rate": 1.2209690690355061e-05,
+      "loss": 0.9121,
+      "step": 18940
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5754300637167111,
+      "learning_rate": 1.2208946504614878e-05,
+      "loss": 0.8525,
+      "step": 18941
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5942169114633231,
+      "learning_rate": 1.220820230601351e-05,
+      "loss": 1.021,
+      "step": 18942
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7867084604898702,
+      "learning_rate": 1.2207458094555289e-05,
+      "loss": 0.9763,
+      "step": 18943
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5970622503365919,
+      "learning_rate": 1.2206713870244549e-05,
+      "loss": 0.8672,
+      "step": 18944
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.20252335929406,
+      "learning_rate": 1.2205969633085621e-05,
+      "loss": 0.8945,
+      "step": 18945
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8604752224932344,
+      "learning_rate": 1.2205225383082844e-05,
+      "loss": 0.7987,
+      "step": 18946
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.728152606362565,
+      "learning_rate": 1.220448112024054e-05,
+      "loss": 0.9651,
+      "step": 18947
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.675752260582192,
+      "learning_rate": 1.2203736844563055e-05,
+      "loss": 0.9623,
+      "step": 18948
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4804819077651565,
+      "learning_rate": 1.2202992556054709e-05,
+      "loss": 0.8379,
+      "step": 18949
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4948049296840242,
+      "learning_rate": 1.2202248254719849e-05,
+      "loss": 0.9039,
+      "step": 18950
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7701792483871404,
+      "learning_rate": 1.2201503940562798e-05,
+      "loss": 0.6839,
+      "step": 18951
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5231935156711915,
+      "learning_rate": 1.22007596135879e-05,
+      "loss": 0.906,
+      "step": 18952
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6284967059762496,
+      "learning_rate": 1.220001527379948e-05,
+      "loss": 0.9587,
+      "step": 18953
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5028233168459515,
+      "learning_rate": 1.2199270921201873e-05,
+      "loss": 0.8926,
+      "step": 18954
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5414226807454132,
+      "learning_rate": 1.2198526555799415e-05,
+      "loss": 0.8953,
+      "step": 18955
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.732699691779257,
+      "learning_rate": 1.219778217759644e-05,
+      "loss": 0.8358,
+      "step": 18956
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.617495259779197,
+      "learning_rate": 1.2197037786597278e-05,
+      "loss": 0.8056,
+      "step": 18957
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8845386248910005,
+      "learning_rate": 1.219629338280627e-05,
+      "loss": 0.8747,
+      "step": 18958
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1166294383940274,
+      "learning_rate": 1.2195548966227749e-05,
+      "loss": 0.8617,
+      "step": 18959
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2625064344758785,
+      "learning_rate": 1.2194804536866043e-05,
+      "loss": 0.919,
+      "step": 18960
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9243641225265649,
+      "learning_rate": 1.219406009472549e-05,
+      "loss": 0.9423,
+      "step": 18961
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7563886832203743,
+      "learning_rate": 1.2193315639810426e-05,
+      "loss": 0.9682,
+      "step": 18962
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5770495266353968,
+      "learning_rate": 1.2192571172125182e-05,
+      "loss": 0.8672,
+      "step": 18963
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6672066313584164,
+      "learning_rate": 1.2191826691674096e-05,
+      "loss": 0.7751,
+      "step": 18964
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9854115254503255,
+      "learning_rate": 1.2191082198461502e-05,
+      "loss": 0.9696,
+      "step": 18965
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.769327091426161,
+      "learning_rate": 1.2190337692491732e-05,
+      "loss": 0.8372,
+      "step": 18966
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8284962108606604,
+      "learning_rate": 1.2189593173769126e-05,
+      "loss": 0.8701,
+      "step": 18967
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8245995381807167,
+      "learning_rate": 1.218884864229801e-05,
+      "loss": 0.8581,
+      "step": 18968
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6329301164577315,
+      "learning_rate": 1.218810409808273e-05,
+      "loss": 0.9022,
+      "step": 18969
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6968968665955455,
+      "learning_rate": 1.2187359541127607e-05,
+      "loss": 0.8265,
+      "step": 18970
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7862202613881053,
+      "learning_rate": 1.2186614971436992e-05,
+      "loss": 0.8743,
+      "step": 18971
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7752305689701162,
+      "learning_rate": 1.2185870389015208e-05,
+      "loss": 0.8163,
+      "step": 18972
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.791583510180377,
+      "learning_rate": 1.21851257938666e-05,
+      "loss": 0.9493,
+      "step": 18973
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.529541640365272,
+      "learning_rate": 1.2184381185995494e-05,
+      "loss": 0.9484,
+      "step": 18974
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8077534132556344,
+      "learning_rate": 1.2183636565406227e-05,
+      "loss": 0.7807,
+      "step": 18975
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.883257982712497,
+      "learning_rate": 1.2182891932103138e-05,
+      "loss": 0.7857,
+      "step": 18976
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8661862780287509,
+      "learning_rate": 1.2182147286090562e-05,
+      "loss": 0.7175,
+      "step": 18977
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8148517488254915,
+      "learning_rate": 1.2181402627372834e-05,
+      "loss": 0.7959,
+      "step": 18978
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.863229246580306,
+      "learning_rate": 1.2180657955954287e-05,
+      "loss": 0.8497,
+      "step": 18979
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8209304481120805,
+      "learning_rate": 1.2179913271839264e-05,
+      "loss": 0.8547,
+      "step": 18980
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.74432738268484,
+      "learning_rate": 1.2179168575032088e-05,
+      "loss": 0.9543,
+      "step": 18981
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7208973054424329,
+      "learning_rate": 1.217842386553711e-05,
+      "loss": 0.6579,
+      "step": 18982
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9364310782761844,
+      "learning_rate": 1.2177679143358655e-05,
+      "loss": 0.8513,
+      "step": 18983
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.704729823002368,
+      "learning_rate": 1.2176934408501065e-05,
+      "loss": 0.8401,
+      "step": 18984
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7324841809935863,
+      "learning_rate": 1.2176189660968671e-05,
+      "loss": 0.8926,
+      "step": 18985
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6781768766050746,
+      "learning_rate": 1.2175444900765811e-05,
+      "loss": 0.8323,
+      "step": 18986
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6215586137423224,
+      "learning_rate": 1.2174700127896826e-05,
+      "loss": 0.8487,
+      "step": 18987
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7173189277029013,
+      "learning_rate": 1.2173955342366046e-05,
+      "loss": 0.9751,
+      "step": 18988
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.320182321484822,
+      "learning_rate": 1.2173210544177809e-05,
+      "loss": 0.8849,
+      "step": 18989
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6455622342665381,
+      "learning_rate": 1.2172465733336453e-05,
+      "loss": 0.7325,
+      "step": 18990
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8091265041661,
+      "learning_rate": 1.2171720909846314e-05,
+      "loss": 0.7924,
+      "step": 18991
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4386116914961635,
+      "learning_rate": 1.217097607371173e-05,
+      "loss": 0.8523,
+      "step": 18992
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.156448708801944,
+      "learning_rate": 1.2170231224937032e-05,
+      "loss": 0.9707,
+      "step": 18993
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6266954193282468,
+      "learning_rate": 1.2169486363526564e-05,
+      "loss": 0.936,
+      "step": 18994
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.781139140619632,
+      "learning_rate": 1.2168741489484656e-05,
+      "loss": 0.8222,
+      "step": 18995
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7032236879243763,
+      "learning_rate": 1.2167996602815652e-05,
+      "loss": 0.8512,
+      "step": 18996
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6345607915942417,
+      "learning_rate": 1.2167251703523886e-05,
+      "loss": 0.8405,
+      "step": 18997
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.747649669763846,
+      "learning_rate": 1.216650679161369e-05,
+      "loss": 0.7556,
+      "step": 18998
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7278373910754175,
+      "learning_rate": 1.2165761867089411e-05,
+      "loss": 0.8945,
+      "step": 18999
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.858432704342256,
+      "learning_rate": 1.2165016929955376e-05,
+      "loss": 0.8933,
+      "step": 19000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7371040513206306,
+      "learning_rate": 1.2164271980215931e-05,
+      "loss": 0.7674,
+      "step": 19001
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.816882550756993,
+      "learning_rate": 1.2163527017875406e-05,
+      "loss": 0.8773,
+      "step": 19002
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.826732846945925,
+      "learning_rate": 1.2162782042938144e-05,
+      "loss": 0.7642,
+      "step": 19003
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.694508622439773,
+      "learning_rate": 1.216203705540848e-05,
+      "loss": 0.6709,
+      "step": 19004
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.306898114402665,
+      "learning_rate": 1.2161292055290751e-05,
+      "loss": 0.8747,
+      "step": 19005
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.357791743213132,
+      "learning_rate": 1.2160547042589294e-05,
+      "loss": 0.9404,
+      "step": 19006
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.562128005449672,
+      "learning_rate": 1.2159802017308451e-05,
+      "loss": 0.9517,
+      "step": 19007
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7559858451697976,
+      "learning_rate": 1.2159056979452555e-05,
+      "loss": 0.96,
+      "step": 19008
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3286717584064824,
+      "learning_rate": 1.2158311929025945e-05,
+      "loss": 0.7973,
+      "step": 19009
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5804886349876355,
+      "learning_rate": 1.215756686603296e-05,
+      "loss": 0.8867,
+      "step": 19010
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6023554110832157,
+      "learning_rate": 1.2156821790477939e-05,
+      "loss": 0.8099,
+      "step": 19011
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6514942657229261,
+      "learning_rate": 1.2156076702365219e-05,
+      "loss": 0.8183,
+      "step": 19012
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7457687348928697,
+      "learning_rate": 1.2155331601699136e-05,
+      "loss": 0.8197,
+      "step": 19013
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1555436793600693,
+      "learning_rate": 1.2154586488484033e-05,
+      "loss": 0.9315,
+      "step": 19014
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.498416411052363,
+      "learning_rate": 1.2153841362724244e-05,
+      "loss": 0.8623,
+      "step": 19015
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.995732697918877,
+      "learning_rate": 1.2153096224424106e-05,
+      "loss": 0.9346,
+      "step": 19016
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8061767228358905,
+      "learning_rate": 1.2152351073587963e-05,
+      "loss": 0.8775,
+      "step": 19017
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6118708172399547,
+      "learning_rate": 1.2151605910220153e-05,
+      "loss": 0.9049,
+      "step": 19018
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.802468888681338,
+      "learning_rate": 1.2150860734325005e-05,
+      "loss": 0.9116,
+      "step": 19019
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5713985754578272,
+      "learning_rate": 1.2150115545906871e-05,
+      "loss": 0.7827,
+      "step": 19020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7229199989384947,
+      "learning_rate": 1.2149370344970079e-05,
+      "loss": 0.6902,
+      "step": 19021
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8484899242063695,
+      "learning_rate": 1.2148625131518976e-05,
+      "loss": 0.8848,
+      "step": 19022
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.778942440965516,
+      "learning_rate": 1.2147879905557896e-05,
+      "loss": 0.8102,
+      "step": 19023
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5559927836235439,
+      "learning_rate": 1.2147134667091181e-05,
+      "loss": 0.8662,
+      "step": 19024
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5688304518027294,
+      "learning_rate": 1.2146389416123166e-05,
+      "loss": 0.8512,
+      "step": 19025
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6608719567840429,
+      "learning_rate": 1.2145644152658191e-05,
+      "loss": 0.823,
+      "step": 19026
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6931002093907367,
+      "learning_rate": 1.2144898876700599e-05,
+      "loss": 0.7473,
+      "step": 19027
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4465543547625053,
+      "learning_rate": 1.2144153588254723e-05,
+      "loss": 0.9659,
+      "step": 19028
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5794575942635969,
+      "learning_rate": 1.2143408287324912e-05,
+      "loss": 0.9535,
+      "step": 19029
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.573395375559583,
+      "learning_rate": 1.2142662973915494e-05,
+      "loss": 0.91,
+      "step": 19030
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7771517898837352,
+      "learning_rate": 1.214191764803082e-05,
+      "loss": 0.9202,
+      "step": 19031
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6142833292342595,
+      "learning_rate": 1.2141172309675218e-05,
+      "loss": 1.0282,
+      "step": 19032
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5135190160154801,
+      "learning_rate": 1.2140426958853034e-05,
+      "loss": 0.8198,
+      "step": 19033
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.565576578858131,
+      "learning_rate": 1.2139681595568606e-05,
+      "loss": 0.8351,
+      "step": 19034
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7916259105778787,
+      "learning_rate": 1.2138936219826273e-05,
+      "loss": 0.9242,
+      "step": 19035
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7221422313160046,
+      "learning_rate": 1.2138190831630381e-05,
+      "loss": 0.9648,
+      "step": 19036
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7920378690706116,
+      "learning_rate": 1.2137445430985263e-05,
+      "loss": 0.8974,
+      "step": 19037
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5225789747235972,
+      "learning_rate": 1.213670001789526e-05,
+      "loss": 0.9093,
+      "step": 19038
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5833380016584668,
+      "learning_rate": 1.2135954592364712e-05,
+      "loss": 0.9065,
+      "step": 19039
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6269492299069332,
+      "learning_rate": 1.2135209154397962e-05,
+      "loss": 0.8649,
+      "step": 19040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7426508102852568,
+      "learning_rate": 1.2134463703999346e-05,
+      "loss": 0.7137,
+      "step": 19041
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.775090091041981,
+      "learning_rate": 1.2133718241173209e-05,
+      "loss": 0.7535,
+      "step": 19042
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8438948643767705,
+      "learning_rate": 1.213297276592389e-05,
+      "loss": 0.8756,
+      "step": 19043
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6075417487424712,
+      "learning_rate": 1.2132227278255727e-05,
+      "loss": 0.9977,
+      "step": 19044
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.817350188302775,
+      "learning_rate": 1.213148177817306e-05,
+      "loss": 0.8141,
+      "step": 19045
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7872294566063491,
+      "learning_rate": 1.2130736265680231e-05,
+      "loss": 0.8771,
+      "step": 19046
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9678522965772367,
+      "learning_rate": 1.2129990740781583e-05,
+      "loss": 0.8916,
+      "step": 19047
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5487226982707725,
+      "learning_rate": 1.2129245203481457e-05,
+      "loss": 0.8845,
+      "step": 19048
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6482433855154908,
+      "learning_rate": 1.2128499653784184e-05,
+      "loss": 0.8731,
+      "step": 19049
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8509314952479994,
+      "learning_rate": 1.2127754091694119e-05,
+      "loss": 0.7985,
+      "step": 19050
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8986600370871547,
+      "learning_rate": 1.2127008517215593e-05,
+      "loss": 0.7975,
+      "step": 19051
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3746097474688497,
+      "learning_rate": 1.2126262930352953e-05,
+      "loss": 0.8047,
+      "step": 19052
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6348853336619944,
+      "learning_rate": 1.2125517331110532e-05,
+      "loss": 1.0022,
+      "step": 19053
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4467153634085683,
+      "learning_rate": 1.2124771719492681e-05,
+      "loss": 0.8798,
+      "step": 19054
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.97091098652131,
+      "learning_rate": 1.2124026095503735e-05,
+      "loss": 0.9096,
+      "step": 19055
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4595836671048288,
+      "learning_rate": 1.2123280459148034e-05,
+      "loss": 0.849,
+      "step": 19056
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8213057092661384,
+      "learning_rate": 1.2122534810429922e-05,
+      "loss": 0.9349,
+      "step": 19057
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7568488667906523,
+      "learning_rate": 1.2121789149353745e-05,
+      "loss": 0.7675,
+      "step": 19058
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7933276476275699,
+      "learning_rate": 1.2121043475923834e-05,
+      "loss": 0.6805,
+      "step": 19059
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.50298943064105,
+      "learning_rate": 1.2120297790144539e-05,
+      "loss": 0.9272,
+      "step": 19060
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7803686958598357,
+      "learning_rate": 1.21195520920202e-05,
+      "loss": 0.9251,
+      "step": 19061
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4900156591356146,
+      "learning_rate": 1.2118806381555157e-05,
+      "loss": 0.892,
+      "step": 19062
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4966323399138959,
+      "learning_rate": 1.2118060658753751e-05,
+      "loss": 0.9445,
+      "step": 19063
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6503346375838563,
+      "learning_rate": 1.2117314923620326e-05,
+      "loss": 0.8822,
+      "step": 19064
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3796679401133924,
+      "learning_rate": 1.2116569176159225e-05,
+      "loss": 0.82,
+      "step": 19065
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.49070620605821,
+      "learning_rate": 1.2115823416374788e-05,
+      "loss": 0.7684,
+      "step": 19066
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6842882556541627,
+      "learning_rate": 1.2115077644271356e-05,
+      "loss": 0.9278,
+      "step": 19067
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4473746840582349,
+      "learning_rate": 1.2114331859853272e-05,
+      "loss": 1.026,
+      "step": 19068
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5255646400692098,
+      "learning_rate": 1.211358606312488e-05,
+      "loss": 0.9073,
+      "step": 19069
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7245515321267628,
+      "learning_rate": 1.2112840254090519e-05,
+      "loss": 0.8968,
+      "step": 19070
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9756792846096674,
+      "learning_rate": 1.2112094432754536e-05,
+      "loss": 0.9117,
+      "step": 19071
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4627425595692094,
+      "learning_rate": 1.2111348599121266e-05,
+      "loss": 0.9015,
+      "step": 19072
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8062149286345496,
+      "learning_rate": 1.211060275319506e-05,
+      "loss": 0.8147,
+      "step": 19073
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9079389243435072,
+      "learning_rate": 1.2109856894980254e-05,
+      "loss": 0.8556,
+      "step": 19074
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.579394740452298,
+      "learning_rate": 1.2109111024481194e-05,
+      "loss": 0.8956,
+      "step": 19075
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.876616059262878,
+      "learning_rate": 1.2108365141702224e-05,
+      "loss": 0.8761,
+      "step": 19076
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4500863248328475,
+      "learning_rate": 1.2107619246647679e-05,
+      "loss": 0.9245,
+      "step": 19077
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5676410862591978,
+      "learning_rate": 1.2106873339321914e-05,
+      "loss": 0.9297,
+      "step": 19078
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7528731664530767,
+      "learning_rate": 1.210612741972926e-05,
+      "loss": 0.7107,
+      "step": 19079
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.416453844973421,
+      "learning_rate": 1.210538148787407e-05,
+      "loss": 0.8163,
+      "step": 19080
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7191592411010286,
+      "learning_rate": 1.210463554376068e-05,
+      "loss": 0.9019,
+      "step": 19081
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7321945275641242,
+      "learning_rate": 1.2103889587393437e-05,
+      "loss": 0.8611,
+      "step": 19082
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5597322685160244,
+      "learning_rate": 1.2103143618776682e-05,
+      "loss": 0.9299,
+      "step": 19083
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6111749566495586,
+      "learning_rate": 1.2102397637914758e-05,
+      "loss": 0.9134,
+      "step": 19084
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.797964727438082,
+      "learning_rate": 1.210165164481201e-05,
+      "loss": 0.7861,
+      "step": 19085
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.822031478743889,
+      "learning_rate": 1.210090563947278e-05,
+      "loss": 0.7943,
+      "step": 19086
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6079020293452067,
+      "learning_rate": 1.2100159621901413e-05,
+      "loss": 0.899,
+      "step": 19087
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8620651323522791,
+      "learning_rate": 1.2099413592102255e-05,
+      "loss": 0.7481,
+      "step": 19088
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8159372288308664,
+      "learning_rate": 1.2098667550079642e-05,
+      "loss": 0.7681,
+      "step": 19089
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.787071029308371,
+      "learning_rate": 1.2097921495837923e-05,
+      "loss": 0.9444,
+      "step": 19090
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9501513389387148,
+      "learning_rate": 1.209717542938144e-05,
+      "loss": 0.9743,
+      "step": 19091
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6320767621418228,
+      "learning_rate": 1.2096429350714538e-05,
+      "loss": 0.8258,
+      "step": 19092
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6005490764940478,
+      "learning_rate": 1.2095683259841562e-05,
+      "loss": 0.8531,
+      "step": 19093
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7285413163461134,
+      "learning_rate": 1.2094937156766855e-05,
+      "loss": 0.6985,
+      "step": 19094
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5299553865667193,
+      "learning_rate": 1.2094191041494759e-05,
+      "loss": 0.9059,
+      "step": 19095
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.660979416718578,
+      "learning_rate": 1.2093444914029622e-05,
+      "loss": 0.9716,
+      "step": 19096
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7006406518176092,
+      "learning_rate": 1.2092698774375786e-05,
+      "loss": 0.8792,
+      "step": 19097
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4906870101377232,
+      "learning_rate": 1.209195262253759e-05,
+      "loss": 0.8705,
+      "step": 19098
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6065488455840389,
+      "learning_rate": 1.2091206458519389e-05,
+      "loss": 0.8994,
+      "step": 19099
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6044306160265676,
+      "learning_rate": 1.2090460282325518e-05,
+      "loss": 0.8716,
+      "step": 19100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7499197738349286,
+      "learning_rate": 1.208971409396033e-05,
+      "loss": 0.9022,
+      "step": 19101
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5114527186669502,
+      "learning_rate": 1.2088967893428161e-05,
+      "loss": 0.6929,
+      "step": 19102
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8147523587579628,
+      "learning_rate": 1.2088221680733362e-05,
+      "loss": 0.8778,
+      "step": 19103
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5335100655426954,
+      "learning_rate": 1.2087475455880272e-05,
+      "loss": 0.9031,
+      "step": 19104
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.388623051486047,
+      "learning_rate": 1.2086729218873242e-05,
+      "loss": 0.9897,
+      "step": 19105
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8222852266773766,
+      "learning_rate": 1.2085982969716615e-05,
+      "loss": 0.832,
+      "step": 19106
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7842255627163583,
+      "learning_rate": 1.208523670841473e-05,
+      "loss": 0.9772,
+      "step": 19107
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6349559345087359,
+      "learning_rate": 1.2084490434971942e-05,
+      "loss": 0.9397,
+      "step": 19108
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6876136697832564,
+      "learning_rate": 1.2083744149392586e-05,
+      "loss": 0.906,
+      "step": 19109
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7558440825960122,
+      "learning_rate": 1.2082997851681015e-05,
+      "loss": 0.827,
+      "step": 19110
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4651105872145742,
+      "learning_rate": 1.2082251541841573e-05,
+      "loss": 0.8428,
+      "step": 19111
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5137949978452792,
+      "learning_rate": 1.2081505219878597e-05,
+      "loss": 0.7988,
+      "step": 19112
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.750256809422099,
+      "learning_rate": 1.2080758885796444e-05,
+      "loss": 0.8319,
+      "step": 19113
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.627888526754419,
+      "learning_rate": 1.208001253959945e-05,
+      "loss": 0.8153,
+      "step": 19114
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8412250833305837,
+      "learning_rate": 1.207926618129197e-05,
+      "loss": 0.8517,
+      "step": 19115
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2367658482668777,
+      "learning_rate": 1.207851981087834e-05,
+      "loss": 0.8383,
+      "step": 19116
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8492794537085508,
+      "learning_rate": 1.2077773428362909e-05,
+      "loss": 0.8476,
+      "step": 19117
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7148414492455908,
+      "learning_rate": 1.2077027033750025e-05,
+      "loss": 0.9625,
+      "step": 19118
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.60167463707066,
+      "learning_rate": 1.207628062704403e-05,
+      "loss": 0.9048,
+      "step": 19119
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4843986117216708,
+      "learning_rate": 1.2075534208249275e-05,
+      "loss": 0.8582,
+      "step": 19120
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9779595391042328,
+      "learning_rate": 1.2074787777370101e-05,
+      "loss": 0.874,
+      "step": 19121
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6544981792694622,
+      "learning_rate": 1.2074041334410858e-05,
+      "loss": 0.9036,
+      "step": 19122
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6624391836138381,
+      "learning_rate": 1.2073294879375883e-05,
+      "loss": 0.8895,
+      "step": 19123
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.664121868455207,
+      "learning_rate": 1.2072548412269536e-05,
+      "loss": 0.9333,
+      "step": 19124
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5515354231315914,
+      "learning_rate": 1.2071801933096153e-05,
+      "loss": 0.9735,
+      "step": 19125
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5195399265498202,
+      "learning_rate": 1.2071055441860082e-05,
+      "loss": 0.9398,
+      "step": 19126
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6378351666060103,
+      "learning_rate": 1.2070308938565674e-05,
+      "loss": 0.87,
+      "step": 19127
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.51757879394117,
+      "learning_rate": 1.2069562423217268e-05,
+      "loss": 0.8603,
+      "step": 19128
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5298313188276993,
+      "learning_rate": 1.2068815895819219e-05,
+      "loss": 0.9251,
+      "step": 19129
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.797694980690013,
+      "learning_rate": 1.2068069356375864e-05,
+      "loss": 0.8663,
+      "step": 19130
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.802077175739465,
+      "learning_rate": 1.2067322804891556e-05,
+      "loss": 0.7537,
+      "step": 19131
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.791073574385891,
+      "learning_rate": 1.206657624137064e-05,
+      "loss": 1.0286,
+      "step": 19132
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8469606853470912,
+      "learning_rate": 1.2065829665817465e-05,
+      "loss": 0.8765,
+      "step": 19133
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7464716828200113,
+      "learning_rate": 1.2065083078236375e-05,
+      "loss": 0.9739,
+      "step": 19134
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4613224251923593,
+      "learning_rate": 1.2064336478631717e-05,
+      "loss": 1.0001,
+      "step": 19135
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4562490162160522,
+      "learning_rate": 1.2063589867007838e-05,
+      "loss": 0.8993,
+      "step": 19136
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5622033663711952,
+      "learning_rate": 1.2062843243369086e-05,
+      "loss": 0.877,
+      "step": 19137
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1336560186983125,
+      "learning_rate": 1.2062096607719809e-05,
+      "loss": 0.8423,
+      "step": 19138
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5126354889247424,
+      "learning_rate": 1.206134996006435e-05,
+      "loss": 0.869,
+      "step": 19139
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.787613071563001,
+      "learning_rate": 1.206060330040706e-05,
+      "loss": 0.7977,
+      "step": 19140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6687379556018747,
+      "learning_rate": 1.2059856628752285e-05,
+      "loss": 0.9669,
+      "step": 19141
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7782222589941146,
+      "learning_rate": 1.2059109945104373e-05,
+      "loss": 0.8162,
+      "step": 19142
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.827227260664717,
+      "learning_rate": 1.2058363249467671e-05,
+      "loss": 0.8001,
+      "step": 19143
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8291822563712963,
+      "learning_rate": 1.2057616541846526e-05,
+      "loss": 0.9157,
+      "step": 19144
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7179979833737973,
+      "learning_rate": 1.2056869822245285e-05,
+      "loss": 0.7847,
+      "step": 19145
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5413289917353705,
+      "learning_rate": 1.20561230906683e-05,
+      "loss": 0.8801,
+      "step": 19146
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8267172172066228,
+      "learning_rate": 1.2055376347119912e-05,
+      "loss": 0.8192,
+      "step": 19147
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.578052189808301,
+      "learning_rate": 1.2054629591604474e-05,
+      "loss": 0.8535,
+      "step": 19148
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7216792881729662,
+      "learning_rate": 1.2053882824126331e-05,
+      "loss": 0.7331,
+      "step": 19149
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8834587203373396,
+      "learning_rate": 1.2053136044689834e-05,
+      "loss": 0.9262,
+      "step": 19150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9226169577468053,
+      "learning_rate": 1.2052389253299326e-05,
+      "loss": 0.8532,
+      "step": 19151
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6151143372176722,
+      "learning_rate": 1.2051642449959162e-05,
+      "loss": 0.8636,
+      "step": 19152
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7758624670424432,
+      "learning_rate": 1.2050895634673681e-05,
+      "loss": 0.8482,
+      "step": 19153
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7773057723802113,
+      "learning_rate": 1.2050148807447242e-05,
+      "loss": 0.9411,
+      "step": 19154
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.748109508066711,
+      "learning_rate": 1.2049401968284182e-05,
+      "loss": 0.7018,
+      "step": 19155
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0420327944636925,
+      "learning_rate": 1.2048655117188859e-05,
+      "loss": 0.9253,
+      "step": 19156
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7850479746632525,
+      "learning_rate": 1.2047908254165614e-05,
+      "loss": 0.7487,
+      "step": 19157
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7619013219072652,
+      "learning_rate": 1.20471613792188e-05,
+      "loss": 0.767,
+      "step": 19158
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6625077808170767,
+      "learning_rate": 1.2046414492352766e-05,
+      "loss": 0.9611,
+      "step": 19159
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.917394864748979,
+      "learning_rate": 1.2045667593571858e-05,
+      "loss": 0.9758,
+      "step": 19160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6735108335216355,
+      "learning_rate": 1.2044920682880425e-05,
+      "loss": 0.8901,
+      "step": 19161
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.603814738838826,
+      "learning_rate": 1.2044173760282818e-05,
+      "loss": 0.8406,
+      "step": 19162
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3235144346169987,
+      "learning_rate": 1.2043426825783383e-05,
+      "loss": 0.834,
+      "step": 19163
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5461168157344274,
+      "learning_rate": 1.204267987938647e-05,
+      "loss": 0.8131,
+      "step": 19164
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.25993264822294,
+      "learning_rate": 1.2041932921096427e-05,
+      "loss": 0.7371,
+      "step": 19165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.613829476935499,
+      "learning_rate": 1.2041185950917604e-05,
+      "loss": 0.7783,
+      "step": 19166
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8797981118414935,
+      "learning_rate": 1.2040438968854354e-05,
+      "loss": 0.8774,
+      "step": 19167
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5651656938188379,
+      "learning_rate": 1.2039691974911019e-05,
+      "loss": 0.8938,
+      "step": 19168
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.344454193695088,
+      "learning_rate": 1.2038944969091953e-05,
+      "loss": 0.9285,
+      "step": 19169
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1480788255802006,
+      "learning_rate": 1.20381979514015e-05,
+      "loss": 0.9202,
+      "step": 19170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.545764256443997,
+      "learning_rate": 1.203745092184402e-05,
+      "loss": 0.8605,
+      "step": 19171
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7499030069510881,
+      "learning_rate": 1.203670388042385e-05,
+      "loss": 0.8689,
+      "step": 19172
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7792106550247315,
+      "learning_rate": 1.203595682714535e-05,
+      "loss": 0.7441,
+      "step": 19173
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8646184086754105,
+      "learning_rate": 1.2035209762012858e-05,
+      "loss": 0.9196,
+      "step": 19174
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7979958444746833,
+      "learning_rate": 1.2034462685030734e-05,
+      "loss": 0.7835,
+      "step": 19175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9776530432378199,
+      "learning_rate": 1.2033715596203326e-05,
+      "loss": 0.9572,
+      "step": 19176
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6081571960633019,
+      "learning_rate": 1.203296849553498e-05,
+      "loss": 0.9506,
+      "step": 19177
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.456565858212386,
+      "learning_rate": 1.203222138303005e-05,
+      "loss": 0.8261,
+      "step": 19178
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.669412120389225,
+      "learning_rate": 1.2031474258692878e-05,
+      "loss": 0.8736,
+      "step": 19179
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.901798459473018,
+      "learning_rate": 1.2030727122527828e-05,
+      "loss": 0.9,
+      "step": 19180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6452072546116987,
+      "learning_rate": 1.2029979974539233e-05,
+      "loss": 0.8927,
+      "step": 19181
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7292989726601309,
+      "learning_rate": 1.202923281473146e-05,
+      "loss": 0.8909,
+      "step": 19182
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7258828542045924,
+      "learning_rate": 1.2028485643108848e-05,
+      "loss": 0.7199,
+      "step": 19183
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7618507093614435,
+      "learning_rate": 1.202773845967575e-05,
+      "loss": 0.9625,
+      "step": 19184
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7889768921472264,
+      "learning_rate": 1.2026991264436516e-05,
+      "loss": 0.912,
+      "step": 19185
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4923901722014443,
+      "learning_rate": 1.2026244057395497e-05,
+      "loss": 0.8901,
+      "step": 19186
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4341672657741094,
+      "learning_rate": 1.2025496838557045e-05,
+      "loss": 0.9514,
+      "step": 19187
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6580595561361462,
+      "learning_rate": 1.202474960792551e-05,
+      "loss": 0.8001,
+      "step": 19188
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5744084863288317,
+      "learning_rate": 1.2024002365505238e-05,
+      "loss": 0.8204,
+      "step": 19189
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.724771427651878,
+      "learning_rate": 1.2023255111300587e-05,
+      "loss": 0.7168,
+      "step": 19190
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6445374564104125,
+      "learning_rate": 1.2022507845315903e-05,
+      "loss": 0.8711,
+      "step": 19191
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6072926830234795,
+      "learning_rate": 1.2021760567555539e-05,
+      "loss": 0.7755,
+      "step": 19192
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.753857923792347,
+      "learning_rate": 1.2021013278023843e-05,
+      "loss": 0.7555,
+      "step": 19193
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2032479354553693,
+      "learning_rate": 1.2020265976725167e-05,
+      "loss": 0.8973,
+      "step": 19194
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2941553757351,
+      "learning_rate": 1.2019518663663865e-05,
+      "loss": 0.866,
+      "step": 19195
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0025274344993753,
+      "learning_rate": 1.2018771338844287e-05,
+      "loss": 0.8801,
+      "step": 19196
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.680960060564954,
+      "learning_rate": 1.201802400227078e-05,
+      "loss": 0.9347,
+      "step": 19197
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9942437245929003,
+      "learning_rate": 1.2017276653947701e-05,
+      "loss": 0.915,
+      "step": 19198
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5990767959836436,
+      "learning_rate": 1.20165292938794e-05,
+      "loss": 0.9051,
+      "step": 19199
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.620634001470352,
+      "learning_rate": 1.2015781922070223e-05,
+      "loss": 0.9464,
+      "step": 19200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5564422087217176,
+      "learning_rate": 1.2015034538524528e-05,
+      "loss": 0.8951,
+      "step": 19201
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.799706658491459,
+      "learning_rate": 1.2014287143246662e-05,
+      "loss": 0.7941,
+      "step": 19202
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7281099857365615,
+      "learning_rate": 1.201353973624098e-05,
+      "loss": 0.6794,
+      "step": 19203
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6806476378074673,
+      "learning_rate": 1.201279231751183e-05,
+      "loss": 0.967,
+      "step": 19204
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8639636535622495,
+      "learning_rate": 1.201204488706357e-05,
+      "loss": 0.8645,
+      "step": 19205
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6917357450584831,
+      "learning_rate": 1.2011297444900545e-05,
+      "loss": 0.8232,
+      "step": 19206
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4799788170849033,
+      "learning_rate": 1.2010549991027109e-05,
+      "loss": 0.8589,
+      "step": 19207
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8734944876386996,
+      "learning_rate": 1.2009802525447616e-05,
+      "loss": 0.9701,
+      "step": 19208
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5751574457276687,
+      "learning_rate": 1.2009055048166415e-05,
+      "loss": 0.8918,
+      "step": 19209
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7483385545166708,
+      "learning_rate": 1.2008307559187862e-05,
+      "loss": 0.8387,
+      "step": 19210
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.901926961884558,
+      "learning_rate": 1.2007560058516301e-05,
+      "loss": 0.9052,
+      "step": 19211
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.14490644439126,
+      "learning_rate": 1.2006812546156097e-05,
+      "loss": 1.0235,
+      "step": 19212
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5111045915934875,
+      "learning_rate": 1.2006065022111591e-05,
+      "loss": 0.9461,
+      "step": 19213
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8083600570162663,
+      "learning_rate": 1.200531748638714e-05,
+      "loss": 0.836,
+      "step": 19214
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4791761097271146,
+      "learning_rate": 1.2004569938987093e-05,
+      "loss": 0.8702,
+      "step": 19215
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0645754315601637,
+      "learning_rate": 1.200382237991581e-05,
+      "loss": 0.7622,
+      "step": 19216
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6570219705909794,
+      "learning_rate": 1.2003074809177635e-05,
+      "loss": 0.8836,
+      "step": 19217
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6885702093316546,
+      "learning_rate": 1.2002327226776926e-05,
+      "loss": 0.7664,
+      "step": 19218
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4967354849627068,
+      "learning_rate": 1.2001579632718034e-05,
+      "loss": 0.8876,
+      "step": 19219
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.148863081260056,
+      "learning_rate": 1.2000832027005314e-05,
+      "loss": 0.8423,
+      "step": 19220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8167523046918537,
+      "learning_rate": 1.200008440964311e-05,
+      "loss": 0.8905,
+      "step": 19221
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8325245720572299,
+      "learning_rate": 1.1999336780635788e-05,
+      "loss": 0.6952,
+      "step": 19222
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.941572090112453,
+      "learning_rate": 1.199858913998769e-05,
+      "loss": 0.7779,
+      "step": 19223
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7170744613443505,
+      "learning_rate": 1.1997841487703175e-05,
+      "loss": 0.8213,
+      "step": 19224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.807565366459705,
+      "learning_rate": 1.1997093823786594e-05,
+      "loss": 1.0354,
+      "step": 19225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7561226248597402,
+      "learning_rate": 1.1996346148242299e-05,
+      "loss": 0.7217,
+      "step": 19226
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.170855244605141,
+      "learning_rate": 1.1995598461074647e-05,
+      "loss": 0.874,
+      "step": 19227
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4469699268969578,
+      "learning_rate": 1.1994850762287988e-05,
+      "loss": 0.8112,
+      "step": 19228
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.828609080987825,
+      "learning_rate": 1.199410305188668e-05,
+      "loss": 0.965,
+      "step": 19229
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7478083770824262,
+      "learning_rate": 1.1993355329875067e-05,
+      "loss": 0.6835,
+      "step": 19230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6636032181950633,
+      "learning_rate": 1.1992607596257513e-05,
+      "loss": 0.9184,
+      "step": 19231
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.496129909137086,
+      "learning_rate": 1.1991859851038362e-05,
+      "loss": 0.75,
+      "step": 19232
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5048246027777408,
+      "learning_rate": 1.1991112094221975e-05,
+      "loss": 0.8418,
+      "step": 19233
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9793011193684478,
+      "learning_rate": 1.19903643258127e-05,
+      "loss": 0.8073,
+      "step": 19234
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7858573395754209,
+      "learning_rate": 1.1989616545814899e-05,
+      "loss": 0.7562,
+      "step": 19235
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5927907686895655,
+      "learning_rate": 1.198886875423292e-05,
+      "loss": 0.942,
+      "step": 19236
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.253054597093482,
+      "learning_rate": 1.1988120951071113e-05,
+      "loss": 0.8889,
+      "step": 19237
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.771953945141259,
+      "learning_rate": 1.198737313633384e-05,
+      "loss": 0.7769,
+      "step": 19238
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.585946014871342,
+      "learning_rate": 1.1986625310025451e-05,
+      "loss": 0.8952,
+      "step": 19239
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.751732829028008,
+      "learning_rate": 1.1985877472150299e-05,
+      "loss": 0.8873,
+      "step": 19240
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7647932557371329,
+      "learning_rate": 1.1985129622712738e-05,
+      "loss": 0.8329,
+      "step": 19241
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.131444627032274,
+      "learning_rate": 1.1984381761717127e-05,
+      "loss": 0.916,
+      "step": 19242
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.661115874009077,
+      "learning_rate": 1.1983633889167815e-05,
+      "loss": 0.9289,
+      "step": 19243
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9111942504239396,
+      "learning_rate": 1.198288600506916e-05,
+      "loss": 0.7586,
+      "step": 19244
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.847641384882033,
+      "learning_rate": 1.1982138109425516e-05,
+      "loss": 0.9015,
+      "step": 19245
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.324474653838885,
+      "learning_rate": 1.1981390202241235e-05,
+      "loss": 0.7971,
+      "step": 19246
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.63093470649996,
+      "learning_rate": 1.1980642283520672e-05,
+      "loss": 0.8905,
+      "step": 19247
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.703010534035725,
+      "learning_rate": 1.1979894353268185e-05,
+      "loss": 0.7223,
+      "step": 19248
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8905128436069696,
+      "learning_rate": 1.1979146411488122e-05,
+      "loss": 0.8566,
+      "step": 19249
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0910812947897215,
+      "learning_rate": 1.1978398458184848e-05,
+      "loss": 0.9726,
+      "step": 19250
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9955612946302406,
+      "learning_rate": 1.1977650493362708e-05,
+      "loss": 0.8587,
+      "step": 19251
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7874504712386653,
+      "learning_rate": 1.1976902517026063e-05,
+      "loss": 0.6762,
+      "step": 19252
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.506933829888821,
+      "learning_rate": 1.1976154529179263e-05,
+      "loss": 0.8388,
+      "step": 19253
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9705078727816874,
+      "learning_rate": 1.1975406529826668e-05,
+      "loss": 0.7836,
+      "step": 19254
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.818487971155155,
+      "learning_rate": 1.1974658518972625e-05,
+      "loss": 0.9625,
+      "step": 19255
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4712474270608367,
+      "learning_rate": 1.1973910496621503e-05,
+      "loss": 0.783,
+      "step": 19256
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.014713959072228,
+      "learning_rate": 1.1973162462777645e-05,
+      "loss": 1.0517,
+      "step": 19257
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.033768280129165,
+      "learning_rate": 1.197241441744541e-05,
+      "loss": 0.8213,
+      "step": 19258
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5235590212486065,
+      "learning_rate": 1.1971666360629157e-05,
+      "loss": 1.0032,
+      "step": 19259
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7657321563687088,
+      "learning_rate": 1.1970918292333234e-05,
+      "loss": 0.7646,
+      "step": 19260
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.687564681182811,
+      "learning_rate": 1.1970170212562004e-05,
+      "loss": 0.8624,
+      "step": 19261
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6132216821037264,
+      "learning_rate": 1.1969422121319815e-05,
+      "loss": 0.8665,
+      "step": 19262
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6000365886073424,
+      "learning_rate": 1.1968674018611033e-05,
+      "loss": 0.9802,
+      "step": 19263
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5219091802821867,
+      "learning_rate": 1.1967925904440005e-05,
+      "loss": 0.9394,
+      "step": 19264
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5995036828846634,
+      "learning_rate": 1.1967177778811087e-05,
+      "loss": 0.946,
+      "step": 19265
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.195407160131043,
+      "learning_rate": 1.1966429641728641e-05,
+      "loss": 0.8726,
+      "step": 19266
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7663537968254819,
+      "learning_rate": 1.1965681493197017e-05,
+      "loss": 0.7395,
+      "step": 19267
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0115867472772115,
+      "learning_rate": 1.1964933333220573e-05,
+      "loss": 0.8066,
+      "step": 19268
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4661729539810688,
+      "learning_rate": 1.1964185161803666e-05,
+      "loss": 0.8727,
+      "step": 19269
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0889302524559574,
+      "learning_rate": 1.196343697895065e-05,
+      "loss": 0.8887,
+      "step": 19270
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5611662556077468,
+      "learning_rate": 1.1962688784665883e-05,
+      "loss": 0.9196,
+      "step": 19271
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5966063077751202,
+      "learning_rate": 1.1961940578953719e-05,
+      "loss": 0.8383,
+      "step": 19272
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6978102607777725,
+      "learning_rate": 1.1961192361818517e-05,
+      "loss": 0.9027,
+      "step": 19273
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.579727009915649,
+      "learning_rate": 1.1960444133264632e-05,
+      "loss": 0.9735,
+      "step": 19274
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6926976270940768,
+      "learning_rate": 1.1959695893296421e-05,
+      "loss": 0.8661,
+      "step": 19275
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7129806810710833,
+      "learning_rate": 1.1958947641918239e-05,
+      "loss": 0.8478,
+      "step": 19276
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5735882470836104,
+      "learning_rate": 1.1958199379134445e-05,
+      "loss": 0.8474,
+      "step": 19277
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7821076343051124,
+      "learning_rate": 1.1957451104949396e-05,
+      "loss": 0.7473,
+      "step": 19278
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7937143218964368,
+      "learning_rate": 1.1956702819367442e-05,
+      "loss": 0.691,
+      "step": 19279
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7805733464481158,
+      "learning_rate": 1.195595452239295e-05,
+      "loss": 0.7656,
+      "step": 19280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6603760677378232,
+      "learning_rate": 1.1955206214030266e-05,
+      "loss": 0.8904,
+      "step": 19281
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4246781580319736,
+      "learning_rate": 1.1954457894283758e-05,
+      "loss": 0.8203,
+      "step": 19282
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7071156882569813,
+      "learning_rate": 1.1953709563157771e-05,
+      "loss": 0.7613,
+      "step": 19283
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.515803869833053,
+      "learning_rate": 1.1952961220656676e-05,
+      "loss": 0.8666,
+      "step": 19284
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8570926391476106,
+      "learning_rate": 1.1952212866784819e-05,
+      "loss": 0.9539,
+      "step": 19285
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4558465032552887,
+      "learning_rate": 1.1951464501546559e-05,
+      "loss": 0.8345,
+      "step": 19286
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5953736560239737,
+      "learning_rate": 1.1950716124946256e-05,
+      "loss": 0.8949,
+      "step": 19287
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.484893581974365,
+      "learning_rate": 1.1949967736988265e-05,
+      "loss": 0.9771,
+      "step": 19288
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7632365733786,
+      "learning_rate": 1.1949219337676948e-05,
+      "loss": 0.9229,
+      "step": 19289
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5255233023363945,
+      "learning_rate": 1.1948470927016656e-05,
+      "loss": 0.9752,
+      "step": 19290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2740398622799733,
+      "learning_rate": 1.1947722505011749e-05,
+      "loss": 0.9158,
+      "step": 19291
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.501502025810659,
+      "learning_rate": 1.1946974071666587e-05,
+      "loss": 0.8303,
+      "step": 19292
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5563314840917852,
+      "learning_rate": 1.1946225626985523e-05,
+      "loss": 0.9443,
+      "step": 19293
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5937752453436966,
+      "learning_rate": 1.194547717097292e-05,
+      "loss": 0.8681,
+      "step": 19294
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8031476884789217,
+      "learning_rate": 1.194472870363313e-05,
+      "loss": 0.8738,
+      "step": 19295
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7103206521952894,
+      "learning_rate": 1.1943980224970516e-05,
+      "loss": 0.8661,
+      "step": 19296
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.615453625787488,
+      "learning_rate": 1.1943231734989434e-05,
+      "loss": 0.7856,
+      "step": 19297
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4878479785361367,
+      "learning_rate": 1.194248323369424e-05,
+      "loss": 0.8758,
+      "step": 19298
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5703604120624466,
+      "learning_rate": 1.1941734721089293e-05,
+      "loss": 0.8082,
+      "step": 19299
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4280220825826484,
+      "learning_rate": 1.1940986197178955e-05,
+      "loss": 0.9826,
+      "step": 19300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8056408272990684,
+      "learning_rate": 1.1940237661967579e-05,
+      "loss": 0.7391,
+      "step": 19301
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8206104471889135,
+      "learning_rate": 1.1939489115459523e-05,
+      "loss": 0.7481,
+      "step": 19302
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.244732592131581,
+      "learning_rate": 1.1938740557659153e-05,
+      "loss": 0.9471,
+      "step": 19303
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7950093603255886,
+      "learning_rate": 1.1937991988570815e-05,
+      "loss": 0.8543,
+      "step": 19304
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.610912189733725,
+      "learning_rate": 1.193724340819888e-05,
+      "loss": 0.8838,
+      "step": 19305
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.976944079005192,
+      "learning_rate": 1.1936494816547695e-05,
+      "loss": 0.8408,
+      "step": 19306
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.697292107946424,
+      "learning_rate": 1.1935746213621629e-05,
+      "loss": 1.0748,
+      "step": 19307
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5356125435962091,
+      "learning_rate": 1.1934997599425034e-05,
+      "loss": 0.9922,
+      "step": 19308
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.673950768365444,
+      "learning_rate": 1.1934248973962266e-05,
+      "loss": 0.8481,
+      "step": 19309
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4771539777377958,
+      "learning_rate": 1.1933500337237693e-05,
+      "loss": 0.8684,
+      "step": 19310
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.503627434362618,
+      "learning_rate": 1.1932751689255666e-05,
+      "loss": 0.8998,
+      "step": 19311
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.546010490171173,
+      "learning_rate": 1.1932003030020552e-05,
+      "loss": 0.8933,
+      "step": 19312
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.64117991216531,
+      "learning_rate": 1.1931254359536698e-05,
+      "loss": 0.8737,
+      "step": 19313
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8269392406499534,
+      "learning_rate": 1.1930505677808476e-05,
+      "loss": 0.9654,
+      "step": 19314
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3657112167685126,
+      "learning_rate": 1.1929756984840236e-05,
+      "loss": 0.8794,
+      "step": 19315
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.736279421472321,
+      "learning_rate": 1.1929008280636338e-05,
+      "loss": 0.8652,
+      "step": 19316
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6618117173687144,
+      "learning_rate": 1.1928259565201145e-05,
+      "loss": 0.8682,
+      "step": 19317
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6680029719977223,
+      "learning_rate": 1.1927510838539014e-05,
+      "loss": 0.8516,
+      "step": 19318
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4793922092512504,
+      "learning_rate": 1.1926762100654306e-05,
+      "loss": 0.8446,
+      "step": 19319
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4437915561634416,
+      "learning_rate": 1.1926013351551375e-05,
+      "loss": 0.8459,
+      "step": 19320
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8193097463635576,
+      "learning_rate": 1.1925264591234588e-05,
+      "loss": 0.7944,
+      "step": 19321
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.436676116213336,
+      "learning_rate": 1.19245158197083e-05,
+      "loss": 0.8922,
+      "step": 19322
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1238596579558777,
+      "learning_rate": 1.1923767036976872e-05,
+      "loss": 0.9533,
+      "step": 19323
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6936471474157644,
+      "learning_rate": 1.1923018243044664e-05,
+      "loss": 0.9715,
+      "step": 19324
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4151622351766593,
+      "learning_rate": 1.1922269437916035e-05,
+      "loss": 0.7996,
+      "step": 19325
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5286685964751203,
+      "learning_rate": 1.1921520621595342e-05,
+      "loss": 1.0473,
+      "step": 19326
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4017044446385454,
+      "learning_rate": 1.1920771794086949e-05,
+      "loss": 0.8727,
+      "step": 19327
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.557469732923206,
+      "learning_rate": 1.1920022955395217e-05,
+      "loss": 1.1133,
+      "step": 19328
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5331563642153123,
+      "learning_rate": 1.1919274105524501e-05,
+      "loss": 0.8587,
+      "step": 19329
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6340523918073322,
+      "learning_rate": 1.1918525244479161e-05,
+      "loss": 0.839,
+      "step": 19330
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1224189818623773,
+      "learning_rate": 1.1917776372263568e-05,
+      "loss": 0.9009,
+      "step": 19331
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4605487649187505,
+      "learning_rate": 1.1917027488882066e-05,
+      "loss": 0.8481,
+      "step": 19332
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8278942543453298,
+      "learning_rate": 1.1916278594339026e-05,
+      "loss": 0.7231,
+      "step": 19333
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6390638430031805,
+      "learning_rate": 1.1915529688638803e-05,
+      "loss": 1.0265,
+      "step": 19334
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.716739392460843,
+      "learning_rate": 1.1914780771785764e-05,
+      "loss": 0.7585,
+      "step": 19335
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6151888818169309,
+      "learning_rate": 1.1914031843784262e-05,
+      "loss": 0.8503,
+      "step": 19336
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1963233467763796,
+      "learning_rate": 1.1913282904638661e-05,
+      "loss": 0.9306,
+      "step": 19337
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3088968512254437,
+      "learning_rate": 1.1912533954353323e-05,
+      "loss": 0.9003,
+      "step": 19338
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6377385762816092,
+      "learning_rate": 1.1911784992932605e-05,
+      "loss": 0.877,
+      "step": 19339
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6338789789408275,
+      "learning_rate": 1.1911036020380868e-05,
+      "loss": 1.0441,
+      "step": 19340
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5126932482500626,
+      "learning_rate": 1.1910287036702478e-05,
+      "loss": 0.8833,
+      "step": 19341
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9251503583806358,
+      "learning_rate": 1.1909538041901789e-05,
+      "loss": 0.9244,
+      "step": 19342
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.811430773758855,
+      "learning_rate": 1.1908789035983166e-05,
+      "loss": 0.6963,
+      "step": 19343
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7357713381009172,
+      "learning_rate": 1.1908040018950968e-05,
+      "loss": 0.9375,
+      "step": 19344
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8117169581270909,
+      "learning_rate": 1.190729099080956e-05,
+      "loss": 0.8702,
+      "step": 19345
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5417010294102047,
+      "learning_rate": 1.1906541951563295e-05,
+      "loss": 0.955,
+      "step": 19346
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7706245144332353,
+      "learning_rate": 1.1905792901216544e-05,
+      "loss": 0.7763,
+      "step": 19347
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.093292048295432,
+      "learning_rate": 1.190504383977366e-05,
+      "loss": 0.8458,
+      "step": 19348
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8903704183182501,
+      "learning_rate": 1.1904294767239009e-05,
+      "loss": 0.8466,
+      "step": 19349
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4847180043544081,
+      "learning_rate": 1.1903545683616953e-05,
+      "loss": 0.7845,
+      "step": 19350
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6291590577050354,
+      "learning_rate": 1.1902796588911847e-05,
+      "loss": 0.93,
+      "step": 19351
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7846289379012186,
+      "learning_rate": 1.1902047483128063e-05,
+      "loss": 0.7272,
+      "step": 19352
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.569861916316345,
+      "learning_rate": 1.190129836626995e-05,
+      "loss": 0.9526,
+      "step": 19353
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5378695024994504,
+      "learning_rate": 1.1900549238341881e-05,
+      "loss": 0.8072,
+      "step": 19354
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7432675526422612,
+      "learning_rate": 1.1899800099348209e-05,
+      "loss": 0.9729,
+      "step": 19355
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7352427540633606,
+      "learning_rate": 1.1899050949293303e-05,
+      "loss": 0.9013,
+      "step": 19356
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3032966522882004,
+      "learning_rate": 1.1898301788181515e-05,
+      "loss": 0.927,
+      "step": 19357
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7319657867142844,
+      "learning_rate": 1.1897552616017216e-05,
+      "loss": 0.9145,
+      "step": 19358
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5423162263977783,
+      "learning_rate": 1.1896803432804768e-05,
+      "loss": 0.9115,
+      "step": 19359
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4588223523804302,
+      "learning_rate": 1.1896054238548527e-05,
+      "loss": 0.8813,
+      "step": 19360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.564105306794996,
+      "learning_rate": 1.189530503325286e-05,
+      "loss": 0.9486,
+      "step": 19361
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.692371421267699,
+      "learning_rate": 1.1894555816922122e-05,
+      "loss": 0.7655,
+      "step": 19362
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6895178285478105,
+      "learning_rate": 1.1893806589560689e-05,
+      "loss": 0.8408,
+      "step": 19363
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8150021198373891,
+      "learning_rate": 1.1893057351172905e-05,
+      "loss": 0.9187,
+      "step": 19364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6713303615646835,
+      "learning_rate": 1.1892308101763148e-05,
+      "loss": 0.9478,
+      "step": 19365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4893612275019452,
+      "learning_rate": 1.1891558841335773e-05,
+      "loss": 0.9726,
+      "step": 19366
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6648329773668034,
+      "learning_rate": 1.189080956989514e-05,
+      "loss": 0.8808,
+      "step": 19367
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7110399666717975,
+      "learning_rate": 1.189006028744562e-05,
+      "loss": 0.9423,
+      "step": 19368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0313692432143426,
+      "learning_rate": 1.1889310993991567e-05,
+      "loss": 0.7576,
+      "step": 19369
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4550552604786222,
+      "learning_rate": 1.188856168953735e-05,
+      "loss": 0.8751,
+      "step": 19370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7784056610136667,
+      "learning_rate": 1.1887812374087328e-05,
+      "loss": 0.8457,
+      "step": 19371
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.13228178022717,
+      "learning_rate": 1.1887063047645863e-05,
+      "loss": 0.848,
+      "step": 19372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7743050457108025,
+      "learning_rate": 1.1886313710217322e-05,
+      "loss": 0.7204,
+      "step": 19373
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7540793982895195,
+      "learning_rate": 1.1885564361806063e-05,
+      "loss": 0.8248,
+      "step": 19374
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.574479950696186,
+      "learning_rate": 1.1884815002416452e-05,
+      "loss": 0.9086,
+      "step": 19375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8019147913411953,
+      "learning_rate": 1.1884065632052852e-05,
+      "loss": 1.023,
+      "step": 19376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.608182911482917,
+      "learning_rate": 1.1883316250719625e-05,
+      "loss": 0.9187,
+      "step": 19377
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.476957053489462,
+      "learning_rate": 1.1882566858421137e-05,
+      "loss": 1.0493,
+      "step": 19378
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0666224245701694,
+      "learning_rate": 1.1881817455161747e-05,
+      "loss": 0.9154,
+      "step": 19379
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6718270058391531,
+      "learning_rate": 1.188106804094582e-05,
+      "loss": 1.0417,
+      "step": 19380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8477388153318786,
+      "learning_rate": 1.1880318615777718e-05,
+      "loss": 0.8887,
+      "step": 19381
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6743839576871669,
+      "learning_rate": 1.1879569179661808e-05,
+      "loss": 0.8624,
+      "step": 19382
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5172583250231566,
+      "learning_rate": 1.1878819732602448e-05,
+      "loss": 0.7295,
+      "step": 19383
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8443911455506842,
+      "learning_rate": 1.1878070274604008e-05,
+      "loss": 0.7021,
+      "step": 19384
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.49374660619474,
+      "learning_rate": 1.1877320805670845e-05,
+      "loss": 0.8433,
+      "step": 19385
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5786426697535818,
+      "learning_rate": 1.1876571325807328e-05,
+      "loss": 0.8756,
+      "step": 19386
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.536064743813778,
+      "learning_rate": 1.1875821835017819e-05,
+      "loss": 0.819,
+      "step": 19387
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7332649629205494,
+      "learning_rate": 1.1875072333306681e-05,
+      "loss": 0.891,
+      "step": 19388
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9420377832345344,
+      "learning_rate": 1.1874322820678277e-05,
+      "loss": 0.8357,
+      "step": 19389
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7561029639209733,
+      "learning_rate": 1.1873573297136973e-05,
+      "loss": 0.7341,
+      "step": 19390
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.572762933450247,
+      "learning_rate": 1.1872823762687133e-05,
+      "loss": 0.8703,
+      "step": 19391
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2491147609870983,
+      "learning_rate": 1.1872074217333118e-05,
+      "loss": 0.9249,
+      "step": 19392
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6182423184487533,
+      "learning_rate": 1.1871324661079295e-05,
+      "loss": 0.8992,
+      "step": 19393
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0485931014288967,
+      "learning_rate": 1.1870575093930027e-05,
+      "loss": 0.8441,
+      "step": 19394
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.01807509568859,
+      "learning_rate": 1.1869825515889682e-05,
+      "loss": 0.8857,
+      "step": 19395
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7378275991892396,
+      "learning_rate": 1.1869075926962616e-05,
+      "loss": 0.9167,
+      "step": 19396
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.734506542831791,
+      "learning_rate": 1.1868326327153202e-05,
+      "loss": 0.8833,
+      "step": 19397
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6934973754757745,
+      "learning_rate": 1.1867576716465799e-05,
+      "loss": 0.7921,
+      "step": 19398
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7618246036942988,
+      "learning_rate": 1.1866827094904773e-05,
+      "loss": 0.9071,
+      "step": 19399
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5956141983653391,
+      "learning_rate": 1.186607746247449e-05,
+      "loss": 0.8977,
+      "step": 19400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5359108205397998,
+      "learning_rate": 1.1865327819179313e-05,
+      "loss": 0.8552,
+      "step": 19401
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.890335334195534,
+      "learning_rate": 1.1864578165023604e-05,
+      "loss": 0.9651,
+      "step": 19402
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8448541781303578,
+      "learning_rate": 1.1863828500011736e-05,
+      "loss": 0.8664,
+      "step": 19403
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7105482214968435,
+      "learning_rate": 1.1863078824148062e-05,
+      "loss": 0.8086,
+      "step": 19404
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.887089918800537,
+      "learning_rate": 1.1862329137436958e-05,
+      "loss": 0.8748,
+      "step": 19405
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6473697609458533,
+      "learning_rate": 1.1861579439882779e-05,
+      "loss": 0.9798,
+      "step": 19406
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6016799395050692,
+      "learning_rate": 1.18608297314899e-05,
+      "loss": 0.8963,
+      "step": 19407
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6294631778504833,
+      "learning_rate": 1.1860080012262683e-05,
+      "loss": 0.9867,
+      "step": 19408
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5838141952036306,
+      "learning_rate": 1.1859330282205483e-05,
+      "loss": 0.9987,
+      "step": 19409
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8543526148803502,
+      "learning_rate": 1.1858580541322682e-05,
+      "loss": 0.9472,
+      "step": 19410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7988761383227878,
+      "learning_rate": 1.1857830789618632e-05,
+      "loss": 0.7364,
+      "step": 19411
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7621477629718203,
+      "learning_rate": 1.1857081027097706e-05,
+      "loss": 0.77,
+      "step": 19412
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9825643260253039,
+      "learning_rate": 1.1856331253764261e-05,
+      "loss": 0.792,
+      "step": 19413
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7297755934021506,
+      "learning_rate": 1.1855581469622673e-05,
+      "loss": 0.8929,
+      "step": 19414
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7505424889411181,
+      "learning_rate": 1.1854831674677298e-05,
+      "loss": 0.7622,
+      "step": 19415
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7264487290986003,
+      "learning_rate": 1.185408186893251e-05,
+      "loss": 0.7791,
+      "step": 19416
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.610510115646491,
+      "learning_rate": 1.185333205239267e-05,
+      "loss": 1.0202,
+      "step": 19417
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1213721653854796,
+      "learning_rate": 1.1852582225062141e-05,
+      "loss": 0.87,
+      "step": 19418
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.474769088559678,
+      "learning_rate": 1.1851832386945294e-05,
+      "loss": 0.7748,
+      "step": 19419
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7288403102918913,
+      "learning_rate": 1.1851082538046492e-05,
+      "loss": 0.6995,
+      "step": 19420
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8376381079671769,
+      "learning_rate": 1.18503326783701e-05,
+      "loss": 0.7166,
+      "step": 19421
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7935273867006136,
+      "learning_rate": 1.1849582807920488e-05,
+      "loss": 0.7143,
+      "step": 19422
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6612076140142842,
+      "learning_rate": 1.1848832926702017e-05,
+      "loss": 0.8924,
+      "step": 19423
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8167860959984603,
+      "learning_rate": 1.1848083034719055e-05,
+      "loss": 1.0286,
+      "step": 19424
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7732155251722788,
+      "learning_rate": 1.1847333131975971e-05,
+      "loss": 0.8418,
+      "step": 19425
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9817924324128826,
+      "learning_rate": 1.1846583218477129e-05,
+      "loss": 0.7388,
+      "step": 19426
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1240713752483527,
+      "learning_rate": 1.1845833294226892e-05,
+      "loss": 0.9342,
+      "step": 19427
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8332122877958913,
+      "learning_rate": 1.184508335922963e-05,
+      "loss": 0.7773,
+      "step": 19428
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6072234891029584,
+      "learning_rate": 1.1844333413489709e-05,
+      "loss": 0.9563,
+      "step": 19429
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6554425969482425,
+      "learning_rate": 1.1843583457011497e-05,
+      "loss": 0.9438,
+      "step": 19430
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5088983085349055,
+      "learning_rate": 1.1842833489799356e-05,
+      "loss": 0.8289,
+      "step": 19431
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7560503798283895,
+      "learning_rate": 1.1842083511857654e-05,
+      "loss": 0.9335,
+      "step": 19432
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.599571622944776,
+      "learning_rate": 1.184133352319076e-05,
+      "loss": 0.8196,
+      "step": 19433
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.761484274669151,
+      "learning_rate": 1.184058352380304e-05,
+      "loss": 0.8595,
+      "step": 19434
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.647694202188808,
+      "learning_rate": 1.183983351369886e-05,
+      "loss": 0.9628,
+      "step": 19435
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8184503222750399,
+      "learning_rate": 1.1839083492882583e-05,
+      "loss": 0.9202,
+      "step": 19436
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8369161848328999,
+      "learning_rate": 1.1838333461358587e-05,
+      "loss": 0.7913,
+      "step": 19437
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.728057799939287,
+      "learning_rate": 1.1837583419131227e-05,
+      "loss": 1.0726,
+      "step": 19438
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6818033216799395,
+      "learning_rate": 1.1836833366204874e-05,
+      "loss": 0.876,
+      "step": 19439
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1134764119080693,
+      "learning_rate": 1.1836083302583896e-05,
+      "loss": 0.7811,
+      "step": 19440
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5722183775392562,
+      "learning_rate": 1.183533322827266e-05,
+      "loss": 0.7835,
+      "step": 19441
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7556025611464823,
+      "learning_rate": 1.1834583143275534e-05,
+      "loss": 0.7099,
+      "step": 19442
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8009406959405777,
+      "learning_rate": 1.1833833047596884e-05,
+      "loss": 0.7408,
+      "step": 19443
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.770238233603683,
+      "learning_rate": 1.1833082941241076e-05,
+      "loss": 0.8069,
+      "step": 19444
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6394193069689738,
+      "learning_rate": 1.1832332824212478e-05,
+      "loss": 0.996,
+      "step": 19445
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6803627163068593,
+      "learning_rate": 1.183158269651546e-05,
+      "loss": 0.8531,
+      "step": 19446
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4603139734660675,
+      "learning_rate": 1.1830832558154388e-05,
+      "loss": 0.8654,
+      "step": 19447
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6306100330649642,
+      "learning_rate": 1.1830082409133627e-05,
+      "loss": 0.9869,
+      "step": 19448
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5806806074225415,
+      "learning_rate": 1.182933224945755e-05,
+      "loss": 0.82,
+      "step": 19449
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9248662408575257,
+      "learning_rate": 1.1828582079130519e-05,
+      "loss": 0.8981,
+      "step": 19450
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9367756005945338,
+      "learning_rate": 1.1827831898156905e-05,
+      "loss": 0.8184,
+      "step": 19451
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.511907961767117,
+      "learning_rate": 1.1827081706541076e-05,
+      "loss": 0.9476,
+      "step": 19452
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.450661582937885,
+      "learning_rate": 1.1826331504287396e-05,
+      "loss": 0.8926,
+      "step": 19453
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8685053872286477,
+      "learning_rate": 1.182558129140024e-05,
+      "loss": 0.9131,
+      "step": 19454
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.90347809765092,
+      "learning_rate": 1.1824831067883969e-05,
+      "loss": 0.8821,
+      "step": 19455
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7589364840418413,
+      "learning_rate": 1.1824080833742952e-05,
+      "loss": 0.7812,
+      "step": 19456
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5997414446164933,
+      "learning_rate": 1.1823330588981562e-05,
+      "loss": 0.8246,
+      "step": 19457
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.071199545039866,
+      "learning_rate": 1.1822580333604163e-05,
+      "loss": 0.9713,
+      "step": 19458
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.799718871033189,
+      "learning_rate": 1.1821830067615128e-05,
+      "loss": 0.8481,
+      "step": 19459
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6083988727911782,
+      "learning_rate": 1.1821079791018815e-05,
+      "loss": 0.7501,
+      "step": 19460
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6423152116665494,
+      "learning_rate": 1.1820329503819604e-05,
+      "loss": 0.7561,
+      "step": 19461
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4992674283417522,
+      "learning_rate": 1.1819579206021854e-05,
+      "loss": 0.8494,
+      "step": 19462
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7440715194709,
+      "learning_rate": 1.1818828897629941e-05,
+      "loss": 0.8611,
+      "step": 19463
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7749173005010401,
+      "learning_rate": 1.181807857864823e-05,
+      "loss": 0.876,
+      "step": 19464
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1378824868246395,
+      "learning_rate": 1.1817328249081088e-05,
+      "loss": 0.9691,
+      "step": 19465
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5576513679410358,
+      "learning_rate": 1.1816577908932887e-05,
+      "loss": 1.0651,
+      "step": 19466
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9493266587439022,
+      "learning_rate": 1.1815827558207996e-05,
+      "loss": 0.8952,
+      "step": 19467
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.675424684451607,
+      "learning_rate": 1.181507719691078e-05,
+      "loss": 0.9799,
+      "step": 19468
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6357966640283002,
+      "learning_rate": 1.181432682504561e-05,
+      "loss": 0.9238,
+      "step": 19469
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7420775208472266,
+      "learning_rate": 1.1813576442616854e-05,
+      "loss": 0.8615,
+      "step": 19470
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.705539015008191,
+      "learning_rate": 1.1812826049628883e-05,
+      "loss": 0.9136,
+      "step": 19471
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4312372641225817,
+      "learning_rate": 1.1812075646086066e-05,
+      "loss": 0.8929,
+      "step": 19472
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5374318042236639,
+      "learning_rate": 1.1811325231992768e-05,
+      "loss": 0.8092,
+      "step": 19473
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9095617977645587,
+      "learning_rate": 1.1810574807353362e-05,
+      "loss": 0.9073,
+      "step": 19474
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6637924180244008,
+      "learning_rate": 1.1809824372172217e-05,
+      "loss": 0.8493,
+      "step": 19475
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6754301253058212,
+      "learning_rate": 1.18090739264537e-05,
+      "loss": 0.9793,
+      "step": 19476
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6226873620270312,
+      "learning_rate": 1.1808323470202185e-05,
+      "loss": 0.8779,
+      "step": 19477
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.021120195479356,
+      "learning_rate": 1.1807573003422035e-05,
+      "loss": 0.8613,
+      "step": 19478
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.395535953235335,
+      "learning_rate": 1.1806822526117626e-05,
+      "loss": 0.8652,
+      "step": 19479
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8482093565656874,
+      "learning_rate": 1.1806072038293322e-05,
+      "loss": 0.8846,
+      "step": 19480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9293513688805168,
+      "learning_rate": 1.1805321539953493e-05,
+      "loss": 0.8879,
+      "step": 19481
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9301011217391173,
+      "learning_rate": 1.1804571031102516e-05,
+      "loss": 0.9057,
+      "step": 19482
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.846474162872791,
+      "learning_rate": 1.1803820511744749e-05,
+      "loss": 0.831,
+      "step": 19483
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5771265728146469,
+      "learning_rate": 1.1803069981884573e-05,
+      "loss": 0.6984,
+      "step": 19484
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7033064887803633,
+      "learning_rate": 1.1802319441526348e-05,
+      "loss": 0.923,
+      "step": 19485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.949069840994881,
+      "learning_rate": 1.1801568890674453e-05,
+      "loss": 0.7731,
+      "step": 19486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5543679933115397,
+      "learning_rate": 1.180081832933325e-05,
+      "loss": 0.9368,
+      "step": 19487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8618135138671021,
+      "learning_rate": 1.1800067757507115e-05,
+      "loss": 0.9359,
+      "step": 19488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5055125959888798,
+      "learning_rate": 1.1799317175200416e-05,
+      "loss": 0.9144,
+      "step": 19489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4970127684945636,
+      "learning_rate": 1.1798566582417521e-05,
+      "loss": 0.9,
+      "step": 19490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.806669461352025,
+      "learning_rate": 1.1797815979162801e-05,
+      "loss": 0.8485,
+      "step": 19491
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5827249066649682,
+      "learning_rate": 1.179706536544063e-05,
+      "loss": 0.8846,
+      "step": 19492
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6453973003449593,
+      "learning_rate": 1.1796314741255372e-05,
+      "loss": 0.7641,
+      "step": 19493
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.574968514596797,
+      "learning_rate": 1.1795564106611401e-05,
+      "loss": 0.8654,
+      "step": 19494
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.612811536996767,
+      "learning_rate": 1.1794813461513092e-05,
+      "loss": 0.9041,
+      "step": 19495
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.873908854053305,
+      "learning_rate": 1.1794062805964808e-05,
+      "loss": 0.7468,
+      "step": 19496
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.586314489237006,
+      "learning_rate": 1.1793312139970923e-05,
+      "loss": 0.9628,
+      "step": 19497
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9744216938753465,
+      "learning_rate": 1.1792561463535805e-05,
+      "loss": 1.0271,
+      "step": 19498
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0099981881875895,
+      "learning_rate": 1.1791810776663829e-05,
+      "loss": 0.7055,
+      "step": 19499
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5413090357193406,
+      "learning_rate": 1.1791060079359362e-05,
+      "loss": 0.8927,
+      "step": 19500
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5933639487066813,
+      "learning_rate": 1.1790309371626775e-05,
+      "loss": 0.9078,
+      "step": 19501
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.034969365113078,
+      "learning_rate": 1.1789558653470443e-05,
+      "loss": 0.8586,
+      "step": 19502
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6595236166994132,
+      "learning_rate": 1.1788807924894734e-05,
+      "loss": 0.7727,
+      "step": 19503
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.662447504161715,
+      "learning_rate": 1.1788057185904015e-05,
+      "loss": 0.9549,
+      "step": 19504
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5379244129955647,
+      "learning_rate": 1.1787306436502664e-05,
+      "loss": 0.7813,
+      "step": 19505
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6616285614963477,
+      "learning_rate": 1.1786555676695049e-05,
+      "loss": 0.8178,
+      "step": 19506
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8569788523703623,
+      "learning_rate": 1.178580490648554e-05,
+      "loss": 0.935,
+      "step": 19507
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7903290081788283,
+      "learning_rate": 1.178505412587851e-05,
+      "loss": 0.7397,
+      "step": 19508
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7818786710890394,
+      "learning_rate": 1.1784303334878331e-05,
+      "loss": 0.7187,
+      "step": 19509
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.914923305930271,
+      "learning_rate": 1.1783552533489372e-05,
+      "loss": 0.8664,
+      "step": 19510
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4977988441637504,
+      "learning_rate": 1.1782801721716004e-05,
+      "loss": 0.8077,
+      "step": 19511
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4271253236349173,
+      "learning_rate": 1.1782050899562605e-05,
+      "loss": 0.7984,
+      "step": 19512
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7592415129810179,
+      "learning_rate": 1.1781300067033534e-05,
+      "loss": 0.7409,
+      "step": 19513
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.584502888502762,
+      "learning_rate": 1.1780549224133177e-05,
+      "loss": 0.8564,
+      "step": 19514
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6581730432173782,
+      "learning_rate": 1.1779798370865893e-05,
+      "loss": 1.0046,
+      "step": 19515
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5212496310263584,
+      "learning_rate": 1.1779047507236064e-05,
+      "loss": 0.9185,
+      "step": 19516
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4160478360935522,
+      "learning_rate": 1.1778296633248056e-05,
+      "loss": 0.8484,
+      "step": 19517
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6887925810053772,
+      "learning_rate": 1.177754574890624e-05,
+      "loss": 0.8267,
+      "step": 19518
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.694939731105114,
+      "learning_rate": 1.1776794854214993e-05,
+      "loss": 0.911,
+      "step": 19519
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7567251256408754,
+      "learning_rate": 1.1776043949178681e-05,
+      "loss": 0.976,
+      "step": 19520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.372752053646052,
+      "learning_rate": 1.177529303380168e-05,
+      "loss": 0.9069,
+      "step": 19521
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7429357054875119,
+      "learning_rate": 1.177454210808836e-05,
+      "loss": 0.9333,
+      "step": 19522
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7745119571318249,
+      "learning_rate": 1.1773791172043094e-05,
+      "loss": 0.9652,
+      "step": 19523
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.883192132339577,
+      "learning_rate": 1.1773040225670257e-05,
+      "loss": 0.9365,
+      "step": 19524
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.801236834456808,
+      "learning_rate": 1.1772289268974215e-05,
+      "loss": 0.732,
+      "step": 19525
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.795673197388128,
+      "learning_rate": 1.1771538301959346e-05,
+      "loss": 0.954,
+      "step": 19526
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5520860955377909,
+      "learning_rate": 1.1770787324630019e-05,
+      "loss": 0.8237,
+      "step": 19527
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6655481911959056,
+      "learning_rate": 1.1770036336990609e-05,
+      "loss": 0.9859,
+      "step": 19528
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.54219713239756,
+      "learning_rate": 1.1769285339045484e-05,
+      "loss": 0.7655,
+      "step": 19529
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.709272954468948,
+      "learning_rate": 1.1768534330799021e-05,
+      "loss": 0.9648,
+      "step": 19530
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7166386408381626,
+      "learning_rate": 1.1767783312255595e-05,
+      "loss": 0.8801,
+      "step": 19531
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8241966823030736,
+      "learning_rate": 1.176703228341957e-05,
+      "loss": 0.758,
+      "step": 19532
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3143966519532055,
+      "learning_rate": 1.1766281244295328e-05,
+      "loss": 0.9171,
+      "step": 19533
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6155815916382377,
+      "learning_rate": 1.1765530194887231e-05,
+      "loss": 0.9195,
+      "step": 19534
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5871733329140127,
+      "learning_rate": 1.1764779135199664e-05,
+      "loss": 0.9397,
+      "step": 19535
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7998675703480734,
+      "learning_rate": 1.1764028065236991e-05,
+      "loss": 0.767,
+      "step": 19536
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3967963355209863,
+      "learning_rate": 1.1763276985003592e-05,
+      "loss": 0.8066,
+      "step": 19537
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.157741287075993,
+      "learning_rate": 1.1762525894503831e-05,
+      "loss": 0.863,
+      "step": 19538
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9205536281666422,
+      "learning_rate": 1.1761774793742092e-05,
+      "loss": 0.9009,
+      "step": 19539
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5591275693922577,
+      "learning_rate": 1.176102368272274e-05,
+      "loss": 0.9053,
+      "step": 19540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8807081588542531,
+      "learning_rate": 1.1760272561450146e-05,
+      "loss": 0.898,
+      "step": 19541
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7020034495410403,
+      "learning_rate": 1.1759521429928693e-05,
+      "loss": 0.9746,
+      "step": 19542
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6627932014046225,
+      "learning_rate": 1.1758770288162746e-05,
+      "loss": 0.7857,
+      "step": 19543
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6210005901287121,
+      "learning_rate": 1.1758019136156684e-05,
+      "loss": 0.8767,
+      "step": 19544
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.596732650142375,
+      "learning_rate": 1.1757267973914876e-05,
+      "loss": 0.7884,
+      "step": 19545
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6670524123837949,
+      "learning_rate": 1.17565168014417e-05,
+      "loss": 0.7689,
+      "step": 19546
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6778012675290956,
+      "learning_rate": 1.1755765618741526e-05,
+      "loss": 0.8798,
+      "step": 19547
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5301410867949015,
+      "learning_rate": 1.175501442581873e-05,
+      "loss": 0.8885,
+      "step": 19548
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.366914005181132,
+      "learning_rate": 1.1754263222677681e-05,
+      "loss": 0.8793,
+      "step": 19549
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6473135195375925,
+      "learning_rate": 1.1753512009322758e-05,
+      "loss": 0.8806,
+      "step": 19550
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1416720349452523,
+      "learning_rate": 1.1752760785758332e-05,
+      "loss": 0.7721,
+      "step": 19551
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1487863857074903,
+      "learning_rate": 1.1752009551988781e-05,
+      "loss": 0.7499,
+      "step": 19552
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.477687337608344,
+      "learning_rate": 1.1751258308018473e-05,
+      "loss": 0.9141,
+      "step": 19553
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7087921483037283,
+      "learning_rate": 1.1750507053851784e-05,
+      "loss": 0.9918,
+      "step": 19554
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.991381384434329,
+      "learning_rate": 1.1749755789493092e-05,
+      "loss": 0.846,
+      "step": 19555
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5818401039806145,
+      "learning_rate": 1.1749004514946765e-05,
+      "loss": 0.9025,
+      "step": 19556
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6484285617331522,
+      "learning_rate": 1.174825323021718e-05,
+      "loss": 0.8382,
+      "step": 19557
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8801867999107298,
+      "learning_rate": 1.1747501935308712e-05,
+      "loss": 1.0084,
+      "step": 19558
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5233100755484177,
+      "learning_rate": 1.1746750630225734e-05,
+      "loss": 0.8751,
+      "step": 19559
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0389240027054107,
+      "learning_rate": 1.1745999314972623e-05,
+      "loss": 0.8335,
+      "step": 19560
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1991075871880246,
+      "learning_rate": 1.174524798955375e-05,
+      "loss": 0.9453,
+      "step": 19561
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4803659676779766,
+      "learning_rate": 1.1744496653973488e-05,
+      "loss": 0.8295,
+      "step": 19562
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8459081021436463,
+      "learning_rate": 1.1743745308236218e-05,
+      "loss": 0.9383,
+      "step": 19563
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7668933756620863,
+      "learning_rate": 1.1742993952346307e-05,
+      "loss": 0.7444,
+      "step": 19564
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3780872989267374,
+      "learning_rate": 1.174224258630814e-05,
+      "loss": 0.8902,
+      "step": 19565
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8384728414510176,
+      "learning_rate": 1.1741491210126077e-05,
+      "loss": 0.9155,
+      "step": 19566
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.519066187062828,
+      "learning_rate": 1.1740739823804506e-05,
+      "loss": 0.7389,
+      "step": 19567
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5371545054294709,
+      "learning_rate": 1.1739988427347795e-05,
+      "loss": 0.8538,
+      "step": 19568
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0351101417670225,
+      "learning_rate": 1.1739237020760319e-05,
+      "loss": 0.9217,
+      "step": 19569
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9915464835534484,
+      "learning_rate": 1.1738485604046457e-05,
+      "loss": 0.9787,
+      "step": 19570
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.633133740625351,
+      "learning_rate": 1.1737734177210581e-05,
+      "loss": 0.8843,
+      "step": 19571
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8225029315104937,
+      "learning_rate": 1.1736982740257065e-05,
+      "loss": 0.8107,
+      "step": 19572
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7426207822628537,
+      "learning_rate": 1.1736231293190286e-05,
+      "loss": 0.8147,
+      "step": 19573
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5838256250665605,
+      "learning_rate": 1.1735479836014621e-05,
+      "loss": 0.905,
+      "step": 19574
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7783153040334632,
+      "learning_rate": 1.1734728368734441e-05,
+      "loss": 0.9035,
+      "step": 19575
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7580624791126273,
+      "learning_rate": 1.1733976891354122e-05,
+      "loss": 0.7136,
+      "step": 19576
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7360477261330423,
+      "learning_rate": 1.1733225403878042e-05,
+      "loss": 0.9427,
+      "step": 19577
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.862333551489734,
+      "learning_rate": 1.1732473906310576e-05,
+      "loss": 0.8165,
+      "step": 19578
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8414228651743292,
+      "learning_rate": 1.1731722398656096e-05,
+      "loss": 0.8857,
+      "step": 19579
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.769357216767121,
+      "learning_rate": 1.1730970880918981e-05,
+      "loss": 0.82,
+      "step": 19580
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.76539609550144,
+      "learning_rate": 1.1730219353103604e-05,
+      "loss": 0.7141,
+      "step": 19581
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5323174659660923,
+      "learning_rate": 1.1729467815214348e-05,
+      "loss": 0.9105,
+      "step": 19582
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5364325171935327,
+      "learning_rate": 1.1728716267255575e-05,
+      "loss": 0.8403,
+      "step": 19583
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5397974561146037,
+      "learning_rate": 1.1727964709231676e-05,
+      "loss": 0.9268,
+      "step": 19584
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9884102372641626,
+      "learning_rate": 1.1727213141147013e-05,
+      "loss": 0.7932,
+      "step": 19585
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0451173664994973,
+      "learning_rate": 1.1726461563005973e-05,
+      "loss": 0.8028,
+      "step": 19586
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7337036265696991,
+      "learning_rate": 1.1725709974812922e-05,
+      "loss": 0.6817,
+      "step": 19587
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.45395553287452,
+      "learning_rate": 1.1724958376572246e-05,
+      "loss": 0.944,
+      "step": 19588
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8765059637110058,
+      "learning_rate": 1.1724206768288316e-05,
+      "loss": 0.9531,
+      "step": 19589
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.854124123429662,
+      "learning_rate": 1.1723455149965504e-05,
+      "loss": 0.846,
+      "step": 19590
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6588007311991386,
+      "learning_rate": 1.1722703521608197e-05,
+      "loss": 0.9369,
+      "step": 19591
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5424095097033454,
+      "learning_rate": 1.1721951883220756e-05,
+      "loss": 0.9345,
+      "step": 19592
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4036685282220698,
+      "learning_rate": 1.1721200234807574e-05,
+      "loss": 0.8016,
+      "step": 19593
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6779012052581967,
+      "learning_rate": 1.1720448576373013e-05,
+      "loss": 0.8715,
+      "step": 19594
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6857744773741319,
+      "learning_rate": 1.1719696907921463e-05,
+      "loss": 0.7974,
+      "step": 19595
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6469593919453998,
+      "learning_rate": 1.171894522945729e-05,
+      "loss": 0.7916,
+      "step": 19596
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7039651026542801,
+      "learning_rate": 1.1718193540984873e-05,
+      "loss": 0.7279,
+      "step": 19597
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.793949869118429,
+      "learning_rate": 1.1717441842508588e-05,
+      "loss": 0.9475,
+      "step": 19598
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6563724426031838,
+      "learning_rate": 1.1716690134032813e-05,
+      "loss": 0.951,
+      "step": 19599
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4102738618749098,
+      "learning_rate": 1.1715938415561928e-05,
+      "loss": 0.8423,
+      "step": 19600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5812107703851743,
+      "learning_rate": 1.1715186687100302e-05,
+      "loss": 0.9423,
+      "step": 19601
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5070060541744912,
+      "learning_rate": 1.1714434948652317e-05,
+      "loss": 0.8062,
+      "step": 19602
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8985200204079604,
+      "learning_rate": 1.171368320022235e-05,
+      "loss": 1.013,
+      "step": 19603
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.086748169862556,
+      "learning_rate": 1.1712931441814776e-05,
+      "loss": 0.9034,
+      "step": 19604
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7768538876128388,
+      "learning_rate": 1.1712179673433974e-05,
+      "loss": 0.9212,
+      "step": 19605
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7069429775351992,
+      "learning_rate": 1.171142789508432e-05,
+      "loss": 0.8356,
+      "step": 19606
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4978129244514486,
+      "learning_rate": 1.1710676106770191e-05,
+      "loss": 0.8972,
+      "step": 19607
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8035257969719265,
+      "learning_rate": 1.1709924308495964e-05,
+      "loss": 0.9625,
+      "step": 19608
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8565823237545118,
+      "learning_rate": 1.1709172500266015e-05,
+      "loss": 0.9176,
+      "step": 19609
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5038370879792173,
+      "learning_rate": 1.1708420682084722e-05,
+      "loss": 0.8671,
+      "step": 19610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5251867749224903,
+      "learning_rate": 1.1707668853956465e-05,
+      "loss": 0.7545,
+      "step": 19611
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5634859933708698,
+      "learning_rate": 1.1706917015885621e-05,
+      "loss": 0.8143,
+      "step": 19612
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6574044226814704,
+      "learning_rate": 1.1706165167876562e-05,
+      "loss": 1.0471,
+      "step": 19613
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6101139521267556,
+      "learning_rate": 1.170541330993367e-05,
+      "loss": 0.9888,
+      "step": 19614
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7695823262000574,
+      "learning_rate": 1.1704661442061323e-05,
+      "loss": 0.9523,
+      "step": 19615
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9008435412742022,
+      "learning_rate": 1.1703909564263898e-05,
+      "loss": 0.6932,
+      "step": 19616
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5298956693908106,
+      "learning_rate": 1.1703157676545768e-05,
+      "loss": 0.7791,
+      "step": 19617
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4505576276169085,
+      "learning_rate": 1.170240577891132e-05,
+      "loss": 0.6575,
+      "step": 19618
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3066987867920106,
+      "learning_rate": 1.1701653871364924e-05,
+      "loss": 0.8855,
+      "step": 19619
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6524219126679032,
+      "learning_rate": 1.1700901953910961e-05,
+      "loss": 0.792,
+      "step": 19620
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.499082031183591,
+      "learning_rate": 1.1700150026553807e-05,
+      "loss": 0.9477,
+      "step": 19621
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1213275074279383,
+      "learning_rate": 1.1699398089297842e-05,
+      "loss": 0.9362,
+      "step": 19622
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5643110942088052,
+      "learning_rate": 1.1698646142147446e-05,
+      "loss": 0.8135,
+      "step": 19623
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4491402946407064,
+      "learning_rate": 1.1697894185106991e-05,
+      "loss": 1.0089,
+      "step": 19624
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8042456596328776,
+      "learning_rate": 1.1697142218180859e-05,
+      "loss": 0.7929,
+      "step": 19625
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9047497108363336,
+      "learning_rate": 1.1696390241373427e-05,
+      "loss": 0.8204,
+      "step": 19626
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.804937814246349,
+      "learning_rate": 1.1695638254689074e-05,
+      "loss": 0.92,
+      "step": 19627
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.814236299571973,
+      "learning_rate": 1.1694886258132182e-05,
+      "loss": 0.7765,
+      "step": 19628
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8612864414353067,
+      "learning_rate": 1.1694134251707121e-05,
+      "loss": 0.8887,
+      "step": 19629
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8244873917939186,
+      "learning_rate": 1.1693382235418276e-05,
+      "loss": 0.9087,
+      "step": 19630
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7314727995129864,
+      "learning_rate": 1.1692630209270024e-05,
+      "loss": 0.7796,
+      "step": 19631
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.521264017905917,
+      "learning_rate": 1.1691878173266743e-05,
+      "loss": 0.8664,
+      "step": 19632
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6152469429697183,
+      "learning_rate": 1.1691126127412814e-05,
+      "loss": 0.8631,
+      "step": 19633
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2251206140697275,
+      "learning_rate": 1.1690374071712608e-05,
+      "loss": 0.9318,
+      "step": 19634
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5901176263199441,
+      "learning_rate": 1.168962200617051e-05,
+      "loss": 0.9387,
+      "step": 19635
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8075551138111146,
+      "learning_rate": 1.1688869930790898e-05,
+      "loss": 0.9988,
+      "step": 19636
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6348616538789653,
+      "learning_rate": 1.1688117845578155e-05,
+      "loss": 0.8255,
+      "step": 19637
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4425141860300041,
+      "learning_rate": 1.168736575053665e-05,
+      "loss": 0.8955,
+      "step": 19638
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4983870376218498,
+      "learning_rate": 1.168661364567077e-05,
+      "loss": 0.8984,
+      "step": 19639
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7038987162821444,
+      "learning_rate": 1.1685861530984893e-05,
+      "loss": 0.8402,
+      "step": 19640
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5809121526371954,
+      "learning_rate": 1.1685109406483392e-05,
+      "loss": 0.8654,
+      "step": 19641
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6356990243426526,
+      "learning_rate": 1.1684357272170655e-05,
+      "loss": 0.8494,
+      "step": 19642
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6514507629052415,
+      "learning_rate": 1.1683605128051053e-05,
+      "loss": 0.8421,
+      "step": 19643
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7567295469454145,
+      "learning_rate": 1.1682852974128974e-05,
+      "loss": 0.786,
+      "step": 19644
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5213871455657204,
+      "learning_rate": 1.1682100810408785e-05,
+      "loss": 0.9151,
+      "step": 19645
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5997319165643538,
+      "learning_rate": 1.1681348636894879e-05,
+      "loss": 0.87,
+      "step": 19646
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4412593771974083,
+      "learning_rate": 1.1680596453591626e-05,
+      "loss": 0.88,
+      "step": 19647
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0764296839027025,
+      "learning_rate": 1.1679844260503407e-05,
+      "loss": 0.9258,
+      "step": 19648
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7406819708143693,
+      "learning_rate": 1.1679092057634605e-05,
+      "loss": 0.8877,
+      "step": 19649
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.547727568816485,
+      "learning_rate": 1.1678339844989597e-05,
+      "loss": 0.9402,
+      "step": 19650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7277815540539518,
+      "learning_rate": 1.1677587622572763e-05,
+      "loss": 0.7982,
+      "step": 19651
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5563284942137636,
+      "learning_rate": 1.1676835390388484e-05,
+      "loss": 0.8793,
+      "step": 19652
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4971279394204373,
+      "learning_rate": 1.1676083148441135e-05,
+      "loss": 0.7898,
+      "step": 19653
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4691817810813,
+      "learning_rate": 1.1675330896735103e-05,
+      "loss": 0.73,
+      "step": 19654
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6840703660846315,
+      "learning_rate": 1.1674578635274763e-05,
+      "loss": 0.977,
+      "step": 19655
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7708986431606644,
+      "learning_rate": 1.1673826364064496e-05,
+      "loss": 0.874,
+      "step": 19656
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.233845957790651,
+      "learning_rate": 1.167307408310868e-05,
+      "loss": 0.938,
+      "step": 19657
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7357486079848325,
+      "learning_rate": 1.1672321792411698e-05,
+      "loss": 0.8021,
+      "step": 19658
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7695438514003867,
+      "learning_rate": 1.167156949197793e-05,
+      "loss": 0.7038,
+      "step": 19659
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7849043835368,
+      "learning_rate": 1.1670817181811756e-05,
+      "loss": 0.668,
+      "step": 19660
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5829332048471525,
+      "learning_rate": 1.1670064861917554e-05,
+      "loss": 0.7573,
+      "step": 19661
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7984828356657034,
+      "learning_rate": 1.1669312532299705e-05,
+      "loss": 0.7445,
+      "step": 19662
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.686516910416376,
+      "learning_rate": 1.1668560192962595e-05,
+      "loss": 0.8762,
+      "step": 19663
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8266943747632738,
+      "learning_rate": 1.1667807843910592e-05,
+      "loss": 0.8238,
+      "step": 19664
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7241315893178133,
+      "learning_rate": 1.166705548514809e-05,
+      "loss": 0.8374,
+      "step": 19665
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7535889315536392,
+      "learning_rate": 1.1666303116679456e-05,
+      "loss": 0.7543,
+      "step": 19666
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6053293094285548,
+      "learning_rate": 1.1665550738509084e-05,
+      "loss": 0.9506,
+      "step": 19667
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.959993107855528,
+      "learning_rate": 1.1664798350641344e-05,
+      "loss": 0.7954,
+      "step": 19668
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9300818412446925,
+      "learning_rate": 1.1664045953080625e-05,
+      "loss": 0.8944,
+      "step": 19669
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8421008118984596,
+      "learning_rate": 1.1663293545831302e-05,
+      "loss": 0.904,
+      "step": 19670
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8566676223507275,
+      "learning_rate": 1.1662541128897758e-05,
+      "loss": 0.7799,
+      "step": 19671
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7335822622115515,
+      "learning_rate": 1.166178870228437e-05,
+      "loss": 0.9701,
+      "step": 19672
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9628654842670485,
+      "learning_rate": 1.1661036265995528e-05,
+      "loss": 0.731,
+      "step": 19673
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4939117309490828,
+      "learning_rate": 1.1660283820035601e-05,
+      "loss": 0.9652,
+      "step": 19674
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.948789165671887,
+      "learning_rate": 1.1659531364408979e-05,
+      "loss": 0.9503,
+      "step": 19675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8525562182680522,
+      "learning_rate": 1.1658778899120038e-05,
+      "loss": 0.7792,
+      "step": 19676
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6878932359938599,
+      "learning_rate": 1.1658026424173163e-05,
+      "loss": 0.8094,
+      "step": 19677
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.346653784321406,
+      "learning_rate": 1.1657273939572734e-05,
+      "loss": 0.9052,
+      "step": 19678
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.557632437189634,
+      "learning_rate": 1.1656521445323129e-05,
+      "loss": 0.8619,
+      "step": 19679
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5777685553951257,
+      "learning_rate": 1.1655768941428734e-05,
+      "loss": 0.779,
+      "step": 19680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7311238104623985,
+      "learning_rate": 1.1655016427893926e-05,
+      "loss": 0.9243,
+      "step": 19681
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.70511281303128,
+      "learning_rate": 1.1654263904723089e-05,
+      "loss": 0.7712,
+      "step": 19682
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4132273771041564,
+      "learning_rate": 1.1653511371920603e-05,
+      "loss": 0.8085,
+      "step": 19683
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7997513096818194,
+      "learning_rate": 1.1652758829490854e-05,
+      "loss": 0.856,
+      "step": 19684
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7513074478964052,
+      "learning_rate": 1.1652006277438214e-05,
+      "loss": 0.8923,
+      "step": 19685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8649242137941553,
+      "learning_rate": 1.1651253715767077e-05,
+      "loss": 0.7291,
+      "step": 19686
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.572516885407298,
+      "learning_rate": 1.1650501144481812e-05,
+      "loss": 0.9192,
+      "step": 19687
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5123137129880482,
+      "learning_rate": 1.164974856358681e-05,
+      "loss": 0.8334,
+      "step": 19688
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7755082777447814,
+      "learning_rate": 1.1648995973086448e-05,
+      "loss": 0.6938,
+      "step": 19689
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6764303738465804,
+      "learning_rate": 1.1648243372985111e-05,
+      "loss": 0.7809,
+      "step": 19690
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4313223872654832,
+      "learning_rate": 1.1647490763287178e-05,
+      "loss": 0.8319,
+      "step": 19691
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7944155755727873,
+      "learning_rate": 1.1646738143997031e-05,
+      "loss": 0.8105,
+      "step": 19692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7522494256383951,
+      "learning_rate": 1.1645985515119059e-05,
+      "loss": 0.7275,
+      "step": 19693
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5443274320618707,
+      "learning_rate": 1.164523287665763e-05,
+      "loss": 0.9,
+      "step": 19694
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6625369274956008,
+      "learning_rate": 1.164448022861714e-05,
+      "loss": 0.8946,
+      "step": 19695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.40807873283967,
+      "learning_rate": 1.1643727571001964e-05,
+      "loss": 0.7676,
+      "step": 19696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6455948590847835,
+      "learning_rate": 1.1642974903816488e-05,
+      "loss": 0.9084,
+      "step": 19697
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7300873135993182,
+      "learning_rate": 1.164222222706509e-05,
+      "loss": 0.7116,
+      "step": 19698
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8106717837981439,
+      "learning_rate": 1.1641469540752153e-05,
+      "loss": 0.8924,
+      "step": 19699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.816369170517364,
+      "learning_rate": 1.164071684488206e-05,
+      "loss": 0.9611,
+      "step": 19700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7404589936298103,
+      "learning_rate": 1.1639964139459196e-05,
+      "loss": 0.8158,
+      "step": 19701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7440187108037345,
+      "learning_rate": 1.1639211424487941e-05,
+      "loss": 0.7514,
+      "step": 19702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8697901729318878,
+      "learning_rate": 1.1638458699972678e-05,
+      "loss": 0.9182,
+      "step": 19703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9498692008624856,
+      "learning_rate": 1.1637705965917791e-05,
+      "loss": 0.9244,
+      "step": 19704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7823850269024766,
+      "learning_rate": 1.163695322232766e-05,
+      "loss": 0.6837,
+      "step": 19705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6367856657105564,
+      "learning_rate": 1.1636200469206671e-05,
+      "loss": 0.8497,
+      "step": 19706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4933042812421775,
+      "learning_rate": 1.1635447706559205e-05,
+      "loss": 0.8035,
+      "step": 19707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5182581648785063,
+      "learning_rate": 1.1634694934389641e-05,
+      "loss": 0.8626,
+      "step": 19708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.656044574032915,
+      "learning_rate": 1.1633942152702367e-05,
+      "loss": 0.9094,
+      "step": 19709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6505825100819598,
+      "learning_rate": 1.1633189361501767e-05,
+      "loss": 0.7741,
+      "step": 19710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5471555377633208,
+      "learning_rate": 1.1632436560792217e-05,
+      "loss": 0.9116,
+      "step": 19711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7252614230776797,
+      "learning_rate": 1.1631683750578112e-05,
+      "loss": 0.8593,
+      "step": 19712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4640408541786656,
+      "learning_rate": 1.163093093086382e-05,
+      "loss": 0.8009,
+      "step": 19713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7760658439199596,
+      "learning_rate": 1.1630178101653736e-05,
+      "loss": 0.7624,
+      "step": 19714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6611754739503988,
+      "learning_rate": 1.1629425262952235e-05,
+      "loss": 0.8577,
+      "step": 19715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7834501506219443,
+      "learning_rate": 1.1628672414763709e-05,
+      "loss": 0.7322,
+      "step": 19716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.834167669991319,
+      "learning_rate": 1.1627919557092532e-05,
+      "loss": 0.8819,
+      "step": 19717
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8494934350766228,
+      "learning_rate": 1.1627166689943098e-05,
+      "loss": 0.8056,
+      "step": 19718
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1785083003100114,
+      "learning_rate": 1.1626413813319779e-05,
+      "loss": 0.9112,
+      "step": 19719
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5902772984574725,
+      "learning_rate": 1.1625660927226966e-05,
+      "loss": 0.9198,
+      "step": 19720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1320489140527012,
+      "learning_rate": 1.162490803166904e-05,
+      "loss": 0.8598,
+      "step": 19721
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0420922186329955,
+      "learning_rate": 1.1624155126650385e-05,
+      "loss": 0.7707,
+      "step": 19722
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5736673030366135,
+      "learning_rate": 1.1623402212175384e-05,
+      "loss": 0.7541,
+      "step": 19723
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6137235520409068,
+      "learning_rate": 1.1622649288248422e-05,
+      "loss": 0.8188,
+      "step": 19724
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3478468376865458,
+      "learning_rate": 1.1621896354873883e-05,
+      "loss": 0.8381,
+      "step": 19725
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9689594733715488,
+      "learning_rate": 1.1621143412056146e-05,
+      "loss": 0.8064,
+      "step": 19726
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4073460389315713,
+      "learning_rate": 1.1620390459799605e-05,
+      "loss": 0.7866,
+      "step": 19727
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6590212503685213,
+      "learning_rate": 1.1619637498108634e-05,
+      "loss": 0.8305,
+      "step": 19728
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.555918451000445,
+      "learning_rate": 1.1618884526987619e-05,
+      "loss": 0.8571,
+      "step": 19729
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5640123298616395,
+      "learning_rate": 1.1618131546440949e-05,
+      "loss": 0.8749,
+      "step": 19730
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7194732497801772,
+      "learning_rate": 1.1617378556473004e-05,
+      "loss": 0.7007,
+      "step": 19731
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8598776289981598,
+      "learning_rate": 1.1616625557088169e-05,
+      "loss": 0.8915,
+      "step": 19732
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6714551031547649,
+      "learning_rate": 1.1615872548290827e-05,
+      "loss": 0.8243,
+      "step": 19733
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9120227088108965,
+      "learning_rate": 1.1615119530085365e-05,
+      "loss": 0.8484,
+      "step": 19734
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7291341362318664,
+      "learning_rate": 1.1614366502476169e-05,
+      "loss": 0.8813,
+      "step": 19735
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.807005138160626,
+      "learning_rate": 1.1613613465467612e-05,
+      "loss": 0.9353,
+      "step": 19736
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4097306314937867,
+      "learning_rate": 1.1612860419064092e-05,
+      "loss": 0.8696,
+      "step": 19737
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4705425867013058,
+      "learning_rate": 1.1612107363269987e-05,
+      "loss": 0.8334,
+      "step": 19738
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8688758936793588,
+      "learning_rate": 1.1611354298089683e-05,
+      "loss": 0.8581,
+      "step": 19739
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5848906404338685,
+      "learning_rate": 1.1610601223527563e-05,
+      "loss": 0.9081,
+      "step": 19740
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7724957220553205,
+      "learning_rate": 1.1609848139588014e-05,
+      "loss": 0.6945,
+      "step": 19741
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9309584708531107,
+      "learning_rate": 1.160909504627542e-05,
+      "loss": 0.9853,
+      "step": 19742
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5661017558658585,
+      "learning_rate": 1.1608341943594162e-05,
+      "loss": 0.856,
+      "step": 19743
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4890217523430633,
+      "learning_rate": 1.160758883154863e-05,
+      "loss": 0.93,
+      "step": 19744
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6198900583016458,
+      "learning_rate": 1.1606835710143207e-05,
+      "loss": 0.8184,
+      "step": 19745
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8439975911569906,
+      "learning_rate": 1.160608257938228e-05,
+      "loss": 1.0504,
+      "step": 19746
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9521262877039238,
+      "learning_rate": 1.1605329439270227e-05,
+      "loss": 0.8439,
+      "step": 19747
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8927377289135352,
+      "learning_rate": 1.1604576289811442e-05,
+      "loss": 0.9233,
+      "step": 19748
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2381569763079314,
+      "learning_rate": 1.1603823131010303e-05,
+      "loss": 0.7015,
+      "step": 19749
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.358987084951402,
+      "learning_rate": 1.16030699628712e-05,
+      "loss": 0.8512,
+      "step": 19750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.193742768870068,
+      "learning_rate": 1.1602316785398515e-05,
+      "loss": 0.8297,
+      "step": 19751
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0370943911772628,
+      "learning_rate": 1.1601563598596635e-05,
+      "loss": 0.8231,
+      "step": 19752
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9644157454705653,
+      "learning_rate": 1.1600810402469944e-05,
+      "loss": 0.8794,
+      "step": 19753
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.736805717329734,
+      "learning_rate": 1.1600057197022826e-05,
+      "loss": 0.8946,
+      "step": 19754
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.778489557808737,
+      "learning_rate": 1.1599303982259673e-05,
+      "loss": 0.8591,
+      "step": 19755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8021232655301456,
+      "learning_rate": 1.1598550758184864e-05,
+      "loss": 0.925,
+      "step": 19756
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6945486519394248,
+      "learning_rate": 1.1597797524802786e-05,
+      "loss": 0.8842,
+      "step": 19757
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7052280939153301,
+      "learning_rate": 1.1597044282117824e-05,
+      "loss": 0.9271,
+      "step": 19758
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.005453875273693,
+      "learning_rate": 1.1596291030134366e-05,
+      "loss": 0.963,
+      "step": 19759
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3895504660009184,
+      "learning_rate": 1.1595537768856794e-05,
+      "loss": 0.9392,
+      "step": 19760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5183378995154289,
+      "learning_rate": 1.15947844982895e-05,
+      "loss": 0.9355,
+      "step": 19761
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4818936080974237,
+      "learning_rate": 1.1594031218436864e-05,
+      "loss": 0.982,
+      "step": 19762
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7153154412337785,
+      "learning_rate": 1.1593277929303275e-05,
+      "loss": 0.676,
+      "step": 19763
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7078840576983165,
+      "learning_rate": 1.1592524630893111e-05,
+      "loss": 0.8713,
+      "step": 19764
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.616981089619901,
+      "learning_rate": 1.1591771323210771e-05,
+      "loss": 0.8105,
+      "step": 19765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5318089426840695,
+      "learning_rate": 1.1591018006260631e-05,
+      "loss": 0.7562,
+      "step": 19766
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5687477340155884,
+      "learning_rate": 1.1590264680047084e-05,
+      "loss": 0.9599,
+      "step": 19767
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8020935401619986,
+      "learning_rate": 1.1589511344574509e-05,
+      "loss": 0.8419,
+      "step": 19768
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9998499866398063,
+      "learning_rate": 1.1588757999847297e-05,
+      "loss": 0.9039,
+      "step": 19769
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.538282780880515,
+      "learning_rate": 1.1588004645869835e-05,
+      "loss": 0.8097,
+      "step": 19770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7943403698030408,
+      "learning_rate": 1.1587251282646505e-05,
+      "loss": 0.731,
+      "step": 19771
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6389356555263435,
+      "learning_rate": 1.1586497910181695e-05,
+      "loss": 0.9474,
+      "step": 19772
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7024109424281377,
+      "learning_rate": 1.1585744528479794e-05,
+      "loss": 0.9363,
+      "step": 19773
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7169855236091323,
+      "learning_rate": 1.1584991137545185e-05,
+      "loss": 0.9225,
+      "step": 19774
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7562691426008141,
+      "learning_rate": 1.1584237737382253e-05,
+      "loss": 0.8448,
+      "step": 19775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5292094530216787,
+      "learning_rate": 1.158348432799539e-05,
+      "loss": 0.8114,
+      "step": 19776
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4809840117846218,
+      "learning_rate": 1.1582730909388982e-05,
+      "loss": 0.8152,
+      "step": 19777
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5157582269245546,
+      "learning_rate": 1.1581977481567413e-05,
+      "loss": 0.8552,
+      "step": 19778
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8381932020746905,
+      "learning_rate": 1.1581224044535067e-05,
+      "loss": 0.9565,
+      "step": 19779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5827106897314676,
+      "learning_rate": 1.1580470598296337e-05,
+      "loss": 0.8825,
+      "step": 19780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8032989170615026,
+      "learning_rate": 1.1579717142855605e-05,
+      "loss": 0.8568,
+      "step": 19781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.969424802104219,
+      "learning_rate": 1.157896367821726e-05,
+      "loss": 0.9374,
+      "step": 19782
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.669910450685055,
+      "learning_rate": 1.1578210204385689e-05,
+      "loss": 0.8288,
+      "step": 19783
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.81463945474899,
+      "learning_rate": 1.157745672136528e-05,
+      "loss": 0.7321,
+      "step": 19784
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7682624559940197,
+      "learning_rate": 1.1576703229160416e-05,
+      "loss": 0.6934,
+      "step": 19785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5214839980139652,
+      "learning_rate": 1.1575949727775487e-05,
+      "loss": 0.8565,
+      "step": 19786
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.523868333244922,
+      "learning_rate": 1.1575196217214879e-05,
+      "loss": 0.7833,
+      "step": 19787
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6258535686754942,
+      "learning_rate": 1.1574442697482982e-05,
+      "loss": 0.7943,
+      "step": 19788
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7603607604447973,
+      "learning_rate": 1.1573689168584182e-05,
+      "loss": 0.8901,
+      "step": 19789
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.499213375373699,
+      "learning_rate": 1.1572935630522864e-05,
+      "loss": 0.9057,
+      "step": 19790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8982962654958715,
+      "learning_rate": 1.1572182083303417e-05,
+      "loss": 0.8295,
+      "step": 19791
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.169395979646538,
+      "learning_rate": 1.1571428526930227e-05,
+      "loss": 1.0023,
+      "step": 19792
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4829204137480019,
+      "learning_rate": 1.1570674961407684e-05,
+      "loss": 0.831,
+      "step": 19793
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8899446692024882,
+      "learning_rate": 1.156992138674017e-05,
+      "loss": 0.8788,
+      "step": 19794
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5047207451840783,
+      "learning_rate": 1.1569167802932084e-05,
+      "loss": 0.8555,
+      "step": 19795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7184756552115772,
+      "learning_rate": 1.15684142099878e-05,
+      "loss": 0.9334,
+      "step": 19796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8373386023586343,
+      "learning_rate": 1.1567660607911718e-05,
+      "loss": 0.7627,
+      "step": 19797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8333004797443897,
+      "learning_rate": 1.1566906996708212e-05,
+      "loss": 0.802,
+      "step": 19798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.468940234768286,
+      "learning_rate": 1.1566153376381683e-05,
+      "loss": 0.8515,
+      "step": 19799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7581195179805019,
+      "learning_rate": 1.1565399746936513e-05,
+      "loss": 0.6815,
+      "step": 19800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7361119353189254,
+      "learning_rate": 1.1564646108377088e-05,
+      "loss": 0.6742,
+      "step": 19801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3730414414917251,
+      "learning_rate": 1.1563892460707799e-05,
+      "loss": 0.7446,
+      "step": 19802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6908325334969643,
+      "learning_rate": 1.1563138803933031e-05,
+      "loss": 0.9619,
+      "step": 19803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7453750741409046,
+      "learning_rate": 1.1562385138057177e-05,
+      "loss": 0.8282,
+      "step": 19804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6646529323988717,
+      "learning_rate": 1.1561631463084621e-05,
+      "loss": 0.919,
+      "step": 19805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.716942809414044,
+      "learning_rate": 1.1560877779019752e-05,
+      "loss": 0.9279,
+      "step": 19806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.743226526991585,
+      "learning_rate": 1.1560124085866957e-05,
+      "loss": 0.7002,
+      "step": 19807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6082317742602392,
+      "learning_rate": 1.1559370383630628e-05,
+      "loss": 0.9175,
+      "step": 19808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5072930369536224,
+      "learning_rate": 1.155861667231515e-05,
+      "loss": 0.9745,
+      "step": 19809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8827271981635088,
+      "learning_rate": 1.1557862951924912e-05,
+      "loss": 0.9411,
+      "step": 19810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7570148304576598,
+      "learning_rate": 1.1557109222464302e-05,
+      "loss": 0.831,
+      "step": 19811
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8553734427995947,
+      "learning_rate": 1.155635548393771e-05,
+      "loss": 0.8168,
+      "step": 19812
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.516146536956088,
+      "learning_rate": 1.1555601736349525e-05,
+      "loss": 0.9732,
+      "step": 19813
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8686445916569256,
+      "learning_rate": 1.1554847979704133e-05,
+      "loss": 1.0387,
+      "step": 19814
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.691017851308433,
+      "learning_rate": 1.1554094214005922e-05,
+      "loss": 0.8093,
+      "step": 19815
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.604592265855881,
+      "learning_rate": 1.1553340439259287e-05,
+      "loss": 0.8101,
+      "step": 19816
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.201331419097739,
+      "learning_rate": 1.1552586655468608e-05,
+      "loss": 0.8346,
+      "step": 19817
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9300418525209906,
+      "learning_rate": 1.1551832862638282e-05,
+      "loss": 0.9919,
+      "step": 19818
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6733963859938177,
+      "learning_rate": 1.1551079060772687e-05,
+      "loss": 0.8058,
+      "step": 19819
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6741611997904142,
+      "learning_rate": 1.1550325249876226e-05,
+      "loss": 0.8686,
+      "step": 19820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8342853106559043,
+      "learning_rate": 1.1549571429953276e-05,
+      "loss": 0.7379,
+      "step": 19821
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8127855117167768,
+      "learning_rate": 1.1548817601008233e-05,
+      "loss": 0.9099,
+      "step": 19822
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6591147509506112,
+      "learning_rate": 1.1548063763045483e-05,
+      "loss": 0.8859,
+      "step": 19823
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5140101447830012,
+      "learning_rate": 1.154730991606941e-05,
+      "loss": 0.9116,
+      "step": 19824
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8106765588389359,
+      "learning_rate": 1.1546556060084418e-05,
+      "loss": 0.7641,
+      "step": 19825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4820505001046331,
+      "learning_rate": 1.1545802195094879e-05,
+      "loss": 0.8092,
+      "step": 19826
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7546190978491766,
+      "learning_rate": 1.1545048321105196e-05,
+      "loss": 0.7395,
+      "step": 19827
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6259258221894246,
+      "learning_rate": 1.1544294438119751e-05,
+      "loss": 0.8766,
+      "step": 19828
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7388026787156696,
+      "learning_rate": 1.1543540546142934e-05,
+      "loss": 0.7387,
+      "step": 19829
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7078227881785522,
+      "learning_rate": 1.1542786645179139e-05,
+      "loss": 0.8248,
+      "step": 19830
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.398422605660684,
+      "learning_rate": 1.1542032735232746e-05,
+      "loss": 0.8313,
+      "step": 19831
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7733671038614063,
+      "learning_rate": 1.1541278816308152e-05,
+      "loss": 0.7706,
+      "step": 19832
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5849350081473135,
+      "learning_rate": 1.1540524888409746e-05,
+      "loss": 0.9104,
+      "step": 19833
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.908543708993366,
+      "learning_rate": 1.1539770951541916e-05,
+      "loss": 0.8643,
+      "step": 19834
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7813414341193818,
+      "learning_rate": 1.1539017005709052e-05,
+      "loss": 0.8157,
+      "step": 19835
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5228615198671227,
+      "learning_rate": 1.1538263050915543e-05,
+      "loss": 0.7838,
+      "step": 19836
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6326837100131357,
+      "learning_rate": 1.1537509087165781e-05,
+      "loss": 0.7968,
+      "step": 19837
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6941323732624431,
+      "learning_rate": 1.1536755114464153e-05,
+      "loss": 0.8897,
+      "step": 19838
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3331984052668115,
+      "learning_rate": 1.1536001132815051e-05,
+      "loss": 0.7363,
+      "step": 19839
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6840731820885753,
+      "learning_rate": 1.1535247142222862e-05,
+      "loss": 0.7519,
+      "step": 19840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.567846015289599,
+      "learning_rate": 1.153449314269198e-05,
+      "loss": 0.9988,
+      "step": 19841
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.600914992065981,
+      "learning_rate": 1.1533739134226797e-05,
+      "loss": 1.0563,
+      "step": 19842
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9565319798806888,
+      "learning_rate": 1.1532985116831692e-05,
+      "loss": 0.8655,
+      "step": 19843
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7719609321139103,
+      "learning_rate": 1.1532231090511069e-05,
+      "loss": 0.9716,
+      "step": 19844
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6465354808767183,
+      "learning_rate": 1.1531477055269305e-05,
+      "loss": 0.8176,
+      "step": 19845
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1614776718421633,
+      "learning_rate": 1.15307230111108e-05,
+      "loss": 0.8644,
+      "step": 19846
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.642697753260951,
+      "learning_rate": 1.152996895803994e-05,
+      "loss": 0.8749,
+      "step": 19847
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0336458135522473,
+      "learning_rate": 1.1529214896061118e-05,
+      "loss": 0.8121,
+      "step": 19848
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4924583762814894,
+      "learning_rate": 1.152846082517872e-05,
+      "loss": 0.859,
+      "step": 19849
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5976651617439348,
+      "learning_rate": 1.1527706745397143e-05,
+      "loss": 0.8557,
+      "step": 19850
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7412147152035213,
+      "learning_rate": 1.1526952656720772e-05,
+      "loss": 0.9616,
+      "step": 19851
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6334713151882239,
+      "learning_rate": 1.1526198559153998e-05,
+      "loss": 0.873,
+      "step": 19852
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.660743375704743,
+      "learning_rate": 1.1525444452701213e-05,
+      "loss": 0.9425,
+      "step": 19853
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5194661063196,
+      "learning_rate": 1.1524690337366808e-05,
+      "loss": 0.8363,
+      "step": 19854
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.605834106418262,
+      "learning_rate": 1.1523936213155174e-05,
+      "loss": 0.8333,
+      "step": 19855
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6194091235318382,
+      "learning_rate": 1.1523182080070701e-05,
+      "loss": 0.849,
+      "step": 19856
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6231251349084062,
+      "learning_rate": 1.1522427938117777e-05,
+      "loss": 0.8664,
+      "step": 19857
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7405944877350212,
+      "learning_rate": 1.1521673787300798e-05,
+      "loss": 0.955,
+      "step": 19858
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4937583204781737,
+      "learning_rate": 1.1520919627624152e-05,
+      "loss": 0.8264,
+      "step": 19859
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8151345066239326,
+      "learning_rate": 1.152016545909223e-05,
+      "loss": 0.9114,
+      "step": 19860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.635350813767822,
+      "learning_rate": 1.1519411281709423e-05,
+      "loss": 0.832,
+      "step": 19861
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.535879847368252,
+      "learning_rate": 1.1518657095480123e-05,
+      "loss": 0.8325,
+      "step": 19862
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7309232770472267,
+      "learning_rate": 1.1517902900408722e-05,
+      "loss": 0.8249,
+      "step": 19863
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.879434995895144,
+      "learning_rate": 1.1517148696499607e-05,
+      "loss": 0.9004,
+      "step": 19864
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.53360220994507,
+      "learning_rate": 1.1516394483757176e-05,
+      "loss": 0.97,
+      "step": 19865
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8291091856462809,
+      "learning_rate": 1.1515640262185813e-05,
+      "loss": 0.7551,
+      "step": 19866
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7178743930772173,
+      "learning_rate": 1.1514886031789913e-05,
+      "loss": 0.831,
+      "step": 19867
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6762869580609192,
+      "learning_rate": 1.1514131792573865e-05,
+      "loss": 0.827,
+      "step": 19868
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5217417610333461,
+      "learning_rate": 1.1513377544542069e-05,
+      "loss": 0.7581,
+      "step": 19869
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.545759985185587,
+      "learning_rate": 1.1512623287698901e-05,
+      "loss": 1.0202,
+      "step": 19870
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7903978278387506,
+      "learning_rate": 1.1511869022048767e-05,
+      "loss": 0.7647,
+      "step": 19871
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5928398611427284,
+      "learning_rate": 1.1511114747596052e-05,
+      "loss": 0.76,
+      "step": 19872
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.395653016154856,
+      "learning_rate": 1.1510360464345147e-05,
+      "loss": 0.8837,
+      "step": 19873
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7328746480607737,
+      "learning_rate": 1.1509606172300447e-05,
+      "loss": 1.0145,
+      "step": 19874
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.814820685724862,
+      "learning_rate": 1.150885187146634e-05,
+      "loss": 0.8053,
+      "step": 19875
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6036617097474308,
+      "learning_rate": 1.1508097561847223e-05,
+      "loss": 0.9078,
+      "step": 19876
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8018809580781943,
+      "learning_rate": 1.150734324344748e-05,
+      "loss": 0.9327,
+      "step": 19877
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8166093824388954,
+      "learning_rate": 1.1506588916271512e-05,
+      "loss": 0.852,
+      "step": 19878
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2355927607612873,
+      "learning_rate": 1.1505834580323704e-05,
+      "loss": 0.7098,
+      "step": 19879
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6816358078163214,
+      "learning_rate": 1.150508023560845e-05,
+      "loss": 1.0306,
+      "step": 19880
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.62637593969522,
+      "learning_rate": 1.1504325882130143e-05,
+      "loss": 0.7996,
+      "step": 19881
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6160401554072181,
+      "learning_rate": 1.1503571519893175e-05,
+      "loss": 0.8354,
+      "step": 19882
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8273058687150137,
+      "learning_rate": 1.1502817148901937e-05,
+      "loss": 0.7207,
+      "step": 19883
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7162557923115955,
+      "learning_rate": 1.1502062769160821e-05,
+      "loss": 0.9105,
+      "step": 19884
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4628717812959375,
+      "learning_rate": 1.1501308380674221e-05,
+      "loss": 0.955,
+      "step": 19885
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.761429442469679,
+      "learning_rate": 1.1500553983446527e-05,
+      "loss": 0.7412,
+      "step": 19886
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6626066813139955,
+      "learning_rate": 1.1499799577482134e-05,
+      "loss": 0.727,
+      "step": 19887
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7646745464172577,
+      "learning_rate": 1.1499045162785434e-05,
+      "loss": 0.9227,
+      "step": 19888
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.086064600214368,
+      "learning_rate": 1.1498290739360815e-05,
+      "loss": 0.8792,
+      "step": 19889
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.507069043259364,
+      "learning_rate": 1.1497536307212675e-05,
+      "loss": 0.915,
+      "step": 19890
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8950106454528861,
+      "learning_rate": 1.1496781866345406e-05,
+      "loss": 0.9766,
+      "step": 19891
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.953065975921572,
+      "learning_rate": 1.1496027416763399e-05,
+      "loss": 0.825,
+      "step": 19892
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5820616277674515,
+      "learning_rate": 1.1495272958471046e-05,
+      "loss": 0.8088,
+      "step": 19893
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.921919800231232,
+      "learning_rate": 1.1494518491472738e-05,
+      "loss": 0.7351,
+      "step": 19894
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2079644673976064,
+      "learning_rate": 1.1493764015772875e-05,
+      "loss": 0.9436,
+      "step": 19895
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.888842427022826,
+      "learning_rate": 1.1493009531375838e-05,
+      "loss": 0.926,
+      "step": 19896
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8567185496508862,
+      "learning_rate": 1.1492255038286034e-05,
+      "loss": 0.8291,
+      "step": 19897
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.029993491608268,
+      "learning_rate": 1.1491500536507845e-05,
+      "loss": 0.8323,
+      "step": 19898
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.678803005049328,
+      "learning_rate": 1.149074602604567e-05,
+      "loss": 0.8653,
+      "step": 19899
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0953234056236796,
+      "learning_rate": 1.1489991506903896e-05,
+      "loss": 0.8676,
+      "step": 19900
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7912803817431955,
+      "learning_rate": 1.1489236979086922e-05,
+      "loss": 0.9367,
+      "step": 19901
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.827450319230305,
+      "learning_rate": 1.1488482442599138e-05,
+      "loss": 0.9252,
+      "step": 19902
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8157080947900361,
+      "learning_rate": 1.148772789744494e-05,
+      "loss": 0.8317,
+      "step": 19903
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9552583141016389,
+      "learning_rate": 1.1486973343628719e-05,
+      "loss": 0.8884,
+      "step": 19904
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.859892690346685,
+      "learning_rate": 1.1486218781154866e-05,
+      "loss": 0.8571,
+      "step": 19905
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6008573396229555,
+      "learning_rate": 1.1485464210027778e-05,
+      "loss": 0.796,
+      "step": 19906
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4990551212081276,
+      "learning_rate": 1.1484709630251847e-05,
+      "loss": 0.9675,
+      "step": 19907
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.643769701388465,
+      "learning_rate": 1.1483955041831465e-05,
+      "loss": 0.7862,
+      "step": 19908
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5094158047543147,
+      "learning_rate": 1.148320044477103e-05,
+      "loss": 0.7506,
+      "step": 19909
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5749496046906164,
+      "learning_rate": 1.1482445839074929e-05,
+      "loss": 0.9778,
+      "step": 19910
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.763939949572326,
+      "learning_rate": 1.1481691224747563e-05,
+      "loss": 0.7511,
+      "step": 19911
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.991817898071483,
+      "learning_rate": 1.1480936601793318e-05,
+      "loss": 1.117,
+      "step": 19912
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8056248899711882,
+      "learning_rate": 1.1480181970216593e-05,
+      "loss": 0.7243,
+      "step": 19913
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2610848103337897,
+      "learning_rate": 1.1479427330021778e-05,
+      "loss": 0.7911,
+      "step": 19914
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7304631697527958,
+      "learning_rate": 1.147867268121327e-05,
+      "loss": 0.7198,
+      "step": 19915
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5947180687972502,
+      "learning_rate": 1.1477918023795464e-05,
+      "loss": 0.863,
+      "step": 19916
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6245489912727666,
+      "learning_rate": 1.1477163357772747e-05,
+      "loss": 0.8746,
+      "step": 19917
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8733113870659013,
+      "learning_rate": 1.1476408683149522e-05,
+      "loss": 0.8557,
+      "step": 19918
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5755291111816656,
+      "learning_rate": 1.1475653999930174e-05,
+      "loss": 0.9414,
+      "step": 19919
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8722009757173612,
+      "learning_rate": 1.1474899308119105e-05,
+      "loss": 1.0498,
+      "step": 19920
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7627092090131662,
+      "learning_rate": 1.1474144607720701e-05,
+      "loss": 0.7764,
+      "step": 19921
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3177232680605266,
+      "learning_rate": 1.1473389898739363e-05,
+      "loss": 0.8343,
+      "step": 19922
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.52349314270893,
+      "learning_rate": 1.1472635181179483e-05,
+      "loss": 0.8156,
+      "step": 19923
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5430878229046487,
+      "learning_rate": 1.1471880455045452e-05,
+      "loss": 0.8993,
+      "step": 19924
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4828294907782003,
+      "learning_rate": 1.1471125720341671e-05,
+      "loss": 0.7731,
+      "step": 19925
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5760904835131437,
+      "learning_rate": 1.1470370977072524e-05,
+      "loss": 0.8569,
+      "step": 19926
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6453886649571863,
+      "learning_rate": 1.1469616225242418e-05,
+      "loss": 0.9843,
+      "step": 19927
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6093161701101075,
+      "learning_rate": 1.1468861464855738e-05,
+      "loss": 0.8498,
+      "step": 19928
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.322144725573969,
+      "learning_rate": 1.1468106695916885e-05,
+      "loss": 0.8487,
+      "step": 19929
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6426506270638894,
+      "learning_rate": 1.1467351918430248e-05,
+      "loss": 0.803,
+      "step": 19930
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8738347899423737,
+      "learning_rate": 1.1466597132400221e-05,
+      "loss": 0.7647,
+      "step": 19931
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.71367711590123,
+      "learning_rate": 1.1465842337831203e-05,
+      "loss": 0.9251,
+      "step": 19932
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.502216726817472,
+      "learning_rate": 1.1465087534727587e-05,
+      "loss": 0.8799,
+      "step": 19933
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.695612049210164,
+      "learning_rate": 1.1464332723093769e-05,
+      "loss": 0.8208,
+      "step": 19934
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7412214832780835,
+      "learning_rate": 1.146357790293414e-05,
+      "loss": 0.7501,
+      "step": 19935
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5182628860651328,
+      "learning_rate": 1.1462823074253098e-05,
+      "loss": 1.0128,
+      "step": 19936
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7401271287908164,
+      "learning_rate": 1.1462068237055037e-05,
+      "loss": 0.8289,
+      "step": 19937
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6983060816717632,
+      "learning_rate": 1.146131339134435e-05,
+      "loss": 0.866,
+      "step": 19938
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4797844914737759,
+      "learning_rate": 1.1460558537125437e-05,
+      "loss": 0.8218,
+      "step": 19939
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8199834386415459,
+      "learning_rate": 1.1459803674402687e-05,
+      "loss": 0.745,
+      "step": 19940
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5248556071520587,
+      "learning_rate": 1.14590488031805e-05,
+      "loss": 0.9162,
+      "step": 19941
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7550628118664964,
+      "learning_rate": 1.1458293923463269e-05,
+      "loss": 0.9745,
+      "step": 19942
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6655194265009112,
+      "learning_rate": 1.1457539035255387e-05,
+      "loss": 1.0129,
+      "step": 19943
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7463845574971127,
+      "learning_rate": 1.1456784138561254e-05,
+      "loss": 0.7165,
+      "step": 19944
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.603266930599086,
+      "learning_rate": 1.1456029233385259e-05,
+      "loss": 0.9657,
+      "step": 19945
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8124763325636468,
+      "learning_rate": 1.1455274319731806e-05,
+      "loss": 0.6619,
+      "step": 19946
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6570155814504726,
+      "learning_rate": 1.1454519397605279e-05,
+      "loss": 0.8317,
+      "step": 19947
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5837866709957844,
+      "learning_rate": 1.1453764467010084e-05,
+      "loss": 0.9086,
+      "step": 19948
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9842096386055974,
+      "learning_rate": 1.145300952795061e-05,
+      "loss": 0.793,
+      "step": 19949
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4191349158266053,
+      "learning_rate": 1.1452254580431256e-05,
+      "loss": 0.8747,
+      "step": 19950
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5008200522370654,
+      "learning_rate": 1.1451499624456415e-05,
+      "loss": 0.8124,
+      "step": 19951
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4819586118569095,
+      "learning_rate": 1.1450744660030484e-05,
+      "loss": 0.8082,
+      "step": 19952
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.413200140204025,
+      "learning_rate": 1.1449989687157859e-05,
+      "loss": 0.9202,
+      "step": 19953
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.442319612758501,
+      "learning_rate": 1.1449234705842933e-05,
+      "loss": 0.9441,
+      "step": 19954
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0409448291681263,
+      "learning_rate": 1.1448479716090103e-05,
+      "loss": 0.8745,
+      "step": 19955
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4661787909215795,
+      "learning_rate": 1.144772471790377e-05,
+      "loss": 0.8705,
+      "step": 19956
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6073714295591224,
+      "learning_rate": 1.1446969711288322e-05,
+      "loss": 0.8191,
+      "step": 19957
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8820861409026945,
+      "learning_rate": 1.144621469624816e-05,
+      "loss": 0.8268,
+      "step": 19958
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7720681181005424,
+      "learning_rate": 1.1445459672787676e-05,
+      "loss": 0.7696,
+      "step": 19959
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9652051276498166,
+      "learning_rate": 1.144470464091127e-05,
+      "loss": 0.9212,
+      "step": 19960
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5959612883105967,
+      "learning_rate": 1.1443949600623334e-05,
+      "loss": 0.8391,
+      "step": 19961
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.844928900368912,
+      "learning_rate": 1.1443194551928267e-05,
+      "loss": 0.9387,
+      "step": 19962
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7666859882708683,
+      "learning_rate": 1.1442439494830467e-05,
+      "loss": 0.891,
+      "step": 19963
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8813404570842018,
+      "learning_rate": 1.1441684429334323e-05,
+      "loss": 0.9405,
+      "step": 19964
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0644597356492156,
+      "learning_rate": 1.144092935544424e-05,
+      "loss": 0.9706,
+      "step": 19965
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5339257610759638,
+      "learning_rate": 1.1440174273164608e-05,
+      "loss": 0.9267,
+      "step": 19966
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7706333054314414,
+      "learning_rate": 1.143941918249983e-05,
+      "loss": 0.6923,
+      "step": 19967
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7134804176477216,
+      "learning_rate": 1.143866408345429e-05,
+      "loss": 0.843,
+      "step": 19968
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9484206770995283,
+      "learning_rate": 1.1437908976032395e-05,
+      "loss": 0.992,
+      "step": 19969
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4584148733853883,
+      "learning_rate": 1.1437153860238541e-05,
+      "loss": 0.8376,
+      "step": 19970
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.561141100815188,
+      "learning_rate": 1.143639873607712e-05,
+      "loss": 0.758,
+      "step": 19971
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7426914839218812,
+      "learning_rate": 1.1435643603552532e-05,
+      "loss": 0.9005,
+      "step": 19972
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.675544226309063,
+      "learning_rate": 1.143488846266917e-05,
+      "loss": 0.8061,
+      "step": 19973
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.462351946197673,
+      "learning_rate": 1.143413331343144e-05,
+      "loss": 0.8101,
+      "step": 19974
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6770188937919148,
+      "learning_rate": 1.1433378155843725e-05,
+      "loss": 0.8279,
+      "step": 19975
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6874538251711149,
+      "learning_rate": 1.1432622989910434e-05,
+      "loss": 0.9099,
+      "step": 19976
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7584831403629017,
+      "learning_rate": 1.1431867815635951e-05,
+      "loss": 0.7232,
+      "step": 19977
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0307697768119137,
+      "learning_rate": 1.1431112633024689e-05,
+      "loss": 0.8221,
+      "step": 19978
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6895069049843283,
+      "learning_rate": 1.1430357442081027e-05,
+      "loss": 1.0128,
+      "step": 19979
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8261435022515087,
+      "learning_rate": 1.1429602242809379e-05,
+      "loss": 0.8747,
+      "step": 19980
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7178415681547086,
+      "learning_rate": 1.142884703521413e-05,
+      "loss": 0.8283,
+      "step": 19981
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5136241569611573,
+      "learning_rate": 1.1428091819299683e-05,
+      "loss": 0.8317,
+      "step": 19982
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5785548073099185,
+      "learning_rate": 1.142733659507043e-05,
+      "loss": 0.9278,
+      "step": 19983
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8263174092247962,
+      "learning_rate": 1.1426581362530775e-05,
+      "loss": 0.8506,
+      "step": 19984
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.681341883937545,
+      "learning_rate": 1.142582612168511e-05,
+      "loss": 0.8036,
+      "step": 19985
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5387410912995274,
+      "learning_rate": 1.1425070872537836e-05,
+      "loss": 0.9242,
+      "step": 19986
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4764240156351782,
+      "learning_rate": 1.1424315615093343e-05,
+      "loss": 0.9296,
+      "step": 19987
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.490519893073459,
+      "learning_rate": 1.142356034935604e-05,
+      "loss": 0.9603,
+      "step": 19988
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0939696168080584,
+      "learning_rate": 1.1422805075330315e-05,
+      "loss": 0.944,
+      "step": 19989
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4653896671525688,
+      "learning_rate": 1.1422049793020566e-05,
+      "loss": 0.9722,
+      "step": 19990
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6244349581215107,
+      "learning_rate": 1.1421294502431197e-05,
+      "loss": 0.9434,
+      "step": 19991
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6517385100774093,
+      "learning_rate": 1.1420539203566598e-05,
+      "loss": 0.9903,
+      "step": 19992
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8281601234507283,
+      "learning_rate": 1.1419783896431174e-05,
+      "loss": 0.8425,
+      "step": 19993
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6374955070705033,
+      "learning_rate": 1.1419028581029315e-05,
+      "loss": 0.8989,
+      "step": 19994
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7692334902247807,
+      "learning_rate": 1.1418273257365426e-05,
+      "loss": 0.7102,
+      "step": 19995
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7247927729957392,
+      "learning_rate": 1.1417517925443897e-05,
+      "loss": 0.757,
+      "step": 19996
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7803568790719009,
+      "learning_rate": 1.1416762585269134e-05,
+      "loss": 0.8663,
+      "step": 19997
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7041777621751353,
+      "learning_rate": 1.1416007236845526e-05,
+      "loss": 0.8192,
+      "step": 19998
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7477839623495661,
+      "learning_rate": 1.141525188017748e-05,
+      "loss": 0.9521,
+      "step": 19999
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7533678752685113,
+      "learning_rate": 1.1414496515269384e-05,
+      "loss": 0.9968,
+      "step": 20000
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7296365031375314,
+      "learning_rate": 1.141374114212565e-05,
+      "loss": 0.8542,
+      "step": 20001
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8331366227252077,
+      "learning_rate": 1.141298576075066e-05,
+      "loss": 0.8198,
+      "step": 20002
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9117537788586993,
+      "learning_rate": 1.1412230371148822e-05,
+      "loss": 0.8095,
+      "step": 20003
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8298253196761827,
+      "learning_rate": 1.1411474973324533e-05,
+      "loss": 0.814,
+      "step": 20004
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9880161575505717,
+      "learning_rate": 1.1410719567282188e-05,
+      "loss": 0.8465,
+      "step": 20005
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6385799610298888,
+      "learning_rate": 1.1409964153026188e-05,
+      "loss": 0.949,
+      "step": 20006
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5009649828803258,
+      "learning_rate": 1.1409208730560927e-05,
+      "loss": 0.7407,
+      "step": 20007
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5225173675053039,
+      "learning_rate": 1.1408453299890811e-05,
+      "loss": 0.9711,
+      "step": 20008
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4935195800092884,
+      "learning_rate": 1.1407697861020232e-05,
+      "loss": 0.7879,
+      "step": 20009
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5405805554774485,
+      "learning_rate": 1.1406942413953593e-05,
+      "loss": 0.8855,
+      "step": 20010
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4878924991134423,
+      "learning_rate": 1.1406186958695287e-05,
+      "loss": 0.8571,
+      "step": 20011
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7696430704876878,
+      "learning_rate": 1.1405431495249714e-05,
+      "loss": 0.8543,
+      "step": 20012
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9143618205646988,
+      "learning_rate": 1.1404676023621278e-05,
+      "loss": 1.0152,
+      "step": 20013
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5558285266125655,
+      "learning_rate": 1.140392054381437e-05,
+      "loss": 0.8471,
+      "step": 20014
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7309955746310002,
+      "learning_rate": 1.1403165055833394e-05,
+      "loss": 0.8744,
+      "step": 20015
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6084504693961648,
+      "learning_rate": 1.1402409559682748e-05,
+      "loss": 1.0134,
+      "step": 20016
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6782651675235194,
+      "learning_rate": 1.1401654055366826e-05,
+      "loss": 0.8605,
+      "step": 20017
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5116084129436014,
+      "learning_rate": 1.1400898542890034e-05,
+      "loss": 0.9515,
+      "step": 20018
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7604636877428205,
+      "learning_rate": 1.1400143022256764e-05,
+      "loss": 0.7237,
+      "step": 20019
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4428698831398237,
+      "learning_rate": 1.139938749347142e-05,
+      "loss": 0.8841,
+      "step": 20020
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.600859487406543,
+      "learning_rate": 1.1398631956538398e-05,
+      "loss": 0.8209,
+      "step": 20021
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.642356964688881,
+      "learning_rate": 1.1397876411462096e-05,
+      "loss": 0.9642,
+      "step": 20022
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9037701282836856,
+      "learning_rate": 1.1397120858246921e-05,
+      "loss": 0.8956,
+      "step": 20023
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6469630955347354,
+      "learning_rate": 1.139636529689726e-05,
+      "loss": 0.9166,
+      "step": 20024
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8734762944415349,
+      "learning_rate": 1.1395609727417522e-05,
+      "loss": 0.946,
+      "step": 20025
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.771732063678891,
+      "learning_rate": 1.13948541498121e-05,
+      "loss": 0.8593,
+      "step": 20026
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5020952168369963,
+      "learning_rate": 1.1394098564085399e-05,
+      "loss": 0.881,
+      "step": 20027
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6515122154465023,
+      "learning_rate": 1.1393342970241809e-05,
+      "loss": 0.9593,
+      "step": 20028
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6054069863811458,
+      "learning_rate": 1.139258736828574e-05,
+      "loss": 0.7784,
+      "step": 20029
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.474499900216771,
+      "learning_rate": 1.1391831758221585e-05,
+      "loss": 0.7922,
+      "step": 20030
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7409125651312657,
+      "learning_rate": 1.1391076140053745e-05,
+      "loss": 0.6856,
+      "step": 20031
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.736506878442236,
+      "learning_rate": 1.139032051378662e-05,
+      "loss": 0.8371,
+      "step": 20032
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7450548834331712,
+      "learning_rate": 1.1389564879424605e-05,
+      "loss": 0.7102,
+      "step": 20033
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7165268383818106,
+      "learning_rate": 1.1388809236972107e-05,
+      "loss": 0.9252,
+      "step": 20034
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5966769700544896,
+      "learning_rate": 1.138805358643352e-05,
+      "loss": 0.8408,
+      "step": 20035
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8199816109863345,
+      "learning_rate": 1.1387297927813248e-05,
+      "loss": 0.8653,
+      "step": 20036
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.104066051210151,
+      "learning_rate": 1.1386542261115685e-05,
+      "loss": 0.8163,
+      "step": 20037
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7947298663326674,
+      "learning_rate": 1.1385786586345238e-05,
+      "loss": 0.843,
+      "step": 20038
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9332426164007457,
+      "learning_rate": 1.13850309035063e-05,
+      "loss": 0.9498,
+      "step": 20039
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7713774824989135,
+      "learning_rate": 1.1384275212603273e-05,
+      "loss": 0.7837,
+      "step": 20040
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7436995309286976,
+      "learning_rate": 1.138351951364056e-05,
+      "loss": 0.768,
+      "step": 20041
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.584106132375453,
+      "learning_rate": 1.1382763806622556e-05,
+      "loss": 0.965,
+      "step": 20042
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6064690157517036,
+      "learning_rate": 1.1382008091553664e-05,
+      "loss": 0.8961,
+      "step": 20043
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6358175904416696,
+      "learning_rate": 1.1381252368438283e-05,
+      "loss": 0.8016,
+      "step": 20044
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7983801538401827,
+      "learning_rate": 1.1380496637280816e-05,
+      "loss": 0.759,
+      "step": 20045
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6165256658757576,
+      "learning_rate": 1.1379740898085657e-05,
+      "loss": 0.7824,
+      "step": 20046
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.399369017656001,
+      "learning_rate": 1.1378985150857211e-05,
+      "loss": 0.8948,
+      "step": 20047
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.321604114482079,
+      "learning_rate": 1.137822939559988e-05,
+      "loss": 0.9997,
+      "step": 20048
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4484461779280453,
+      "learning_rate": 1.1377473632318057e-05,
+      "loss": 0.8611,
+      "step": 20049
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.749193218531454,
+      "learning_rate": 1.1376717861016149e-05,
+      "loss": 0.9705,
+      "step": 20050
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6448830082670696,
+      "learning_rate": 1.137596208169855e-05,
+      "loss": 0.9306,
+      "step": 20051
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8087735329396828,
+      "learning_rate": 1.1375206294369669e-05,
+      "loss": 0.8711,
+      "step": 20052
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1883310866778625,
+      "learning_rate": 1.1374450499033898e-05,
+      "loss": 0.9143,
+      "step": 20053
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5154454039634013,
+      "learning_rate": 1.1373694695695644e-05,
+      "loss": 1.0292,
+      "step": 20054
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4875158534878734,
+      "learning_rate": 1.1372938884359302e-05,
+      "loss": 0.7863,
+      "step": 20055
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.094371725210933,
+      "learning_rate": 1.1372183065029273e-05,
+      "loss": 0.9108,
+      "step": 20056
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.499714284175747,
+      "learning_rate": 1.1371427237709964e-05,
+      "loss": 0.9396,
+      "step": 20057
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8107882943659468,
+      "learning_rate": 1.1370671402405768e-05,
+      "loss": 0.8837,
+      "step": 20058
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5277157639184853,
+      "learning_rate": 1.1369915559121094e-05,
+      "loss": 0.8838,
+      "step": 20059
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8736542512752126,
+      "learning_rate": 1.1369159707860336e-05,
+      "loss": 0.9176,
+      "step": 20060
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9018725072467206,
+      "learning_rate": 1.1368403848627896e-05,
+      "loss": 0.885,
+      "step": 20061
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.544246300510382,
+      "learning_rate": 1.1367647981428173e-05,
+      "loss": 0.8583,
+      "step": 20062
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5952085176540376,
+      "learning_rate": 1.1366892106265575e-05,
+      "loss": 1.0079,
+      "step": 20063
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6506691272524887,
+      "learning_rate": 1.1366136223144495e-05,
+      "loss": 0.842,
+      "step": 20064
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5036505490295462,
+      "learning_rate": 1.136538033206934e-05,
+      "loss": 1.0066,
+      "step": 20065
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4628375343625584,
+      "learning_rate": 1.1364624433044507e-05,
+      "loss": 0.9272,
+      "step": 20066
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.677023400493113,
+      "learning_rate": 1.1363868526074397e-05,
+      "loss": 0.8346,
+      "step": 20067
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8237168946073794,
+      "learning_rate": 1.1363112611163415e-05,
+      "loss": 0.8067,
+      "step": 20068
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7692399217244075,
+      "learning_rate": 1.1362356688315959e-05,
+      "loss": 0.7384,
+      "step": 20069
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5250557756783252,
+      "learning_rate": 1.136160075753643e-05,
+      "loss": 0.9488,
+      "step": 20070
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7565090398297513,
+      "learning_rate": 1.1360844818829234e-05,
+      "loss": 0.7192,
+      "step": 20071
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5048537485342672,
+      "learning_rate": 1.1360088872198764e-05,
+      "loss": 0.9293,
+      "step": 20072
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5453369955324432,
+      "learning_rate": 1.1359332917649427e-05,
+      "loss": 0.7504,
+      "step": 20073
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4916223831257722,
+      "learning_rate": 1.135857695518563e-05,
+      "loss": 0.8469,
+      "step": 20074
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.574320923114633,
+      "learning_rate": 1.1357820984811758e-05,
+      "loss": 0.8186,
+      "step": 20075
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.149237975028731,
+      "learning_rate": 1.135706500653223e-05,
+      "loss": 0.7986,
+      "step": 20076
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8790010312934544,
+      "learning_rate": 1.1356309020351435e-05,
+      "loss": 0.9788,
+      "step": 20077
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6385229591359318,
+      "learning_rate": 1.1355553026273783e-05,
+      "loss": 0.9246,
+      "step": 20078
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.698034784175329,
+      "learning_rate": 1.1354797024303668e-05,
+      "loss": 0.9833,
+      "step": 20079
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6509736933895365,
+      "learning_rate": 1.1354041014445499e-05,
+      "loss": 0.9524,
+      "step": 20080
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.730523280596124,
+      "learning_rate": 1.1353284996703672e-05,
+      "loss": 0.9076,
+      "step": 20081
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.550472856545373,
+      "learning_rate": 1.1352528971082594e-05,
+      "loss": 0.94,
+      "step": 20082
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4617618330709354,
+      "learning_rate": 1.1351772937586663e-05,
+      "loss": 0.9379,
+      "step": 20083
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.740779892805053,
+      "learning_rate": 1.1351016896220282e-05,
+      "loss": 1.0183,
+      "step": 20084
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.98944192846243,
+      "learning_rate": 1.1350260846987852e-05,
+      "loss": 0.8152,
+      "step": 20085
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5348094983741944,
+      "learning_rate": 1.1349504789893776e-05,
+      "loss": 0.8581,
+      "step": 20086
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5450974669359503,
+      "learning_rate": 1.1348748724942458e-05,
+      "loss": 0.7117,
+      "step": 20087
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5644482770364276,
+      "learning_rate": 1.1347992652138298e-05,
+      "loss": 1.0009,
+      "step": 20088
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8224050031191825,
+      "learning_rate": 1.1347236571485696e-05,
+      "loss": 0.6924,
+      "step": 20089
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4743250342396963,
+      "learning_rate": 1.1346480482989055e-05,
+      "loss": 0.7037,
+      "step": 20090
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7489121937027494,
+      "learning_rate": 1.1345724386652781e-05,
+      "loss": 0.7314,
+      "step": 20091
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.709498380508593,
+      "learning_rate": 1.1344968282481273e-05,
+      "loss": 0.944,
+      "step": 20092
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5676580093997843,
+      "learning_rate": 1.1344212170478932e-05,
+      "loss": 0.8863,
+      "step": 20093
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6331554591243502,
+      "learning_rate": 1.1343456050650163e-05,
+      "loss": 0.9082,
+      "step": 20094
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4673916904187754,
+      "learning_rate": 1.134269992299937e-05,
+      "loss": 0.9864,
+      "step": 20095
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5752291914268461,
+      "learning_rate": 1.1341943787530951e-05,
+      "loss": 0.8556,
+      "step": 20096
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4923164346650186,
+      "learning_rate": 1.1341187644249311e-05,
+      "loss": 0.9574,
+      "step": 20097
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4993228815187747,
+      "learning_rate": 1.134043149315885e-05,
+      "loss": 0.8236,
+      "step": 20098
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.681795930595709,
+      "learning_rate": 1.1339675334263976e-05,
+      "loss": 0.8301,
+      "step": 20099
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7290860945568989,
+      "learning_rate": 1.1338919167569085e-05,
+      "loss": 0.846,
+      "step": 20100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9632661662010196,
+      "learning_rate": 1.1338162993078585e-05,
+      "loss": 0.9757,
+      "step": 20101
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.634509311930549,
+      "learning_rate": 1.1337406810796872e-05,
+      "loss": 0.8888,
+      "step": 20102
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8489741697823894,
+      "learning_rate": 1.133665062072836e-05,
+      "loss": 0.8184,
+      "step": 20103
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.591474704194511,
+      "learning_rate": 1.1335894422877441e-05,
+      "loss": 0.982,
+      "step": 20104
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5584095201266763,
+      "learning_rate": 1.1335138217248518e-05,
+      "loss": 0.8907,
+      "step": 20105
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2511810773021073,
+      "learning_rate": 1.1334382003846006e-05,
+      "loss": 0.8315,
+      "step": 20106
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.958494726424024,
+      "learning_rate": 1.133362578267429e-05,
+      "loss": 0.9224,
+      "step": 20107
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2074984422531998,
+      "learning_rate": 1.133286955373779e-05,
+      "loss": 0.7187,
+      "step": 20108
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.703155728395006,
+      "learning_rate": 1.1332113317040898e-05,
+      "loss": 0.9427,
+      "step": 20109
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5889437569615816,
+      "learning_rate": 1.1331357072588024e-05,
+      "loss": 0.9273,
+      "step": 20110
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.550379068281583,
+      "learning_rate": 1.1330600820383564e-05,
+      "loss": 0.8894,
+      "step": 20111
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7725381687689584,
+      "learning_rate": 1.1329844560431926e-05,
+      "loss": 0.7264,
+      "step": 20112
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.584398318807647,
+      "learning_rate": 1.132908829273751e-05,
+      "loss": 0.9628,
+      "step": 20113
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7423019890929974,
+      "learning_rate": 1.1328332017304724e-05,
+      "loss": 0.9571,
+      "step": 20114
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.92375252527954,
+      "learning_rate": 1.1327575734137968e-05,
+      "loss": 0.8689,
+      "step": 20115
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6194307911510517,
+      "learning_rate": 1.1326819443241645e-05,
+      "loss": 0.901,
+      "step": 20116
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7314156496565776,
+      "learning_rate": 1.1326063144620158e-05,
+      "loss": 0.905,
+      "step": 20117
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.751105847947157,
+      "learning_rate": 1.1325306838277914e-05,
+      "loss": 0.7904,
+      "step": 20118
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6645970643213364,
+      "learning_rate": 1.1324550524219313e-05,
+      "loss": 0.9323,
+      "step": 20119
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0941162064067824,
+      "learning_rate": 1.1323794202448759e-05,
+      "loss": 0.9427,
+      "step": 20120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7696381351869709,
+      "learning_rate": 1.1323037872970657e-05,
+      "loss": 0.7069,
+      "step": 20121
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.899951815854348,
+      "learning_rate": 1.1322281535789408e-05,
+      "loss": 0.8142,
+      "step": 20122
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.972266783086978,
+      "learning_rate": 1.1321525190909419e-05,
+      "loss": 0.9748,
+      "step": 20123
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6811996075607483,
+      "learning_rate": 1.1320768838335092e-05,
+      "loss": 0.8328,
+      "step": 20124
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.703513193047584,
+      "learning_rate": 1.1320012478070831e-05,
+      "loss": 0.9728,
+      "step": 20125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5968479442237773,
+      "learning_rate": 1.1319256110121036e-05,
+      "loss": 0.743,
+      "step": 20126
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6500616684101823,
+      "learning_rate": 1.131849973449012e-05,
+      "loss": 0.9191,
+      "step": 20127
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.180390646554525,
+      "learning_rate": 1.1317743351182475e-05,
+      "loss": 0.7733,
+      "step": 20128
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7324344175576106,
+      "learning_rate": 1.1316986960202516e-05,
+      "loss": 0.6922,
+      "step": 20129
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0577936424126406,
+      "learning_rate": 1.1316230561554637e-05,
+      "loss": 0.8878,
+      "step": 20130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7731223701433219,
+      "learning_rate": 1.131547415524325e-05,
+      "loss": 0.928,
+      "step": 20131
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7160018085259847,
+      "learning_rate": 1.1314717741272757e-05,
+      "loss": 0.9865,
+      "step": 20132
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1679934206919573,
+      "learning_rate": 1.1313961319647561e-05,
+      "loss": 0.9665,
+      "step": 20133
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.577936184279289,
+      "learning_rate": 1.1313204890372065e-05,
+      "loss": 0.8116,
+      "step": 20134
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5083181515527329,
+      "learning_rate": 1.1312448453450672e-05,
+      "loss": 0.824,
+      "step": 20135
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8070909848697393,
+      "learning_rate": 1.1311692008887793e-05,
+      "loss": 0.7092,
+      "step": 20136
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8152945184462165,
+      "learning_rate": 1.1310935556687825e-05,
+      "loss": 0.7821,
+      "step": 20137
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.38460080149644,
+      "learning_rate": 1.1310179096855176e-05,
+      "loss": 0.9062,
+      "step": 20138
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6324047763353833,
+      "learning_rate": 1.1309422629394251e-05,
+      "loss": 0.8593,
+      "step": 20139
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.487647479536861,
+      "learning_rate": 1.1308666154309451e-05,
+      "loss": 0.9153,
+      "step": 20140
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6469860320929262,
+      "learning_rate": 1.1307909671605182e-05,
+      "loss": 0.8184,
+      "step": 20141
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8035700997959174,
+      "learning_rate": 1.1307153181285848e-05,
+      "loss": 0.9488,
+      "step": 20142
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6503393364165666,
+      "learning_rate": 1.1306396683355858e-05,
+      "loss": 0.868,
+      "step": 20143
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4061393420445703,
+      "learning_rate": 1.130564017781961e-05,
+      "loss": 0.8637,
+      "step": 20144
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7794351121351628,
+      "learning_rate": 1.1304883664681512e-05,
+      "loss": 0.6663,
+      "step": 20145
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6682756483116352,
+      "learning_rate": 1.1304127143945971e-05,
+      "loss": 0.7994,
+      "step": 20146
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8756442283254826,
+      "learning_rate": 1.1303370615617384e-05,
+      "loss": 0.8185,
+      "step": 20147
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5741282669942045,
+      "learning_rate": 1.1302614079700165e-05,
+      "loss": 0.8098,
+      "step": 20148
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6242082088069565,
+      "learning_rate": 1.130185753619871e-05,
+      "loss": 0.7757,
+      "step": 20149
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9389695866288292,
+      "learning_rate": 1.1301100985117433e-05,
+      "loss": 0.7746,
+      "step": 20150
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6021527163193985,
+      "learning_rate": 1.130034442646073e-05,
+      "loss": 0.8501,
+      "step": 20151
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9572312809957033,
+      "learning_rate": 1.1299587860233014e-05,
+      "loss": 0.8859,
+      "step": 20152
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0189781633647406,
+      "learning_rate": 1.1298831286438683e-05,
+      "loss": 0.9974,
+      "step": 20153
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5032956802888402,
+      "learning_rate": 1.1298074705082145e-05,
+      "loss": 0.9385,
+      "step": 20154
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.52959800214092,
+      "learning_rate": 1.1297318116167808e-05,
+      "loss": 0.7743,
+      "step": 20155
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.431935940144541,
+      "learning_rate": 1.129656151970007e-05,
+      "loss": 0.8286,
+      "step": 20156
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5883982149592863,
+      "learning_rate": 1.1295804915683343e-05,
+      "loss": 0.9742,
+      "step": 20157
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8439217775880836,
+      "learning_rate": 1.1295048304122029e-05,
+      "loss": 0.9378,
+      "step": 20158
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.549612431253319,
+      "learning_rate": 1.1294291685020535e-05,
+      "loss": 0.8536,
+      "step": 20159
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.920010729523101,
+      "learning_rate": 1.1293535058383262e-05,
+      "loss": 0.9471,
+      "step": 20160
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5267704309875287,
+      "learning_rate": 1.1292778424214623e-05,
+      "loss": 0.9418,
+      "step": 20161
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.116293100068908,
+      "learning_rate": 1.1292021782519015e-05,
+      "loss": 0.8609,
+      "step": 20162
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6288085791438423,
+      "learning_rate": 1.1291265133300849e-05,
+      "loss": 0.9906,
+      "step": 20163
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4994728025451771,
+      "learning_rate": 1.1290508476564528e-05,
+      "loss": 0.8734,
+      "step": 20164
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.538263939574793,
+      "learning_rate": 1.1289751812314458e-05,
+      "loss": 0.8144,
+      "step": 20165
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7557786667529235,
+      "learning_rate": 1.1288995140555044e-05,
+      "loss": 0.6986,
+      "step": 20166
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1121503062968916,
+      "learning_rate": 1.1288238461290693e-05,
+      "loss": 0.9618,
+      "step": 20167
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.969123826542206,
+      "learning_rate": 1.128748177452581e-05,
+      "loss": 0.89,
+      "step": 20168
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7603289808140934,
+      "learning_rate": 1.12867250802648e-05,
+      "loss": 0.7743,
+      "step": 20169
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7904240650535407,
+      "learning_rate": 1.1285968378512069e-05,
+      "loss": 0.6942,
+      "step": 20170
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8524578977652948,
+      "learning_rate": 1.1285211669272023e-05,
+      "loss": 0.889,
+      "step": 20171
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7598058830977842,
+      "learning_rate": 1.1284454952549067e-05,
+      "loss": 0.7461,
+      "step": 20172
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.350011248211703,
+      "learning_rate": 1.1283698228347609e-05,
+      "loss": 0.7906,
+      "step": 20173
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.080060087129128,
+      "learning_rate": 1.1282941496672053e-05,
+      "loss": 0.9127,
+      "step": 20174
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.5925662280168185,
+      "learning_rate": 1.1282184757526806e-05,
+      "loss": 0.8802,
+      "step": 20175
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9047117566586191,
+      "learning_rate": 1.1281428010916276e-05,
+      "loss": 0.9109,
+      "step": 20176
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5550820631566904,
+      "learning_rate": 1.128067125684486e-05,
+      "loss": 0.8909,
+      "step": 20177
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9820990144500115,
+      "learning_rate": 1.1279914495316975e-05,
+      "loss": 0.9374,
+      "step": 20178
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4952660496388908,
+      "learning_rate": 1.127915772633702e-05,
+      "loss": 0.7989,
+      "step": 20179
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.686688104031446,
+      "learning_rate": 1.1278400949909407e-05,
+      "loss": 0.8368,
+      "step": 20180
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4417723141640195,
+      "learning_rate": 1.1277644166038535e-05,
+      "loss": 0.8956,
+      "step": 20181
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8335234983005118,
+      "learning_rate": 1.127688737472882e-05,
+      "loss": 0.9521,
+      "step": 20182
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.657927095662369,
+      "learning_rate": 1.1276130575984658e-05,
+      "loss": 0.7676,
+      "step": 20183
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7275384829971976,
+      "learning_rate": 1.1275373769810461e-05,
+      "loss": 0.7948,
+      "step": 20184
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.371098638060373,
+      "learning_rate": 1.1274616956210633e-05,
+      "loss": 0.8898,
+      "step": 20185
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.06402840609793,
+      "learning_rate": 1.1273860135189583e-05,
+      "loss": 0.8615,
+      "step": 20186
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7417583012405338,
+      "learning_rate": 1.1273103306751713e-05,
+      "loss": 0.6729,
+      "step": 20187
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.846237432544169,
+      "learning_rate": 1.1272346470901436e-05,
+      "loss": 0.6726,
+      "step": 20188
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7559394304011187,
+      "learning_rate": 1.1271589627643154e-05,
+      "loss": 0.8533,
+      "step": 20189
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.456491567059257,
+      "learning_rate": 1.1270832776981272e-05,
+      "loss": 0.927,
+      "step": 20190
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7363329104786944,
+      "learning_rate": 1.1270075918920202e-05,
+      "loss": 0.8157,
+      "step": 20191
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5737965309102726,
+      "learning_rate": 1.126931905346435e-05,
+      "loss": 0.7104,
+      "step": 20192
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9448543179947386,
+      "learning_rate": 1.1268562180618116e-05,
+      "loss": 0.8107,
+      "step": 20193
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8762608802119753,
+      "learning_rate": 1.1267805300385914e-05,
+      "loss": 0.8899,
+      "step": 20194
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8217161396128159,
+      "learning_rate": 1.1267048412772147e-05,
+      "loss": 0.7582,
+      "step": 20195
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.741109677902333,
+      "learning_rate": 1.1266291517781225e-05,
+      "loss": 0.8609,
+      "step": 20196
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7534460626031907,
+      "learning_rate": 1.1265534615417553e-05,
+      "loss": 0.7176,
+      "step": 20197
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.548612090994168,
+      "learning_rate": 1.1264777705685535e-05,
+      "loss": 0.9287,
+      "step": 20198
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.681844652994417,
+      "learning_rate": 1.1264020788589582e-05,
+      "loss": 0.8479,
+      "step": 20199
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7518913114508987,
+      "learning_rate": 1.12632638641341e-05,
+      "loss": 0.8722,
+      "step": 20200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.652140812045073,
+      "learning_rate": 1.1262506932323496e-05,
+      "loss": 0.8755,
+      "step": 20201
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7952075412540816,
+      "learning_rate": 1.1261749993162175e-05,
+      "loss": 0.9627,
+      "step": 20202
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7432296128196338,
+      "learning_rate": 1.1260993046654546e-05,
+      "loss": 0.8655,
+      "step": 20203
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5691636787639711,
+      "learning_rate": 1.1260236092805022e-05,
+      "loss": 0.8622,
+      "step": 20204
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7944925749175538,
+      "learning_rate": 1.1259479131617999e-05,
+      "loss": 0.7969,
+      "step": 20205
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.681830824987752,
+      "learning_rate": 1.1258722163097893e-05,
+      "loss": 0.9725,
+      "step": 20206
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6747083765884028,
+      "learning_rate": 1.1257965187249102e-05,
+      "loss": 0.8224,
+      "step": 20207
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.502382142958547,
+      "learning_rate": 1.1257208204076046e-05,
+      "loss": 1.0841,
+      "step": 20208
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7880158826980885,
+      "learning_rate": 1.1256451213583121e-05,
+      "loss": 0.7211,
+      "step": 20209
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.800164865177909,
+      "learning_rate": 1.1255694215774743e-05,
+      "loss": 0.7932,
+      "step": 20210
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7325559664009929,
+      "learning_rate": 1.1254937210655316e-05,
+      "loss": 0.7087,
+      "step": 20211
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.394693105974517,
+      "learning_rate": 1.1254180198229245e-05,
+      "loss": 0.9234,
+      "step": 20212
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5365416842766493,
+      "learning_rate": 1.1253423178500943e-05,
+      "loss": 0.9513,
+      "step": 20213
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.636794350237761,
+      "learning_rate": 1.125266615147481e-05,
+      "loss": 0.8102,
+      "step": 20214
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6542779465465627,
+      "learning_rate": 1.125190911715526e-05,
+      "loss": 0.8133,
+      "step": 20215
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9948739470792751,
+      "learning_rate": 1.1251152075546698e-05,
+      "loss": 0.8761,
+      "step": 20216
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6969563172824536,
+      "learning_rate": 1.1250395026653536e-05,
+      "loss": 0.8956,
+      "step": 20217
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.972240185590071,
+      "learning_rate": 1.1249637970480174e-05,
+      "loss": 0.9394,
+      "step": 20218
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8434361299521238,
+      "learning_rate": 1.1248880907031025e-05,
+      "loss": 0.9658,
+      "step": 20219
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7036922689838192,
+      "learning_rate": 1.1248123836310499e-05,
+      "loss": 1.063,
+      "step": 20220
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5162298862371675,
+      "learning_rate": 1.1247366758322997e-05,
+      "loss": 0.818,
+      "step": 20221
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5517162928081571,
+      "learning_rate": 1.1246609673072932e-05,
+      "loss": 0.848,
+      "step": 20222
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8418687734060095,
+      "learning_rate": 1.124585258056471e-05,
+      "loss": 0.8999,
+      "step": 20223
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7488531330808927,
+      "learning_rate": 1.1245095480802743e-05,
+      "loss": 0.9847,
+      "step": 20224
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.773548147374692,
+      "learning_rate": 1.1244338373791432e-05,
+      "loss": 0.7516,
+      "step": 20225
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8171098884415489,
+      "learning_rate": 1.1243581259535193e-05,
+      "loss": 0.8584,
+      "step": 20226
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5900531252067067,
+      "learning_rate": 1.1242824138038431e-05,
+      "loss": 0.8401,
+      "step": 20227
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7921139705603941,
+      "learning_rate": 1.1242067009305546e-05,
+      "loss": 0.8742,
+      "step": 20228
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8046873028337589,
+      "learning_rate": 1.1241309873340962e-05,
+      "loss": 0.9087,
+      "step": 20229
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4101727874789756,
+      "learning_rate": 1.1240552730149073e-05,
+      "loss": 0.7425,
+      "step": 20230
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5626545347036418,
+      "learning_rate": 1.1239795579734296e-05,
+      "loss": 0.8202,
+      "step": 20231
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6844062692995827,
+      "learning_rate": 1.1239038422101033e-05,
+      "loss": 0.8233,
+      "step": 20232
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6916331185920679,
+      "learning_rate": 1.1238281257253703e-05,
+      "loss": 0.8917,
+      "step": 20233
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6554846101082028,
+      "learning_rate": 1.1237524085196703e-05,
+      "loss": 0.9253,
+      "step": 20234
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.035469596339316,
+      "learning_rate": 1.1236766905934447e-05,
+      "loss": 0.8842,
+      "step": 20235
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8078029534441296,
+      "learning_rate": 1.123600971947134e-05,
+      "loss": 0.9126,
+      "step": 20236
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.72037110603948,
+      "learning_rate": 1.1235252525811797e-05,
+      "loss": 0.7787,
+      "step": 20237
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7503686411973239,
+      "learning_rate": 1.1234495324960221e-05,
+      "loss": 0.6965,
+      "step": 20238
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.621866966941399,
+      "learning_rate": 1.123373811692102e-05,
+      "loss": 0.8601,
+      "step": 20239
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.732745741742998,
+      "learning_rate": 1.123298090169861e-05,
+      "loss": 0.9849,
+      "step": 20240
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.783505526226835,
+      "learning_rate": 1.1232223679297391e-05,
+      "loss": 0.8381,
+      "step": 20241
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.878105247890018,
+      "learning_rate": 1.1231466449721777e-05,
+      "loss": 0.8577,
+      "step": 20242
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.576772493491271,
+      "learning_rate": 1.1230709212976176e-05,
+      "loss": 0.8375,
+      "step": 20243
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7999214885007799,
+      "learning_rate": 1.1229951969064996e-05,
+      "loss": 0.6721,
+      "step": 20244
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.245862241623525,
+      "learning_rate": 1.1229194717992644e-05,
+      "loss": 0.9861,
+      "step": 20245
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9446663535639912,
+      "learning_rate": 1.1228437459763535e-05,
+      "loss": 0.8623,
+      "step": 20246
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4829114531460172,
+      "learning_rate": 1.1227680194382072e-05,
+      "loss": 0.9106,
+      "step": 20247
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0573597848659873,
+      "learning_rate": 1.1226922921852668e-05,
+      "loss": 0.8524,
+      "step": 20248
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5567673508709021,
+      "learning_rate": 1.1226165642179727e-05,
+      "loss": 0.9402,
+      "step": 20249
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7429852709185998,
+      "learning_rate": 1.1225408355367665e-05,
+      "loss": 0.9541,
+      "step": 20250
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5008798167390809,
+      "learning_rate": 1.1224651061420884e-05,
+      "loss": 0.906,
+      "step": 20251
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4711687789619308,
+      "learning_rate": 1.1223893760343798e-05,
+      "loss": 0.8475,
+      "step": 20252
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6061961466673962,
+      "learning_rate": 1.1223136452140815e-05,
+      "loss": 0.8401,
+      "step": 20253
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7814373495611932,
+      "learning_rate": 1.1222379136816347e-05,
+      "loss": 0.726,
+      "step": 20254
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5950844055018178,
+      "learning_rate": 1.12216218143748e-05,
+      "loss": 0.8088,
+      "step": 20255
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6604340954694092,
+      "learning_rate": 1.122086448482058e-05,
+      "loss": 0.864,
+      "step": 20256
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.696087277331619,
+      "learning_rate": 1.1220107148158104e-05,
+      "loss": 0.7889,
+      "step": 20257
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.17401741586214,
+      "learning_rate": 1.1219349804391774e-05,
+      "loss": 0.7915,
+      "step": 20258
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8709370297598131,
+      "learning_rate": 1.121859245352601e-05,
+      "loss": 1.0317,
+      "step": 20259
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5097535771799724,
+      "learning_rate": 1.1217835095565208e-05,
+      "loss": 0.924,
+      "step": 20260
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5362598253533595,
+      "learning_rate": 1.1217077730513792e-05,
+      "loss": 0.9469,
+      "step": 20261
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.78690412222418,
+      "learning_rate": 1.1216320358376158e-05,
+      "loss": 0.7954,
+      "step": 20262
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7536922377911927,
+      "learning_rate": 1.1215562979156725e-05,
+      "loss": 0.7218,
+      "step": 20263
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7673463399030712,
+      "learning_rate": 1.1214805592859898e-05,
+      "loss": 0.7103,
+      "step": 20264
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5704702299512023,
+      "learning_rate": 1.1214048199490089e-05,
+      "loss": 0.9751,
+      "step": 20265
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5027987944928305,
+      "learning_rate": 1.1213290799051706e-05,
+      "loss": 0.9966,
+      "step": 20266
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.680327883755817,
+      "learning_rate": 1.121253339154916e-05,
+      "loss": 0.791,
+      "step": 20267
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7735054896179075,
+      "learning_rate": 1.1211775976986861e-05,
+      "loss": 0.7045,
+      "step": 20268
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5128225449227983,
+      "learning_rate": 1.121101855536922e-05,
+      "loss": 0.8395,
+      "step": 20269
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4183115930361654,
+      "learning_rate": 1.1210261126700643e-05,
+      "loss": 0.8345,
+      "step": 20270
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.635719084932388,
+      "learning_rate": 1.1209503690985544e-05,
+      "loss": 0.9765,
+      "step": 20271
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4079495303871954,
+      "learning_rate": 1.1208746248228332e-05,
+      "loss": 0.7232,
+      "step": 20272
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9088758160032455,
+      "learning_rate": 1.1207988798433416e-05,
+      "loss": 0.8692,
+      "step": 20273
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6116387302687898,
+      "learning_rate": 1.1207231341605208e-05,
+      "loss": 1.0274,
+      "step": 20274
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.497439653506529,
+      "learning_rate": 1.1206473877748115e-05,
+      "loss": 0.8536,
+      "step": 20275
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.662690575736708,
+      "learning_rate": 1.120571640686655e-05,
+      "loss": 0.903,
+      "step": 20276
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4801696838604224,
+      "learning_rate": 1.1204958928964922e-05,
+      "loss": 0.9302,
+      "step": 20277
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5411687060531523,
+      "learning_rate": 1.1204201444047644e-05,
+      "loss": 0.8795,
+      "step": 20278
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.395162982282887,
+      "learning_rate": 1.120344395211912e-05,
+      "loss": 0.887,
+      "step": 20279
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4831489501668371,
+      "learning_rate": 1.1202686453183769e-05,
+      "loss": 0.8688,
+      "step": 20280
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7527092093192065,
+      "learning_rate": 1.120192894724599e-05,
+      "loss": 0.8358,
+      "step": 20281
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8054388201803235,
+      "learning_rate": 1.1201171434310207e-05,
+      "loss": 0.8063,
+      "step": 20282
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5190748897003195,
+      "learning_rate": 1.1200413914380819e-05,
+      "loss": 0.9207,
+      "step": 20283
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8464484887689048,
+      "learning_rate": 1.1199656387462244e-05,
+      "loss": 0.8078,
+      "step": 20284
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4587884065024161,
+      "learning_rate": 1.1198898853558889e-05,
+      "loss": 0.9251,
+      "step": 20285
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7808896473122702,
+      "learning_rate": 1.1198141312675164e-05,
+      "loss": 0.8955,
+      "step": 20286
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2959600905403796,
+      "learning_rate": 1.1197383764815482e-05,
+      "loss": 0.75,
+      "step": 20287
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.536737768130564,
+      "learning_rate": 1.119662620998425e-05,
+      "loss": 0.8522,
+      "step": 20288
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7425053767834704,
+      "learning_rate": 1.1195868648185886e-05,
+      "loss": 0.8787,
+      "step": 20289
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6878122046701607,
+      "learning_rate": 1.1195111079424791e-05,
+      "loss": 0.9482,
+      "step": 20290
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5476448216216727,
+      "learning_rate": 1.1194353503705387e-05,
+      "loss": 0.8779,
+      "step": 20291
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5186229563397784,
+      "learning_rate": 1.1193595921032074e-05,
+      "loss": 0.8361,
+      "step": 20292
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6247399577631596,
+      "learning_rate": 1.119283833140927e-05,
+      "loss": 0.8264,
+      "step": 20293
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6176404948074936,
+      "learning_rate": 1.1192080734841385e-05,
+      "loss": 0.7577,
+      "step": 20294
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7187895443061668,
+      "learning_rate": 1.1191323131332825e-05,
+      "loss": 0.6758,
+      "step": 20295
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7566529931181102,
+      "learning_rate": 1.1190565520888005e-05,
+      "loss": 0.71,
+      "step": 20296
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5503923398745403,
+      "learning_rate": 1.1189807903511336e-05,
+      "loss": 0.8946,
+      "step": 20297
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.641059552094596,
+      "learning_rate": 1.1189050279207229e-05,
+      "loss": 0.996,
+      "step": 20298
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7803464281981148,
+      "learning_rate": 1.1188292647980099e-05,
+      "loss": 0.6711,
+      "step": 20299
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.01359463777956,
+      "learning_rate": 1.1187535009834347e-05,
+      "loss": 0.8529,
+      "step": 20300
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.913175918103345,
+      "learning_rate": 1.1186777364774393e-05,
+      "loss": 1.0265,
+      "step": 20301
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7432506618577475,
+      "learning_rate": 1.1186019712804644e-05,
+      "loss": 0.683,
+      "step": 20302
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7555179194654638,
+      "learning_rate": 1.1185262053929513e-05,
+      "loss": 0.9003,
+      "step": 20303
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.605144415908284,
+      "learning_rate": 1.1184504388153413e-05,
+      "loss": 0.8251,
+      "step": 20304
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5080141263423084,
+      "learning_rate": 1.118374671548075e-05,
+      "loss": 0.9006,
+      "step": 20305
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5504846673062165,
+      "learning_rate": 1.1182989035915943e-05,
+      "loss": 0.8468,
+      "step": 20306
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4980205784355327,
+      "learning_rate": 1.1182231349463394e-05,
+      "loss": 0.8683,
+      "step": 20307
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7859673473086048,
+      "learning_rate": 1.1181473656127526e-05,
+      "loss": 0.737,
+      "step": 20308
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3078293986549667,
+      "learning_rate": 1.118071595591274e-05,
+      "loss": 0.9033,
+      "step": 20309
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.785650305507859,
+      "learning_rate": 1.1179958248823455e-05,
+      "loss": 0.8611,
+      "step": 20310
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.273649513155275,
+      "learning_rate": 1.1179200534864077e-05,
+      "loss": 0.9126,
+      "step": 20311
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6979453381023268,
+      "learning_rate": 1.1178442814039021e-05,
+      "loss": 0.8083,
+      "step": 20312
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.215575890803779,
+      "learning_rate": 1.1177685086352699e-05,
+      "loss": 0.8047,
+      "step": 20313
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5013363578141956,
+      "learning_rate": 1.1176927351809519e-05,
+      "loss": 0.8767,
+      "step": 20314
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7563806906496133,
+      "learning_rate": 1.1176169610413895e-05,
+      "loss": 0.9638,
+      "step": 20315
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.587886507563719,
+      "learning_rate": 1.1175411862170242e-05,
+      "loss": 0.7053,
+      "step": 20316
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.514372145501967,
+      "learning_rate": 1.117465410708297e-05,
+      "loss": 0.9008,
+      "step": 20317
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.707942689895442,
+      "learning_rate": 1.1173896345156486e-05,
+      "loss": 0.8726,
+      "step": 20318
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6204339858062926,
+      "learning_rate": 1.1173138576395208e-05,
+      "loss": 0.862,
+      "step": 20319
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5882463906186335,
+      "learning_rate": 1.1172380800803544e-05,
+      "loss": 0.9403,
+      "step": 20320
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6076039015878836,
+      "learning_rate": 1.117162301838591e-05,
+      "loss": 0.8284,
+      "step": 20321
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6123103593178356,
+      "learning_rate": 1.1170865229146715e-05,
+      "loss": 0.837,
+      "step": 20322
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8893828724920536,
+      "learning_rate": 1.1170107433090371e-05,
+      "loss": 0.806,
+      "step": 20323
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7122936138464704,
+      "learning_rate": 1.1169349630221294e-05,
+      "loss": 0.9125,
+      "step": 20324
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.642818350297602,
+      "learning_rate": 1.116859182054389e-05,
+      "loss": 0.958,
+      "step": 20325
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.563049973968072,
+      "learning_rate": 1.1167834004062576e-05,
+      "loss": 0.9584,
+      "step": 20326
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4542653147794218,
+      "learning_rate": 1.1167076180781764e-05,
+      "loss": 0.831,
+      "step": 20327
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.160234184213588,
+      "learning_rate": 1.116631835070586e-05,
+      "loss": 0.9242,
+      "step": 20328
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.026666660486072,
+      "learning_rate": 1.1165560513839289e-05,
+      "loss": 0.809,
+      "step": 20329
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7596883160834147,
+      "learning_rate": 1.1164802670186448e-05,
+      "loss": 0.6776,
+      "step": 20330
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.651897199490809,
+      "learning_rate": 1.1164044819751763e-05,
+      "loss": 0.881,
+      "step": 20331
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5312929951764784,
+      "learning_rate": 1.1163286962539637e-05,
+      "loss": 0.7885,
+      "step": 20332
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.625522947006211,
+      "learning_rate": 1.1162529098554489e-05,
+      "loss": 0.9839,
+      "step": 20333
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.786614235846975,
+      "learning_rate": 1.1161771227800727e-05,
+      "loss": 0.8564,
+      "step": 20334
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7066129543896087,
+      "learning_rate": 1.1161013350282763e-05,
+      "loss": 0.9246,
+      "step": 20335
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.616947445406381,
+      "learning_rate": 1.1160255466005014e-05,
+      "loss": 0.8967,
+      "step": 20336
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.690778263529688,
+      "learning_rate": 1.1159497574971888e-05,
+      "loss": 0.8344,
+      "step": 20337
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8842431682466336,
+      "learning_rate": 1.1158739677187802e-05,
+      "loss": 1.0086,
+      "step": 20338
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8148607367452431,
+      "learning_rate": 1.1157981772657165e-05,
+      "loss": 0.7019,
+      "step": 20339
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8708417662491361,
+      "learning_rate": 1.1157223861384394e-05,
+      "loss": 0.8925,
+      "step": 20340
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.985706249492532,
+      "learning_rate": 1.1156465943373897e-05,
+      "loss": 0.9746,
+      "step": 20341
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9860416547054458,
+      "learning_rate": 1.1155708018630094e-05,
+      "loss": 0.822,
+      "step": 20342
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5462247298912164,
+      "learning_rate": 1.1154950087157389e-05,
+      "loss": 0.8091,
+      "step": 20343
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.643638780419317,
+      "learning_rate": 1.1154192148960199e-05,
+      "loss": 0.8563,
+      "step": 20344
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6454074634642548,
+      "learning_rate": 1.1153434204042936e-05,
+      "loss": 0.858,
+      "step": 20345
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5832223555254978,
+      "learning_rate": 1.1152676252410015e-05,
+      "loss": 0.776,
+      "step": 20346
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5138150351259332,
+      "learning_rate": 1.115191829406585e-05,
+      "loss": 0.903,
+      "step": 20347
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.80468355385903,
+      "learning_rate": 1.1151160329014849e-05,
+      "loss": 0.7508,
+      "step": 20348
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7718714341408733,
+      "learning_rate": 1.1150402357261429e-05,
+      "loss": 0.7025,
+      "step": 20349
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6052753774062893,
+      "learning_rate": 1.1149644378810003e-05,
+      "loss": 0.8179,
+      "step": 20350
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7703176075049273,
+      "learning_rate": 1.1148886393664984e-05,
+      "loss": 0.7243,
+      "step": 20351
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.655280029760585,
+      "learning_rate": 1.1148128401830783e-05,
+      "loss": 0.9465,
+      "step": 20352
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2997061706305497,
+      "learning_rate": 1.1147370403311815e-05,
+      "loss": 0.9922,
+      "step": 20353
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5411821825763785,
+      "learning_rate": 1.1146612398112492e-05,
+      "loss": 0.9013,
+      "step": 20354
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5806968770160035,
+      "learning_rate": 1.1145854386237231e-05,
+      "loss": 0.8465,
+      "step": 20355
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7109490639424254,
+      "learning_rate": 1.1145096367690445e-05,
+      "loss": 0.8635,
+      "step": 20356
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.784953707367561,
+      "learning_rate": 1.1144338342476544e-05,
+      "loss": 0.7627,
+      "step": 20357
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5862720497533949,
+      "learning_rate": 1.114358031059994e-05,
+      "loss": 0.7904,
+      "step": 20358
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.871758999554641,
+      "learning_rate": 1.1142822272065053e-05,
+      "loss": 0.9036,
+      "step": 20359
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.579369109784929,
+      "learning_rate": 1.1142064226876289e-05,
+      "loss": 0.8881,
+      "step": 20360
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7945023712127937,
+      "learning_rate": 1.1141306175038069e-05,
+      "loss": 0.8119,
+      "step": 20361
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5509762482233886,
+      "learning_rate": 1.1140548116554799e-05,
+      "loss": 0.9833,
+      "step": 20362
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7784414163511395,
+      "learning_rate": 1.1139790051430901e-05,
+      "loss": 0.6932,
+      "step": 20363
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7820311789016654,
+      "learning_rate": 1.1139031979670785e-05,
+      "loss": 0.886,
+      "step": 20364
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7995427423757475,
+      "learning_rate": 1.1138273901278863e-05,
+      "loss": 0.9614,
+      "step": 20365
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.457251375137564,
+      "learning_rate": 1.1137515816259548e-05,
+      "loss": 0.8618,
+      "step": 20366
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7937734375017071,
+      "learning_rate": 1.1136757724617256e-05,
+      "loss": 0.8444,
+      "step": 20367
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5617035255875586,
+      "learning_rate": 1.1135999626356403e-05,
+      "loss": 0.8493,
+      "step": 20368
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.596281778482196,
+      "learning_rate": 1.1135241521481397e-05,
+      "loss": 0.8366,
+      "step": 20369
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6567436834281544,
+      "learning_rate": 1.1134483409996657e-05,
+      "loss": 0.8826,
+      "step": 20370
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5514431341067765,
+      "learning_rate": 1.1133725291906596e-05,
+      "loss": 0.8451,
+      "step": 20371
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7225669104858845,
+      "learning_rate": 1.1132967167215627e-05,
+      "loss": 0.7049,
+      "step": 20372
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4493639981962687,
+      "learning_rate": 1.1132209035928166e-05,
+      "loss": 0.8575,
+      "step": 20373
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7590535873321342,
+      "learning_rate": 1.1131450898048624e-05,
+      "loss": 0.8629,
+      "step": 20374
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3718623948165733,
+      "learning_rate": 1.1130692753581416e-05,
+      "loss": 0.8045,
+      "step": 20375
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1112861586259872,
+      "learning_rate": 1.1129934602530959e-05,
+      "loss": 0.8921,
+      "step": 20376
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7258636085218382,
+      "learning_rate": 1.1129176444901663e-05,
+      "loss": 0.6506,
+      "step": 20377
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5820666024259709,
+      "learning_rate": 1.1128418280697948e-05,
+      "loss": 0.8415,
+      "step": 20378
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6755989801815176,
+      "learning_rate": 1.1127660109924218e-05,
+      "loss": 0.8416,
+      "step": 20379
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.722573687272766,
+      "learning_rate": 1.11269019325849e-05,
+      "loss": 0.7193,
+      "step": 20380
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.675416360083304,
+      "learning_rate": 1.1126143748684397e-05,
+      "loss": 0.8609,
+      "step": 20381
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5191679240490934,
+      "learning_rate": 1.1125385558227136e-05,
+      "loss": 0.8928,
+      "step": 20382
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5522843275494316,
+      "learning_rate": 1.1124627361217514e-05,
+      "loss": 0.804,
+      "step": 20383
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6479030473386973,
+      "learning_rate": 1.1123869157659962e-05,
+      "loss": 0.7973,
+      "step": 20384
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.599007459581555,
+      "learning_rate": 1.1123110947558888e-05,
+      "loss": 0.8106,
+      "step": 20385
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7001488649546688,
+      "learning_rate": 1.1122352730918704e-05,
+      "loss": 0.9548,
+      "step": 20386
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7285866059395347,
+      "learning_rate": 1.112159450774383e-05,
+      "loss": 0.7115,
+      "step": 20387
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6384825220022674,
+      "learning_rate": 1.1120836278038674e-05,
+      "loss": 0.951,
+      "step": 20388
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7889742392862733,
+      "learning_rate": 1.1120078041807658e-05,
+      "loss": 0.8941,
+      "step": 20389
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5142636823076054,
+      "learning_rate": 1.1119319799055188e-05,
+      "loss": 0.6971,
+      "step": 20390
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5573186795568303,
+      "learning_rate": 1.111856154978569e-05,
+      "loss": 0.9607,
+      "step": 20391
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3288841083068155,
+      "learning_rate": 1.1117803294003568e-05,
+      "loss": 0.9018,
+      "step": 20392
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7071986511207522,
+      "learning_rate": 1.1117045031713242e-05,
+      "loss": 0.8588,
+      "step": 20393
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7127588906789337,
+      "learning_rate": 1.1116286762919129e-05,
+      "loss": 0.7882,
+      "step": 20394
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.525841774946471,
+      "learning_rate": 1.1115528487625638e-05,
+      "loss": 0.9467,
+      "step": 20395
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5767027658884676,
+      "learning_rate": 1.1114770205837189e-05,
+      "loss": 0.9856,
+      "step": 20396
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7286498676995418,
+      "learning_rate": 1.1114011917558194e-05,
+      "loss": 0.7022,
+      "step": 20397
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6684194084856705,
+      "learning_rate": 1.1113253622793067e-05,
+      "loss": 0.9286,
+      "step": 20398
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6459643077235429,
+      "learning_rate": 1.1112495321546228e-05,
+      "loss": 0.8977,
+      "step": 20399
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9361117223665851,
+      "learning_rate": 1.1111737013822089e-05,
+      "loss": 0.8078,
+      "step": 20400
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6450884695036287,
+      "learning_rate": 1.1110978699625061e-05,
+      "loss": 0.9416,
+      "step": 20401
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.59300468772116,
+      "learning_rate": 1.1110220378959568e-05,
+      "loss": 0.9398,
+      "step": 20402
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9087607355919232,
+      "learning_rate": 1.1109462051830018e-05,
+      "loss": 0.8697,
+      "step": 20403
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0867195781808237,
+      "learning_rate": 1.110870371824083e-05,
+      "loss": 0.9168,
+      "step": 20404
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6968494893591861,
+      "learning_rate": 1.1107945378196417e-05,
+      "loss": 0.9457,
+      "step": 20405
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6317513211802457,
+      "learning_rate": 1.1107187031701194e-05,
+      "loss": 1.0295,
+      "step": 20406
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7364631880959613,
+      "learning_rate": 1.110642867875958e-05,
+      "loss": 0.7778,
+      "step": 20407
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7844906175451931,
+      "learning_rate": 1.110567031937599e-05,
+      "loss": 0.7392,
+      "step": 20408
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.639506649811617,
+      "learning_rate": 1.1104911953554831e-05,
+      "loss": 0.9056,
+      "step": 20409
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.613289765449944,
+      "learning_rate": 1.1104153581300532e-05,
+      "loss": 0.8688,
+      "step": 20410
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6921817316626118,
+      "learning_rate": 1.1103395202617495e-05,
+      "loss": 0.8656,
+      "step": 20411
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.48728585206841,
+      "learning_rate": 1.1102636817510147e-05,
+      "loss": 0.9135,
+      "step": 20412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5836922780322857,
+      "learning_rate": 1.1101878425982892e-05,
+      "loss": 0.8425,
+      "step": 20413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6897651417314565,
+      "learning_rate": 1.1101120028040158e-05,
+      "loss": 0.9524,
+      "step": 20414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8904863215059518,
+      "learning_rate": 1.1100361623686352e-05,
+      "loss": 0.7858,
+      "step": 20415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7580328588978973,
+      "learning_rate": 1.1099603212925893e-05,
+      "loss": 0.7899,
+      "step": 20416
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9973626670979605,
+      "learning_rate": 1.1098844795763195e-05,
+      "loss": 1.0386,
+      "step": 20417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.775779605855723,
+      "learning_rate": 1.1098086372202674e-05,
+      "loss": 0.8676,
+      "step": 20418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.969668902037601,
+      "learning_rate": 1.109732794224875e-05,
+      "loss": 0.9286,
+      "step": 20419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.590312831282997,
+      "learning_rate": 1.1096569505905832e-05,
+      "loss": 0.7633,
+      "step": 20420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6891960795884495,
+      "learning_rate": 1.109581106317834e-05,
+      "loss": 0.8491,
+      "step": 20421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6223988915406498,
+      "learning_rate": 1.109505261407069e-05,
+      "loss": 0.8576,
+      "step": 20422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6391185762539517,
+      "learning_rate": 1.1094294158587295e-05,
+      "loss": 0.8719,
+      "step": 20423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7062433158957064,
+      "learning_rate": 1.1093535696732574e-05,
+      "loss": 1.0351,
+      "step": 20424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.449526322353309,
+      "learning_rate": 1.1092777228510943e-05,
+      "loss": 0.9623,
+      "step": 20425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.660465485298454,
+      "learning_rate": 1.1092018753926817e-05,
+      "loss": 0.8713,
+      "step": 20426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8633417438587638,
+      "learning_rate": 1.109126027298461e-05,
+      "loss": 0.8258,
+      "step": 20427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6131815710328719,
+      "learning_rate": 1.109050178568874e-05,
+      "loss": 0.8535,
+      "step": 20428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8706212886986733,
+      "learning_rate": 1.1089743292043629e-05,
+      "loss": 0.7421,
+      "step": 20429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6907765419087306,
+      "learning_rate": 1.1088984792053681e-05,
+      "loss": 0.8421,
+      "step": 20430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7443061375308156,
+      "learning_rate": 1.1088226285723323e-05,
+      "loss": 0.7763,
+      "step": 20431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7733610597281737,
+      "learning_rate": 1.1087467773056962e-05,
+      "loss": 0.9263,
+      "step": 20432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7772551228935762,
+      "learning_rate": 1.1086709254059025e-05,
+      "loss": 0.752,
+      "step": 20433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.696852465191334,
+      "learning_rate": 1.1085950728733919e-05,
+      "loss": 0.8872,
+      "step": 20434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4847846607009416,
+      "learning_rate": 1.1085192197086063e-05,
+      "loss": 0.9752,
+      "step": 20435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6406432084260965,
+      "learning_rate": 1.108443365911988e-05,
+      "loss": 0.8531,
+      "step": 20436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5998818166751088,
+      "learning_rate": 1.1083675114839775e-05,
+      "loss": 0.8944,
+      "step": 20437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4697835268818331,
+      "learning_rate": 1.1082916564250175e-05,
+      "loss": 0.9725,
+      "step": 20438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7165304925343257,
+      "learning_rate": 1.1082158007355487e-05,
+      "loss": 0.8633,
+      "step": 20439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.669219347206408,
+      "learning_rate": 1.108139944416014e-05,
+      "loss": 0.7142,
+      "step": 20440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8008256090194272,
+      "learning_rate": 1.1080640874668535e-05,
+      "loss": 0.7526,
+      "step": 20441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6525481196626617,
+      "learning_rate": 1.1079882298885102e-05,
+      "loss": 0.8699,
+      "step": 20442
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8942292388722415,
+      "learning_rate": 1.107912371681425e-05,
+      "loss": 0.8914,
+      "step": 20443
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.440415102898689,
+      "learning_rate": 1.1078365128460398e-05,
+      "loss": 0.8054,
+      "step": 20444
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7603680369628802,
+      "learning_rate": 1.1077606533827965e-05,
+      "loss": 0.9694,
+      "step": 20445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6948325626928418,
+      "learning_rate": 1.1076847932921364e-05,
+      "loss": 0.8869,
+      "step": 20446
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.799896078407742,
+      "learning_rate": 1.1076089325745013e-05,
+      "loss": 0.9738,
+      "step": 20447
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9233410435900657,
+      "learning_rate": 1.1075330712303329e-05,
+      "loss": 0.8544,
+      "step": 20448
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2447011544613216,
+      "learning_rate": 1.107457209260073e-05,
+      "loss": 0.7365,
+      "step": 20449
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6585214006340883,
+      "learning_rate": 1.1073813466641633e-05,
+      "loss": 0.8425,
+      "step": 20450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.624780685849894,
+      "learning_rate": 1.1073054834430452e-05,
+      "loss": 0.9426,
+      "step": 20451
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5628106016591148,
+      "learning_rate": 1.1072296195971606e-05,
+      "loss": 0.8956,
+      "step": 20452
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.931437996985502,
+      "learning_rate": 1.1071537551269513e-05,
+      "loss": 0.8346,
+      "step": 20453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.863696201400093,
+      "learning_rate": 1.107077890032859e-05,
+      "loss": 0.8794,
+      "step": 20454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5266205534053061,
+      "learning_rate": 1.1070020243153253e-05,
+      "loss": 0.9479,
+      "step": 20455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7284043774999103,
+      "learning_rate": 1.1069261579747919e-05,
+      "loss": 0.8578,
+      "step": 20456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8011321180482467,
+      "learning_rate": 1.1068502910117006e-05,
+      "loss": 0.7104,
+      "step": 20457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2125537180571357,
+      "learning_rate": 1.1067744234264928e-05,
+      "loss": 0.766,
+      "step": 20458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.206978045408816,
+      "learning_rate": 1.106698555219611e-05,
+      "loss": 0.9046,
+      "step": 20459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5516121447022286,
+      "learning_rate": 1.106622686391496e-05,
+      "loss": 0.9877,
+      "step": 20460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4233125302064338,
+      "learning_rate": 1.1065468169425902e-05,
+      "loss": 0.7124,
+      "step": 20461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7792194875795575,
+      "learning_rate": 1.1064709468733348e-05,
+      "loss": 0.9201,
+      "step": 20462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8192632467918002,
+      "learning_rate": 1.1063950761841722e-05,
+      "loss": 0.9387,
+      "step": 20463
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7168432557888516,
+      "learning_rate": 1.1063192048755436e-05,
+      "loss": 0.7915,
+      "step": 20464
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.635122141074231,
+      "learning_rate": 1.106243332947891e-05,
+      "loss": 0.818,
+      "step": 20465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7486517564578155,
+      "learning_rate": 1.106167460401656e-05,
+      "loss": 0.9378,
+      "step": 20466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.518423275401708,
+      "learning_rate": 1.1060915872372804e-05,
+      "loss": 1.0007,
+      "step": 20467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5428179451991846,
+      "learning_rate": 1.1060157134552061e-05,
+      "loss": 0.7693,
+      "step": 20468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7723000831567083,
+      "learning_rate": 1.105939839055875e-05,
+      "loss": 0.7519,
+      "step": 20469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6403524494563904,
+      "learning_rate": 1.1058639640397282e-05,
+      "loss": 0.8656,
+      "step": 20470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6189123199509898,
+      "learning_rate": 1.1057880884072078e-05,
+      "loss": 0.9164,
+      "step": 20471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5441054691688845,
+      "learning_rate": 1.1057122121587561e-05,
+      "loss": 0.8437,
+      "step": 20472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.787127984352621,
+      "learning_rate": 1.1056363352948141e-05,
+      "loss": 0.9294,
+      "step": 20473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7655160078860784,
+      "learning_rate": 1.105560457815824e-05,
+      "loss": 0.7471,
+      "step": 20474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7581453187269471,
+      "learning_rate": 1.1054845797222276e-05,
+      "loss": 0.7935,
+      "step": 20475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6204413784455598,
+      "learning_rate": 1.1054087010144662e-05,
+      "loss": 0.835,
+      "step": 20476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.888653985493399,
+      "learning_rate": 1.1053328216929823e-05,
+      "loss": 0.7962,
+      "step": 20477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9626069631095822,
+      "learning_rate": 1.1052569417582174e-05,
+      "loss": 0.8743,
+      "step": 20478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6644197595766175,
+      "learning_rate": 1.1051810612106134e-05,
+      "loss": 0.9874,
+      "step": 20479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6338321154044229,
+      "learning_rate": 1.1051051800506118e-05,
+      "loss": 0.7802,
+      "step": 20480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8043671715951182,
+      "learning_rate": 1.1050292982786543e-05,
+      "loss": 0.8568,
+      "step": 20481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.181966046145459,
+      "learning_rate": 1.1049534158951834e-05,
+      "loss": 0.9657,
+      "step": 20482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.909000209219228,
+      "learning_rate": 1.10487753290064e-05,
+      "loss": 0.951,
+      "step": 20483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.077966103291599,
+      "learning_rate": 1.1048016492954668e-05,
+      "loss": 0.9254,
+      "step": 20484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4665434987118862,
+      "learning_rate": 1.104725765080105e-05,
+      "loss": 0.8899,
+      "step": 20485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.64423959365334,
+      "learning_rate": 1.1046498802549968e-05,
+      "loss": 0.9374,
+      "step": 20486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5028257759903725,
+      "learning_rate": 1.104573994820584e-05,
+      "loss": 0.9899,
+      "step": 20487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6045260506527281,
+      "learning_rate": 1.104498108777308e-05,
+      "loss": 0.8792,
+      "step": 20488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6796783822666035,
+      "learning_rate": 1.104422222125611e-05,
+      "loss": 0.9521,
+      "step": 20489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.679755618878615,
+      "learning_rate": 1.1043463348659346e-05,
+      "loss": 0.838,
+      "step": 20490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3288503111718644,
+      "learning_rate": 1.1042704469987213e-05,
+      "loss": 0.8784,
+      "step": 20491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7272753165787036,
+      "learning_rate": 1.1041945585244118e-05,
+      "loss": 0.868,
+      "step": 20492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0583578342339703,
+      "learning_rate": 1.1041186694434492e-05,
+      "loss": 0.8697,
+      "step": 20493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6677816398272374,
+      "learning_rate": 1.1040427797562744e-05,
+      "loss": 0.8044,
+      "step": 20494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6584388939810304,
+      "learning_rate": 1.1039668894633298e-05,
+      "loss": 0.847,
+      "step": 20495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6325856479462695,
+      "learning_rate": 1.1038909985650568e-05,
+      "loss": 0.9173,
+      "step": 20496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3610767628769644,
+      "learning_rate": 1.1038151070618976e-05,
+      "loss": 0.8742,
+      "step": 20497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7322173996451192,
+      "learning_rate": 1.103739214954294e-05,
+      "loss": 0.7146,
+      "step": 20498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5233285542307953,
+      "learning_rate": 1.103663322242688e-05,
+      "loss": 0.7455,
+      "step": 20499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6794386775401675,
+      "learning_rate": 1.1035874289275212e-05,
+      "loss": 0.8636,
+      "step": 20500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6515826951623076,
+      "learning_rate": 1.1035115350092356e-05,
+      "loss": 0.7842,
+      "step": 20501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7578420371359293,
+      "learning_rate": 1.1034356404882728e-05,
+      "loss": 0.6879,
+      "step": 20502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.39851766330652,
+      "learning_rate": 1.1033597453650752e-05,
+      "loss": 0.9443,
+      "step": 20503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5042744898834637,
+      "learning_rate": 1.1032838496400846e-05,
+      "loss": 0.8716,
+      "step": 20504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6477339781139848,
+      "learning_rate": 1.1032079533137426e-05,
+      "loss": 0.8671,
+      "step": 20505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9496391190158973,
+      "learning_rate": 1.1031320563864912e-05,
+      "loss": 0.803,
+      "step": 20506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.026253808265068,
+      "learning_rate": 1.1030561588587722e-05,
+      "loss": 0.8718,
+      "step": 20507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8381707875358388,
+      "learning_rate": 1.102980260731028e-05,
+      "loss": 0.8679,
+      "step": 20508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8247800824527842,
+      "learning_rate": 1.1029043620036996e-05,
+      "loss": 0.7146,
+      "step": 20509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5678864875995284,
+      "learning_rate": 1.1028284626772299e-05,
+      "loss": 0.8546,
+      "step": 20510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.64930477783937,
+      "learning_rate": 1.1027525627520599e-05,
+      "loss": 0.8386,
+      "step": 20511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6648302151601337,
+      "learning_rate": 1.1026766622286322e-05,
+      "loss": 0.811,
+      "step": 20512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8466736465861635,
+      "learning_rate": 1.1026007611073883e-05,
+      "loss": 0.9559,
+      "step": 20513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5848819958109415,
+      "learning_rate": 1.1025248593887707e-05,
+      "loss": 0.9546,
+      "step": 20514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8467269720851633,
+      "learning_rate": 1.1024489570732207e-05,
+      "loss": 0.9093,
+      "step": 20515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.306889803706258,
+      "learning_rate": 1.10237305416118e-05,
+      "loss": 0.8195,
+      "step": 20516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6700889388808033,
+      "learning_rate": 1.1022971506530915e-05,
+      "loss": 0.8018,
+      "step": 20517
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5782048234832888,
+      "learning_rate": 1.1022212465493965e-05,
+      "loss": 0.8143,
+      "step": 20518
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7162767573506401,
+      "learning_rate": 1.102145341850537e-05,
+      "loss": 0.8509,
+      "step": 20519
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7956179305581552,
+      "learning_rate": 1.1020694365569548e-05,
+      "loss": 0.9307,
+      "step": 20520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.691324187452232,
+      "learning_rate": 1.1019935306690921e-05,
+      "loss": 0.6485,
+      "step": 20521
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8735794993129644,
+      "learning_rate": 1.101917624187391e-05,
+      "loss": 0.919,
+      "step": 20522
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8073607207727809,
+      "learning_rate": 1.101841717112293e-05,
+      "loss": 0.9099,
+      "step": 20523
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7511844137826684,
+      "learning_rate": 1.1017658094442403e-05,
+      "loss": 0.8249,
+      "step": 20524
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6326842536809458,
+      "learning_rate": 1.1016899011836749e-05,
+      "loss": 0.9534,
+      "step": 20525
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5134468427784982,
+      "learning_rate": 1.1016139923310388e-05,
+      "loss": 0.9114,
+      "step": 20526
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7464295410482051,
+      "learning_rate": 1.1015380828867736e-05,
+      "loss": 0.8481,
+      "step": 20527
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0439059966361817,
+      "learning_rate": 1.1014621728513218e-05,
+      "loss": 0.973,
+      "step": 20528
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5324218809300771,
+      "learning_rate": 1.101386262225125e-05,
+      "loss": 0.9532,
+      "step": 20529
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7525693713054031,
+      "learning_rate": 1.1013103510086252e-05,
+      "loss": 0.8186,
+      "step": 20530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6935898360073438,
+      "learning_rate": 1.1012344392022647e-05,
+      "loss": 0.8889,
+      "step": 20531
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8259674904738452,
+      "learning_rate": 1.1011585268064849e-05,
+      "loss": 0.752,
+      "step": 20532
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.997267350401591,
+      "learning_rate": 1.1010826138217284e-05,
+      "loss": 0.9426,
+      "step": 20533
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5013152146603468,
+      "learning_rate": 1.1010067002484368e-05,
+      "loss": 0.8024,
+      "step": 20534
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4599851970150293,
+      "learning_rate": 1.1009307860870522e-05,
+      "loss": 0.8694,
+      "step": 20535
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.949278695852461,
+      "learning_rate": 1.1008548713380167e-05,
+      "loss": 0.7643,
+      "step": 20536
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5211403653176467,
+      "learning_rate": 1.1007789560017722e-05,
+      "loss": 0.9229,
+      "step": 20537
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.395299432993757,
+      "learning_rate": 1.1007030400787608e-05,
+      "loss": 0.9378,
+      "step": 20538
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.992081533651281,
+      "learning_rate": 1.1006271235694242e-05,
+      "loss": 0.7737,
+      "step": 20539
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.121841542861454,
+      "learning_rate": 1.1005512064742048e-05,
+      "loss": 0.9744,
+      "step": 20540
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6498162809535937,
+      "learning_rate": 1.1004752887935442e-05,
+      "loss": 0.9113,
+      "step": 20541
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.535688084513882,
+      "learning_rate": 1.100399370527885e-05,
+      "loss": 0.8324,
+      "step": 20542
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.12219317268618,
+      "learning_rate": 1.1003234516776685e-05,
+      "loss": 1.0021,
+      "step": 20543
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7965495417502826,
+      "learning_rate": 1.1002475322433374e-05,
+      "loss": 0.9671,
+      "step": 20544
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4696880762076707,
+      "learning_rate": 1.1001716122253335e-05,
+      "loss": 0.9145,
+      "step": 20545
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8141724271008155,
+      "learning_rate": 1.1000956916240985e-05,
+      "loss": 0.802,
+      "step": 20546
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7079310816729434,
+      "learning_rate": 1.1000197704400748e-05,
+      "loss": 0.8579,
+      "step": 20547
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4505981178101033,
+      "learning_rate": 1.0999438486737043e-05,
+      "loss": 0.8066,
+      "step": 20548
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6517774846706794,
+      "learning_rate": 1.099867926325429e-05,
+      "loss": 0.8358,
+      "step": 20549
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6334002996213088,
+      "learning_rate": 1.0997920033956912e-05,
+      "loss": 0.934,
+      "step": 20550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.542023268102395,
+      "learning_rate": 1.0997160798849327e-05,
+      "loss": 0.7518,
+      "step": 20551
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.659659112096927,
+      "learning_rate": 1.0996401557935957e-05,
+      "loss": 0.7649,
+      "step": 20552
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.080894912724547,
+      "learning_rate": 1.099564231122122e-05,
+      "loss": 0.9459,
+      "step": 20553
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6182953740643284,
+      "learning_rate": 1.0994883058709538e-05,
+      "loss": 0.7856,
+      "step": 20554
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6154676708761382,
+      "learning_rate": 1.0994123800405332e-05,
+      "loss": 0.9729,
+      "step": 20555
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2258935969151445,
+      "learning_rate": 1.0993364536313022e-05,
+      "loss": 0.9043,
+      "step": 20556
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4893700189023857,
+      "learning_rate": 1.0992605266437032e-05,
+      "loss": 0.8475,
+      "step": 20557
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.769627321872246,
+      "learning_rate": 1.0991845990781777e-05,
+      "loss": 0.7737,
+      "step": 20558
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7704396972175387,
+      "learning_rate": 1.0991086709351686e-05,
+      "loss": 0.8228,
+      "step": 20559
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.932728345505028,
+      "learning_rate": 1.0990327422151168e-05,
+      "loss": 0.741,
+      "step": 20560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0208556379186424,
+      "learning_rate": 1.0989568129184654e-05,
+      "loss": 0.7366,
+      "step": 20561
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3702251397703322,
+      "learning_rate": 1.0988808830456557e-05,
+      "loss": 0.8118,
+      "step": 20562
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4257829882091875,
+      "learning_rate": 1.0988049525971307e-05,
+      "loss": 0.8102,
+      "step": 20563
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6700156635861776,
+      "learning_rate": 1.0987290215733315e-05,
+      "loss": 0.8048,
+      "step": 20564
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.567178317051542,
+      "learning_rate": 1.098653089974701e-05,
+      "loss": 0.9999,
+      "step": 20565
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6650567399202456,
+      "learning_rate": 1.0985771578016808e-05,
+      "loss": 0.8194,
+      "step": 20566
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4229528010488648,
+      "learning_rate": 1.0985012250547133e-05,
+      "loss": 0.8154,
+      "step": 20567
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.097265370913677,
+      "learning_rate": 1.0984252917342406e-05,
+      "loss": 0.7917,
+      "step": 20568
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.174379686401279,
+      "learning_rate": 1.0983493578407044e-05,
+      "loss": 0.9848,
+      "step": 20569
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4433139043683305,
+      "learning_rate": 1.0982734233745473e-05,
+      "loss": 0.8628,
+      "step": 20570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5998240573794338,
+      "learning_rate": 1.098197488336211e-05,
+      "loss": 0.8332,
+      "step": 20571
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.670598921909606,
+      "learning_rate": 1.0981215527261383e-05,
+      "loss": 0.7829,
+      "step": 20572
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6857708499194004,
+      "learning_rate": 1.0980456165447704e-05,
+      "loss": 0.8281,
+      "step": 20573
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7635049507366846,
+      "learning_rate": 1.09796967979255e-05,
+      "loss": 0.7233,
+      "step": 20574
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7918628642971888,
+      "learning_rate": 1.0978937424699191e-05,
+      "loss": 0.7087,
+      "step": 20575
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4495281815300751,
+      "learning_rate": 1.0978178045773197e-05,
+      "loss": 0.8605,
+      "step": 20576
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5281278169080401,
+      "learning_rate": 1.0977418661151943e-05,
+      "loss": 0.7979,
+      "step": 20577
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6626396953733698,
+      "learning_rate": 1.097665927083985e-05,
+      "loss": 0.8447,
+      "step": 20578
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.909061531431448,
+      "learning_rate": 1.0975899874841334e-05,
+      "loss": 0.8574,
+      "step": 20579
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7163340553766362,
+      "learning_rate": 1.0975140473160824e-05,
+      "loss": 0.964,
+      "step": 20580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6929621088165487,
+      "learning_rate": 1.0974381065802731e-05,
+      "loss": 0.9549,
+      "step": 20581
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.542961235458147,
+      "learning_rate": 1.0973621652771487e-05,
+      "loss": 0.7884,
+      "step": 20582
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.725162283387043,
+      "learning_rate": 1.0972862234071508e-05,
+      "loss": 0.9202,
+      "step": 20583
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8559826304294367,
+      "learning_rate": 1.0972102809707219e-05,
+      "loss": 0.9351,
+      "step": 20584
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6920206580620154,
+      "learning_rate": 1.0971343379683037e-05,
+      "loss": 0.8509,
+      "step": 20585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6040285750528664,
+      "learning_rate": 1.0970583944003388e-05,
+      "loss": 0.8819,
+      "step": 20586
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4440420256249986,
+      "learning_rate": 1.0969824502672694e-05,
+      "loss": 0.8176,
+      "step": 20587
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6029545590165766,
+      "learning_rate": 1.0969065055695371e-05,
+      "loss": 0.8432,
+      "step": 20588
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9308674001982757,
+      "learning_rate": 1.0968305603075849e-05,
+      "loss": 0.8688,
+      "step": 20589
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.995971171296378,
+      "learning_rate": 1.096754614481854e-05,
+      "loss": 0.868,
+      "step": 20590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7115440485770355,
+      "learning_rate": 1.0966786680927875e-05,
+      "loss": 0.8359,
+      "step": 20591
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.826516171676189,
+      "learning_rate": 1.0966027211408268e-05,
+      "loss": 0.871,
+      "step": 20592
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5412284025560021,
+      "learning_rate": 1.096526773626415e-05,
+      "loss": 0.9599,
+      "step": 20593
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7645001167685566,
+      "learning_rate": 1.0964508255499933e-05,
+      "loss": 0.7956,
+      "step": 20594
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7207395411582516,
+      "learning_rate": 1.0963748769120046e-05,
+      "loss": 0.792,
+      "step": 20595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5431554953296043,
+      "learning_rate": 1.0962989277128909e-05,
+      "loss": 0.7704,
+      "step": 20596
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7885356157519392,
+      "learning_rate": 1.0962229779530941e-05,
+      "loss": 0.7264,
+      "step": 20597
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.23571187520689,
+      "learning_rate": 1.0961470276330568e-05,
+      "loss": 0.9203,
+      "step": 20598
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6923015955941496,
+      "learning_rate": 1.0960710767532209e-05,
+      "loss": 0.8771,
+      "step": 20599
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.55235652785362,
+      "learning_rate": 1.095995125314029e-05,
+      "loss": 0.995,
+      "step": 20600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6136956699726879,
+      "learning_rate": 1.095919173315923e-05,
+      "loss": 0.9649,
+      "step": 20601
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4273320550885118,
+      "learning_rate": 1.0958432207593453e-05,
+      "loss": 0.8234,
+      "step": 20602
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.668642772358669,
+      "learning_rate": 1.095767267644738e-05,
+      "loss": 0.8587,
+      "step": 20603
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8179953813555343,
+      "learning_rate": 1.095691313972543e-05,
+      "loss": 1.0219,
+      "step": 20604
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6129795131191271,
+      "learning_rate": 1.0956153597432033e-05,
+      "loss": 0.7424,
+      "step": 20605
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.558072276809395,
+      "learning_rate": 1.0955394049571606e-05,
+      "loss": 0.7229,
+      "step": 20606
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.782197136710956,
+      "learning_rate": 1.0954634496148573e-05,
+      "loss": 0.6875,
+      "step": 20607
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4424367720356515,
+      "learning_rate": 1.0953874937167355e-05,
+      "loss": 0.8666,
+      "step": 20608
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8236502875852503,
+      "learning_rate": 1.0953115372632375e-05,
+      "loss": 0.908,
+      "step": 20609
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7657596072720299,
+      "learning_rate": 1.0952355802548058e-05,
+      "loss": 0.7099,
+      "step": 20610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.71331460075315,
+      "learning_rate": 1.095159622691882e-05,
+      "loss": 0.8833,
+      "step": 20611
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8013245226026717,
+      "learning_rate": 1.095083664574909e-05,
+      "loss": 0.9351,
+      "step": 20612
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.606051219774423,
+      "learning_rate": 1.0950077059043285e-05,
+      "loss": 0.9291,
+      "step": 20613
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5937822041153373,
+      "learning_rate": 1.0949317466805834e-05,
+      "loss": 0.8983,
+      "step": 20614
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.627695630765971,
+      "learning_rate": 1.0948557869041151e-05,
+      "loss": 0.8584,
+      "step": 20615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8007451638583927,
+      "learning_rate": 1.094779826575367e-05,
+      "loss": 0.8919,
+      "step": 20616
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5961630908023219,
+      "learning_rate": 1.0947038656947806e-05,
+      "loss": 0.8245,
+      "step": 20617
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6324912679182648,
+      "learning_rate": 1.0946279042627981e-05,
+      "loss": 0.9652,
+      "step": 20618
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9355022556575594,
+      "learning_rate": 1.0945519422798622e-05,
+      "loss": 0.8068,
+      "step": 20619
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3953747280649933,
+      "learning_rate": 1.0944759797464146e-05,
+      "loss": 0.8282,
+      "step": 20620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7952817213619142,
+      "learning_rate": 1.0944000166628986e-05,
+      "loss": 0.7433,
+      "step": 20621
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9273120457423758,
+      "learning_rate": 1.094324053029755e-05,
+      "loss": 0.9551,
+      "step": 20622
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.810995837873978,
+      "learning_rate": 1.0942480888474276e-05,
+      "loss": 0.8692,
+      "step": 20623
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5255795412893445,
+      "learning_rate": 1.0941721241163577e-05,
+      "loss": 0.9141,
+      "step": 20624
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5843083558397755,
+      "learning_rate": 1.0940961588369879e-05,
+      "loss": 0.8524,
+      "step": 20625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2566142710481056,
+      "learning_rate": 1.0940201930097603e-05,
+      "loss": 1.0029,
+      "step": 20626
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8591624110901466,
+      "learning_rate": 1.0939442266351174e-05,
+      "loss": 0.8947,
+      "step": 20627
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7999483842100588,
+      "learning_rate": 1.0938682597135017e-05,
+      "loss": 0.7016,
+      "step": 20628
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8753791717816906,
+      "learning_rate": 1.0937922922453553e-05,
+      "loss": 0.864,
+      "step": 20629
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8925465112535507,
+      "learning_rate": 1.0937163242311201e-05,
+      "loss": 0.8622,
+      "step": 20630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7724852472975974,
+      "learning_rate": 1.0936403556712392e-05,
+      "loss": 0.7806,
+      "step": 20631
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8062396777028322,
+      "learning_rate": 1.0935643865661543e-05,
+      "loss": 0.8695,
+      "step": 20632
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5994974075041408,
+      "learning_rate": 1.0934884169163078e-05,
+      "loss": 0.8802,
+      "step": 20633
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5968480574424397,
+      "learning_rate": 1.0934124467221424e-05,
+      "loss": 0.9166,
+      "step": 20634
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7832474097389497,
+      "learning_rate": 1.0933364759841001e-05,
+      "loss": 0.7543,
+      "step": 20635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7197288966205745,
+      "learning_rate": 1.0932605047026232e-05,
+      "loss": 0.6911,
+      "step": 20636
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5112702803151592,
+      "learning_rate": 1.0931845328781541e-05,
+      "loss": 1.016,
+      "step": 20637
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7703984437713205,
+      "learning_rate": 1.0931085605111354e-05,
+      "loss": 0.8867,
+      "step": 20638
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5070365701810038,
+      "learning_rate": 1.0930325876020086e-05,
+      "loss": 0.8406,
+      "step": 20639
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5160782302806788,
+      "learning_rate": 1.0929566141512173e-05,
+      "loss": 0.8949,
+      "step": 20640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.128974584452834,
+      "learning_rate": 1.0928806401592026e-05,
+      "loss": 0.9008,
+      "step": 20641
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.887685716636584,
+      "learning_rate": 1.092804665626408e-05,
+      "loss": 0.9908,
+      "step": 20642
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7614522942625153,
+      "learning_rate": 1.0927286905532749e-05,
+      "loss": 0.8685,
+      "step": 20643
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3443583909779613,
+      "learning_rate": 1.0926527149402461e-05,
+      "loss": 0.8775,
+      "step": 20644
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8421347855564274,
+      "learning_rate": 1.0925767387877638e-05,
+      "loss": 0.9597,
+      "step": 20645
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7598176710590951,
+      "learning_rate": 1.0925007620962703e-05,
+      "loss": 0.7944,
+      "step": 20646
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5364678745403724,
+      "learning_rate": 1.0924247848662084e-05,
+      "loss": 0.8345,
+      "step": 20647
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6561517656097986,
+      "learning_rate": 1.0923488070980197e-05,
+      "loss": 0.9411,
+      "step": 20648
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1240377672578035,
+      "learning_rate": 1.0922728287921474e-05,
+      "loss": 0.9489,
+      "step": 20649
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7597649923515934,
+      "learning_rate": 1.0921968499490332e-05,
+      "loss": 0.6645,
+      "step": 20650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.075165967341111,
+      "learning_rate": 1.09212087056912e-05,
+      "loss": 0.8546,
+      "step": 20651
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1073170643099477,
+      "learning_rate": 1.0920448906528498e-05,
+      "loss": 0.93,
+      "step": 20652
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.514917498367453,
+      "learning_rate": 1.091968910200665e-05,
+      "loss": 1.001,
+      "step": 20653
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.542681875431613,
+      "learning_rate": 1.0918929292130082e-05,
+      "loss": 0.841,
+      "step": 20654
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.605241991246326,
+      "learning_rate": 1.0918169476903217e-05,
+      "loss": 0.8802,
+      "step": 20655
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5775317638578346,
+      "learning_rate": 1.0917409656330477e-05,
+      "loss": 0.8766,
+      "step": 20656
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.72221097897191,
+      "learning_rate": 1.091664983041629e-05,
+      "loss": 0.9046,
+      "step": 20657
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2808076523753495,
+      "learning_rate": 1.0915889999165077e-05,
+      "loss": 0.9066,
+      "step": 20658
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7430206780807227,
+      "learning_rate": 1.091513016258126e-05,
+      "loss": 0.8275,
+      "step": 20659
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9908497956663012,
+      "learning_rate": 1.0914370320669267e-05,
+      "loss": 0.9705,
+      "step": 20660
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4886848527709688,
+      "learning_rate": 1.0913610473433525e-05,
+      "loss": 0.836,
+      "step": 20661
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0692631891601687,
+      "learning_rate": 1.0912850620878445e-05,
+      "loss": 0.9432,
+      "step": 20662
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3917209324130067,
+      "learning_rate": 1.0912090763008468e-05,
+      "loss": 0.8374,
+      "step": 20663
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4634576541577897,
+      "learning_rate": 1.0911330899828003e-05,
+      "loss": 0.8129,
+      "step": 20664
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.569710574070737,
+      "learning_rate": 1.0910571031341485e-05,
+      "loss": 0.8993,
+      "step": 20665
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8569029475723051,
+      "learning_rate": 1.0909811157553331e-05,
+      "loss": 0.7085,
+      "step": 20666
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.670211819959687,
+      "learning_rate": 1.090905127846797e-05,
+      "loss": 0.9066,
+      "step": 20667
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5220961819113359,
+      "learning_rate": 1.0908291394089828e-05,
+      "loss": 0.86,
+      "step": 20668
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6840782939879653,
+      "learning_rate": 1.0907531504423321e-05,
+      "loss": 0.9329,
+      "step": 20669
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.678572363602756,
+      "learning_rate": 1.0906771609472882e-05,
+      "loss": 0.8281,
+      "step": 20670
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7415476287687064,
+      "learning_rate": 1.0906011709242927e-05,
+      "loss": 0.9101,
+      "step": 20671
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.53883570244161,
+      "learning_rate": 1.0905251803737891e-05,
+      "loss": 0.7652,
+      "step": 20672
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.889752465976827,
+      "learning_rate": 1.0904491892962186e-05,
+      "loss": 0.9285,
+      "step": 20673
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.748791947023184,
+      "learning_rate": 1.0903731976920249e-05,
+      "loss": 0.6877,
+      "step": 20674
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7341288902584876,
+      "learning_rate": 1.0902972055616494e-05,
+      "loss": 0.686,
+      "step": 20675
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.078219295097642,
+      "learning_rate": 1.0902212129055353e-05,
+      "loss": 0.837,
+      "step": 20676
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.62572191399015,
+      "learning_rate": 1.0901452197241245e-05,
+      "loss": 0.8643,
+      "step": 20677
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8116528592577816,
+      "learning_rate": 1.0900692260178596e-05,
+      "loss": 0.7452,
+      "step": 20678
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0857565843000625,
+      "learning_rate": 1.0899932317871832e-05,
+      "loss": 0.8497,
+      "step": 20679
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6707396190544626,
+      "learning_rate": 1.0899172370325379e-05,
+      "loss": 0.8668,
+      "step": 20680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5407907934011036,
+      "learning_rate": 1.0898412417543658e-05,
+      "loss": 0.955,
+      "step": 20681
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.780199269025818,
+      "learning_rate": 1.0897652459531098e-05,
+      "loss": 0.8878,
+      "step": 20682
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3153365032558937,
+      "learning_rate": 1.089689249629212e-05,
+      "loss": 0.9139,
+      "step": 20683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6544060075400797,
+      "learning_rate": 1.0896132527831149e-05,
+      "loss": 0.8545,
+      "step": 20684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.620234020427021,
+      "learning_rate": 1.0895372554152611e-05,
+      "loss": 0.9704,
+      "step": 20685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4970844660089615,
+      "learning_rate": 1.089461257526093e-05,
+      "loss": 0.944,
+      "step": 20686
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6352454533498026,
+      "learning_rate": 1.0893852591160533e-05,
+      "loss": 0.9361,
+      "step": 20687
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5885210559592824,
+      "learning_rate": 1.0893092601855841e-05,
+      "loss": 0.8754,
+      "step": 20688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.034770268961468,
+      "learning_rate": 1.0892332607351283e-05,
+      "loss": 0.8339,
+      "step": 20689
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6775208361319833,
+      "learning_rate": 1.0891572607651281e-05,
+      "loss": 0.8054,
+      "step": 20690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6014107969771398,
+      "learning_rate": 1.0890812602760263e-05,
+      "loss": 0.8695,
+      "step": 20691
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3112583168640413,
+      "learning_rate": 1.089005259268265e-05,
+      "loss": 0.7038,
+      "step": 20692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.546898596297959,
+      "learning_rate": 1.0889292577422869e-05,
+      "loss": 0.8924,
+      "step": 20693
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4889632929502514,
+      "learning_rate": 1.0888532556985346e-05,
+      "loss": 1.0288,
+      "step": 20694
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8252558597100277,
+      "learning_rate": 1.0887772531374506e-05,
+      "loss": 0.7339,
+      "step": 20695
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5809886840841214,
+      "learning_rate": 1.0887012500594771e-05,
+      "loss": 1.0181,
+      "step": 20696
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5361423816491988,
+      "learning_rate": 1.0886252464650569e-05,
+      "loss": 0.7992,
+      "step": 20697
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3632844564321536,
+      "learning_rate": 1.0885492423546324e-05,
+      "loss": 0.8454,
+      "step": 20698
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.504290657584246,
+      "learning_rate": 1.0884732377286464e-05,
+      "loss": 0.9515,
+      "step": 20699
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7660108160970405,
+      "learning_rate": 1.0883972325875411e-05,
+      "loss": 1.007,
+      "step": 20700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.743491078922215,
+      "learning_rate": 1.088321226931759e-05,
+      "loss": 0.9144,
+      "step": 20701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1396550271356283,
+      "learning_rate": 1.0882452207617428e-05,
+      "loss": 0.7887,
+      "step": 20702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5140840862225988,
+      "learning_rate": 1.0881692140779352e-05,
+      "loss": 0.9622,
+      "step": 20703
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6502857097489165,
+      "learning_rate": 1.0880932068807783e-05,
+      "loss": 0.8604,
+      "step": 20704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8167035207265351,
+      "learning_rate": 1.0880171991707149e-05,
+      "loss": 0.7727,
+      "step": 20705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8050693851706373,
+      "learning_rate": 1.0879411909481874e-05,
+      "loss": 0.7058,
+      "step": 20706
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.687905704744995,
+      "learning_rate": 1.0878651822136386e-05,
+      "loss": 0.9859,
+      "step": 20707
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6499447162343044,
+      "learning_rate": 1.087789172967511e-05,
+      "loss": 0.8933,
+      "step": 20708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5363052673246487,
+      "learning_rate": 1.0877131632102468e-05,
+      "loss": 0.851,
+      "step": 20709
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8483924283730273,
+      "learning_rate": 1.087637152942289e-05,
+      "loss": 0.9134,
+      "step": 20710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.568912376561982,
+      "learning_rate": 1.0875611421640797e-05,
+      "loss": 0.8485,
+      "step": 20711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.47244269729918,
+      "learning_rate": 1.0874851308760622e-05,
+      "loss": 0.7396,
+      "step": 20712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4340923774993493,
+      "learning_rate": 1.087409119078678e-05,
+      "loss": 0.9001,
+      "step": 20713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6891157399443228,
+      "learning_rate": 1.0873331067723704e-05,
+      "loss": 0.9507,
+      "step": 20714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0277456567258825,
+      "learning_rate": 1.0872570939575818e-05,
+      "loss": 0.9885,
+      "step": 20715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5856956048395998,
+      "learning_rate": 1.087181080634755e-05,
+      "loss": 0.7674,
+      "step": 20716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6367897779021208,
+      "learning_rate": 1.087105066804332e-05,
+      "loss": 0.9279,
+      "step": 20717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.760025707748924,
+      "learning_rate": 1.087029052466756e-05,
+      "loss": 0.8185,
+      "step": 20718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9392854045484198,
+      "learning_rate": 1.0869530376224693e-05,
+      "loss": 0.9206,
+      "step": 20719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6235120234060387,
+      "learning_rate": 1.0868770222719143e-05,
+      "loss": 0.8368,
+      "step": 20720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5178379517760636,
+      "learning_rate": 1.0868010064155338e-05,
+      "loss": 0.8751,
+      "step": 20721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6547501135537828,
+      "learning_rate": 1.0867249900537703e-05,
+      "loss": 0.9172,
+      "step": 20722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7169179647702573,
+      "learning_rate": 1.0866489731870668e-05,
+      "loss": 0.8421,
+      "step": 20723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9199373062500553,
+      "learning_rate": 1.0865729558158652e-05,
+      "loss": 0.9593,
+      "step": 20724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.588671274920196,
+      "learning_rate": 1.0864969379406088e-05,
+      "loss": 0.9954,
+      "step": 20725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9154861534132857,
+      "learning_rate": 1.0864209195617395e-05,
+      "loss": 0.7844,
+      "step": 20726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5786289488606062,
+      "learning_rate": 1.0863449006797005e-05,
+      "loss": 0.7729,
+      "step": 20727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6046402003373823,
+      "learning_rate": 1.0862688812949338e-05,
+      "loss": 0.8155,
+      "step": 20728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8062896772019357,
+      "learning_rate": 1.0861928614078827e-05,
+      "loss": 0.9396,
+      "step": 20729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7014205482404472,
+      "learning_rate": 1.0861168410189894e-05,
+      "loss": 0.9542,
+      "step": 20730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7967496145226785,
+      "learning_rate": 1.0860408201286964e-05,
+      "loss": 0.7825,
+      "step": 20731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.456986575083432,
+      "learning_rate": 1.0859647987374467e-05,
+      "loss": 0.8029,
+      "step": 20732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4676731672447096,
+      "learning_rate": 1.0858887768456824e-05,
+      "loss": 0.9079,
+      "step": 20733
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5520720302342859,
+      "learning_rate": 1.0858127544538468e-05,
+      "loss": 0.7916,
+      "step": 20734
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8030305323580375,
+      "learning_rate": 1.085736731562382e-05,
+      "loss": 0.7511,
+      "step": 20735
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.037287869175786,
+      "learning_rate": 1.085660708171731e-05,
+      "loss": 0.9161,
+      "step": 20736
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7517924517194234,
+      "learning_rate": 1.085584684282336e-05,
+      "loss": 0.7286,
+      "step": 20737
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.888574848905277,
+      "learning_rate": 1.0855086598946401e-05,
+      "loss": 0.9421,
+      "step": 20738
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.424963729103932,
+      "learning_rate": 1.0854326350090855e-05,
+      "loss": 0.9206,
+      "step": 20739
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5902934700727196,
+      "learning_rate": 1.0853566096261155e-05,
+      "loss": 0.8085,
+      "step": 20740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5952353951969658,
+      "learning_rate": 1.0852805837461718e-05,
+      "loss": 0.8757,
+      "step": 20741
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7985486859465565,
+      "learning_rate": 1.0852045573696979e-05,
+      "loss": 0.7577,
+      "step": 20742
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6423938721730016,
+      "learning_rate": 1.0851285304971354e-05,
+      "loss": 0.7933,
+      "step": 20743
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6776100334206787,
+      "learning_rate": 1.0850525031289286e-05,
+      "loss": 0.969,
+      "step": 20744
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8725193868496492,
+      "learning_rate": 1.0849764752655185e-05,
+      "loss": 0.6745,
+      "step": 20745
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.537154310584036,
+      "learning_rate": 1.0849004469073487e-05,
+      "loss": 0.921,
+      "step": 20746
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.775451847963637,
+      "learning_rate": 1.0848244180548618e-05,
+      "loss": 0.6966,
+      "step": 20747
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6423982125243135,
+      "learning_rate": 1.0847483887085e-05,
+      "loss": 0.797,
+      "step": 20748
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8808792164026857,
+      "learning_rate": 1.0846723588687065e-05,
+      "loss": 0.8148,
+      "step": 20749
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5862120183179198,
+      "learning_rate": 1.0845963285359233e-05,
+      "loss": 0.8643,
+      "step": 20750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8968371285809893,
+      "learning_rate": 1.084520297710594e-05,
+      "loss": 0.9148,
+      "step": 20751
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8776764405713835,
+      "learning_rate": 1.0844442663931604e-05,
+      "loss": 0.9417,
+      "step": 20752
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6169482926845387,
+      "learning_rate": 1.0843682345840655e-05,
+      "loss": 0.8043,
+      "step": 20753
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.720452161655374,
+      "learning_rate": 1.0842922022837524e-05,
+      "loss": 0.8463,
+      "step": 20754
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7634747550826894,
+      "learning_rate": 1.0842161694926632e-05,
+      "loss": 0.7117,
+      "step": 20755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5964064137110294,
+      "learning_rate": 1.0841401362112408e-05,
+      "loss": 0.8248,
+      "step": 20756
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.903015454747804,
+      "learning_rate": 1.0840641024399278e-05,
+      "loss": 0.88,
+      "step": 20757
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7688592334410718,
+      "learning_rate": 1.083988068179167e-05,
+      "loss": 0.7771,
+      "step": 20758
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.820180083241534,
+      "learning_rate": 1.0839120334294013e-05,
+      "loss": 0.8756,
+      "step": 20759
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5672698798931988,
+      "learning_rate": 1.083835998191073e-05,
+      "loss": 0.8329,
+      "step": 20760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7731174362307913,
+      "learning_rate": 1.0837599624646251e-05,
+      "loss": 0.7329,
+      "step": 20761
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7481798834319016,
+      "learning_rate": 1.0836839262504998e-05,
+      "loss": 0.8662,
+      "step": 20762
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2570334106678547,
+      "learning_rate": 1.0836078895491408e-05,
+      "loss": 0.9434,
+      "step": 20763
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9573928385109916,
+      "learning_rate": 1.0835318523609897e-05,
+      "loss": 0.8462,
+      "step": 20764
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7678151427137213,
+      "learning_rate": 1.08345581468649e-05,
+      "loss": 0.6874,
+      "step": 20765
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6926898029196484,
+      "learning_rate": 1.0833797765260838e-05,
+      "loss": 0.8663,
+      "step": 20766
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6817516701519593,
+      "learning_rate": 1.0833037378802145e-05,
+      "loss": 0.9888,
+      "step": 20767
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6859630739819358,
+      "learning_rate": 1.0832276987493243e-05,
+      "loss": 0.9345,
+      "step": 20768
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6328128834018416,
+      "learning_rate": 1.083151659133856e-05,
+      "loss": 0.8503,
+      "step": 20769
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4546025541898848,
+      "learning_rate": 1.0830756190342528e-05,
+      "loss": 0.8321,
+      "step": 20770
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.614949731938825,
+      "learning_rate": 1.0829995784509565e-05,
+      "loss": 0.9429,
+      "step": 20771
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4654151183650659,
+      "learning_rate": 1.0829235373844109e-05,
+      "loss": 0.8148,
+      "step": 20772
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5877752768967472,
+      "learning_rate": 1.0828474958350577e-05,
+      "loss": 0.8558,
+      "step": 20773
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5581192002099324,
+      "learning_rate": 1.0827714538033407e-05,
+      "loss": 0.9828,
+      "step": 20774
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7313549115438682,
+      "learning_rate": 1.0826954112897014e-05,
+      "loss": 0.9307,
+      "step": 20775
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7553248900087994,
+      "learning_rate": 1.0826193682945839e-05,
+      "loss": 0.7917,
+      "step": 20776
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7694537648639169,
+      "learning_rate": 1.0825433248184301e-05,
+      "loss": 0.7512,
+      "step": 20777
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7530001905403494,
+      "learning_rate": 1.0824672808616827e-05,
+      "loss": 0.952,
+      "step": 20778
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1347467408780303,
+      "learning_rate": 1.0823912364247848e-05,
+      "loss": 0.9006,
+      "step": 20779
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.624940557643498,
+      "learning_rate": 1.0823151915081791e-05,
+      "loss": 0.8794,
+      "step": 20780
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7485256374482373,
+      "learning_rate": 1.0822391461123083e-05,
+      "loss": 0.7128,
+      "step": 20781
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6116750941991147,
+      "learning_rate": 1.0821631002376152e-05,
+      "loss": 0.9092,
+      "step": 20782
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.451242309715871,
+      "learning_rate": 1.0820870538845423e-05,
+      "loss": 0.8586,
+      "step": 20783
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.727939884764859,
+      "learning_rate": 1.0820110070535326e-05,
+      "loss": 0.7306,
+      "step": 20784
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.327110741834155,
+      "learning_rate": 1.081934959745029e-05,
+      "loss": 0.7602,
+      "step": 20785
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7996605092794153,
+      "learning_rate": 1.081858911959474e-05,
+      "loss": 0.8953,
+      "step": 20786
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7953377504327483,
+      "learning_rate": 1.0817828636973106e-05,
+      "loss": 0.8096,
+      "step": 20787
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.439078553936952,
+      "learning_rate": 1.0817068149589816e-05,
+      "loss": 0.8273,
+      "step": 20788
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6111481453342509,
+      "learning_rate": 1.0816307657449294e-05,
+      "loss": 0.9078,
+      "step": 20789
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7844490412090683,
+      "learning_rate": 1.0815547160555973e-05,
+      "loss": 0.8958,
+      "step": 20790
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6722701343122435,
+      "learning_rate": 1.0814786658914279e-05,
+      "loss": 0.8343,
+      "step": 20791
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.600215413597511,
+      "learning_rate": 1.0814026152528634e-05,
+      "loss": 0.8652,
+      "step": 20792
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6371205551590697,
+      "learning_rate": 1.0813265641403476e-05,
+      "loss": 0.7397,
+      "step": 20793
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.656969791789697,
+      "learning_rate": 1.0812505125543223e-05,
+      "loss": 1.0597,
+      "step": 20794
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.931380865288011,
+      "learning_rate": 1.0811744604952312e-05,
+      "loss": 0.8343,
+      "step": 20795
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5338666473292712,
+      "learning_rate": 1.0810984079635164e-05,
+      "loss": 0.8632,
+      "step": 20796
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8119257484592541,
+      "learning_rate": 1.0810223549596215e-05,
+      "loss": 0.7511,
+      "step": 20797
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5306350987600492,
+      "learning_rate": 1.0809463014839884e-05,
+      "loss": 0.8748,
+      "step": 20798
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9734629109505772,
+      "learning_rate": 1.0808702475370603e-05,
+      "loss": 0.8813,
+      "step": 20799
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.59360060620927,
+      "learning_rate": 1.0807941931192802e-05,
+      "loss": 0.839,
+      "step": 20800
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9450952974205094,
+      "learning_rate": 1.0807181382310904e-05,
+      "loss": 0.7923,
+      "step": 20801
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8813231541416475,
+      "learning_rate": 1.0806420828729345e-05,
+      "loss": 0.8279,
+      "step": 20802
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.490209696949414,
+      "learning_rate": 1.0805660270452542e-05,
+      "loss": 0.8883,
+      "step": 20803
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7738298622651865,
+      "learning_rate": 1.0804899707484938e-05,
+      "loss": 0.8103,
+      "step": 20804
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8199807335927163,
+      "learning_rate": 1.0804139139830947e-05,
+      "loss": 0.9219,
+      "step": 20805
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7353701053266085,
+      "learning_rate": 1.0803378567495006e-05,
+      "loss": 0.7,
+      "step": 20806
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5720343492800268,
+      "learning_rate": 1.0802617990481542e-05,
+      "loss": 0.9417,
+      "step": 20807
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4166432624100846,
+      "learning_rate": 1.0801857408794978e-05,
+      "loss": 0.8789,
+      "step": 20808
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7669595202152764,
+      "learning_rate": 1.080109682243975e-05,
+      "loss": 0.8042,
+      "step": 20809
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7601777207881861,
+      "learning_rate": 1.0800336231420278e-05,
+      "loss": 0.8309,
+      "step": 20810
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7858460360780979,
+      "learning_rate": 1.0799575635741e-05,
+      "loss": 0.8256,
+      "step": 20811
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.053340172508877,
+      "learning_rate": 1.079881503540634e-05,
+      "loss": 0.9553,
+      "step": 20812
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4180818595670437,
+      "learning_rate": 1.079805443042072e-05,
+      "loss": 0.824,
+      "step": 20813
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8922823582305858,
+      "learning_rate": 1.0797293820788578e-05,
+      "loss": 0.8202,
+      "step": 20814
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.54311951516737,
+      "learning_rate": 1.079653320651434e-05,
+      "loss": 0.885,
+      "step": 20815
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.481624552056782,
+      "learning_rate": 1.0795772587602432e-05,
+      "loss": 0.8691,
+      "step": 20816
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7489189564427754,
+      "learning_rate": 1.0795011964057285e-05,
+      "loss": 0.7453,
+      "step": 20817
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6072928803031765,
+      "learning_rate": 1.0794251335883326e-05,
+      "loss": 0.9101,
+      "step": 20818
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.890451947573665,
+      "learning_rate": 1.0793490703084985e-05,
+      "loss": 0.8885,
+      "step": 20819
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9102006280679056,
+      "learning_rate": 1.0792730065666688e-05,
+      "loss": 0.9416,
+      "step": 20820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6744853574774383,
+      "learning_rate": 1.0791969423632869e-05,
+      "loss": 0.7531,
+      "step": 20821
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.812570031969033,
+      "learning_rate": 1.0791208776987949e-05,
+      "loss": 0.8641,
+      "step": 20822
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.549034848749549,
+      "learning_rate": 1.0790448125736366e-05,
+      "loss": 0.7711,
+      "step": 20823
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.868877841132654,
+      "learning_rate": 1.0789687469882538e-05,
+      "loss": 0.911,
+      "step": 20824
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.784844004922259,
+      "learning_rate": 1.0788926809430904e-05,
+      "loss": 0.8453,
+      "step": 20825
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5938786963887281,
+      "learning_rate": 1.0788166144385888e-05,
+      "loss": 0.7642,
+      "step": 20826
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7955842284853822,
+      "learning_rate": 1.0787405474751917e-05,
+      "loss": 0.8136,
+      "step": 20827
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7466663358112573,
+      "learning_rate": 1.0786644800533423e-05,
+      "loss": 0.8734,
+      "step": 20828
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7794636274448489,
+      "learning_rate": 1.0785884121734836e-05,
+      "loss": 0.7005,
+      "step": 20829
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7655332961071855,
+      "learning_rate": 1.0785123438360578e-05,
+      "loss": 0.7765,
+      "step": 20830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.657316494073292,
+      "learning_rate": 1.0784362750415088e-05,
+      "loss": 0.9889,
+      "step": 20831
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4769341521285462,
+      "learning_rate": 1.0783602057902786e-05,
+      "loss": 0.94,
+      "step": 20832
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6914181427022694,
+      "learning_rate": 1.0782841360828107e-05,
+      "loss": 0.9107,
+      "step": 20833
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0958354557432086,
+      "learning_rate": 1.0782080659195475e-05,
+      "loss": 0.7723,
+      "step": 20834
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6283055767123737,
+      "learning_rate": 1.0781319953009324e-05,
+      "loss": 0.9768,
+      "step": 20835
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7377557594780502,
+      "learning_rate": 1.078055924227408e-05,
+      "loss": 0.905,
+      "step": 20836
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5952653946571964,
+      "learning_rate": 1.0779798526994173e-05,
+      "loss": 0.8866,
+      "step": 20837
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0941337385141767,
+      "learning_rate": 1.0779037807174032e-05,
+      "loss": 0.8724,
+      "step": 20838
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8551054970225132,
+      "learning_rate": 1.077827708281809e-05,
+      "loss": 0.715,
+      "step": 20839
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6394796066391286,
+      "learning_rate": 1.0777516353930767e-05,
+      "loss": 0.759,
+      "step": 20840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9398138427452551,
+      "learning_rate": 1.0776755620516501e-05,
+      "loss": 0.8382,
+      "step": 20841
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8818109651224053,
+      "learning_rate": 1.0775994882579718e-05,
+      "loss": 0.8423,
+      "step": 20842
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.780615200556561,
+      "learning_rate": 1.0775234140124845e-05,
+      "loss": 0.8595,
+      "step": 20843
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7083912146338092,
+      "learning_rate": 1.0774473393156315e-05,
+      "loss": 0.8157,
+      "step": 20844
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9786235012588138,
+      "learning_rate": 1.0773712641678553e-05,
+      "loss": 0.8554,
+      "step": 20845
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.605657658273022,
+      "learning_rate": 1.0772951885695994e-05,
+      "loss": 0.9452,
+      "step": 20846
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7167710285409744,
+      "learning_rate": 1.0772191125213063e-05,
+      "loss": 0.7352,
+      "step": 20847
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5532727475530081,
+      "learning_rate": 1.0771430360234194e-05,
+      "loss": 0.8935,
+      "step": 20848
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7057721446553153,
+      "learning_rate": 1.0770669590763809e-05,
+      "loss": 0.9704,
+      "step": 20849
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7128625562665531,
+      "learning_rate": 1.0769908816806343e-05,
+      "loss": 0.99,
+      "step": 20850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9912115624518076,
+      "learning_rate": 1.0769148038366223e-05,
+      "loss": 0.9055,
+      "step": 20851
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7545952186450747,
+      "learning_rate": 1.076838725544788e-05,
+      "loss": 0.7037,
+      "step": 20852
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0816752306523654,
+      "learning_rate": 1.0767626468055747e-05,
+      "loss": 0.9702,
+      "step": 20853
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7234845451243412,
+      "learning_rate": 1.0766865676194245e-05,
+      "loss": 0.7229,
+      "step": 20854
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.153413853146304,
+      "learning_rate": 1.0766104879867811e-05,
+      "loss": 0.6296,
+      "step": 20855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.790656188199441,
+      "learning_rate": 1.076534407908087e-05,
+      "loss": 0.9483,
+      "step": 20856
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7887784491174201,
+      "learning_rate": 1.0764583273837855e-05,
+      "loss": 0.959,
+      "step": 20857
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.968124474632756,
+      "learning_rate": 1.0763822464143191e-05,
+      "loss": 0.9178,
+      "step": 20858
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.566484572799923,
+      "learning_rate": 1.0763061650001314e-05,
+      "loss": 0.8257,
+      "step": 20859
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7337568608967024,
+      "learning_rate": 1.076230083141665e-05,
+      "loss": 0.8605,
+      "step": 20860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.063126408732457,
+      "learning_rate": 1.076154000839363e-05,
+      "loss": 0.795,
+      "step": 20861
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5916227564335084,
+      "learning_rate": 1.076077918093668e-05,
+      "loss": 0.8622,
+      "step": 20862
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7754807244291304,
+      "learning_rate": 1.0760018349050236e-05,
+      "loss": 0.7966,
+      "step": 20863
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5892931713912455,
+      "learning_rate": 1.0759257512738723e-05,
+      "loss": 0.8837,
+      "step": 20864
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4763116339212872,
+      "learning_rate": 1.0758496672006571e-05,
+      "loss": 0.7675,
+      "step": 20865
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7100267661249722,
+      "learning_rate": 1.0757735826858213e-05,
+      "loss": 0.8626,
+      "step": 20866
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6606420976752194,
+      "learning_rate": 1.0756974977298078e-05,
+      "loss": 0.8028,
+      "step": 20867
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7618331825898529,
+      "learning_rate": 1.0756214123330593e-05,
+      "loss": 0.7536,
+      "step": 20868
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0045334629018488,
+      "learning_rate": 1.0755453264960192e-05,
+      "loss": 1.0582,
+      "step": 20869
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2056864720457408,
+      "learning_rate": 1.0754692402191303e-05,
+      "loss": 0.7269,
+      "step": 20870
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4418077368066813,
+      "learning_rate": 1.0753931535028353e-05,
+      "loss": 0.9825,
+      "step": 20871
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6038414994851005,
+      "learning_rate": 1.0753170663475778e-05,
+      "loss": 0.8451,
+      "step": 20872
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.005077382871523,
+      "learning_rate": 1.0752409787538e-05,
+      "loss": 0.9052,
+      "step": 20873
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5659460439773172,
+      "learning_rate": 1.0751648907219461e-05,
+      "loss": 0.8962,
+      "step": 20874
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.693354679426197,
+      "learning_rate": 1.075088802252458e-05,
+      "loss": 0.8065,
+      "step": 20875
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5303050654821222,
+      "learning_rate": 1.0750127133457792e-05,
+      "loss": 0.8564,
+      "step": 20876
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.566266663953381,
+      "learning_rate": 1.0749366240023526e-05,
+      "loss": 0.8315,
+      "step": 20877
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5154429137321632,
+      "learning_rate": 1.0748605342226213e-05,
+      "loss": 0.8985,
+      "step": 20878
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.494424080140688,
+      "learning_rate": 1.0747844440070284e-05,
+      "loss": 0.9709,
+      "step": 20879
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9853717312108519,
+      "learning_rate": 1.0747083533560166e-05,
+      "loss": 0.9437,
+      "step": 20880
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.745598359193529,
+      "learning_rate": 1.0746322622700289e-05,
+      "loss": 0.8977,
+      "step": 20881
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6025724165997417,
+      "learning_rate": 1.0745561707495088e-05,
+      "loss": 0.7833,
+      "step": 20882
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8823538433774053,
+      "learning_rate": 1.074480078794899e-05,
+      "loss": 0.7731,
+      "step": 20883
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.305722389867889,
+      "learning_rate": 1.0744039864066427e-05,
+      "loss": 0.8355,
+      "step": 20884
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.68961293267122,
+      "learning_rate": 1.0743278935851828e-05,
+      "loss": 0.9472,
+      "step": 20885
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7561969371992554,
+      "learning_rate": 1.0742518003309624e-05,
+      "loss": 0.6942,
+      "step": 20886
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4507410422398097,
+      "learning_rate": 1.0741757066444244e-05,
+      "loss": 0.874,
+      "step": 20887
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.265641385630125,
+      "learning_rate": 1.0740996125260122e-05,
+      "loss": 0.8351,
+      "step": 20888
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6155482982306588,
+      "learning_rate": 1.0740235179761682e-05,
+      "loss": 0.9437,
+      "step": 20889
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5916322240663991,
+      "learning_rate": 1.0739474229953362e-05,
+      "loss": 0.8322,
+      "step": 20890
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.514322012440531,
+      "learning_rate": 1.073871327583959e-05,
+      "loss": 0.8101,
+      "step": 20891
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.210515402984382,
+      "learning_rate": 1.0737952317424789e-05,
+      "loss": 0.8627,
+      "step": 20892
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5190273941163404,
+      "learning_rate": 1.0737191354713403e-05,
+      "loss": 0.8081,
+      "step": 20893
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5223737602289538,
+      "learning_rate": 1.0736430387709851e-05,
+      "loss": 0.8448,
+      "step": 20894
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6701624061987108,
+      "learning_rate": 1.0735669416418571e-05,
+      "loss": 0.8276,
+      "step": 20895
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.103057842063493,
+      "learning_rate": 1.0734908440843987e-05,
+      "loss": 0.9087,
+      "step": 20896
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.688558327078628,
+      "learning_rate": 1.0734147460990539e-05,
+      "loss": 0.8276,
+      "step": 20897
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.405616609353082,
+      "learning_rate": 1.0733386476862649e-05,
+      "loss": 0.8542,
+      "step": 20898
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5996154235136284,
+      "learning_rate": 1.0732625488464749e-05,
+      "loss": 0.7607,
+      "step": 20899
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.920918421542563,
+      "learning_rate": 1.0731864495801272e-05,
+      "loss": 0.9903,
+      "step": 20900
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5364763369665544,
+      "learning_rate": 1.0731103498876647e-05,
+      "loss": 0.7931,
+      "step": 20901
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.83064258762873,
+      "learning_rate": 1.0730342497695312e-05,
+      "loss": 0.9183,
+      "step": 20902
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6722683587597162,
+      "learning_rate": 1.0729581492261684e-05,
+      "loss": 0.8578,
+      "step": 20903
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5394101468074701,
+      "learning_rate": 1.072882048258021e-05,
+      "loss": 0.8207,
+      "step": 20904
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7153864702960817,
+      "learning_rate": 1.0728059468655303e-05,
+      "loss": 0.7931,
+      "step": 20905
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7767197068518024,
+      "learning_rate": 1.0727298450491413e-05,
+      "loss": 0.9074,
+      "step": 20906
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9995677238290004,
+      "learning_rate": 1.0726537428092956e-05,
+      "loss": 0.8608,
+      "step": 20907
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9146576910905315,
+      "learning_rate": 1.0725776401464366e-05,
+      "loss": 0.7817,
+      "step": 20908
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8216176933418415,
+      "learning_rate": 1.0725015370610078e-05,
+      "loss": 0.8953,
+      "step": 20909
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5229439992737253,
+      "learning_rate": 1.0724254335534524e-05,
+      "loss": 0.8151,
+      "step": 20910
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.640242867088388,
+      "learning_rate": 1.0723493296242129e-05,
+      "loss": 0.9243,
+      "step": 20911
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.611482281716169,
+      "learning_rate": 1.0722732252737327e-05,
+      "loss": 0.9068,
+      "step": 20912
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.17880917832693,
+      "learning_rate": 1.072197120502455e-05,
+      "loss": 0.9773,
+      "step": 20913
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4582962142457965,
+      "learning_rate": 1.0721210153108226e-05,
+      "loss": 0.8848,
+      "step": 20914
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1983575268229907,
+      "learning_rate": 1.072044909699279e-05,
+      "loss": 0.881,
+      "step": 20915
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6517652961073257,
+      "learning_rate": 1.0719688036682672e-05,
+      "loss": 0.8954,
+      "step": 20916
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7636987817703658,
+      "learning_rate": 1.07189269721823e-05,
+      "loss": 0.8624,
+      "step": 20917
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.047408828415218,
+      "learning_rate": 1.0718165903496109e-05,
+      "loss": 0.8351,
+      "step": 20918
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.599885532194717,
+      "learning_rate": 1.0717404830628529e-05,
+      "loss": 0.8517,
+      "step": 20919
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.686370344011247,
+      "learning_rate": 1.071664375358399e-05,
+      "loss": 0.9075,
+      "step": 20920
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7230768526892437,
+      "learning_rate": 1.0715882672366926e-05,
+      "loss": 0.9874,
+      "step": 20921
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0933178805878385,
+      "learning_rate": 1.0715121586981763e-05,
+      "loss": 1.0304,
+      "step": 20922
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4270100015343132,
+      "learning_rate": 1.071436049743294e-05,
+      "loss": 0.8502,
+      "step": 20923
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5998965352833765,
+      "learning_rate": 1.0713599403724879e-05,
+      "loss": 0.8013,
+      "step": 20924
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5890580721262364,
+      "learning_rate": 1.0712838305862023e-05,
+      "loss": 0.8481,
+      "step": 20925
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.770168887150953,
+      "learning_rate": 1.071207720384879e-05,
+      "loss": 0.8518,
+      "step": 20926
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.577987300199751,
+      "learning_rate": 1.0711316097689623e-05,
+      "loss": 0.8473,
+      "step": 20927
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7261003749374767,
+      "learning_rate": 1.0710554987388946e-05,
+      "loss": 0.774,
+      "step": 20928
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.847676791739997,
+      "learning_rate": 1.0709793872951193e-05,
+      "loss": 0.7799,
+      "step": 20929
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5864666073930092,
+      "learning_rate": 1.0709032754380797e-05,
+      "loss": 0.905,
+      "step": 20930
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6307817669779519,
+      "learning_rate": 1.0708271631682186e-05,
+      "loss": 0.763,
+      "step": 20931
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8120248257174036,
+      "learning_rate": 1.0707510504859795e-05,
+      "loss": 0.9633,
+      "step": 20932
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7006891961324426,
+      "learning_rate": 1.070674937391805e-05,
+      "loss": 0.8081,
+      "step": 20933
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7045991272258572,
+      "learning_rate": 1.0705988238861391e-05,
+      "loss": 0.7076,
+      "step": 20934
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7062572541932528,
+      "learning_rate": 1.0705227099694242e-05,
+      "loss": 0.8431,
+      "step": 20935
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6799385081470843,
+      "learning_rate": 1.070446595642104e-05,
+      "loss": 0.9404,
+      "step": 20936
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.706722935374691,
+      "learning_rate": 1.0703704809046211e-05,
+      "loss": 0.8763,
+      "step": 20937
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.57983004598697,
+      "learning_rate": 1.0702943657574193e-05,
+      "loss": 0.935,
+      "step": 20938
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.677457233809996,
+      "learning_rate": 1.0702182502009412e-05,
+      "loss": 0.7406,
+      "step": 20939
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.818553993155371,
+      "learning_rate": 1.0701421342356304e-05,
+      "loss": 0.7238,
+      "step": 20940
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7555974220247058,
+      "learning_rate": 1.0700660178619297e-05,
+      "loss": 0.6798,
+      "step": 20941
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7982964155371441,
+      "learning_rate": 1.069989901080283e-05,
+      "loss": 0.887,
+      "step": 20942
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4711021916529452,
+      "learning_rate": 1.0699137838911323e-05,
+      "loss": 0.8017,
+      "step": 20943
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5652409470498199,
+      "learning_rate": 1.0698376662949217e-05,
+      "loss": 0.8338,
+      "step": 20944
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7651594756237032,
+      "learning_rate": 1.0697615482920937e-05,
+      "loss": 0.7678,
+      "step": 20945
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6769508073426649,
+      "learning_rate": 1.0696854298830925e-05,
+      "loss": 0.9118,
+      "step": 20946
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.086508192605957,
+      "learning_rate": 1.0696093110683603e-05,
+      "loss": 0.9801,
+      "step": 20947
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.932272571215264,
+      "learning_rate": 1.069533191848341e-05,
+      "loss": 0.955,
+      "step": 20948
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6539747848477822,
+      "learning_rate": 1.0694570722234769e-05,
+      "loss": 0.8633,
+      "step": 20949
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4812270134019865,
+      "learning_rate": 1.0693809521942118e-05,
+      "loss": 0.8516,
+      "step": 20950
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.019685747774531,
+      "learning_rate": 1.0693048317609893e-05,
+      "loss": 0.8955,
+      "step": 20951
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7746820475778688,
+      "learning_rate": 1.0692287109242516e-05,
+      "loss": 0.8215,
+      "step": 20952
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0950876513283334,
+      "learning_rate": 1.069152589684443e-05,
+      "loss": 0.8002,
+      "step": 20953
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4162609005068456,
+      "learning_rate": 1.0690764680420056e-05,
+      "loss": 0.865,
+      "step": 20954
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9472347543918094,
+      "learning_rate": 1.0690003459973836e-05,
+      "loss": 0.9356,
+      "step": 20955
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4808029008680474,
+      "learning_rate": 1.0689242235510195e-05,
+      "loss": 0.8742,
+      "step": 20956
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1690335556230442,
+      "learning_rate": 1.0688481007033566e-05,
+      "loss": 0.9858,
+      "step": 20957
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4436245185039338,
+      "learning_rate": 1.0687719774548384e-05,
+      "loss": 0.7754,
+      "step": 20958
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.434960528022441,
+      "learning_rate": 1.068695853805908e-05,
+      "loss": 0.9702,
+      "step": 20959
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.360945421914972,
+      "learning_rate": 1.0686197297570087e-05,
+      "loss": 0.9085,
+      "step": 20960
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.086156345641532,
+      "learning_rate": 1.0685436053085833e-05,
+      "loss": 0.7793,
+      "step": 20961
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4953382256727938,
+      "learning_rate": 1.0684674804610756e-05,
+      "loss": 0.8967,
+      "step": 20962
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.970782488719627,
+      "learning_rate": 1.0683913552149286e-05,
+      "loss": 0.8667,
+      "step": 20963
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9432875319840246,
+      "learning_rate": 1.0683152295705853e-05,
+      "loss": 0.871,
+      "step": 20964
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6552746996762238,
+      "learning_rate": 1.0682391035284892e-05,
+      "loss": 0.7774,
+      "step": 20965
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4677740865937885,
+      "learning_rate": 1.0681629770890832e-05,
+      "loss": 0.9726,
+      "step": 20966
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2356414348140277,
+      "learning_rate": 1.068086850252811e-05,
+      "loss": 0.8463,
+      "step": 20967
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.57475799792263,
+      "learning_rate": 1.0680107230201155e-05,
+      "loss": 0.8617,
+      "step": 20968
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5670802758035844,
+      "learning_rate": 1.06793459539144e-05,
+      "loss": 0.8747,
+      "step": 20969
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8389974233102295,
+      "learning_rate": 1.0678584673672279e-05,
+      "loss": 0.8295,
+      "step": 20970
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6982810004769184,
+      "learning_rate": 1.0677823389479223e-05,
+      "loss": 0.7626,
+      "step": 20971
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5572670806758675,
+      "learning_rate": 1.0677062101339666e-05,
+      "loss": 0.8691,
+      "step": 20972
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7617593756477612,
+      "learning_rate": 1.0676300809258035e-05,
+      "loss": 0.8009,
+      "step": 20973
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9106763859275178,
+      "learning_rate": 1.0675539513238774e-05,
+      "loss": 0.7708,
+      "step": 20974
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1581410934468384,
+      "learning_rate": 1.06747782132863e-05,
+      "loss": 0.9075,
+      "step": 20975
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5112806720514165,
+      "learning_rate": 1.0674016909405059e-05,
+      "loss": 0.7412,
+      "step": 20976
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5730915893234838,
+      "learning_rate": 1.0673255601599472e-05,
+      "loss": 0.9953,
+      "step": 20977
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4506251682651736,
+      "learning_rate": 1.0672494289873985e-05,
+      "loss": 0.824,
+      "step": 20978
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7939351102453736,
+      "learning_rate": 1.067173297423302e-05,
+      "loss": 0.9226,
+      "step": 20979
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0648360388286298,
+      "learning_rate": 1.0670971654681014e-05,
+      "loss": 0.8873,
+      "step": 20980
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.722731970681374,
+      "learning_rate": 1.0670210331222397e-05,
+      "loss": 0.9725,
+      "step": 20981
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7548101463494985,
+      "learning_rate": 1.0669449003861603e-05,
+      "loss": 0.8472,
+      "step": 20982
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8779024289909706,
+      "learning_rate": 1.0668687672603066e-05,
+      "loss": 0.7238,
+      "step": 20983
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6109755290300802,
+      "learning_rate": 1.0667926337451217e-05,
+      "loss": 0.8119,
+      "step": 20984
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7857447470678819,
+      "learning_rate": 1.0667164998410491e-05,
+      "loss": 0.7636,
+      "step": 20985
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9600668137534647,
+      "learning_rate": 1.0666403655485318e-05,
+      "loss": 0.8317,
+      "step": 20986
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6220977307637157,
+      "learning_rate": 1.066564230868013e-05,
+      "loss": 0.8508,
+      "step": 20987
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8243956137424655,
+      "learning_rate": 1.0664880957999365e-05,
+      "loss": 0.8604,
+      "step": 20988
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.584158560162457,
+      "learning_rate": 1.0664119603447451e-05,
+      "loss": 0.9053,
+      "step": 20989
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5678466642683866,
+      "learning_rate": 1.0663358245028822e-05,
+      "loss": 0.8887,
+      "step": 20990
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8843416551529297,
+      "learning_rate": 1.0662596882747913e-05,
+      "loss": 0.728,
+      "step": 20991
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.007077220478652,
+      "learning_rate": 1.0661835516609153e-05,
+      "loss": 0.9294,
+      "step": 20992
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6074598185938203,
+      "learning_rate": 1.066107414661698e-05,
+      "loss": 0.8567,
+      "step": 20993
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4428427436668876,
+      "learning_rate": 1.066031277277582e-05,
+      "loss": 0.8237,
+      "step": 20994
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7761780057364648,
+      "learning_rate": 1.0659551395090115e-05,
+      "loss": 0.6948,
+      "step": 20995
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6601282216784485,
+      "learning_rate": 1.0658790013564289e-05,
+      "loss": 0.8302,
+      "step": 20996
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7205584959301674,
+      "learning_rate": 1.0658028628202782e-05,
+      "loss": 0.8401,
+      "step": 20997
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6932911690007142,
+      "learning_rate": 1.0657267239010021e-05,
+      "loss": 0.8197,
+      "step": 20998
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.109368433006648,
+      "learning_rate": 1.0656505845990442e-05,
+      "loss": 0.8805,
+      "step": 20999
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0280990142380846,
+      "learning_rate": 1.0655744449148482e-05,
+      "loss": 0.8672,
+      "step": 21000
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7474313173435166,
+      "learning_rate": 1.0654983048488566e-05,
+      "loss": 0.6643,
+      "step": 21001
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.702274459469771,
+      "learning_rate": 1.0654221644015135e-05,
+      "loss": 0.9134,
+      "step": 21002
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5309597841677156,
+      "learning_rate": 1.0653460235732613e-05,
+      "loss": 0.9173,
+      "step": 21003
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.0980944013679315,
+      "learning_rate": 1.0652698823645444e-05,
+      "loss": 0.8025,
+      "step": 21004
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8881170783418864,
+      "learning_rate": 1.0651937407758052e-05,
+      "loss": 0.8331,
+      "step": 21005
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8572513903708323,
+      "learning_rate": 1.0651175988074877e-05,
+      "loss": 1.01,
+      "step": 21006
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.640574399594974,
+      "learning_rate": 1.0650414564600349e-05,
+      "loss": 0.951,
+      "step": 21007
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5765773865397577,
+      "learning_rate": 1.06496531373389e-05,
+      "loss": 0.8386,
+      "step": 21008
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7354664800620592,
+      "learning_rate": 1.0648891706294967e-05,
+      "loss": 0.7577,
+      "step": 21009
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7346267685338875,
+      "learning_rate": 1.0648130271472978e-05,
+      "loss": 0.7477,
+      "step": 21010
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8510045214484392,
+      "learning_rate": 1.0647368832877371e-05,
+      "loss": 0.8436,
+      "step": 21011
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7708102953869301,
+      "learning_rate": 1.0646607390512578e-05,
+      "loss": 0.7306,
+      "step": 21012
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5910185570877742,
+      "learning_rate": 1.064584594438303e-05,
+      "loss": 0.744,
+      "step": 21013
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4694790961717048,
+      "learning_rate": 1.0645084494493166e-05,
+      "loss": 0.8915,
+      "step": 21014
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8801079215388168,
+      "learning_rate": 1.0644323040847414e-05,
+      "loss": 0.9932,
+      "step": 21015
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6950432847010501,
+      "learning_rate": 1.0643561583450208e-05,
+      "loss": 0.8087,
+      "step": 21016
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4935067135207931,
+      "learning_rate": 1.0642800122305983e-05,
+      "loss": 0.8862,
+      "step": 21017
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5523536782804843,
+      "learning_rate": 1.0642038657419174e-05,
+      "loss": 0.8149,
+      "step": 21018
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8167111806809985,
+      "learning_rate": 1.064127718879421e-05,
+      "loss": 0.7844,
+      "step": 21019
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7878732118525367,
+      "learning_rate": 1.0640515716435529e-05,
+      "loss": 0.8664,
+      "step": 21020
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7897892374314797,
+      "learning_rate": 1.0639754240347564e-05,
+      "loss": 0.9297,
+      "step": 21021
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.102971722788658,
+      "learning_rate": 1.0638992760534743e-05,
+      "loss": 0.8954,
+      "step": 21022
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8366776753313974,
+      "learning_rate": 1.063823127700151e-05,
+      "loss": 0.8391,
+      "step": 21023
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.704859456835255,
+      "learning_rate": 1.0637469789752286e-05,
+      "loss": 0.9929,
+      "step": 21024
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.644556248985522,
+      "learning_rate": 1.0636708298791515e-05,
+      "loss": 1.0154,
+      "step": 21025
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6872854407861804,
+      "learning_rate": 1.0635946804123625e-05,
+      "loss": 1.0658,
+      "step": 21026
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5634052530934375,
+      "learning_rate": 1.0635185305753054e-05,
+      "loss": 0.8698,
+      "step": 21027
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.765376420781081,
+      "learning_rate": 1.0634423803684231e-05,
+      "loss": 0.861,
+      "step": 21028
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4614315615747429,
+      "learning_rate": 1.0633662297921595e-05,
+      "loss": 0.9081,
+      "step": 21029
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5708404437031585,
+      "learning_rate": 1.0632900788469572e-05,
+      "loss": 0.8917,
+      "step": 21030
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.593950529804848,
+      "learning_rate": 1.06321392753326e-05,
+      "loss": 0.9487,
+      "step": 21031
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6026261631862848,
+      "learning_rate": 1.0631377758515116e-05,
+      "loss": 0.8389,
+      "step": 21032
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7398006654154431,
+      "learning_rate": 1.0630616238021549e-05,
+      "loss": 0.6631,
+      "step": 21033
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8130642140349018,
+      "learning_rate": 1.0629854713856336e-05,
+      "loss": 0.8608,
+      "step": 21034
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7901899917466617,
+      "learning_rate": 1.0629093186023906e-05,
+      "loss": 0.9398,
+      "step": 21035
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.715236544275807,
+      "learning_rate": 1.06283316545287e-05,
+      "loss": 0.8239,
+      "step": 21036
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8116464716234864,
+      "learning_rate": 1.0627570119375148e-05,
+      "loss": 0.9028,
+      "step": 21037
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.092444977198058,
+      "learning_rate": 1.0626808580567683e-05,
+      "loss": 0.8932,
+      "step": 21038
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5471117144534798,
+      "learning_rate": 1.0626047038110741e-05,
+      "loss": 0.8463,
+      "step": 21039
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6113200492548143,
+      "learning_rate": 1.0625285492008752e-05,
+      "loss": 0.7963,
+      "step": 21040
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.799324884264848,
+      "learning_rate": 1.0624523942266157e-05,
+      "loss": 0.7144,
+      "step": 21041
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5707309970015424,
+      "learning_rate": 1.0623762388887382e-05,
+      "loss": 0.877,
+      "step": 21042
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6300531473351216,
+      "learning_rate": 1.0623000831876868e-05,
+      "loss": 0.9575,
+      "step": 21043
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6467075964975377,
+      "learning_rate": 1.0622239271239048e-05,
+      "loss": 0.8159,
+      "step": 21044
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6586589953243345,
+      "learning_rate": 1.0621477706978347e-05,
+      "loss": 0.7844,
+      "step": 21045
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.756829493792052,
+      "learning_rate": 1.062071613909921e-05,
+      "loss": 0.7875,
+      "step": 21046
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5584803891777073,
+      "learning_rate": 1.0619954567606066e-05,
+      "loss": 1.0306,
+      "step": 21047
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8092696793337715,
+      "learning_rate": 1.0619192992503351e-05,
+      "loss": 0.8271,
+      "step": 21048
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6245159044704855,
+      "learning_rate": 1.0618431413795498e-05,
+      "loss": 0.941,
+      "step": 21049
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5980315546927024,
+      "learning_rate": 1.0617669831486944e-05,
+      "loss": 0.9033,
+      "step": 21050
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9228937875983707,
+      "learning_rate": 1.0616908245582121e-05,
+      "loss": 0.9612,
+      "step": 21051
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4722773875429955,
+      "learning_rate": 1.0616146656085457e-05,
+      "loss": 0.8939,
+      "step": 21052
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0458848702571513,
+      "learning_rate": 1.0615385063001398e-05,
+      "loss": 0.9047,
+      "step": 21053
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8103962226475374,
+      "learning_rate": 1.0614623466334368e-05,
+      "loss": 0.7363,
+      "step": 21054
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6502997648816586,
+      "learning_rate": 1.0613861866088809e-05,
+      "loss": 0.9521,
+      "step": 21055
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0234237258432644,
+      "learning_rate": 1.0613100262269148e-05,
+      "loss": 0.7721,
+      "step": 21056
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.746142791893492,
+      "learning_rate": 1.0612338654879827e-05,
+      "loss": 0.7123,
+      "step": 21057
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5320961914714495,
+      "learning_rate": 1.0611577043925276e-05,
+      "loss": 0.8026,
+      "step": 21058
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5297442118199776,
+      "learning_rate": 1.0610815429409928e-05,
+      "loss": 0.8466,
+      "step": 21059
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6068959598675283,
+      "learning_rate": 1.0610053811338222e-05,
+      "loss": 0.8492,
+      "step": 21060
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8606353739769048,
+      "learning_rate": 1.0609292189714586e-05,
+      "loss": 0.8316,
+      "step": 21061
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4823369638669879,
+      "learning_rate": 1.060853056454346e-05,
+      "loss": 1.0444,
+      "step": 21062
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.709979853053645,
+      "learning_rate": 1.0607768935829277e-05,
+      "loss": 0.7508,
+      "step": 21063
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6962455176425981,
+      "learning_rate": 1.0607007303576469e-05,
+      "loss": 0.9223,
+      "step": 21064
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4000918886986247,
+      "learning_rate": 1.0606245667789473e-05,
+      "loss": 0.9423,
+      "step": 21065
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.561095528267441,
+      "learning_rate": 1.0605484028472723e-05,
+      "loss": 0.8658,
+      "step": 21066
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.051070288969408,
+      "learning_rate": 1.0604722385630652e-05,
+      "loss": 0.9931,
+      "step": 21067
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7639255412221965,
+      "learning_rate": 1.0603960739267698e-05,
+      "loss": 0.7588,
+      "step": 21068
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3453700597425233,
+      "learning_rate": 1.0603199089388293e-05,
+      "loss": 0.8675,
+      "step": 21069
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6025770016604202,
+      "learning_rate": 1.060243743599687e-05,
+      "loss": 0.8448,
+      "step": 21070
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7473197622813165,
+      "learning_rate": 1.0601675779097868e-05,
+      "loss": 0.6909,
+      "step": 21071
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5279511118892597,
+      "learning_rate": 1.060091411869572e-05,
+      "loss": 0.8425,
+      "step": 21072
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6844340431168647,
+      "learning_rate": 1.0600152454794853e-05,
+      "loss": 0.8291,
+      "step": 21073
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.731259284848296,
+      "learning_rate": 1.0599390787399717e-05,
+      "loss": 0.7923,
+      "step": 21074
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7625503486033554,
+      "learning_rate": 1.0598629116514733e-05,
+      "loss": 0.676,
+      "step": 21075
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5360300338109352,
+      "learning_rate": 1.0597867442144343e-05,
+      "loss": 0.7711,
+      "step": 21076
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0253422601101927,
+      "learning_rate": 1.0597105764292976e-05,
+      "loss": 0.9562,
+      "step": 21077
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3771786457519701,
+      "learning_rate": 1.0596344082965076e-05,
+      "loss": 0.8484,
+      "step": 21078
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5720364116436008,
+      "learning_rate": 1.0595582398165067e-05,
+      "loss": 0.918,
+      "step": 21079
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.769502912651028,
+      "learning_rate": 1.059482070989739e-05,
+      "loss": 0.6958,
+      "step": 21080
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7365309457834369,
+      "learning_rate": 1.059405901816648e-05,
+      "loss": 0.7152,
+      "step": 21081
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5101518098899236,
+      "learning_rate": 1.059329732297677e-05,
+      "loss": 0.7955,
+      "step": 21082
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4472214327297805,
+      "learning_rate": 1.0592535624332694e-05,
+      "loss": 0.9395,
+      "step": 21083
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6119365053334545,
+      "learning_rate": 1.0591773922238685e-05,
+      "loss": 0.9882,
+      "step": 21084
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.563914308336679,
+      "learning_rate": 1.0591012216699189e-05,
+      "loss": 0.8921,
+      "step": 21085
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7673702449132351,
+      "learning_rate": 1.0590250507718624e-05,
+      "loss": 0.7316,
+      "step": 21086
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9089711936720901,
+      "learning_rate": 1.058948879530144e-05,
+      "loss": 1.015,
+      "step": 21087
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4708583215675766,
+      "learning_rate": 1.0588727079452063e-05,
+      "loss": 0.7888,
+      "step": 21088
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.115104526175894,
+      "learning_rate": 1.058796536017493e-05,
+      "loss": 0.8561,
+      "step": 21089
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.549677819867726,
+      "learning_rate": 1.058720363747448e-05,
+      "loss": 0.9086,
+      "step": 21090
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5888865179893283,
+      "learning_rate": 1.0586441911355138e-05,
+      "loss": 0.8707,
+      "step": 21091
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6651511538438444,
+      "learning_rate": 1.0585680181821351e-05,
+      "loss": 0.95,
+      "step": 21092
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5397291014036676,
+      "learning_rate": 1.0584918448877547e-05,
+      "loss": 0.7142,
+      "step": 21093
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5338892432427513,
+      "learning_rate": 1.0584156712528161e-05,
+      "loss": 0.8976,
+      "step": 21094
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.60887388103411,
+      "learning_rate": 1.0583394972777632e-05,
+      "loss": 0.8161,
+      "step": 21095
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6712970595016907,
+      "learning_rate": 1.0582633229630393e-05,
+      "loss": 0.8704,
+      "step": 21096
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.520478926701969,
+      "learning_rate": 1.0581871483090878e-05,
+      "loss": 0.9002,
+      "step": 21097
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6298247920391833,
+      "learning_rate": 1.0581109733163522e-05,
+      "loss": 0.9351,
+      "step": 21098
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6655630806934916,
+      "learning_rate": 1.0580347979852763e-05,
+      "loss": 0.9011,
+      "step": 21099
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6730022343428201,
+      "learning_rate": 1.0579586223163032e-05,
+      "loss": 0.8364,
+      "step": 21100
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8192826075988358,
+      "learning_rate": 1.0578824463098769e-05,
+      "loss": 0.6704,
+      "step": 21101
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7361594969644873,
+      "learning_rate": 1.0578062699664406e-05,
+      "loss": 0.828,
+      "step": 21102
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.683865169489703,
+      "learning_rate": 1.0577300932864378e-05,
+      "loss": 0.9024,
+      "step": 21103
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2282357810961324,
+      "learning_rate": 1.0576539162703123e-05,
+      "loss": 0.7685,
+      "step": 21104
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5981158986012114,
+      "learning_rate": 1.0575777389185071e-05,
+      "loss": 0.9379,
+      "step": 21105
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6049295530999372,
+      "learning_rate": 1.0575015612314665e-05,
+      "loss": 0.8219,
+      "step": 21106
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4829807991424373,
+      "learning_rate": 1.0574253832096332e-05,
+      "loss": 0.78,
+      "step": 21107
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9582004572599105,
+      "learning_rate": 1.0573492048534515e-05,
+      "loss": 0.805,
+      "step": 21108
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4600730486352274,
+      "learning_rate": 1.0572730261633644e-05,
+      "loss": 1.0205,
+      "step": 21109
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8502116463147442,
+      "learning_rate": 1.0571968471398157e-05,
+      "loss": 0.7873,
+      "step": 21110
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.632090410978752,
+      "learning_rate": 1.0571206677832488e-05,
+      "loss": 0.8938,
+      "step": 21111
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6752152373129237,
+      "learning_rate": 1.0570444880941074e-05,
+      "loss": 0.85,
+      "step": 21112
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4711523779640454,
+      "learning_rate": 1.0569683080728345e-05,
+      "loss": 0.9283,
+      "step": 21113
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.623770727638417,
+      "learning_rate": 1.0568921277198744e-05,
+      "loss": 0.8492,
+      "step": 21114
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.913238796602055,
+      "learning_rate": 1.0568159470356702e-05,
+      "loss": 0.6753,
+      "step": 21115
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6532341606838192,
+      "learning_rate": 1.0567397660206658e-05,
+      "loss": 0.9373,
+      "step": 21116
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7526534265167725,
+      "learning_rate": 1.0566635846753044e-05,
+      "loss": 0.7535,
+      "step": 21117
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.831991905911648,
+      "learning_rate": 1.0565874030000296e-05,
+      "loss": 0.7563,
+      "step": 21118
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5333689022569357,
+      "learning_rate": 1.056511220995285e-05,
+      "loss": 0.8482,
+      "step": 21119
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8912776953699686,
+      "learning_rate": 1.0564350386615143e-05,
+      "loss": 0.9192,
+      "step": 21120
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7321121400822663,
+      "learning_rate": 1.0563588559991608e-05,
+      "loss": 0.7905,
+      "step": 21121
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.66552621535034,
+      "learning_rate": 1.0562826730086684e-05,
+      "loss": 0.9657,
+      "step": 21122
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.092165860599864,
+      "learning_rate": 1.0562064896904804e-05,
+      "loss": 0.8619,
+      "step": 21123
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.558024417288204,
+      "learning_rate": 1.0561303060450404e-05,
+      "loss": 0.8974,
+      "step": 21124
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7851908779085555,
+      "learning_rate": 1.056054122072792e-05,
+      "loss": 0.7782,
+      "step": 21125
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.475623059504033,
+      "learning_rate": 1.0559779377741786e-05,
+      "loss": 0.8941,
+      "step": 21126
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7413694708657095,
+      "learning_rate": 1.055901753149644e-05,
+      "loss": 0.8884,
+      "step": 21127
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.604667453021967,
+      "learning_rate": 1.0558255681996316e-05,
+      "loss": 0.8893,
+      "step": 21128
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5069411189529995,
+      "learning_rate": 1.0557493829245854e-05,
+      "loss": 0.8676,
+      "step": 21129
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5542539120246457,
+      "learning_rate": 1.0556731973249486e-05,
+      "loss": 0.8917,
+      "step": 21130
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.909985980265421,
+      "learning_rate": 1.0555970114011644e-05,
+      "loss": 0.8959,
+      "step": 21131
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.603875843180325,
+      "learning_rate": 1.055520825153677e-05,
+      "loss": 0.9105,
+      "step": 21132
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5158587077315226,
+      "learning_rate": 1.0554446385829297e-05,
+      "loss": 1.1377,
+      "step": 21133
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.580798169588363,
+      "learning_rate": 1.0553684516893665e-05,
+      "loss": 0.9016,
+      "step": 21134
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7365315773670482,
+      "learning_rate": 1.0552922644734302e-05,
+      "loss": 0.7124,
+      "step": 21135
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8559448257906543,
+      "learning_rate": 1.0552160769355653e-05,
+      "loss": 0.7907,
+      "step": 21136
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7827026867621459,
+      "learning_rate": 1.0551398890762148e-05,
+      "loss": 0.7419,
+      "step": 21137
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6536746226697925,
+      "learning_rate": 1.0550637008958222e-05,
+      "loss": 1.0129,
+      "step": 21138
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8487393569465995,
+      "learning_rate": 1.0549875123948314e-05,
+      "loss": 0.8742,
+      "step": 21139
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.515925873527868,
+      "learning_rate": 1.0549113235736858e-05,
+      "loss": 0.9224,
+      "step": 21140
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4678555809309004,
+      "learning_rate": 1.0548351344328293e-05,
+      "loss": 0.8535,
+      "step": 21141
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4822704292169344,
+      "learning_rate": 1.0547589449727051e-05,
+      "loss": 0.8313,
+      "step": 21142
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6505612529210654,
+      "learning_rate": 1.054682755193757e-05,
+      "loss": 0.8443,
+      "step": 21143
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8589888551647629,
+      "learning_rate": 1.0546065650964288e-05,
+      "loss": 0.9481,
+      "step": 21144
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7243290307598838,
+      "learning_rate": 1.0545303746811637e-05,
+      "loss": 0.9186,
+      "step": 21145
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4347770981431494,
+      "learning_rate": 1.0544541839484055e-05,
+      "loss": 0.7908,
+      "step": 21146
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7635214898312355,
+      "learning_rate": 1.0543779928985978e-05,
+      "loss": 0.6924,
+      "step": 21147
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.904279898359969,
+      "learning_rate": 1.0543018015321842e-05,
+      "loss": 0.8918,
+      "step": 21148
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.018496683864238,
+      "learning_rate": 1.0542256098496085e-05,
+      "loss": 0.9004,
+      "step": 21149
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5307634944160178,
+      "learning_rate": 1.0541494178513138e-05,
+      "loss": 0.9825,
+      "step": 21150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6148569490868159,
+      "learning_rate": 1.0540732255377442e-05,
+      "loss": 0.8209,
+      "step": 21151
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.761484683927302,
+      "learning_rate": 1.0539970329093432e-05,
+      "loss": 0.756,
+      "step": 21152
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7405682653312613,
+      "learning_rate": 1.0539208399665546e-05,
+      "loss": 0.7806,
+      "step": 21153
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4790689789027398,
+      "learning_rate": 1.0538446467098214e-05,
+      "loss": 0.9456,
+      "step": 21154
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8024869056022739,
+      "learning_rate": 1.053768453139588e-05,
+      "loss": 0.88,
+      "step": 21155
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7705455189731925,
+      "learning_rate": 1.0536922592562973e-05,
+      "loss": 0.9047,
+      "step": 21156
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7329901357599751,
+      "learning_rate": 1.0536160650603935e-05,
+      "loss": 0.6955,
+      "step": 21157
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6123353736719517,
+      "learning_rate": 1.0535398705523197e-05,
+      "loss": 0.8555,
+      "step": 21158
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6258370693687272,
+      "learning_rate": 1.0534636757325203e-05,
+      "loss": 0.8079,
+      "step": 21159
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.503476489789765,
+      "learning_rate": 1.0533874806014382e-05,
+      "loss": 0.8621,
+      "step": 21160
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.684127450154533,
+      "learning_rate": 1.053311285159517e-05,
+      "loss": 1.0349,
+      "step": 21161
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5454818535696044,
+      "learning_rate": 1.053235089407201e-05,
+      "loss": 0.9109,
+      "step": 21162
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7432219433090637,
+      "learning_rate": 1.0531588933449332e-05,
+      "loss": 0.9631,
+      "step": 21163
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7133993597785653,
+      "learning_rate": 1.0530826969731577e-05,
+      "loss": 0.8306,
+      "step": 21164
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.751663574678955,
+      "learning_rate": 1.053006500292318e-05,
+      "loss": 0.7339,
+      "step": 21165
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.558351336573403,
+      "learning_rate": 1.0529303033028575e-05,
+      "loss": 0.9581,
+      "step": 21166
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5410541643412519,
+      "learning_rate": 1.05285410600522e-05,
+      "loss": 0.8459,
+      "step": 21167
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2653788364005467,
+      "learning_rate": 1.0527779083998492e-05,
+      "loss": 0.9146,
+      "step": 21168
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.360235001456729,
+      "learning_rate": 1.0527017104871885e-05,
+      "loss": 0.7812,
+      "step": 21169
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.886981039857169,
+      "learning_rate": 1.0526255122676823e-05,
+      "loss": 0.8454,
+      "step": 21170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7795677530423442,
+      "learning_rate": 1.0525493137417733e-05,
+      "loss": 0.8564,
+      "step": 21171
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.654894436975137,
+      "learning_rate": 1.0524731149099057e-05,
+      "loss": 0.8453,
+      "step": 21172
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.477700856654528,
+      "learning_rate": 1.0523969157725229e-05,
+      "loss": 0.9429,
+      "step": 21173
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5862429984613935,
+      "learning_rate": 1.0523207163300688e-05,
+      "loss": 0.8155,
+      "step": 21174
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.715993450627016,
+      "learning_rate": 1.0522445165829867e-05,
+      "loss": 0.9016,
+      "step": 21175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5340665792297679,
+      "learning_rate": 1.0521683165317209e-05,
+      "loss": 0.8256,
+      "step": 21176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6009285486418314,
+      "learning_rate": 1.0520921161767142e-05,
+      "loss": 0.9251,
+      "step": 21177
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7483354107915028,
+      "learning_rate": 1.0520159155184111e-05,
+      "loss": 0.8845,
+      "step": 21178
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5598916230447735,
+      "learning_rate": 1.0519397145572542e-05,
+      "loss": 0.8574,
+      "step": 21179
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7469962680687985,
+      "learning_rate": 1.0518635132936888e-05,
+      "loss": 0.8655,
+      "step": 21180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6746055462216785,
+      "learning_rate": 1.051787311728157e-05,
+      "loss": 1.0269,
+      "step": 21181
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.301179158223289,
+      "learning_rate": 1.0517111098611028e-05,
+      "loss": 0.8041,
+      "step": 21182
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.661940775604163,
+      "learning_rate": 1.0516349076929708e-05,
+      "loss": 0.8135,
+      "step": 21183
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7116471037114331,
+      "learning_rate": 1.0515587052242035e-05,
+      "loss": 0.9062,
+      "step": 21184
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7673476915457667,
+      "learning_rate": 1.0514825024552456e-05,
+      "loss": 0.8921,
+      "step": 21185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6678103152489476,
+      "learning_rate": 1.0514062993865396e-05,
+      "loss": 0.8637,
+      "step": 21186
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7788440049052353,
+      "learning_rate": 1.0513300960185307e-05,
+      "loss": 0.7234,
+      "step": 21187
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6554968211453955,
+      "learning_rate": 1.0512538923516613e-05,
+      "loss": 0.8619,
+      "step": 21188
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.628315693243579,
+      "learning_rate": 1.0511776883863753e-05,
+      "loss": 0.8739,
+      "step": 21189
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6358791767590366,
+      "learning_rate": 1.0511014841231168e-05,
+      "loss": 0.8543,
+      "step": 21190
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4184129612711311,
+      "learning_rate": 1.0510252795623294e-05,
+      "loss": 0.7508,
+      "step": 21191
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6071167514823184,
+      "learning_rate": 1.0509490747044566e-05,
+      "loss": 0.9219,
+      "step": 21192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.802209272786225,
+      "learning_rate": 1.050872869549942e-05,
+      "loss": 0.9049,
+      "step": 21193
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.800356874383796,
+      "learning_rate": 1.0507966640992295e-05,
+      "loss": 0.8305,
+      "step": 21194
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9451820469859071,
+      "learning_rate": 1.0507204583527629e-05,
+      "loss": 0.8066,
+      "step": 21195
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.542862470107483,
+      "learning_rate": 1.0506442523109857e-05,
+      "loss": 0.9099,
+      "step": 21196
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7623375736506433,
+      "learning_rate": 1.0505680459743414e-05,
+      "loss": 0.9066,
+      "step": 21197
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.699549158148408,
+      "learning_rate": 1.050491839343274e-05,
+      "loss": 0.8281,
+      "step": 21198
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.546730884288475,
+      "learning_rate": 1.0504156324182274e-05,
+      "loss": 0.9049,
+      "step": 21199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5922153190031412,
+      "learning_rate": 1.0503394251996446e-05,
+      "loss": 0.9523,
+      "step": 21200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4116132977298435,
+      "learning_rate": 1.0502632176879701e-05,
+      "loss": 0.8744,
+      "step": 21201
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.713903188531107,
+      "learning_rate": 1.0501870098836473e-05,
+      "loss": 0.8933,
+      "step": 21202
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5941210708972027,
+      "learning_rate": 1.0501108017871193e-05,
+      "loss": 0.8956,
+      "step": 21203
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6843716706418754,
+      "learning_rate": 1.050034593398831e-05,
+      "loss": 0.9521,
+      "step": 21204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7716086436805867,
+      "learning_rate": 1.0499583847192249e-05,
+      "loss": 0.8193,
+      "step": 21205
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8282985892257613,
+      "learning_rate": 1.0498821757487456e-05,
+      "loss": 0.7806,
+      "step": 21206
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6240082389792176,
+      "learning_rate": 1.0498059664878361e-05,
+      "loss": 0.8224,
+      "step": 21207
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.637244307760087,
+      "learning_rate": 1.0497297569369412e-05,
+      "loss": 0.8388,
+      "step": 21208
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5158773312736749,
+      "learning_rate": 1.0496535470965033e-05,
+      "loss": 1.0889,
+      "step": 21209
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6539779808072104,
+      "learning_rate": 1.049577336966967e-05,
+      "loss": 0.8337,
+      "step": 21210
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4922083401036086,
+      "learning_rate": 1.0495011265487758e-05,
+      "loss": 0.7076,
+      "step": 21211
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7536982909711625,
+      "learning_rate": 1.0494249158423733e-05,
+      "loss": 0.9465,
+      "step": 21212
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7121362089972987,
+      "learning_rate": 1.0493487048482031e-05,
+      "loss": 0.9672,
+      "step": 21213
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7562019880952615,
+      "learning_rate": 1.049272493566709e-05,
+      "loss": 0.6881,
+      "step": 21214
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.295969266019012,
+      "learning_rate": 1.0491962819983351e-05,
+      "loss": 0.7679,
+      "step": 21215
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.501025461116298,
+      "learning_rate": 1.0491200701435252e-05,
+      "loss": 0.8205,
+      "step": 21216
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6118253302377674,
+      "learning_rate": 1.0490438580027223e-05,
+      "loss": 0.9161,
+      "step": 21217
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6349042148434803,
+      "learning_rate": 1.0489676455763704e-05,
+      "loss": 0.8639,
+      "step": 21218
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.513843351887841,
+      "learning_rate": 1.0488914328649136e-05,
+      "loss": 0.9639,
+      "step": 21219
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4734973866177656,
+      "learning_rate": 1.0488152198687955e-05,
+      "loss": 0.8631,
+      "step": 21220
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.682314736623875,
+      "learning_rate": 1.0487390065884596e-05,
+      "loss": 0.8028,
+      "step": 21221
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4460797701766055,
+      "learning_rate": 1.0486627930243497e-05,
+      "loss": 0.8582,
+      "step": 21222
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6768446738659477,
+      "learning_rate": 1.0485865791769096e-05,
+      "loss": 0.9485,
+      "step": 21223
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4466216569325772,
+      "learning_rate": 1.0485103650465832e-05,
+      "loss": 0.8134,
+      "step": 21224
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.482778445246265,
+      "learning_rate": 1.0484341506338143e-05,
+      "loss": 0.874,
+      "step": 21225
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6735658064682086,
+      "learning_rate": 1.0483579359390459e-05,
+      "loss": 0.8698,
+      "step": 21226
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7182424006388357,
+      "learning_rate": 1.0482817209627228e-05,
+      "loss": 0.782,
+      "step": 21227
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7631492392524226,
+      "learning_rate": 1.0482055057052878e-05,
+      "loss": 0.866,
+      "step": 21228
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6399260144192025,
+      "learning_rate": 1.0481292901671853e-05,
+      "loss": 0.8314,
+      "step": 21229
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.004738901426218,
+      "learning_rate": 1.0480530743488586e-05,
+      "loss": 0.8926,
+      "step": 21230
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5700019039892004,
+      "learning_rate": 1.0479768582507519e-05,
+      "loss": 0.7607,
+      "step": 21231
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7647205318528838,
+      "learning_rate": 1.0479006418733091e-05,
+      "loss": 0.8498,
+      "step": 21232
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4171233947860173,
+      "learning_rate": 1.047824425216973e-05,
+      "loss": 0.9456,
+      "step": 21233
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8926517384110009,
+      "learning_rate": 1.0477482082821884e-05,
+      "loss": 0.8321,
+      "step": 21234
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.416420296402692,
+      "learning_rate": 1.0476719910693981e-05,
+      "loss": 0.9506,
+      "step": 21235
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4813520557601785,
+      "learning_rate": 1.0475957735790469e-05,
+      "loss": 0.738,
+      "step": 21236
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2351383371285163,
+      "learning_rate": 1.0475195558115776e-05,
+      "loss": 0.8769,
+      "step": 21237
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5528169796109348,
+      "learning_rate": 1.0474433377674349e-05,
+      "loss": 0.9219,
+      "step": 21238
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6646605822178226,
+      "learning_rate": 1.0473671194470618e-05,
+      "loss": 0.9996,
+      "step": 21239
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6520696326790723,
+      "learning_rate": 1.0472909008509023e-05,
+      "loss": 0.7941,
+      "step": 21240
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5507080069799475,
+      "learning_rate": 1.0472146819794003e-05,
+      "loss": 0.8259,
+      "step": 21241
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.529565646311186,
+      "learning_rate": 1.0471384628329993e-05,
+      "loss": 0.8468,
+      "step": 21242
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0475873780432186,
+      "learning_rate": 1.0470622434121434e-05,
+      "loss": 0.861,
+      "step": 21243
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.808347349032198,
+      "learning_rate": 1.0469860237172759e-05,
+      "loss": 0.9071,
+      "step": 21244
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.773493246452049,
+      "learning_rate": 1.0469098037488415e-05,
+      "loss": 0.8453,
+      "step": 21245
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5659006319359066,
+      "learning_rate": 1.0468335835072829e-05,
+      "loss": 0.932,
+      "step": 21246
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4402946811180455,
+      "learning_rate": 1.0467573629930445e-05,
+      "loss": 0.8241,
+      "step": 21247
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7315873717040748,
+      "learning_rate": 1.0466811422065699e-05,
+      "loss": 0.8261,
+      "step": 21248
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.884227410490399,
+      "learning_rate": 1.046604921148303e-05,
+      "loss": 0.744,
+      "step": 21249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6038214815493055,
+      "learning_rate": 1.0465286998186872e-05,
+      "loss": 0.8029,
+      "step": 21250
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5036038470721744,
+      "learning_rate": 1.0464524782181669e-05,
+      "loss": 0.9246,
+      "step": 21251
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6650777991954369,
+      "learning_rate": 1.0463762563471854e-05,
+      "loss": 0.956,
+      "step": 21252
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6439735136211264,
+      "learning_rate": 1.0463000342061868e-05,
+      "loss": 0.8678,
+      "step": 21253
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5765497578636698,
+      "learning_rate": 1.0462238117956145e-05,
+      "loss": 0.8643,
+      "step": 21254
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5377826445761613,
+      "learning_rate": 1.046147589115913e-05,
+      "loss": 0.9086,
+      "step": 21255
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7778008207038654,
+      "learning_rate": 1.046071366167525e-05,
+      "loss": 0.9656,
+      "step": 21256
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7820275945081112,
+      "learning_rate": 1.0459951429508955e-05,
+      "loss": 0.6927,
+      "step": 21257
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.821884172845473,
+      "learning_rate": 1.0459189194664671e-05,
+      "loss": 0.7515,
+      "step": 21258
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5469768333473484,
+      "learning_rate": 1.045842695714685e-05,
+      "loss": 0.8825,
+      "step": 21259
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.497892870052387,
+      "learning_rate": 1.0457664716959916e-05,
+      "loss": 0.8879,
+      "step": 21260
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.35430519989343,
+      "learning_rate": 1.0456902474108315e-05,
+      "loss": 0.7513,
+      "step": 21261
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.729812058303291,
+      "learning_rate": 1.0456140228596484e-05,
+      "loss": 0.8337,
+      "step": 21262
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.798454909190282,
+      "learning_rate": 1.0455377980428858e-05,
+      "loss": 0.8622,
+      "step": 21263
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0454342621386563,
+      "learning_rate": 1.045461572960988e-05,
+      "loss": 0.8385,
+      "step": 21264
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6884775164153614,
+      "learning_rate": 1.0453853476143982e-05,
+      "loss": 0.9265,
+      "step": 21265
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.752834971251204,
+      "learning_rate": 1.0453091220035607e-05,
+      "loss": 0.8328,
+      "step": 21266
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4457956186049075,
+      "learning_rate": 1.0452328961289192e-05,
+      "loss": 0.7388,
+      "step": 21267
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5721521306057615,
+      "learning_rate": 1.0451566699909174e-05,
+      "loss": 0.7969,
+      "step": 21268
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6060074979218144,
+      "learning_rate": 1.0450804435899993e-05,
+      "loss": 0.8584,
+      "step": 21269
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5023065534740487,
+      "learning_rate": 1.0450042169266083e-05,
+      "loss": 0.8835,
+      "step": 21270
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8761605237593204,
+      "learning_rate": 1.0449279900011888e-05,
+      "loss": 0.9849,
+      "step": 21271
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6494294343215696,
+      "learning_rate": 1.044851762814184e-05,
+      "loss": 0.8783,
+      "step": 21272
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8959936963615616,
+      "learning_rate": 1.0447755353660384e-05,
+      "loss": 0.8898,
+      "step": 21273
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5030871763698017,
+      "learning_rate": 1.0446993076571952e-05,
+      "loss": 0.9332,
+      "step": 21274
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4275043930328868,
+      "learning_rate": 1.0446230796880987e-05,
+      "loss": 0.9482,
+      "step": 21275
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8898669794006215,
+      "learning_rate": 1.0445468514591925e-05,
+      "loss": 0.8692,
+      "step": 21276
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5585718312461097,
+      "learning_rate": 1.0444706229709201e-05,
+      "loss": 0.9229,
+      "step": 21277
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5202467674749343,
+      "learning_rate": 1.044394394223726e-05,
+      "loss": 0.9584,
+      "step": 21278
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6321090631471906,
+      "learning_rate": 1.0443181652180534e-05,
+      "loss": 0.8094,
+      "step": 21279
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8004767724585679,
+      "learning_rate": 1.0442419359543465e-05,
+      "loss": 0.7249,
+      "step": 21280
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5874730440648037,
+      "learning_rate": 1.044165706433049e-05,
+      "loss": 0.8792,
+      "step": 21281
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5544653110106,
+      "learning_rate": 1.0440894766546051e-05,
+      "loss": 0.942,
+      "step": 21282
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4954032985869685,
+      "learning_rate": 1.0440132466194582e-05,
+      "loss": 0.837,
+      "step": 21283
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.581838386146551,
+      "learning_rate": 1.0439370163280518e-05,
+      "loss": 0.8456,
+      "step": 21284
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0617616601269555,
+      "learning_rate": 1.0438607857808307e-05,
+      "loss": 0.7748,
+      "step": 21285
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5929641952010754,
+      "learning_rate": 1.0437845549782377e-05,
+      "loss": 0.8873,
+      "step": 21286
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7526988197554192,
+      "learning_rate": 1.0437083239207177e-05,
+      "loss": 0.8773,
+      "step": 21287
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6020070970624425,
+      "learning_rate": 1.0436320926087135e-05,
+      "loss": 0.7723,
+      "step": 21288
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5116516605187724,
+      "learning_rate": 1.0435558610426701e-05,
+      "loss": 0.8206,
+      "step": 21289
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5881955648090529,
+      "learning_rate": 1.0434796292230303e-05,
+      "loss": 0.996,
+      "step": 21290
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5559874407547833,
+      "learning_rate": 1.0434033971502381e-05,
+      "loss": 0.7962,
+      "step": 21291
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7875360028090714,
+      "learning_rate": 1.043327164824738e-05,
+      "loss": 0.7575,
+      "step": 21292
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7003316900408738,
+      "learning_rate": 1.0432509322469732e-05,
+      "loss": 0.8541,
+      "step": 21293
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7886172709139251,
+      "learning_rate": 1.043174699417388e-05,
+      "loss": 0.7889,
+      "step": 21294
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7580357268615151,
+      "learning_rate": 1.0430984663364256e-05,
+      "loss": 0.7608,
+      "step": 21295
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7401162460072443,
+      "learning_rate": 1.0430222330045306e-05,
+      "loss": 0.9361,
+      "step": 21296
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4969950341790443,
+      "learning_rate": 1.0429459994221465e-05,
+      "loss": 0.8244,
+      "step": 21297
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.674911336384759,
+      "learning_rate": 1.0428697655897172e-05,
+      "loss": 0.8788,
+      "step": 21298
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.638523863428847,
+      "learning_rate": 1.0427935315076864e-05,
+      "loss": 0.9379,
+      "step": 21299
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6873708127839038,
+      "learning_rate": 1.0427172971764984e-05,
+      "loss": 0.8527,
+      "step": 21300
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.809186634136146,
+      "learning_rate": 1.0426410625965965e-05,
+      "loss": 0.7991,
+      "step": 21301
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.520822783400068,
+      "learning_rate": 1.0425648277684251e-05,
+      "loss": 0.898,
+      "step": 21302
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5603035412711546,
+      "learning_rate": 1.0424885926924275e-05,
+      "loss": 0.8297,
+      "step": 21303
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.755805075588733,
+      "learning_rate": 1.042412357369048e-05,
+      "loss": 0.8469,
+      "step": 21304
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7871437097303708,
+      "learning_rate": 1.0423361217987303e-05,
+      "loss": 0.9219,
+      "step": 21305
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5900007932700886,
+      "learning_rate": 1.0422598859819186e-05,
+      "loss": 0.9032,
+      "step": 21306
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.228642114716493,
+      "learning_rate": 1.0421836499190559e-05,
+      "loss": 0.9096,
+      "step": 21307
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.472739904837777,
+      "learning_rate": 1.0421074136105872e-05,
+      "loss": 0.7753,
+      "step": 21308
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.163094112618134,
+      "learning_rate": 1.0420311770569553e-05,
+      "loss": 0.8848,
+      "step": 21309
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7253478387867011,
+      "learning_rate": 1.0419549402586049e-05,
+      "loss": 0.6959,
+      "step": 21310
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5149912102748666,
+      "learning_rate": 1.0418787032159797e-05,
+      "loss": 0.8031,
+      "step": 21311
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8559431088794,
+      "learning_rate": 1.041802465929523e-05,
+      "loss": 0.7787,
+      "step": 21312
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.783060753296023,
+      "learning_rate": 1.0417262283996793e-05,
+      "loss": 0.801,
+      "step": 21313
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4186701252390408,
+      "learning_rate": 1.0416499906268922e-05,
+      "loss": 0.7163,
+      "step": 21314
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.665498513811963,
+      "learning_rate": 1.0415737526116059e-05,
+      "loss": 0.7983,
+      "step": 21315
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.435385646735263,
+      "learning_rate": 1.0414975143542636e-05,
+      "loss": 0.8532,
+      "step": 21316
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.787349361892877,
+      "learning_rate": 1.0414212758553102e-05,
+      "loss": 0.7926,
+      "step": 21317
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8564320926537954,
+      "learning_rate": 1.0413450371151888e-05,
+      "loss": 0.9151,
+      "step": 21318
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.804072629728545,
+      "learning_rate": 1.0412687981343435e-05,
+      "loss": 0.6908,
+      "step": 21319
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5699225031392485,
+      "learning_rate": 1.0411925589132182e-05,
+      "loss": 0.7656,
+      "step": 21320
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5898390636694282,
+      "learning_rate": 1.0411163194522566e-05,
+      "loss": 0.902,
+      "step": 21321
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7505634631011591,
+      "learning_rate": 1.041040079751903e-05,
+      "loss": 0.7373,
+      "step": 21322
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5301314415982883,
+      "learning_rate": 1.0409638398126009e-05,
+      "loss": 0.7938,
+      "step": 21323
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5299269641188038,
+      "learning_rate": 1.0408875996347945e-05,
+      "loss": 0.8292,
+      "step": 21324
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7393654203874491,
+      "learning_rate": 1.0408113592189275e-05,
+      "loss": 0.8937,
+      "step": 21325
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.707525330224961,
+      "learning_rate": 1.0407351185654437e-05,
+      "loss": 0.9202,
+      "step": 21326
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8085877912772308,
+      "learning_rate": 1.0406588776747875e-05,
+      "loss": 0.7267,
+      "step": 21327
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2220552080721476,
+      "learning_rate": 1.040582636547402e-05,
+      "loss": 0.8347,
+      "step": 21328
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5265066620217016,
+      "learning_rate": 1.0405063951837318e-05,
+      "loss": 0.8882,
+      "step": 21329
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6560525069316316,
+      "learning_rate": 1.0404301535842205e-05,
+      "loss": 0.9231,
+      "step": 21330
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7250181122454222,
+      "learning_rate": 1.040353911749312e-05,
+      "loss": 0.8216,
+      "step": 21331
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6707614070383938,
+      "learning_rate": 1.0402776696794506e-05,
+      "loss": 0.8645,
+      "step": 21332
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.71758905906992,
+      "learning_rate": 1.040201427375079e-05,
+      "loss": 0.9032,
+      "step": 21333
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.491785224212202,
+      "learning_rate": 1.0401251848366429e-05,
+      "loss": 0.8754,
+      "step": 21334
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5454037159594207,
+      "learning_rate": 1.0400489420645848e-05,
+      "loss": 0.8185,
+      "step": 21335
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.117879193722375,
+      "learning_rate": 1.0399726990593493e-05,
+      "loss": 0.8969,
+      "step": 21336
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.705979365815494,
+      "learning_rate": 1.0398964558213796e-05,
+      "loss": 0.8254,
+      "step": 21337
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.828162543801919,
+      "learning_rate": 1.0398202123511207e-05,
+      "loss": 0.7308,
+      "step": 21338
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.682997342312902,
+      "learning_rate": 1.0397439686490154e-05,
+      "loss": 0.8573,
+      "step": 21339
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5214032350764255,
+      "learning_rate": 1.0396677247155087e-05,
+      "loss": 0.9307,
+      "step": 21340
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6734224950714187,
+      "learning_rate": 1.0395914805510437e-05,
+      "loss": 0.789,
+      "step": 21341
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.774431641123307,
+      "learning_rate": 1.0395152361560645e-05,
+      "loss": 0.7421,
+      "step": 21342
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5766394462595692,
+      "learning_rate": 1.039438991531015e-05,
+      "loss": 0.8316,
+      "step": 21343
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8426756179644126,
+      "learning_rate": 1.0393627466763395e-05,
+      "loss": 0.9511,
+      "step": 21344
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9698713740047713,
+      "learning_rate": 1.0392865015924813e-05,
+      "loss": 0.7422,
+      "step": 21345
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5344034448133186,
+      "learning_rate": 1.0392102562798847e-05,
+      "loss": 0.8069,
+      "step": 21346
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8325680779469098,
+      "learning_rate": 1.0391340107389938e-05,
+      "loss": 0.8974,
+      "step": 21347
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7874211847442618,
+      "learning_rate": 1.039057764970252e-05,
+      "loss": 0.7108,
+      "step": 21348
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6403834394317425,
+      "learning_rate": 1.0389815189741038e-05,
+      "loss": 0.867,
+      "step": 21349
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5141335321914726,
+      "learning_rate": 1.0389052727509928e-05,
+      "loss": 0.888,
+      "step": 21350
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4984324784587535,
+      "learning_rate": 1.0388290263013628e-05,
+      "loss": 0.9894,
+      "step": 21351
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8622388630396236,
+      "learning_rate": 1.038752779625658e-05,
+      "loss": 0.8733,
+      "step": 21352
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.38849041968993,
+      "learning_rate": 1.0386765327243222e-05,
+      "loss": 0.8437,
+      "step": 21353
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2006501192253767,
+      "learning_rate": 1.0386002855977995e-05,
+      "loss": 0.934,
+      "step": 21354
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.552309359163137,
+      "learning_rate": 1.038524038246534e-05,
+      "loss": 0.8634,
+      "step": 21355
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7995003604986338,
+      "learning_rate": 1.0384477906709687e-05,
+      "loss": 0.7301,
+      "step": 21356
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7429333916582863,
+      "learning_rate": 1.0383715428715489e-05,
+      "loss": 0.6699,
+      "step": 21357
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4631245180061077,
+      "learning_rate": 1.038295294848717e-05,
+      "loss": 0.8411,
+      "step": 21358
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.585227681591031,
+      "learning_rate": 1.0382190466029186e-05,
+      "loss": 0.9278,
+      "step": 21359
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7325095432583599,
+      "learning_rate": 1.0381427981345961e-05,
+      "loss": 0.7093,
+      "step": 21360
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7023086123511901,
+      "learning_rate": 1.0380665494441949e-05,
+      "loss": 0.8851,
+      "step": 21361
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6047339197437014,
+      "learning_rate": 1.0379903005321576e-05,
+      "loss": 0.7914,
+      "step": 21362
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4822972737925053,
+      "learning_rate": 1.037914051398929e-05,
+      "loss": 0.8979,
+      "step": 21363
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6718479944562024,
+      "learning_rate": 1.0378378020449527e-05,
+      "loss": 0.8753,
+      "step": 21364
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.634039724143322,
+      "learning_rate": 1.0377615524706726e-05,
+      "loss": 0.8655,
+      "step": 21365
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6069451042875516,
+      "learning_rate": 1.0376853026765332e-05,
+      "loss": 1.0218,
+      "step": 21366
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6069669208950843,
+      "learning_rate": 1.0376090526629773e-05,
+      "loss": 0.8492,
+      "step": 21367
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.536972568261959,
+      "learning_rate": 1.0375328024304505e-05,
+      "loss": 0.8758,
+      "step": 21368
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7097593800716062,
+      "learning_rate": 1.0374565519793953e-05,
+      "loss": 0.8146,
+      "step": 21369
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7522541224767931,
+      "learning_rate": 1.0373803013102563e-05,
+      "loss": 0.7116,
+      "step": 21370
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5176234488035472,
+      "learning_rate": 1.0373040504234775e-05,
+      "loss": 0.7873,
+      "step": 21371
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5847790036196228,
+      "learning_rate": 1.0372277993195026e-05,
+      "loss": 0.9484,
+      "step": 21372
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8989820810562468,
+      "learning_rate": 1.0371515479987755e-05,
+      "loss": 0.8637,
+      "step": 21373
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9375409831524912,
+      "learning_rate": 1.0370752964617405e-05,
+      "loss": 0.8271,
+      "step": 21374
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.501288124109198,
+      "learning_rate": 1.0369990447088415e-05,
+      "loss": 0.9833,
+      "step": 21375
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7647371661597115,
+      "learning_rate": 1.0369227927405223e-05,
+      "loss": 0.7477,
+      "step": 21376
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5583352681681448,
+      "learning_rate": 1.0368465405572265e-05,
+      "loss": 0.8448,
+      "step": 21377
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6221900625131194,
+      "learning_rate": 1.0367702881593988e-05,
+      "loss": 0.9932,
+      "step": 21378
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7920584750495542,
+      "learning_rate": 1.0366940355474829e-05,
+      "loss": 0.7651,
+      "step": 21379
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.810132659212968,
+      "learning_rate": 1.0366177827219227e-05,
+      "loss": 0.7282,
+      "step": 21380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8019308511034072,
+      "learning_rate": 1.036541529683162e-05,
+      "loss": 0.6538,
+      "step": 21381
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7094997054444043,
+      "learning_rate": 1.036465276431645e-05,
+      "loss": 0.8869,
+      "step": 21382
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.482373418027004,
+      "learning_rate": 1.036389022967816e-05,
+      "loss": 0.8793,
+      "step": 21383
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5141703114558072,
+      "learning_rate": 1.036312769292118e-05,
+      "loss": 0.9087,
+      "step": 21384
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7380447473368448,
+      "learning_rate": 1.0362365154049959e-05,
+      "loss": 0.8781,
+      "step": 21385
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3892111066858877,
+      "learning_rate": 1.036160261306893e-05,
+      "loss": 0.8959,
+      "step": 21386
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6682279052753737,
+      "learning_rate": 1.036084006998254e-05,
+      "loss": 0.9635,
+      "step": 21387
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7642832999594316,
+      "learning_rate": 1.0360077524795222e-05,
+      "loss": 0.6925,
+      "step": 21388
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5214979220272014,
+      "learning_rate": 1.0359314977511423e-05,
+      "loss": 0.9798,
+      "step": 21389
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6571877660604788,
+      "learning_rate": 1.0358552428135576e-05,
+      "loss": 0.9882,
+      "step": 21390
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7812774031273684,
+      "learning_rate": 1.0357789876672122e-05,
+      "loss": 0.7887,
+      "step": 21391
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7933358506686211,
+      "learning_rate": 1.0357027323125503e-05,
+      "loss": 0.6553,
+      "step": 21392
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.751996411627358,
+      "learning_rate": 1.0356264767500158e-05,
+      "loss": 0.7335,
+      "step": 21393
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.78536949904636,
+      "learning_rate": 1.0355502209800525e-05,
+      "loss": 0.9724,
+      "step": 21394
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5231528331420652,
+      "learning_rate": 1.0354739650031047e-05,
+      "loss": 0.8867,
+      "step": 21395
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.484985968001301,
+      "learning_rate": 1.0353977088196164e-05,
+      "loss": 0.8302,
+      "step": 21396
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.048126509421457,
+      "learning_rate": 1.0353214524300312e-05,
+      "loss": 0.9572,
+      "step": 21397
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7030088366167537,
+      "learning_rate": 1.0352451958347936e-05,
+      "loss": 0.9191,
+      "step": 21398
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7662231771719831,
+      "learning_rate": 1.035168939034347e-05,
+      "loss": 0.7003,
+      "step": 21399
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4735703737713364,
+      "learning_rate": 1.0350926820291359e-05,
+      "loss": 0.9805,
+      "step": 21400
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.095864537855366,
+      "learning_rate": 1.035016424819604e-05,
+      "loss": 0.9005,
+      "step": 21401
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5250779625497473,
+      "learning_rate": 1.0349401674061952e-05,
+      "loss": 0.9582,
+      "step": 21402
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.03551923266874,
+      "learning_rate": 1.034863909789354e-05,
+      "loss": 0.8007,
+      "step": 21403
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.490200969698481,
+      "learning_rate": 1.034787651969524e-05,
+      "loss": 0.9518,
+      "step": 21404
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6210410107369775,
+      "learning_rate": 1.0347113939471491e-05,
+      "loss": 0.7863,
+      "step": 21405
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4705456948495994,
+      "learning_rate": 1.0346351357226737e-05,
+      "loss": 0.7563,
+      "step": 21406
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6351226461080368,
+      "learning_rate": 1.0345588772965413e-05,
+      "loss": 0.7876,
+      "step": 21407
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5766385740523106,
+      "learning_rate": 1.0344826186691968e-05,
+      "loss": 0.9012,
+      "step": 21408
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.874685792468082,
+      "learning_rate": 1.0344063598410829e-05,
+      "loss": 0.8256,
+      "step": 21409
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5946030580798465,
+      "learning_rate": 1.0343301008126447e-05,
+      "loss": 0.9262,
+      "step": 21410
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6692404562417031,
+      "learning_rate": 1.0342538415843255e-05,
+      "loss": 1.0195,
+      "step": 21411
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1578104895624888,
+      "learning_rate": 1.0341775821565699e-05,
+      "loss": 0.7961,
+      "step": 21412
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6969403816184714,
+      "learning_rate": 1.0341013225298213e-05,
+      "loss": 0.8732,
+      "step": 21413
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.073821671043593,
+      "learning_rate": 1.0340250627045238e-05,
+      "loss": 0.8107,
+      "step": 21414
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8011376867187459,
+      "learning_rate": 1.0339488026811222e-05,
+      "loss": 0.8043,
+      "step": 21415
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.65795475025434,
+      "learning_rate": 1.0338725424600596e-05,
+      "loss": 0.8813,
+      "step": 21416
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.872132064023377,
+      "learning_rate": 1.0337962820417804e-05,
+      "loss": 1.0038,
+      "step": 21417
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5575407716114333,
+      "learning_rate": 1.0337200214267282e-05,
+      "loss": 0.8324,
+      "step": 21418
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6204234997406497,
+      "learning_rate": 1.033643760615348e-05,
+      "loss": 0.8195,
+      "step": 21419
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5297532382100034,
+      "learning_rate": 1.033567499608083e-05,
+      "loss": 0.7341,
+      "step": 21420
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6832858168981466,
+      "learning_rate": 1.033491238405377e-05,
+      "loss": 0.8094,
+      "step": 21421
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4786330447552518,
+      "learning_rate": 1.0334149770076747e-05,
+      "loss": 0.8671,
+      "step": 21422
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5799879253836195,
+      "learning_rate": 1.0333387154154197e-05,
+      "loss": 0.8649,
+      "step": 21423
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7853529120244868,
+      "learning_rate": 1.0332624536290563e-05,
+      "loss": 0.7596,
+      "step": 21424
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8108647070210321,
+      "learning_rate": 1.0331861916490283e-05,
+      "loss": 0.7817,
+      "step": 21425
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6774769395457696,
+      "learning_rate": 1.0331099294757798e-05,
+      "loss": 0.913,
+      "step": 21426
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5555763268717202,
+      "learning_rate": 1.0330336671097547e-05,
+      "loss": 0.9389,
+      "step": 21427
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5486546486689114,
+      "learning_rate": 1.0329574045513973e-05,
+      "loss": 1.0547,
+      "step": 21428
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.508552984560923,
+      "learning_rate": 1.0328811418011514e-05,
+      "loss": 0.8171,
+      "step": 21429
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8689641615257915,
+      "learning_rate": 1.0328048788594609e-05,
+      "loss": 0.8675,
+      "step": 21430
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.102652853352893,
+      "learning_rate": 1.0327286157267701e-05,
+      "loss": 0.808,
+      "step": 21431
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7826220112209021,
+      "learning_rate": 1.032652352403523e-05,
+      "loss": 0.866,
+      "step": 21432
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8204603398163794,
+      "learning_rate": 1.0325760888901638e-05,
+      "loss": 0.7442,
+      "step": 21433
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7799610149468691,
+      "learning_rate": 1.032499825187136e-05,
+      "loss": 0.7328,
+      "step": 21434
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6295344290807259,
+      "learning_rate": 1.0324235612948839e-05,
+      "loss": 0.9226,
+      "step": 21435
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8971525189533052,
+      "learning_rate": 1.0323472972138518e-05,
+      "loss": 0.86,
+      "step": 21436
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5500579030628516,
+      "learning_rate": 1.0322710329444831e-05,
+      "loss": 0.9912,
+      "step": 21437
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3540229523605727,
+      "learning_rate": 1.0321947684872227e-05,
+      "loss": 0.9533,
+      "step": 21438
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5228103816804586,
+      "learning_rate": 1.0321185038425137e-05,
+      "loss": 0.8937,
+      "step": 21439
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.024646927803771,
+      "learning_rate": 1.032042239010801e-05,
+      "loss": 0.8761,
+      "step": 21440
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.012629625649098,
+      "learning_rate": 1.031965973992528e-05,
+      "loss": 0.8397,
+      "step": 21441
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5959684348624485,
+      "learning_rate": 1.0318897087881392e-05,
+      "loss": 0.9085,
+      "step": 21442
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4523129641560348,
+      "learning_rate": 1.0318134433980781e-05,
+      "loss": 0.851,
+      "step": 21443
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.61743235508885,
+      "learning_rate": 1.0317371778227893e-05,
+      "loss": 0.9045,
+      "step": 21444
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.552223297609032,
+      "learning_rate": 1.0316609120627165e-05,
+      "loss": 0.9005,
+      "step": 21445
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.445662030841537,
+      "learning_rate": 1.0315846461183039e-05,
+      "loss": 0.8055,
+      "step": 21446
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.630953323455005,
+      "learning_rate": 1.0315083799899957e-05,
+      "loss": 0.7938,
+      "step": 21447
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4909945030271001,
+      "learning_rate": 1.0314321136782354e-05,
+      "loss": 0.7709,
+      "step": 21448
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5198257631357028,
+      "learning_rate": 1.0313558471834674e-05,
+      "loss": 1.0016,
+      "step": 21449
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8705050052006553,
+      "learning_rate": 1.031279580506136e-05,
+      "loss": 0.8753,
+      "step": 21450
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7059071511464559,
+      "learning_rate": 1.0312033136466848e-05,
+      "loss": 0.9602,
+      "step": 21451
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7685778064963693,
+      "learning_rate": 1.0311270466055578e-05,
+      "loss": 0.8371,
+      "step": 21452
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7901687378976314,
+      "learning_rate": 1.0310507793831997e-05,
+      "loss": 0.7139,
+      "step": 21453
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.924156343229844,
+      "learning_rate": 1.0309745119800539e-05,
+      "loss": 0.8692,
+      "step": 21454
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5540842100011014,
+      "learning_rate": 1.030898244396565e-05,
+      "loss": 0.8725,
+      "step": 21455
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7679928097443819,
+      "learning_rate": 1.0308219766331762e-05,
+      "loss": 0.7479,
+      "step": 21456
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.525798213923217,
+      "learning_rate": 1.0307457086903325e-05,
+      "loss": 0.8023,
+      "step": 21457
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7298947662340444,
+      "learning_rate": 1.0306694405684771e-05,
+      "loss": 0.725,
+      "step": 21458
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5762911103389348,
+      "learning_rate": 1.030593172268055e-05,
+      "loss": 0.9072,
+      "step": 21459
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7422986180404842,
+      "learning_rate": 1.0305169037895095e-05,
+      "loss": 0.9183,
+      "step": 21460
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.937171505045185,
+      "learning_rate": 1.0304406351332852e-05,
+      "loss": 1.0042,
+      "step": 21461
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7483977499609333,
+      "learning_rate": 1.0303643662998255e-05,
+      "loss": 0.7681,
+      "step": 21462
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.909396084518182,
+      "learning_rate": 1.0302880972895751e-05,
+      "loss": 0.8402,
+      "step": 21463
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8030962126597938,
+      "learning_rate": 1.030211828102978e-05,
+      "loss": 0.833,
+      "step": 21464
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.487560028903939,
+      "learning_rate": 1.0301355587404775e-05,
+      "loss": 0.8879,
+      "step": 21465
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6630228013853385,
+      "learning_rate": 1.0300592892025187e-05,
+      "loss": 0.8673,
+      "step": 21466
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6146298683682505,
+      "learning_rate": 1.029983019489545e-05,
+      "loss": 0.958,
+      "step": 21467
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8399117873016357,
+      "learning_rate": 1.029906749602001e-05,
+      "loss": 0.8368,
+      "step": 21468
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.602936821945854,
+      "learning_rate": 1.0298304795403298e-05,
+      "loss": 0.8034,
+      "step": 21469
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7552558163326146,
+      "learning_rate": 1.0297542093049767e-05,
+      "loss": 0.7272,
+      "step": 21470
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.326854710210165,
+      "learning_rate": 1.0296779388963849e-05,
+      "loss": 0.8567,
+      "step": 21471
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.741673419437297,
+      "learning_rate": 1.0296016683149988e-05,
+      "loss": 0.7525,
+      "step": 21472
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5949834970241104,
+      "learning_rate": 1.0295253975612624e-05,
+      "loss": 0.8402,
+      "step": 21473
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5018658479907925,
+      "learning_rate": 1.02944912663562e-05,
+      "loss": 0.9564,
+      "step": 21474
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6339388472365821,
+      "learning_rate": 1.0293728555385149e-05,
+      "loss": 0.7624,
+      "step": 21475
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.537052564967732,
+      "learning_rate": 1.0292965842703922e-05,
+      "loss": 0.8605,
+      "step": 21476
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6263195034172768,
+      "learning_rate": 1.0292203128316953e-05,
+      "loss": 0.8712,
+      "step": 21477
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8436800408306013,
+      "learning_rate": 1.0291440412228688e-05,
+      "loss": 0.9725,
+      "step": 21478
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7816112142035561,
+      "learning_rate": 1.029067769444356e-05,
+      "loss": 0.7402,
+      "step": 21479
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9948993086063738,
+      "learning_rate": 1.0289914974966018e-05,
+      "loss": 0.8892,
+      "step": 21480
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5669831302230222,
+      "learning_rate": 1.0289152253800499e-05,
+      "loss": 0.974,
+      "step": 21481
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5526728365769968,
+      "learning_rate": 1.0288389530951444e-05,
+      "loss": 0.8057,
+      "step": 21482
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6444330123817879,
+      "learning_rate": 1.0287626806423292e-05,
+      "loss": 0.8624,
+      "step": 21483
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.782503676608747,
+      "learning_rate": 1.0286864080220486e-05,
+      "loss": 0.7829,
+      "step": 21484
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8581454031374074,
+      "learning_rate": 1.028610135234747e-05,
+      "loss": 0.7742,
+      "step": 21485
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4239636839151333,
+      "learning_rate": 1.0285338622808676e-05,
+      "loss": 0.7932,
+      "step": 21486
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5487060725402348,
+      "learning_rate": 1.0284575891608556e-05,
+      "loss": 1.0126,
+      "step": 21487
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.37993792047054,
+      "learning_rate": 1.0283813158751539e-05,
+      "loss": 0.9312,
+      "step": 21488
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.750016836784388,
+      "learning_rate": 1.0283050424242076e-05,
+      "loss": 0.6833,
+      "step": 21489
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.615763643515268,
+      "learning_rate": 1.02822876880846e-05,
+      "loss": 0.9367,
+      "step": 21490
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7577750931839269,
+      "learning_rate": 1.0281524950283562e-05,
+      "loss": 0.8813,
+      "step": 21491
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.987262539656815,
+      "learning_rate": 1.0280762210843391e-05,
+      "loss": 0.9154,
+      "step": 21492
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6334594719677897,
+      "learning_rate": 1.0279999469768535e-05,
+      "loss": 0.7186,
+      "step": 21493
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.415525855586338,
+      "learning_rate": 1.0279236727063436e-05,
+      "loss": 0.8979,
+      "step": 21494
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7686330532366916,
+      "learning_rate": 1.0278473982732529e-05,
+      "loss": 0.7367,
+      "step": 21495
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.490857081848831,
+      "learning_rate": 1.0277711236780261e-05,
+      "loss": 0.9252,
+      "step": 21496
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7875105509034634,
+      "learning_rate": 1.0276948489211068e-05,
+      "loss": 0.7875,
+      "step": 21497
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.609909652363044,
+      "learning_rate": 1.0276185740029393e-05,
+      "loss": 0.796,
+      "step": 21498
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8084625100358125,
+      "learning_rate": 1.027542298923968e-05,
+      "loss": 0.9148,
+      "step": 21499
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8269279084138423,
+      "learning_rate": 1.0274660236846364e-05,
+      "loss": 0.9082,
+      "step": 21500
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.83278079019822,
+      "learning_rate": 1.0273897482853891e-05,
+      "loss": 0.8239,
+      "step": 21501
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4446711364652096,
+      "learning_rate": 1.02731347272667e-05,
+      "loss": 0.7632,
+      "step": 21502
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6098555614627952,
+      "learning_rate": 1.0272371970089232e-05,
+      "loss": 0.8623,
+      "step": 21503
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6503796306574023,
+      "learning_rate": 1.0271609211325928e-05,
+      "loss": 0.8354,
+      "step": 21504
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.854096920426312,
+      "learning_rate": 1.0270846450981228e-05,
+      "loss": 0.9424,
+      "step": 21505
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.478940742390325,
+      "learning_rate": 1.0270083689059578e-05,
+      "loss": 0.9432,
+      "step": 21506
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6311330050173816,
+      "learning_rate": 1.0269320925565409e-05,
+      "loss": 0.7647,
+      "step": 21507
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5720868512765267,
+      "learning_rate": 1.0268558160503174e-05,
+      "loss": 0.7677,
+      "step": 21508
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.581199953599987,
+      "learning_rate": 1.0267795393877305e-05,
+      "loss": 0.8372,
+      "step": 21509
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0809512921581375,
+      "learning_rate": 1.026703262569225e-05,
+      "loss": 0.9509,
+      "step": 21510
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.538536316406451,
+      "learning_rate": 1.0266269855952442e-05,
+      "loss": 0.8728,
+      "step": 21511
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6945637847078336,
+      "learning_rate": 1.026550708466233e-05,
+      "loss": 0.8769,
+      "step": 21512
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.205797161075929,
+      "learning_rate": 1.026474431182635e-05,
+      "loss": 0.8835,
+      "step": 21513
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6474820961297298,
+      "learning_rate": 1.0263981537448944e-05,
+      "loss": 0.9178,
+      "step": 21514
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7728162873611656,
+      "learning_rate": 1.0263218761534557e-05,
+      "loss": 0.6701,
+      "step": 21515
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4441928672569608,
+      "learning_rate": 1.0262455984087626e-05,
+      "loss": 0.8632,
+      "step": 21516
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8834611650492827,
+      "learning_rate": 1.0261693205112592e-05,
+      "loss": 0.8587,
+      "step": 21517
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6611477786648061,
+      "learning_rate": 1.0260930424613895e-05,
+      "loss": 0.6344,
+      "step": 21518
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6177317831712772,
+      "learning_rate": 1.0260167642595984e-05,
+      "loss": 0.8638,
+      "step": 21519
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.548779001500217,
+      "learning_rate": 1.0259404859063288e-05,
+      "loss": 0.9744,
+      "step": 21520
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5582630351409483,
+      "learning_rate": 1.0258642074020261e-05,
+      "loss": 1.0226,
+      "step": 21521
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8838224437407403,
+      "learning_rate": 1.0257879287471337e-05,
+      "loss": 0.8848,
+      "step": 21522
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.331330406812713,
+      "learning_rate": 1.0257116499420955e-05,
+      "loss": 0.8727,
+      "step": 21523
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8771978791907256,
+      "learning_rate": 1.0256353709873563e-05,
+      "loss": 0.816,
+      "step": 21524
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6587756583023583,
+      "learning_rate": 1.0255590918833596e-05,
+      "loss": 0.8326,
+      "step": 21525
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8695785644294929,
+      "learning_rate": 1.0254828126305498e-05,
+      "loss": 0.9773,
+      "step": 21526
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8743118063769335,
+      "learning_rate": 1.025406533229371e-05,
+      "loss": 0.8013,
+      "step": 21527
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4864518617254263,
+      "learning_rate": 1.0253302536802674e-05,
+      "loss": 0.8597,
+      "step": 21528
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.496569616805326,
+      "learning_rate": 1.025253973983683e-05,
+      "loss": 0.9801,
+      "step": 21529
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4329967239207775,
+      "learning_rate": 1.025177694140062e-05,
+      "loss": 0.783,
+      "step": 21530
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5946471405451241,
+      "learning_rate": 1.0251014141498486e-05,
+      "loss": 0.8986,
+      "step": 21531
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7151371491814793,
+      "learning_rate": 1.0250251340134865e-05,
+      "loss": 0.8665,
+      "step": 21532
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6317419597781293,
+      "learning_rate": 1.0249488537314203e-05,
+      "loss": 1.0343,
+      "step": 21533
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5226233922073544,
+      "learning_rate": 1.0248725733040941e-05,
+      "loss": 0.8394,
+      "step": 21534
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5115964170284886,
+      "learning_rate": 1.0247962927319518e-05,
+      "loss": 0.866,
+      "step": 21535
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7224074512509269,
+      "learning_rate": 1.0247200120154377e-05,
+      "loss": 0.7749,
+      "step": 21536
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6730796127855134,
+      "learning_rate": 1.0246437311549957e-05,
+      "loss": 0.8335,
+      "step": 21537
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3933749438450844,
+      "learning_rate": 1.0245674501510703e-05,
+      "loss": 0.9047,
+      "step": 21538
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4957031789376103,
+      "learning_rate": 1.0244911690041051e-05,
+      "loss": 0.9109,
+      "step": 21539
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.533742998980566,
+      "learning_rate": 1.024414887714545e-05,
+      "loss": 0.9633,
+      "step": 21540
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.62648832135089,
+      "learning_rate": 1.0243386062828333e-05,
+      "loss": 0.7187,
+      "step": 21541
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6804129572704207,
+      "learning_rate": 1.024262324709415e-05,
+      "loss": 1.0252,
+      "step": 21542
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8468382562055438,
+      "learning_rate": 1.0241860429947334e-05,
+      "loss": 0.7982,
+      "step": 21543
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7131645149563428,
+      "learning_rate": 1.024109761139233e-05,
+      "loss": 0.8135,
+      "step": 21544
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5927579788452444,
+      "learning_rate": 1.0240334791433581e-05,
+      "loss": 0.8479,
+      "step": 21545
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5965076999894807,
+      "learning_rate": 1.0239571970075526e-05,
+      "loss": 0.9089,
+      "step": 21546
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4894739520241038,
+      "learning_rate": 1.0238809147322606e-05,
+      "loss": 0.8689,
+      "step": 21547
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7708985028757824,
+      "learning_rate": 1.0238046323179265e-05,
+      "loss": 0.9796,
+      "step": 21548
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6322278038117237,
+      "learning_rate": 1.0237283497649942e-05,
+      "loss": 0.921,
+      "step": 21549
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.063767549488954,
+      "learning_rate": 1.0236520670739081e-05,
+      "loss": 0.7815,
+      "step": 21550
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7583547407337152,
+      "learning_rate": 1.023575784245112e-05,
+      "loss": 0.765,
+      "step": 21551
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6914469887506793,
+      "learning_rate": 1.0234995012790503e-05,
+      "loss": 0.8282,
+      "step": 21552
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4828351808769256,
+      "learning_rate": 1.023423218176167e-05,
+      "loss": 0.8156,
+      "step": 21553
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8889699578055752,
+      "learning_rate": 1.0233469349369062e-05,
+      "loss": 0.8208,
+      "step": 21554
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.521910631257128,
+      "learning_rate": 1.0232706515617123e-05,
+      "loss": 0.8176,
+      "step": 21555
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7079492696139926,
+      "learning_rate": 1.0231943680510295e-05,
+      "loss": 0.8833,
+      "step": 21556
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7559309575974008,
+      "learning_rate": 1.0231180844053017e-05,
+      "loss": 0.771,
+      "step": 21557
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6547215673896838,
+      "learning_rate": 1.0230418006249728e-05,
+      "loss": 1.0049,
+      "step": 21558
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6495348185015497,
+      "learning_rate": 1.0229655167104876e-05,
+      "loss": 0.9274,
+      "step": 21559
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.776393379190168,
+      "learning_rate": 1.0228892326622892e-05,
+      "loss": 0.8228,
+      "step": 21560
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3331480680304955,
+      "learning_rate": 1.022812948480823e-05,
+      "loss": 0.7363,
+      "step": 21561
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.883833342359796,
+      "learning_rate": 1.0227366641665327e-05,
+      "loss": 0.9182,
+      "step": 21562
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9918623676618263,
+      "learning_rate": 1.022660379719862e-05,
+      "loss": 0.7761,
+      "step": 21563
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6976253016409129,
+      "learning_rate": 1.0225840951412557e-05,
+      "loss": 0.852,
+      "step": 21564
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7037015908018889,
+      "learning_rate": 1.0225078104311572e-05,
+      "loss": 0.8209,
+      "step": 21565
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9735966565841028,
+      "learning_rate": 1.0224315255900116e-05,
+      "loss": 0.9848,
+      "step": 21566
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9547965830293945,
+      "learning_rate": 1.0223552406182622e-05,
+      "loss": 0.6378,
+      "step": 21567
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9373475488443013,
+      "learning_rate": 1.0222789555163538e-05,
+      "loss": 0.9853,
+      "step": 21568
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.659103192384134,
+      "learning_rate": 1.02220267028473e-05,
+      "loss": 0.8584,
+      "step": 21569
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5903560266508694,
+      "learning_rate": 1.0221263849238357e-05,
+      "loss": 1.0243,
+      "step": 21570
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7373265469623259,
+      "learning_rate": 1.022050099434114e-05,
+      "loss": 0.7536,
+      "step": 21571
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5603448637979587,
+      "learning_rate": 1.02197381381601e-05,
+      "loss": 0.9776,
+      "step": 21572
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.742365889974328,
+      "learning_rate": 1.0218975280699675e-05,
+      "loss": 0.8656,
+      "step": 21573
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.456173146233232,
+      "learning_rate": 1.0218212421964304e-05,
+      "loss": 0.7902,
+      "step": 21574
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5482946942744151,
+      "learning_rate": 1.0217449561958432e-05,
+      "loss": 0.9339,
+      "step": 21575
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6588261337879555,
+      "learning_rate": 1.02166867006865e-05,
+      "loss": 0.8594,
+      "step": 21576
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7046129086792154,
+      "learning_rate": 1.021592383815295e-05,
+      "loss": 0.8431,
+      "step": 21577
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0283037393134604,
+      "learning_rate": 1.0215160974362224e-05,
+      "loss": 0.9426,
+      "step": 21578
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7914172993089403,
+      "learning_rate": 1.0214398109318761e-05,
+      "loss": 0.6981,
+      "step": 21579
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5689300417676322,
+      "learning_rate": 1.0213635243027004e-05,
+      "loss": 0.789,
+      "step": 21580
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6458673280579617,
+      "learning_rate": 1.0212872375491396e-05,
+      "loss": 0.8515,
+      "step": 21581
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0811763192783332,
+      "learning_rate": 1.021210950671638e-05,
+      "loss": 0.8926,
+      "step": 21582
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9227690881696897,
+      "learning_rate": 1.021134663670639e-05,
+      "loss": 0.8819,
+      "step": 21583
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.767292106863545,
+      "learning_rate": 1.0210583765465875e-05,
+      "loss": 0.6917,
+      "step": 21584
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6255744992575274,
+      "learning_rate": 1.0209820892999275e-05,
+      "loss": 0.8414,
+      "step": 21585
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7861974302384528,
+      "learning_rate": 1.0209058019311034e-05,
+      "loss": 0.7076,
+      "step": 21586
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6845393363551047,
+      "learning_rate": 1.0208295144405592e-05,
+      "loss": 0.8856,
+      "step": 21587
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6026663428919297,
+      "learning_rate": 1.0207532268287384e-05,
+      "loss": 0.9692,
+      "step": 21588
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6460671306630017,
+      "learning_rate": 1.020676939096086e-05,
+      "loss": 1.0382,
+      "step": 21589
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4438816246523776,
+      "learning_rate": 1.0206006512430457e-05,
+      "loss": 0.8066,
+      "step": 21590
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6376912032397415,
+      "learning_rate": 1.0205243632700622e-05,
+      "loss": 0.9647,
+      "step": 21591
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6273074200798923,
+      "learning_rate": 1.0204480751775792e-05,
+      "loss": 0.962,
+      "step": 21592
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7094960617322539,
+      "learning_rate": 1.0203717869660413e-05,
+      "loss": 0.8677,
+      "step": 21593
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.573705508649346,
+      "learning_rate": 1.0202954986358921e-05,
+      "loss": 0.8689,
+      "step": 21594
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6109220654008536,
+      "learning_rate": 1.0202192101875763e-05,
+      "loss": 0.8993,
+      "step": 21595
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6391510778710656,
+      "learning_rate": 1.0201429216215378e-05,
+      "loss": 0.9722,
+      "step": 21596
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.625112409272121,
+      "learning_rate": 1.0200666329382205e-05,
+      "loss": 0.7812,
+      "step": 21597
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7411215487473017,
+      "learning_rate": 1.0199903441380692e-05,
+      "loss": 0.9565,
+      "step": 21598
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5046358125901804,
+      "learning_rate": 1.0199140552215276e-05,
+      "loss": 0.7959,
+      "step": 21599
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6331424849363292,
+      "learning_rate": 1.0198377661890404e-05,
+      "loss": 0.7484,
+      "step": 21600
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6509093247709474,
+      "learning_rate": 1.0197614770410513e-05,
+      "loss": 0.9932,
+      "step": 21601
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7376540363343977,
+      "learning_rate": 1.0196851877780045e-05,
+      "loss": 0.7217,
+      "step": 21602
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.520133059345237,
+      "learning_rate": 1.0196088984003444e-05,
+      "loss": 0.891,
+      "step": 21603
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7833294651929578,
+      "learning_rate": 1.0195326089085152e-05,
+      "loss": 0.7092,
+      "step": 21604
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4760507689371138,
+      "learning_rate": 1.0194563193029608e-05,
+      "loss": 0.888,
+      "step": 21605
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6591410782035192,
+      "learning_rate": 1.0193800295841257e-05,
+      "loss": 0.8106,
+      "step": 21606
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.48172839187895,
+      "learning_rate": 1.0193037397524538e-05,
+      "loss": 0.7903,
+      "step": 21607
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6352065733545922,
+      "learning_rate": 1.0192274498083895e-05,
+      "loss": 0.8488,
+      "step": 21608
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7487788295572144,
+      "learning_rate": 1.0191511597523767e-05,
+      "loss": 0.8292,
+      "step": 21609
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.504924255665461,
+      "learning_rate": 1.01907486958486e-05,
+      "loss": 0.7406,
+      "step": 21610
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7611567721707446,
+      "learning_rate": 1.0189985793062833e-05,
+      "loss": 0.8918,
+      "step": 21611
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9483372334968103,
+      "learning_rate": 1.0189222889170906e-05,
+      "loss": 0.7747,
+      "step": 21612
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9168277563744163,
+      "learning_rate": 1.0188459984177268e-05,
+      "loss": 0.97,
+      "step": 21613
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.580670935125233,
+      "learning_rate": 1.0187697078086354e-05,
+      "loss": 0.7874,
+      "step": 21614
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6499641695071137,
+      "learning_rate": 1.0186934170902611e-05,
+      "loss": 0.7843,
+      "step": 21615
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6966296191883004,
+      "learning_rate": 1.0186171262630473e-05,
+      "loss": 0.9071,
+      "step": 21616
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8002531068335659,
+      "learning_rate": 1.018540835327439e-05,
+      "loss": 0.8158,
+      "step": 21617
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9597471622323404,
+      "learning_rate": 1.0184645442838802e-05,
+      "loss": 0.8298,
+      "step": 21618
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5868018594469218,
+      "learning_rate": 1.0183882531328148e-05,
+      "loss": 0.8304,
+      "step": 21619
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6139049599806414,
+      "learning_rate": 1.0183119618746871e-05,
+      "loss": 0.8254,
+      "step": 21620
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.795199046329761,
+      "learning_rate": 1.0182356705099416e-05,
+      "loss": 0.8765,
+      "step": 21621
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6477128958442442,
+      "learning_rate": 1.0181593790390221e-05,
+      "loss": 0.886,
+      "step": 21622
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7782797940545108,
+      "learning_rate": 1.0180830874623729e-05,
+      "loss": 0.9108,
+      "step": 21623
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5890711855334514,
+      "learning_rate": 1.0180067957804383e-05,
+      "loss": 0.8806,
+      "step": 21624
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8079156170072637,
+      "learning_rate": 1.0179305039936624e-05,
+      "loss": 1.0195,
+      "step": 21625
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.768937084798378,
+      "learning_rate": 1.0178542121024895e-05,
+      "loss": 0.8387,
+      "step": 21626
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7571188697942788,
+      "learning_rate": 1.0177779201073636e-05,
+      "loss": 0.7394,
+      "step": 21627
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0480148390922883,
+      "learning_rate": 1.017701628008729e-05,
+      "loss": 0.8288,
+      "step": 21628
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.666349294583804,
+      "learning_rate": 1.0176253358070301e-05,
+      "loss": 0.9048,
+      "step": 21629
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.724236536823466,
+      "learning_rate": 1.017549043502711e-05,
+      "loss": 0.8567,
+      "step": 21630
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3819429511791697,
+      "learning_rate": 1.0174727510962154e-05,
+      "loss": 0.822,
+      "step": 21631
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.580222031724976,
+      "learning_rate": 1.0173964585879884e-05,
+      "loss": 0.8429,
+      "step": 21632
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.501491917680089,
+      "learning_rate": 1.0173201659784732e-05,
+      "loss": 0.8585,
+      "step": 21633
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5432366892248892,
+      "learning_rate": 1.017243873268115e-05,
+      "loss": 0.8294,
+      "step": 21634
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5885603758264475,
+      "learning_rate": 1.017167580457357e-05,
+      "loss": 0.9797,
+      "step": 21635
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8137552344652135,
+      "learning_rate": 1.0170912875466445e-05,
+      "loss": 0.8666,
+      "step": 21636
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7761882366179944,
+      "learning_rate": 1.0170149945364204e-05,
+      "loss": 1.0316,
+      "step": 21637
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7487032435882459,
+      "learning_rate": 1.0169387014271303e-05,
+      "loss": 0.7188,
+      "step": 21638
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0857526492392644,
+      "learning_rate": 1.0168624082192173e-05,
+      "loss": 0.774,
+      "step": 21639
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8206282651353438,
+      "learning_rate": 1.0167861149131265e-05,
+      "loss": 0.8777,
+      "step": 21640
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8054883152127768,
+      "learning_rate": 1.016709821509301e-05,
+      "loss": 0.8058,
+      "step": 21641
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5603108537170522,
+      "learning_rate": 1.0166335280081861e-05,
+      "loss": 0.9962,
+      "step": 21642
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.777544680047807,
+      "learning_rate": 1.0165572344102252e-05,
+      "loss": 0.9728,
+      "step": 21643
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9599492691210172,
+      "learning_rate": 1.0164809407158632e-05,
+      "loss": 0.8768,
+      "step": 21644
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5477137513100767,
+      "learning_rate": 1.0164046469255437e-05,
+      "loss": 0.8899,
+      "step": 21645
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.516050792583434,
+      "learning_rate": 1.0163283530397112e-05,
+      "loss": 0.7933,
+      "step": 21646
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.550051013470634,
+      "learning_rate": 1.01625205905881e-05,
+      "loss": 0.8763,
+      "step": 21647
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5810520895784332,
+      "learning_rate": 1.0161757649832836e-05,
+      "loss": 0.9116,
+      "step": 21648
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8021610084143809,
+      "learning_rate": 1.0160994708135775e-05,
+      "loss": 0.7082,
+      "step": 21649
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5227524244249826,
+      "learning_rate": 1.0160231765501345e-05,
+      "loss": 0.7879,
+      "step": 21650
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3063866371292663,
+      "learning_rate": 1.0159468821934001e-05,
+      "loss": 0.7996,
+      "step": 21651
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5925752066840357,
+      "learning_rate": 1.0158705877438178e-05,
+      "loss": 0.8877,
+      "step": 21652
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7545124379493922,
+      "learning_rate": 1.0157942932018317e-05,
+      "loss": 0.8531,
+      "step": 21653
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5677549794724108,
+      "learning_rate": 1.0157179985678862e-05,
+      "loss": 0.8646,
+      "step": 21654
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4406240348311024,
+      "learning_rate": 1.0156417038424259e-05,
+      "loss": 0.9527,
+      "step": 21655
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6243112412508014,
+      "learning_rate": 1.0155654090258942e-05,
+      "loss": 0.8997,
+      "step": 21656
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.555370494684097,
+      "learning_rate": 1.015489114118736e-05,
+      "loss": 0.8202,
+      "step": 21657
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8855514440084518,
+      "learning_rate": 1.0154128191213952e-05,
+      "loss": 0.7912,
+      "step": 21658
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.715885517436341,
+      "learning_rate": 1.015336524034316e-05,
+      "loss": 0.9279,
+      "step": 21659
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8927764453278266,
+      "learning_rate": 1.015260228857943e-05,
+      "loss": 0.8758,
+      "step": 21660
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4500917990584394,
+      "learning_rate": 1.0151839335927199e-05,
+      "loss": 0.917,
+      "step": 21661
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5363757212130937,
+      "learning_rate": 1.015107638239091e-05,
+      "loss": 0.7894,
+      "step": 21662
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9092921069040125,
+      "learning_rate": 1.015031342797501e-05,
+      "loss": 0.9584,
+      "step": 21663
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6839008315041464,
+      "learning_rate": 1.0149550472683933e-05,
+      "loss": 0.939,
+      "step": 21664
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.547552985017612,
+      "learning_rate": 1.0148787516522129e-05,
+      "loss": 0.9565,
+      "step": 21665
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4843646671851558,
+      "learning_rate": 1.0148024559494038e-05,
+      "loss": 0.8492,
+      "step": 21666
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6496561227037025,
+      "learning_rate": 1.0147261601604099e-05,
+      "loss": 0.8913,
+      "step": 21667
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6328737096323347,
+      "learning_rate": 1.014649864285676e-05,
+      "loss": 0.8954,
+      "step": 21668
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6556178967258475,
+      "learning_rate": 1.0145735683256453e-05,
+      "loss": 0.8791,
+      "step": 21669
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7733087667233816,
+      "learning_rate": 1.0144972722807633e-05,
+      "loss": 0.7892,
+      "step": 21670
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0346558742922416,
+      "learning_rate": 1.0144209761514729e-05,
+      "loss": 0.8733,
+      "step": 21671
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6568243361590183,
+      "learning_rate": 1.0143446799382197e-05,
+      "loss": 0.7482,
+      "step": 21672
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9511477504526147,
+      "learning_rate": 1.0142683836414472e-05,
+      "loss": 0.8844,
+      "step": 21673
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.658612068289357,
+      "learning_rate": 1.0141920872615992e-05,
+      "loss": 0.8525,
+      "step": 21674
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6800034296590685,
+      "learning_rate": 1.0141157907991206e-05,
+      "loss": 0.8634,
+      "step": 21675
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6905749596941317,
+      "learning_rate": 1.0140394942544554e-05,
+      "loss": 0.8498,
+      "step": 21676
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5959173900786414,
+      "learning_rate": 1.0139631976280477e-05,
+      "loss": 0.8495,
+      "step": 21677
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5672394712953621,
+      "learning_rate": 1.013886900920342e-05,
+      "loss": 0.7766,
+      "step": 21678
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7623162968883406,
+      "learning_rate": 1.0138106041317823e-05,
+      "loss": 0.7297,
+      "step": 21679
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5327537984886386,
+      "learning_rate": 1.013734307262813e-05,
+      "loss": 0.7205,
+      "step": 21680
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7199180819859516,
+      "learning_rate": 1.013658010313878e-05,
+      "loss": 0.785,
+      "step": 21681
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5817331611640872,
+      "learning_rate": 1.0135817132854218e-05,
+      "loss": 0.8174,
+      "step": 21682
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.540686466369039,
+      "learning_rate": 1.0135054161778886e-05,
+      "loss": 0.8095,
+      "step": 21683
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5852257918200554,
+      "learning_rate": 1.0134291189917227e-05,
+      "loss": 0.8664,
+      "step": 21684
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5151284353144738,
+      "learning_rate": 1.013352821727368e-05,
+      "loss": 0.914,
+      "step": 21685
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7065931392970657,
+      "learning_rate": 1.0132765243852691e-05,
+      "loss": 0.9358,
+      "step": 21686
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.40069716902002,
+      "learning_rate": 1.0132002269658703e-05,
+      "loss": 0.8845,
+      "step": 21687
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6880514506981592,
+      "learning_rate": 1.0131239294696151e-05,
+      "loss": 0.9211,
+      "step": 21688
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8177303629530843,
+      "learning_rate": 1.0130476318969487e-05,
+      "loss": 0.8847,
+      "step": 21689
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.784210491659116,
+      "learning_rate": 1.0129713342483145e-05,
+      "loss": 0.7382,
+      "step": 21690
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5223205766197725,
+      "learning_rate": 1.0128950365241575e-05,
+      "loss": 0.9044,
+      "step": 21691
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5569642030837545,
+      "learning_rate": 1.0128187387249208e-05,
+      "loss": 0.8037,
+      "step": 21692
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8167458927307505,
+      "learning_rate": 1.01274244085105e-05,
+      "loss": 0.884,
+      "step": 21693
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5563956662718175,
+      "learning_rate": 1.0126661429029885e-05,
+      "loss": 0.8436,
+      "step": 21694
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6052993875208752,
+      "learning_rate": 1.0125898448811803e-05,
+      "loss": 0.7322,
+      "step": 21695
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6618695803683938,
+      "learning_rate": 1.0125135467860706e-05,
+      "loss": 0.9423,
+      "step": 21696
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.926165509527701,
+      "learning_rate": 1.0124372486181027e-05,
+      "loss": 0.7901,
+      "step": 21697
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6598299262026386,
+      "learning_rate": 1.0123609503777216e-05,
+      "loss": 0.8855,
+      "step": 21698
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6800289488470344,
+      "learning_rate": 1.0122846520653707e-05,
+      "loss": 0.8435,
+      "step": 21699
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6757273796996845,
+      "learning_rate": 1.0122083536814951e-05,
+      "loss": 0.8617,
+      "step": 21700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7876820262698934,
+      "learning_rate": 1.0121320552265384e-05,
+      "loss": 0.9534,
+      "step": 21701
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.58250485828763,
+      "learning_rate": 1.0120557567009449e-05,
+      "loss": 0.9271,
+      "step": 21702
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.610954953894264,
+      "learning_rate": 1.011979458105159e-05,
+      "loss": 0.9494,
+      "step": 21703
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5393651442183465,
+      "learning_rate": 1.0119031594396249e-05,
+      "loss": 0.8587,
+      "step": 21704
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.579210481016133,
+      "learning_rate": 1.011826860704787e-05,
+      "loss": 0.7979,
+      "step": 21705
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.350445873196079,
+      "learning_rate": 1.0117505619010891e-05,
+      "loss": 0.8755,
+      "step": 21706
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3976548301001963,
+      "learning_rate": 1.011674263028976e-05,
+      "loss": 0.8586,
+      "step": 21707
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6114047699638878,
+      "learning_rate": 1.0115979640888916e-05,
+      "loss": 0.8138,
+      "step": 21708
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8914011920940401,
+      "learning_rate": 1.01152166508128e-05,
+      "loss": 0.7674,
+      "step": 21709
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.236822080016877,
+      "learning_rate": 1.0114453660065856e-05,
+      "loss": 0.8812,
+      "step": 21710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4338878050788464,
+      "learning_rate": 1.0113690668652529e-05,
+      "loss": 0.7779,
+      "step": 21711
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4885034131734352,
+      "learning_rate": 1.0112927676577258e-05,
+      "loss": 0.99,
+      "step": 21712
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.542536453684377,
+      "learning_rate": 1.0112164683844487e-05,
+      "loss": 0.8848,
+      "step": 21713
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7322208902248242,
+      "learning_rate": 1.0111401690458655e-05,
+      "loss": 0.7907,
+      "step": 21714
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7322043432634273,
+      "learning_rate": 1.0110638696424211e-05,
+      "loss": 0.871,
+      "step": 21715
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5675275013889032,
+      "learning_rate": 1.010987570174559e-05,
+      "loss": 0.7863,
+      "step": 21716
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.813166492465954,
+      "learning_rate": 1.0109112706427242e-05,
+      "loss": 0.7144,
+      "step": 21717
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.537892911659265,
+      "learning_rate": 1.01083497104736e-05,
+      "loss": 0.9606,
+      "step": 21718
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2225421509621004,
+      "learning_rate": 1.0107586713889117e-05,
+      "loss": 0.9453,
+      "step": 21719
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.579763046309954,
+      "learning_rate": 1.0106823716678225e-05,
+      "loss": 0.9085,
+      "step": 21720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8546004200049215,
+      "learning_rate": 1.0106060718845377e-05,
+      "loss": 0.8663,
+      "step": 21721
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.5878834205000665,
+      "learning_rate": 1.0105297720395003e-05,
+      "loss": 0.8019,
+      "step": 21722
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5595192768767228,
+      "learning_rate": 1.0104534721331558e-05,
+      "loss": 0.8146,
+      "step": 21723
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.593915202912874,
+      "learning_rate": 1.0103771721659476e-05,
+      "loss": 0.7759,
+      "step": 21724
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6258932480001105,
+      "learning_rate": 1.0103008721383202e-05,
+      "loss": 0.781,
+      "step": 21725
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7716309618273696,
+      "learning_rate": 1.010224572050718e-05,
+      "loss": 0.8833,
+      "step": 21726
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6346622340025507,
+      "learning_rate": 1.0101482719035851e-05,
+      "loss": 0.9158,
+      "step": 21727
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.831175650288649,
+      "learning_rate": 1.0100719716973657e-05,
+      "loss": 0.8444,
+      "step": 21728
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6941650000706017,
+      "learning_rate": 1.009995671432504e-05,
+      "loss": 0.8179,
+      "step": 21729
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6782499146529988,
+      "learning_rate": 1.0099193711094446e-05,
+      "loss": 0.8566,
+      "step": 21730
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7628865445046162,
+      "learning_rate": 1.0098430707286313e-05,
+      "loss": 0.8453,
+      "step": 21731
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7669944940862428,
+      "learning_rate": 1.0097667702905085e-05,
+      "loss": 0.7218,
+      "step": 21732
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7843190853268022,
+      "learning_rate": 1.0096904697955205e-05,
+      "loss": 0.7494,
+      "step": 21733
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0788187016091615,
+      "learning_rate": 1.0096141692441117e-05,
+      "loss": 0.8331,
+      "step": 21734
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.137343615021249,
+      "learning_rate": 1.0095378686367258e-05,
+      "loss": 0.9137,
+      "step": 21735
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6847098533049847,
+      "learning_rate": 1.0094615679738078e-05,
+      "loss": 0.7685,
+      "step": 21736
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5989947072709179,
+      "learning_rate": 1.0093852672558013e-05,
+      "loss": 0.8255,
+      "step": 21737
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7678767018299555,
+      "learning_rate": 1.009308966483151e-05,
+      "loss": 0.8011,
+      "step": 21738
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5529234965741285,
+      "learning_rate": 1.0092326656563006e-05,
+      "loss": 0.8079,
+      "step": 21739
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5257411050561631,
+      "learning_rate": 1.0091563647756953e-05,
+      "loss": 0.9002,
+      "step": 21740
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.596850154027534,
+      "learning_rate": 1.0090800638417779e-05,
+      "loss": 0.832,
+      "step": 21741
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5373159757127182,
+      "learning_rate": 1.0090037628549943e-05,
+      "loss": 0.8846,
+      "step": 21742
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6656245767244948,
+      "learning_rate": 1.0089274618157874e-05,
+      "loss": 0.8623,
+      "step": 21743
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.596445095475899,
+      "learning_rate": 1.0088511607246021e-05,
+      "loss": 0.865,
+      "step": 21744
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7256657738484937,
+      "learning_rate": 1.008774859581883e-05,
+      "loss": 0.8462,
+      "step": 21745
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7292496241968003,
+      "learning_rate": 1.0086985583880731e-05,
+      "loss": 0.9061,
+      "step": 21746
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4734976856559265,
+      "learning_rate": 1.0086222571436182e-05,
+      "loss": 0.7635,
+      "step": 21747
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7222105935583225,
+      "learning_rate": 1.0085459558489612e-05,
+      "loss": 0.7363,
+      "step": 21748
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6781120883327942,
+      "learning_rate": 1.0084696545045474e-05,
+      "loss": 0.9811,
+      "step": 21749
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5393815848962886,
+      "learning_rate": 1.0083933531108202e-05,
+      "loss": 0.7449,
+      "step": 21750
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7587807473188257,
+      "learning_rate": 1.0083170516682244e-05,
+      "loss": 0.7058,
+      "step": 21751
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6817775518895424,
+      "learning_rate": 1.0082407501772043e-05,
+      "loss": 0.9068,
+      "step": 21752
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.901527089063246,
+      "learning_rate": 1.0081644486382036e-05,
+      "loss": 0.9871,
+      "step": 21753
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6590163167900567,
+      "learning_rate": 1.008088147051667e-05,
+      "loss": 1.0329,
+      "step": 21754
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.590897991866097,
+      "learning_rate": 1.0080118454180387e-05,
+      "loss": 0.9564,
+      "step": 21755
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7053959390779434,
+      "learning_rate": 1.007935543737763e-05,
+      "loss": 0.8615,
+      "step": 21756
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7593010173684037,
+      "learning_rate": 1.0078592420112838e-05,
+      "loss": 0.7266,
+      "step": 21757
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.75425009650721,
+      "learning_rate": 1.0077829402390458e-05,
+      "loss": 0.7238,
+      "step": 21758
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.851345428253561,
+      "learning_rate": 1.007706638421493e-05,
+      "loss": 0.7982,
+      "step": 21759
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7773949427886483,
+      "learning_rate": 1.0076303365590696e-05,
+      "loss": 0.7213,
+      "step": 21760
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6187261027904065,
+      "learning_rate": 1.0075540346522202e-05,
+      "loss": 0.9344,
+      "step": 21761
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.658731304789479,
+      "learning_rate": 1.0074777327013887e-05,
+      "loss": 0.8036,
+      "step": 21762
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.121214839389335,
+      "learning_rate": 1.0074014307070194e-05,
+      "loss": 0.7632,
+      "step": 21763
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.718138699692215,
+      "learning_rate": 1.0073251286695567e-05,
+      "loss": 0.9253,
+      "step": 21764
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6206221359655255,
+      "learning_rate": 1.007248826589445e-05,
+      "loss": 1.0125,
+      "step": 21765
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6708977433374046,
+      "learning_rate": 1.0071725244671281e-05,
+      "loss": 0.9308,
+      "step": 21766
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7949802098443135,
+      "learning_rate": 1.0070962223030502e-05,
+      "loss": 0.9062,
+      "step": 21767
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.766558891352242,
+      "learning_rate": 1.0070199200976564e-05,
+      "loss": 0.9796,
+      "step": 21768
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.852078604027192,
+      "learning_rate": 1.00694361785139e-05,
+      "loss": 0.8459,
+      "step": 21769
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1924244473296364,
+      "learning_rate": 1.006867315564696e-05,
+      "loss": 0.8698,
+      "step": 21770
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.108111054562879,
+      "learning_rate": 1.0067910132380179e-05,
+      "loss": 0.8652,
+      "step": 21771
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.651073491914873,
+      "learning_rate": 1.006714710871801e-05,
+      "loss": 1.0019,
+      "step": 21772
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.193436485263264,
+      "learning_rate": 1.0066384084664882e-05,
+      "loss": 0.8198,
+      "step": 21773
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8912492088125858,
+      "learning_rate": 1.006562106022525e-05,
+      "loss": 0.8816,
+      "step": 21774
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7479321113133734,
+      "learning_rate": 1.0064858035403549e-05,
+      "loss": 0.7277,
+      "step": 21775
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6961232973491949,
+      "learning_rate": 1.0064095010204225e-05,
+      "loss": 0.8943,
+      "step": 21776
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7499542466650491,
+      "learning_rate": 1.0063331984631719e-05,
+      "loss": 0.7382,
+      "step": 21777
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9362450185573863,
+      "learning_rate": 1.0062568958690475e-05,
+      "loss": 0.8422,
+      "step": 21778
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.736913145981142,
+      "learning_rate": 1.0061805932384933e-05,
+      "loss": 0.8119,
+      "step": 21779
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7608961538815564,
+      "learning_rate": 1.0061042905719538e-05,
+      "loss": 0.6772,
+      "step": 21780
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.605728486865367,
+      "learning_rate": 1.0060279878698734e-05,
+      "loss": 0.875,
+      "step": 21781
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6305286606807912,
+      "learning_rate": 1.0059516851326958e-05,
+      "loss": 0.9168,
+      "step": 21782
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.949427816159664,
+      "learning_rate": 1.0058753823608658e-05,
+      "loss": 0.8802,
+      "step": 21783
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.26464810873628,
+      "learning_rate": 1.0057990795548274e-05,
+      "loss": 0.9785,
+      "step": 21784
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.777579337481158,
+      "learning_rate": 1.0057227767150249e-05,
+      "loss": 0.7773,
+      "step": 21785
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5753159819341414,
+      "learning_rate": 1.0056464738419028e-05,
+      "loss": 0.91,
+      "step": 21786
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7126100193590084,
+      "learning_rate": 1.005570170935905e-05,
+      "loss": 0.8184,
+      "step": 21787
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8334972502226046,
+      "learning_rate": 1.0054938679974757e-05,
+      "loss": 0.7947,
+      "step": 21788
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0472298602425893,
+      "learning_rate": 1.0054175650270596e-05,
+      "loss": 0.8961,
+      "step": 21789
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5344566035013676,
+      "learning_rate": 1.0053412620251006e-05,
+      "loss": 0.9764,
+      "step": 21790
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2941600883285624,
+      "learning_rate": 1.0052649589920433e-05,
+      "loss": 0.8339,
+      "step": 21791
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5744435876224088,
+      "learning_rate": 1.0051886559283315e-05,
+      "loss": 0.9815,
+      "step": 21792
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.281913005684803,
+      "learning_rate": 1.0051123528344096e-05,
+      "loss": 0.8092,
+      "step": 21793
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5775186600745177,
+      "learning_rate": 1.0050360497107221e-05,
+      "loss": 0.8729,
+      "step": 21794
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5482403832790588,
+      "learning_rate": 1.0049597465577133e-05,
+      "loss": 1.0401,
+      "step": 21795
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8232370285143549,
+      "learning_rate": 1.0048834433758275e-05,
+      "loss": 0.8422,
+      "step": 21796
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.808783900377266,
+      "learning_rate": 1.004807140165508e-05,
+      "loss": 0.7691,
+      "step": 21797
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8499848372946415,
+      "learning_rate": 1.0047308369272004e-05,
+      "loss": 0.8414,
+      "step": 21798
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6849402746524853,
+      "learning_rate": 1.0046545336613481e-05,
+      "loss": 0.9689,
+      "step": 21799
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7998665646118328,
+      "learning_rate": 1.0045782303683958e-05,
+      "loss": 0.814,
+      "step": 21800
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6693887119126067,
+      "learning_rate": 1.0045019270487873e-05,
+      "loss": 0.8979,
+      "step": 21801
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6255818573671297,
+      "learning_rate": 1.0044256237029674e-05,
+      "loss": 0.7536,
+      "step": 21802
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7285606705223158,
+      "learning_rate": 1.0043493203313803e-05,
+      "loss": 0.7962,
+      "step": 21803
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.922678726599728,
+      "learning_rate": 1.0042730169344698e-05,
+      "loss": 0.8154,
+      "step": 21804
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6906108882185031,
+      "learning_rate": 1.0041967135126805e-05,
+      "loss": 0.8784,
+      "step": 21805
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5737639380942945,
+      "learning_rate": 1.0041204100664566e-05,
+      "loss": 0.8456,
+      "step": 21806
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7670388388796225,
+      "learning_rate": 1.0040441065962423e-05,
+      "loss": 0.8648,
+      "step": 21807
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9539877308747915,
+      "learning_rate": 1.0039678031024819e-05,
+      "loss": 0.8621,
+      "step": 21808
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7055497912506261,
+      "learning_rate": 1.0038914995856199e-05,
+      "loss": 0.8614,
+      "step": 21809
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6394809749350212,
+      "learning_rate": 1.0038151960461002e-05,
+      "loss": 0.8841,
+      "step": 21810
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.530547618989492,
+      "learning_rate": 1.0037388924843674e-05,
+      "loss": 0.8737,
+      "step": 21811
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7845698555405685,
+      "learning_rate": 1.0036625889008656e-05,
+      "loss": 0.7116,
+      "step": 21812
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9658797931271197,
+      "learning_rate": 1.0035862852960388e-05,
+      "loss": 0.8384,
+      "step": 21813
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7890013684582335,
+      "learning_rate": 1.0035099816703315e-05,
+      "loss": 0.7334,
+      "step": 21814
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.630734759787147,
+      "learning_rate": 1.0034336780241882e-05,
+      "loss": 0.8633,
+      "step": 21815
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5594709359438172,
+      "learning_rate": 1.0033573743580527e-05,
+      "loss": 0.8601,
+      "step": 21816
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6167305851245062,
+      "learning_rate": 1.0032810706723699e-05,
+      "loss": 0.9321,
+      "step": 21817
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7791542171445287,
+      "learning_rate": 1.0032047669675832e-05,
+      "loss": 0.8993,
+      "step": 21818
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4449641352586964,
+      "learning_rate": 1.0031284632441378e-05,
+      "loss": 0.8653,
+      "step": 21819
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4957924190307919,
+      "learning_rate": 1.003052159502477e-05,
+      "loss": 0.9914,
+      "step": 21820
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7794323454790072,
+      "learning_rate": 1.002975855743046e-05,
+      "loss": 0.7235,
+      "step": 21821
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5570593820807273,
+      "learning_rate": 1.0028995519662883e-05,
+      "loss": 0.7365,
+      "step": 21822
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4498157566047687,
+      "learning_rate": 1.0028232481726488e-05,
+      "loss": 0.8379,
+      "step": 21823
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9295837045275832,
+      "learning_rate": 1.0027469443625712e-05,
+      "loss": 0.9494,
+      "step": 21824
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6787532515488337,
+      "learning_rate": 1.0026706405365001e-05,
+      "loss": 0.9859,
+      "step": 21825
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5203320587363625,
+      "learning_rate": 1.0025943366948795e-05,
+      "loss": 0.9018,
+      "step": 21826
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4898091720238227,
+      "learning_rate": 1.002518032838154e-05,
+      "loss": 0.8834,
+      "step": 21827
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6935764564606144,
+      "learning_rate": 1.0024417289667678e-05,
+      "loss": 0.8507,
+      "step": 21828
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2056327492333136,
+      "learning_rate": 1.0023654250811652e-05,
+      "loss": 0.7081,
+      "step": 21829
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.441203656737497,
+      "learning_rate": 1.00228912118179e-05,
+      "loss": 0.849,
+      "step": 21830
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5992982972584333,
+      "learning_rate": 1.0022128172690868e-05,
+      "loss": 0.704,
+      "step": 21831
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6326369149537814,
+      "learning_rate": 1.0021365133435004e-05,
+      "loss": 0.8907,
+      "step": 21832
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1052639786800373,
+      "learning_rate": 1.002060209405474e-05,
+      "loss": 0.7764,
+      "step": 21833
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7573712182633134,
+      "learning_rate": 1.0019839054554528e-05,
+      "loss": 0.7584,
+      "step": 21834
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6095138422895359,
+      "learning_rate": 1.0019076014938805e-05,
+      "loss": 0.8506,
+      "step": 21835
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5020057101804312,
+      "learning_rate": 1.0018312975212015e-05,
+      "loss": 0.8805,
+      "step": 21836
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6553573845925849,
+      "learning_rate": 1.0017549935378601e-05,
+      "loss": 0.9134,
+      "step": 21837
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.764816527127129,
+      "learning_rate": 1.0016786895443005e-05,
+      "loss": 0.9422,
+      "step": 21838
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5013347679099678,
+      "learning_rate": 1.0016023855409671e-05,
+      "loss": 0.7389,
+      "step": 21839
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.748751095520832,
+      "learning_rate": 1.0015260815283043e-05,
+      "loss": 0.9049,
+      "step": 21840
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.791803716927055,
+      "learning_rate": 1.001449777506756e-05,
+      "loss": 0.8613,
+      "step": 21841
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7781686448286624,
+      "learning_rate": 1.0013734734767666e-05,
+      "loss": 0.8732,
+      "step": 21842
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4174115259759528,
+      "learning_rate": 1.0012971694387803e-05,
+      "loss": 0.9606,
+      "step": 21843
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8442479738834707,
+      "learning_rate": 1.0012208653932416e-05,
+      "loss": 0.8811,
+      "step": 21844
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6335906198880137,
+      "learning_rate": 1.0011445613405948e-05,
+      "loss": 0.8626,
+      "step": 21845
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4062482793627855,
+      "learning_rate": 1.0010682572812837e-05,
+      "loss": 0.7996,
+      "step": 21846
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6099087509611558,
+      "learning_rate": 1.0009919532157532e-05,
+      "loss": 0.8674,
+      "step": 21847
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7039257337485387,
+      "learning_rate": 1.000915649144447e-05,
+      "loss": 0.886,
+      "step": 21848
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7419189839873086,
+      "learning_rate": 1.0008393450678098e-05,
+      "loss": 0.8859,
+      "step": 21849
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7580012592642668,
+      "learning_rate": 1.0007630409862854e-05,
+      "loss": 0.7109,
+      "step": 21850
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5572650844419051,
+      "learning_rate": 1.0006867369003185e-05,
+      "loss": 0.8609,
+      "step": 21851
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5828474332804612,
+      "learning_rate": 1.000610432810353e-05,
+      "loss": 0.7923,
+      "step": 21852
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7314536624343991,
+      "learning_rate": 1.0005341287168338e-05,
+      "loss": 0.6932,
+      "step": 21853
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7964989358285222,
+      "learning_rate": 1.0004578246202046e-05,
+      "loss": 0.7617,
+      "step": 21854
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6279888061176353,
+      "learning_rate": 1.0003815205209095e-05,
+      "loss": 0.7848,
+      "step": 21855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.466236948784687,
+      "learning_rate": 1.0003052164193931e-05,
+      "loss": 0.7552,
+      "step": 21856
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7339437494027274,
+      "learning_rate": 1.0002289123161e-05,
+      "loss": 0.9579,
+      "step": 21857
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7558462003184712,
+      "learning_rate": 1.0001526082114738e-05,
+      "loss": 0.9102,
+      "step": 21858
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6792633319667494,
+      "learning_rate": 1.0000763041059591e-05,
+      "loss": 0.8976,
+      "step": 21859
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6912426794420096,
+      "learning_rate": 1e-05,
+      "loss": 0.6901,
+      "step": 21860
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6289823385904734,
+      "learning_rate": 9.999236958940412e-06,
+      "loss": 0.9506,
+      "step": 21861
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7832673823824348,
+      "learning_rate": 9.998473917885268e-06,
+      "loss": 0.9365,
+      "step": 21862
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7159338062691045,
+      "learning_rate": 9.997710876839005e-06,
+      "loss": 0.9735,
+      "step": 21863
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.784124646704149,
+      "learning_rate": 9.99694783580607e-06,
+      "loss": 0.8221,
+      "step": 21864
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.558833509000656,
+      "learning_rate": 9.996184794790907e-06,
+      "loss": 0.8213,
+      "step": 21865
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7144568215935145,
+      "learning_rate": 9.99542175379796e-06,
+      "loss": 0.8551,
+      "step": 21866
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6407908414356676,
+      "learning_rate": 9.994658712831666e-06,
+      "loss": 0.8259,
+      "step": 21867
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6380666130683892,
+      "learning_rate": 9.99389567189647e-06,
+      "loss": 0.8956,
+      "step": 21868
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4495572933367453,
+      "learning_rate": 9.99313263099682e-06,
+      "loss": 0.9071,
+      "step": 21869
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5928836290818584,
+      "learning_rate": 9.99236959013715e-06,
+      "loss": 0.9531,
+      "step": 21870
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.607686455498206,
+      "learning_rate": 9.991606549321904e-06,
+      "loss": 0.909,
+      "step": 21871
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.247780003791083,
+      "learning_rate": 9.99084350855553e-06,
+      "loss": 0.7271,
+      "step": 21872
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9458352512754966,
+      "learning_rate": 9.990080467842472e-06,
+      "loss": 0.8302,
+      "step": 21873
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7019620566005698,
+      "learning_rate": 9.989317427187165e-06,
+      "loss": 1.0026,
+      "step": 21874
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4625948300120206,
+      "learning_rate": 9.988554386594052e-06,
+      "loss": 0.9229,
+      "step": 21875
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1195234644799483,
+      "learning_rate": 9.987791346067586e-06,
+      "loss": 0.7989,
+      "step": 21876
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6656150433414298,
+      "learning_rate": 9.987028305612199e-06,
+      "loss": 0.9447,
+      "step": 21877
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2189096447598,
+      "learning_rate": 9.986265265232337e-06,
+      "loss": 0.9892,
+      "step": 21878
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6732967993899486,
+      "learning_rate": 9.985502224932442e-06,
+      "loss": 0.8476,
+      "step": 21879
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7038512478167376,
+      "learning_rate": 9.98473918471696e-06,
+      "loss": 0.9297,
+      "step": 21880
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8489013467611564,
+      "learning_rate": 9.983976144590332e-06,
+      "loss": 0.8886,
+      "step": 21881
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7375008047603404,
+      "learning_rate": 9.983213104556995e-06,
+      "loss": 0.9458,
+      "step": 21882
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4411818355198687,
+      "learning_rate": 9.982450064621405e-06,
+      "loss": 0.7104,
+      "step": 21883
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7454345422703956,
+      "learning_rate": 9.98168702478799e-06,
+      "loss": 0.8212,
+      "step": 21884
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6262608534312475,
+      "learning_rate": 9.980923985061199e-06,
+      "loss": 0.8474,
+      "step": 21885
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4542037290797816,
+      "learning_rate": 9.980160945445474e-06,
+      "loss": 0.7691,
+      "step": 21886
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7006578302300717,
+      "learning_rate": 9.979397905945264e-06,
+      "loss": 0.8177,
+      "step": 21887
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7318460628314467,
+      "learning_rate": 9.978634866565e-06,
+      "loss": 0.8186,
+      "step": 21888
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.718048340565004,
+      "learning_rate": 9.977871827309132e-06,
+      "loss": 0.7856,
+      "step": 21889
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6828780182420744,
+      "learning_rate": 9.977108788182104e-06,
+      "loss": 0.9283,
+      "step": 21890
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6237045943966169,
+      "learning_rate": 9.976345749188353e-06,
+      "loss": 0.8771,
+      "step": 21891
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5715435901884978,
+      "learning_rate": 9.975582710332325e-06,
+      "loss": 0.8318,
+      "step": 21892
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.791019292713136,
+      "learning_rate": 9.974819671618462e-06,
+      "loss": 0.7679,
+      "step": 21893
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9275462059168875,
+      "learning_rate": 9.97405663305121e-06,
+      "loss": 0.8594,
+      "step": 21894
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.815558159691927,
+      "learning_rate": 9.973293594635002e-06,
+      "loss": 0.8234,
+      "step": 21895
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.514180167478662,
+      "learning_rate": 9.972530556374291e-06,
+      "loss": 0.8121,
+      "step": 21896
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6612074154987466,
+      "learning_rate": 9.971767518273517e-06,
+      "loss": 0.8582,
+      "step": 21897
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.565814430702309,
+      "learning_rate": 9.97100448033712e-06,
+      "loss": 0.9489,
+      "step": 21898
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.663530238215812,
+      "learning_rate": 9.970241442569543e-06,
+      "loss": 0.8314,
+      "step": 21899
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4590657760173766,
+      "learning_rate": 9.96947840497523e-06,
+      "loss": 0.8961,
+      "step": 21900
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6814439369995395,
+      "learning_rate": 9.968715367558627e-06,
+      "loss": 1.0,
+      "step": 21901
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.690052041679457,
+      "learning_rate": 9.967952330324171e-06,
+      "loss": 0.9433,
+      "step": 21902
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6441972146350783,
+      "learning_rate": 9.967189293276303e-06,
+      "loss": 0.9144,
+      "step": 21903
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7475464041329538,
+      "learning_rate": 9.966426256419475e-06,
+      "loss": 0.8645,
+      "step": 21904
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.980688288543041,
+      "learning_rate": 9.965663219758121e-06,
+      "loss": 0.7757,
+      "step": 21905
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4838565442537337,
+      "learning_rate": 9.964900183296688e-06,
+      "loss": 0.8741,
+      "step": 21906
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.339099248725121,
+      "learning_rate": 9.964137147039613e-06,
+      "loss": 0.935,
+      "step": 21907
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5225834469743749,
+      "learning_rate": 9.96337411099135e-06,
+      "loss": 0.822,
+      "step": 21908
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.785264481629427,
+      "learning_rate": 9.962611075156329e-06,
+      "loss": 0.8698,
+      "step": 21909
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.77342937249748,
+      "learning_rate": 9.961848039539e-06,
+      "loss": 0.9074,
+      "step": 21910
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7327237299160654,
+      "learning_rate": 9.961085004143806e-06,
+      "loss": 0.7222,
+      "step": 21911
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7610942570041959,
+      "learning_rate": 9.960321968975182e-06,
+      "loss": 0.7848,
+      "step": 21912
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8008686414867155,
+      "learning_rate": 9.959558934037578e-06,
+      "loss": 0.7193,
+      "step": 21913
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8183292708288283,
+      "learning_rate": 9.958795899335438e-06,
+      "loss": 0.7584,
+      "step": 21914
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7947993516884089,
+      "learning_rate": 9.9580328648732e-06,
+      "loss": 0.7704,
+      "step": 21915
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4780558739752914,
+      "learning_rate": 9.957269830655305e-06,
+      "loss": 0.8369,
+      "step": 21916
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7815769941372772,
+      "learning_rate": 9.9565067966862e-06,
+      "loss": 0.9169,
+      "step": 21917
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7673122637787124,
+      "learning_rate": 9.955743762970329e-06,
+      "loss": 0.7672,
+      "step": 21918
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6457596206642044,
+      "learning_rate": 9.95498072951213e-06,
+      "loss": 0.8676,
+      "step": 21919
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5752908713302143,
+      "learning_rate": 9.954217696316043e-06,
+      "loss": 0.8209,
+      "step": 21920
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0297772570634827,
+      "learning_rate": 9.95345466338652e-06,
+      "loss": 0.825,
+      "step": 21921
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6926700401062984,
+      "learning_rate": 9.952691630728001e-06,
+      "loss": 0.8263,
+      "step": 21922
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.104027241843911,
+      "learning_rate": 9.951928598344923e-06,
+      "loss": 0.9533,
+      "step": 21923
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.603774476499243,
+      "learning_rate": 9.95116556624173e-06,
+      "loss": 0.8341,
+      "step": 21924
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6379093209186053,
+      "learning_rate": 9.950402534422869e-06,
+      "loss": 0.7826,
+      "step": 21925
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.670603906977267,
+      "learning_rate": 9.94963950289278e-06,
+      "loss": 0.8182,
+      "step": 21926
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6711447685616059,
+      "learning_rate": 9.948876471655906e-06,
+      "loss": 0.9411,
+      "step": 21927
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4689084772877934,
+      "learning_rate": 9.948113440716687e-06,
+      "loss": 0.8368,
+      "step": 21928
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4936819077898966,
+      "learning_rate": 9.947350410079572e-06,
+      "loss": 0.8377,
+      "step": 21929
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7083786924147784,
+      "learning_rate": 9.946587379748997e-06,
+      "loss": 0.856,
+      "step": 21930
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8124562157699545,
+      "learning_rate": 9.945824349729405e-06,
+      "loss": 0.7416,
+      "step": 21931
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8857407791198122,
+      "learning_rate": 9.945061320025243e-06,
+      "loss": 1.0722,
+      "step": 21932
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4792559206286033,
+      "learning_rate": 9.944298290640955e-06,
+      "loss": 0.8698,
+      "step": 21933
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7171566247089707,
+      "learning_rate": 9.943535261580977e-06,
+      "loss": 0.8706,
+      "step": 21934
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8145514067719586,
+      "learning_rate": 9.942772232849751e-06,
+      "loss": 0.7772,
+      "step": 21935
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5899589971699812,
+      "learning_rate": 9.942009204451731e-06,
+      "loss": 0.8056,
+      "step": 21936
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9402614228016628,
+      "learning_rate": 9.941246176391345e-06,
+      "loss": 0.8862,
+      "step": 21937
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5991696637181283,
+      "learning_rate": 9.940483148673045e-06,
+      "loss": 0.6961,
+      "step": 21938
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4161385261817725,
+      "learning_rate": 9.939720121301268e-06,
+      "loss": 1.0146,
+      "step": 21939
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3283181961355344,
+      "learning_rate": 9.938957094280464e-06,
+      "loss": 0.7845,
+      "step": 21940
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7714472749321283,
+      "learning_rate": 9.93819406761507e-06,
+      "loss": 0.7338,
+      "step": 21941
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6070553997284425,
+      "learning_rate": 9.937431041309528e-06,
+      "loss": 0.7849,
+      "step": 21942
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7423276113544797,
+      "learning_rate": 9.936668015368286e-06,
+      "loss": 0.8831,
+      "step": 21943
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7313044030668645,
+      "learning_rate": 9.935904989795778e-06,
+      "loss": 0.8599,
+      "step": 21944
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.715354709861492,
+      "learning_rate": 9.935141964596453e-06,
+      "loss": 0.8847,
+      "step": 21945
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7274575416605205,
+      "learning_rate": 9.934378939774753e-06,
+      "loss": 0.7016,
+      "step": 21946
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5627944858768685,
+      "learning_rate": 9.933615915335121e-06,
+      "loss": 0.808,
+      "step": 21947
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7094949842322518,
+      "learning_rate": 9.932852891281994e-06,
+      "loss": 0.8812,
+      "step": 21948
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5008733729780113,
+      "learning_rate": 9.932089867619821e-06,
+      "loss": 0.811,
+      "step": 21949
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0013013398877315,
+      "learning_rate": 9.931326844353044e-06,
+      "loss": 0.7786,
+      "step": 21950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8274459071554452,
+      "learning_rate": 9.930563821486101e-06,
+      "loss": 0.8635,
+      "step": 21951
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7331890712982176,
+      "learning_rate": 9.929800799023438e-06,
+      "loss": 0.8853,
+      "step": 21952
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.23611709245301,
+      "learning_rate": 9.929037776969498e-06,
+      "loss": 0.8104,
+      "step": 21953
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7204008458662892,
+      "learning_rate": 9.928274755328724e-06,
+      "loss": 0.9047,
+      "step": 21954
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7380701506642595,
+      "learning_rate": 9.927511734105555e-06,
+      "loss": 0.8932,
+      "step": 21955
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6168550101342598,
+      "learning_rate": 9.926748713304433e-06,
+      "loss": 0.8331,
+      "step": 21956
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.108571943457281,
+      "learning_rate": 9.925985692929809e-06,
+      "loss": 0.8385,
+      "step": 21957
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8262560104365595,
+      "learning_rate": 9.925222672986116e-06,
+      "loss": 0.7337,
+      "step": 21958
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.457151229409821,
+      "learning_rate": 9.924459653477802e-06,
+      "loss": 0.8819,
+      "step": 21959
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6722590265704131,
+      "learning_rate": 9.923696634409304e-06,
+      "loss": 0.8691,
+      "step": 21960
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7431320311871792,
+      "learning_rate": 9.922933615785073e-06,
+      "loss": 0.7141,
+      "step": 21961
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9099695226347098,
+      "learning_rate": 9.922170597609545e-06,
+      "loss": 0.9432,
+      "step": 21962
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.095207322035934,
+      "learning_rate": 9.921407579887164e-06,
+      "loss": 1.0117,
+      "step": 21963
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6632787881541702,
+      "learning_rate": 9.920644562622375e-06,
+      "loss": 0.8712,
+      "step": 21964
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5765845280850197,
+      "learning_rate": 9.919881545819614e-06,
+      "loss": 0.901,
+      "step": 21965
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.750780164336429,
+      "learning_rate": 9.919118529483333e-06,
+      "loss": 0.7153,
+      "step": 21966
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7495329533125555,
+      "learning_rate": 9.918355513617964e-06,
+      "loss": 0.7426,
+      "step": 21967
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7303604330629745,
+      "learning_rate": 9.917592498227964e-06,
+      "loss": 0.8704,
+      "step": 21968
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8110867485115603,
+      "learning_rate": 9.916829483317758e-06,
+      "loss": 0.7214,
+      "step": 21969
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7521195912101307,
+      "learning_rate": 9.9160664688918e-06,
+      "loss": 0.6963,
+      "step": 21970
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7840183165764574,
+      "learning_rate": 9.915303454954533e-06,
+      "loss": 0.7196,
+      "step": 21971
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8188035935481571,
+      "learning_rate": 9.914540441510392e-06,
+      "loss": 0.7018,
+      "step": 21972
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5759907529997286,
+      "learning_rate": 9.913777428563823e-06,
+      "loss": 0.7588,
+      "step": 21973
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7871906627378394,
+      "learning_rate": 9.91301441611927e-06,
+      "loss": 0.8044,
+      "step": 21974
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4802009623729286,
+      "learning_rate": 9.912251404181177e-06,
+      "loss": 0.9616,
+      "step": 21975
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8110926790736561,
+      "learning_rate": 9.911488392753982e-06,
+      "loss": 0.8954,
+      "step": 21976
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4884610423002982,
+      "learning_rate": 9.910725381842127e-06,
+      "loss": 0.9292,
+      "step": 21977
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6740654295658943,
+      "learning_rate": 9.909962371450062e-06,
+      "loss": 0.827,
+      "step": 21978
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5107544379386457,
+      "learning_rate": 9.909199361582223e-06,
+      "loss": 0.8734,
+      "step": 21979
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4636428435706565,
+      "learning_rate": 9.908436352243052e-06,
+      "loss": 0.705,
+      "step": 21980
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.577407001392402,
+      "learning_rate": 9.907673343436994e-06,
+      "loss": 0.8509,
+      "step": 21981
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5650450813666459,
+      "learning_rate": 9.906910335168495e-06,
+      "loss": 0.7903,
+      "step": 21982
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.635033842591341,
+      "learning_rate": 9.90614732744199e-06,
+      "loss": 0.8802,
+      "step": 21983
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.77861574972165,
+      "learning_rate": 9.905384320261924e-06,
+      "loss": 0.6757,
+      "step": 21984
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5313943832971921,
+      "learning_rate": 9.904621313632747e-06,
+      "loss": 0.841,
+      "step": 21985
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7750856743157801,
+      "learning_rate": 9.903858307558887e-06,
+      "loss": 0.9012,
+      "step": 21986
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.755723122114638,
+      "learning_rate": 9.903095302044798e-06,
+      "loss": 1.0374,
+      "step": 21987
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6645804610212245,
+      "learning_rate": 9.902332297094915e-06,
+      "loss": 0.9574,
+      "step": 21988
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4623688131557813,
+      "learning_rate": 9.901569292713689e-06,
+      "loss": 0.8188,
+      "step": 21989
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.092989744746987,
+      "learning_rate": 9.900806288905557e-06,
+      "loss": 0.7953,
+      "step": 21990
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4996289407832606,
+      "learning_rate": 9.900043285674961e-06,
+      "loss": 0.8675,
+      "step": 21991
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7119015365703762,
+      "learning_rate": 9.899280283026346e-06,
+      "loss": 0.9894,
+      "step": 21992
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5736090067552388,
+      "learning_rate": 9.898517280964152e-06,
+      "loss": 0.8516,
+      "step": 21993
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4605788693588013,
+      "learning_rate": 9.897754279492822e-06,
+      "loss": 0.8678,
+      "step": 21994
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0356698533873914,
+      "learning_rate": 9.8969912786168e-06,
+      "loss": 0.9186,
+      "step": 21995
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5657715901112979,
+      "learning_rate": 9.896228278340529e-06,
+      "loss": 0.8398,
+      "step": 21996
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5360714462841656,
+      "learning_rate": 9.895465278668445e-06,
+      "loss": 0.8249,
+      "step": 21997
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.969143277962997,
+      "learning_rate": 9.894702279604997e-06,
+      "loss": 0.8856,
+      "step": 21998
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6030302160631231,
+      "learning_rate": 9.89393928115463e-06,
+      "loss": 0.8334,
+      "step": 21999
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.517159712520526,
+      "learning_rate": 9.893176283321779e-06,
+      "loss": 0.8554,
+      "step": 22000
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4916800266332988,
+      "learning_rate": 9.892413286110886e-06,
+      "loss": 1.0477,
+      "step": 22001
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1686710612889564,
+      "learning_rate": 9.8916502895264e-06,
+      "loss": 0.82,
+      "step": 22002
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.669483155557792,
+      "learning_rate": 9.890887293572763e-06,
+      "loss": 0.952,
+      "step": 22003
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7924408866599535,
+      "learning_rate": 9.890124298254412e-06,
+      "loss": 0.6723,
+      "step": 22004
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7870352551360278,
+      "learning_rate": 9.88936130357579e-06,
+      "loss": 0.8942,
+      "step": 22005
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6455216420813594,
+      "learning_rate": 9.888598309541347e-06,
+      "loss": 0.8783,
+      "step": 22006
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8710531371132475,
+      "learning_rate": 9.887835316155516e-06,
+      "loss": 0.8613,
+      "step": 22007
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7816374648289206,
+      "learning_rate": 9.887072323422743e-06,
+      "loss": 0.7306,
+      "step": 22008
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7809663101049094,
+      "learning_rate": 9.886309331347471e-06,
+      "loss": 0.7348,
+      "step": 22009
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7472870247293637,
+      "learning_rate": 9.885546339934145e-06,
+      "loss": 0.9028,
+      "step": 22010
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.93396598723121,
+      "learning_rate": 9.884783349187202e-06,
+      "loss": 0.7871,
+      "step": 22011
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.539764354340255,
+      "learning_rate": 9.884020359111087e-06,
+      "loss": 0.7396,
+      "step": 22012
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7793577165496596,
+      "learning_rate": 9.883257369710244e-06,
+      "loss": 0.6858,
+      "step": 22013
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.749250373762449,
+      "learning_rate": 9.88249438098911e-06,
+      "loss": 0.7189,
+      "step": 22014
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7466161885885196,
+      "learning_rate": 9.881731392952134e-06,
+      "loss": 0.7014,
+      "step": 22015
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.435994524560895,
+      "learning_rate": 9.880968405603751e-06,
+      "loss": 0.8188,
+      "step": 22016
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.22845315605373,
+      "learning_rate": 9.880205418948414e-06,
+      "loss": 0.8519,
+      "step": 22017
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.444950303380993,
+      "learning_rate": 9.879442432990553e-06,
+      "loss": 0.8019,
+      "step": 22018
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5262841322121936,
+      "learning_rate": 9.87867944773462e-06,
+      "loss": 0.8269,
+      "step": 22019
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4534784770429239,
+      "learning_rate": 9.877916463185054e-06,
+      "loss": 0.7573,
+      "step": 22020
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4592375856546547,
+      "learning_rate": 9.877153479346296e-06,
+      "loss": 0.9599,
+      "step": 22021
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8060227081887507,
+      "learning_rate": 9.876390496222787e-06,
+      "loss": 0.7431,
+      "step": 22022
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.748250396697735,
+      "learning_rate": 9.875627513818973e-06,
+      "loss": 0.8901,
+      "step": 22023
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8138989257194835,
+      "learning_rate": 9.874864532139299e-06,
+      "loss": 0.8122,
+      "step": 22024
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8677306823515054,
+      "learning_rate": 9.874101551188198e-06,
+      "loss": 0.8585,
+      "step": 22025
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6662052419845388,
+      "learning_rate": 9.873338570970119e-06,
+      "loss": 0.9885,
+      "step": 22026
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4557918992080516,
+      "learning_rate": 9.872575591489505e-06,
+      "loss": 0.8668,
+      "step": 22027
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4373346233930866,
+      "learning_rate": 9.871812612750794e-06,
+      "loss": 0.8924,
+      "step": 22028
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6300604140406258,
+      "learning_rate": 9.87104963475843e-06,
+      "loss": 0.9461,
+      "step": 22029
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7362371319765892,
+      "learning_rate": 9.870286657516856e-06,
+      "loss": 0.8467,
+      "step": 22030
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.343345302956341,
+      "learning_rate": 9.869523681030518e-06,
+      "loss": 0.9606,
+      "step": 22031
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8532467416630019,
+      "learning_rate": 9.868760705303852e-06,
+      "loss": 0.9163,
+      "step": 22032
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8584581141811578,
+      "learning_rate": 9.8679977303413e-06,
+      "loss": 0.9312,
+      "step": 22033
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8123948266359321,
+      "learning_rate": 9.867234756147314e-06,
+      "loss": 0.6799,
+      "step": 22034
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7465134125417747,
+      "learning_rate": 9.866471782726323e-06,
+      "loss": 0.8358,
+      "step": 22035
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4715152513662348,
+      "learning_rate": 9.865708810082776e-06,
+      "loss": 0.95,
+      "step": 22036
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.446537943240741,
+      "learning_rate": 9.864945838221114e-06,
+      "loss": 0.8751,
+      "step": 22037
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7670570270120247,
+      "learning_rate": 9.864182867145784e-06,
+      "loss": 0.8278,
+      "step": 22038
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.595000691031975,
+      "learning_rate": 9.863419896861222e-06,
+      "loss": 0.8505,
+      "step": 22039
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.54381715548741,
+      "learning_rate": 9.862656927371873e-06,
+      "loss": 0.9594,
+      "step": 22040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6852950017619883,
+      "learning_rate": 9.861893958682183e-06,
+      "loss": 0.8313,
+      "step": 22041
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6787214569066782,
+      "learning_rate": 9.861130990796583e-06,
+      "loss": 1.0046,
+      "step": 22042
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6473295369119,
+      "learning_rate": 9.860368023719525e-06,
+      "loss": 0.9403,
+      "step": 22043
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7834478621541502,
+      "learning_rate": 9.85960505745545e-06,
+      "loss": 0.7258,
+      "step": 22044
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.71294025168943,
+      "learning_rate": 9.858842092008799e-06,
+      "loss": 0.8557,
+      "step": 22045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6071908041522587,
+      "learning_rate": 9.85807912738401e-06,
+      "loss": 0.8454,
+      "step": 22046
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7634058864925624,
+      "learning_rate": 9.857316163585533e-06,
+      "loss": 0.7497,
+      "step": 22047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5747592465851676,
+      "learning_rate": 9.856553200617805e-06,
+      "loss": 0.8593,
+      "step": 22048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.840452178905495,
+      "learning_rate": 9.855790238485273e-06,
+      "loss": 0.8118,
+      "step": 22049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5199574737551518,
+      "learning_rate": 9.85502727719237e-06,
+      "loss": 0.8186,
+      "step": 22050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5406021415504072,
+      "learning_rate": 9.854264316743549e-06,
+      "loss": 0.9107,
+      "step": 22051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5832525555867598,
+      "learning_rate": 9.853501357143247e-06,
+      "loss": 0.824,
+      "step": 22052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4274790246209224,
+      "learning_rate": 9.852738398395904e-06,
+      "loss": 0.9177,
+      "step": 22053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.017212533027335,
+      "learning_rate": 9.851975440505964e-06,
+      "loss": 0.8634,
+      "step": 22054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.851240408126718,
+      "learning_rate": 9.851212483477871e-06,
+      "loss": 0.8974,
+      "step": 22055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.170460015116881,
+      "learning_rate": 9.850449527316069e-06,
+      "loss": 0.9669,
+      "step": 22056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5303104457932482,
+      "learning_rate": 9.849686572024995e-06,
+      "loss": 0.8567,
+      "step": 22057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6704486039744215,
+      "learning_rate": 9.84892361760909e-06,
+      "loss": 0.8897,
+      "step": 22058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8240410876627688,
+      "learning_rate": 9.848160664072806e-06,
+      "loss": 0.8842,
+      "step": 22059
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4609391543068546,
+      "learning_rate": 9.847397711420574e-06,
+      "loss": 0.8845,
+      "step": 22060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7413682904296631,
+      "learning_rate": 9.846634759656843e-06,
+      "loss": 0.722,
+      "step": 22061
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7370894799600805,
+      "learning_rate": 9.84587180878605e-06,
+      "loss": 0.8881,
+      "step": 22062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8787805823069286,
+      "learning_rate": 9.845108858812645e-06,
+      "loss": 0.9664,
+      "step": 22063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8433040498773778,
+      "learning_rate": 9.844345909741061e-06,
+      "loss": 0.826,
+      "step": 22064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3521949259334765,
+      "learning_rate": 9.843582961575743e-06,
+      "loss": 0.9074,
+      "step": 22065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6381958300848423,
+      "learning_rate": 9.842820014321143e-06,
+      "loss": 0.8168,
+      "step": 22066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6570912476883208,
+      "learning_rate": 9.842057067981686e-06,
+      "loss": 0.8534,
+      "step": 22067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.567997417652067,
+      "learning_rate": 9.841294122561826e-06,
+      "loss": 0.7661,
+      "step": 22068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0765573303804676,
+      "learning_rate": 9.840531178066e-06,
+      "loss": 0.8933,
+      "step": 22069
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5850583436271362,
+      "learning_rate": 9.839768234498658e-06,
+      "loss": 0.812,
+      "step": 22070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6926440962040645,
+      "learning_rate": 9.839005291864229e-06,
+      "loss": 0.8552,
+      "step": 22071
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.602580592687665,
+      "learning_rate": 9.838242350167164e-06,
+      "loss": 0.8454,
+      "step": 22072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6100514582595997,
+      "learning_rate": 9.837479409411907e-06,
+      "loss": 0.8279,
+      "step": 22073
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6711143369879837,
+      "learning_rate": 9.836716469602891e-06,
+      "loss": 0.9109,
+      "step": 22074
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7858455752493907,
+      "learning_rate": 9.835953530744567e-06,
+      "loss": 0.899,
+      "step": 22075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6082547766160977,
+      "learning_rate": 9.835190592841371e-06,
+      "loss": 0.9803,
+      "step": 22076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7487840990855327,
+      "learning_rate": 9.834427655897751e-06,
+      "loss": 0.6309,
+      "step": 22077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.740702070726888,
+      "learning_rate": 9.83366471991814e-06,
+      "loss": 0.8429,
+      "step": 22078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7429307746008328,
+      "learning_rate": 9.83290178490699e-06,
+      "loss": 0.7067,
+      "step": 22079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5009848078511798,
+      "learning_rate": 9.83213885086874e-06,
+      "loss": 0.8537,
+      "step": 22080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6788859566009697,
+      "learning_rate": 9.831375917807829e-06,
+      "loss": 0.7206,
+      "step": 22081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8561034311213451,
+      "learning_rate": 9.830612985728699e-06,
+      "loss": 0.8442,
+      "step": 22082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7730946936515326,
+      "learning_rate": 9.829850054635794e-06,
+      "loss": 0.8113,
+      "step": 22083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5554703295604542,
+      "learning_rate": 9.829087124533559e-06,
+      "loss": 0.8183,
+      "step": 22084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8133232918427,
+      "learning_rate": 9.828324195426431e-06,
+      "loss": 0.9312,
+      "step": 22085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7553679295342891,
+      "learning_rate": 9.827561267318852e-06,
+      "loss": 0.7082,
+      "step": 22086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7532758241447324,
+      "learning_rate": 9.82679834021527e-06,
+      "loss": 0.7722,
+      "step": 22087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4707735842939378,
+      "learning_rate": 9.826035414120122e-06,
+      "loss": 0.8906,
+      "step": 22088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6662896347862057,
+      "learning_rate": 9.825272489037847e-06,
+      "loss": 0.7864,
+      "step": 22089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8032450903432589,
+      "learning_rate": 9.824509564972892e-06,
+      "loss": 0.7099,
+      "step": 22090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7834071418888495,
+      "learning_rate": 9.823746641929702e-06,
+      "loss": 0.8481,
+      "step": 22091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7733027435698322,
+      "learning_rate": 9.822983719912712e-06,
+      "loss": 0.7633,
+      "step": 22092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6970099545114443,
+      "learning_rate": 9.822220798926367e-06,
+      "loss": 0.8569,
+      "step": 22093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.546369504260628,
+      "learning_rate": 9.82145787897511e-06,
+      "loss": 0.7935,
+      "step": 22094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6993706327863163,
+      "learning_rate": 9.82069496006338e-06,
+      "loss": 0.9074,
+      "step": 22095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5520048470557912,
+      "learning_rate": 9.81993204219562e-06,
+      "loss": 0.8728,
+      "step": 22096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6138986102167612,
+      "learning_rate": 9.819169125376275e-06,
+      "loss": 0.7995,
+      "step": 22097
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5318724653595075,
+      "learning_rate": 9.818406209609784e-06,
+      "loss": 0.9155,
+      "step": 22098
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7819753014943192,
+      "learning_rate": 9.817643294900587e-06,
+      "loss": 0.7981,
+      "step": 22099
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.084718834809789,
+      "learning_rate": 9.81688038125313e-06,
+      "loss": 0.8578,
+      "step": 22100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.703370684245349,
+      "learning_rate": 9.816117468671858e-06,
+      "loss": 0.9217,
+      "step": 22101
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.587308107476178,
+      "learning_rate": 9.815354557161203e-06,
+      "loss": 1.0024,
+      "step": 22102
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6032708510244695,
+      "learning_rate": 9.814591646725611e-06,
+      "loss": 0.7767,
+      "step": 22103
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.783085336398448,
+      "learning_rate": 9.813828737369527e-06,
+      "loss": 0.6957,
+      "step": 22104
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8674643971459974,
+      "learning_rate": 9.813065829097394e-06,
+      "loss": 0.7903,
+      "step": 22105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5023129994956257,
+      "learning_rate": 9.81230292191365e-06,
+      "loss": 0.7866,
+      "step": 22106
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.491321921434214,
+      "learning_rate": 9.811540015822733e-06,
+      "loss": 0.9197,
+      "step": 22107
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5983401766109697,
+      "learning_rate": 9.810777110829096e-06,
+      "loss": 0.8991,
+      "step": 22108
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.307534001197095,
+      "learning_rate": 9.81001420693717e-06,
+      "loss": 0.9194,
+      "step": 22109
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5094389286722922,
+      "learning_rate": 9.809251304151404e-06,
+      "loss": 0.8443,
+      "step": 22110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.608573013812087,
+      "learning_rate": 9.808488402476234e-06,
+      "loss": 0.8476,
+      "step": 22111
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.850688660533536,
+      "learning_rate": 9.80772550191611e-06,
+      "loss": 0.9577,
+      "step": 22112
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4989359743986608,
+      "learning_rate": 9.806962602475466e-06,
+      "loss": 0.8664,
+      "step": 22113
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.858180176152865,
+      "learning_rate": 9.806199704158745e-06,
+      "loss": 0.9276,
+      "step": 22114
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.552589448679472,
+      "learning_rate": 9.805436806970397e-06,
+      "loss": 0.8995,
+      "step": 22115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5604957268864557,
+      "learning_rate": 9.804673910914851e-06,
+      "loss": 0.9171,
+      "step": 22116
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5833965003728574,
+      "learning_rate": 9.803911015996557e-06,
+      "loss": 1.0136,
+      "step": 22117
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6203952746487298,
+      "learning_rate": 9.803148122219955e-06,
+      "loss": 0.8423,
+      "step": 22118
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8916639193469094,
+      "learning_rate": 9.802385229589493e-06,
+      "loss": 0.7973,
+      "step": 22119
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5593304255709572,
+      "learning_rate": 9.801622338109599e-06,
+      "loss": 0.7577,
+      "step": 22120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6497447611944989,
+      "learning_rate": 9.800859447784726e-06,
+      "loss": 0.8089,
+      "step": 22121
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9502886339937353,
+      "learning_rate": 9.800096558619313e-06,
+      "loss": 0.7954,
+      "step": 22122
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6691604666385103,
+      "learning_rate": 9.799333670617797e-06,
+      "loss": 0.88,
+      "step": 22123
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8654991668624732,
+      "learning_rate": 9.798570783784627e-06,
+      "loss": 0.7942,
+      "step": 22124
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2762785940065773,
+      "learning_rate": 9.79780789812424e-06,
+      "loss": 0.925,
+      "step": 22125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.508978129241158,
+      "learning_rate": 9.797045013641084e-06,
+      "loss": 0.8807,
+      "step": 22126
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6157292554740177,
+      "learning_rate": 9.79628213033959e-06,
+      "loss": 0.8769,
+      "step": 22127
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4444426931305838,
+      "learning_rate": 9.79551924822421e-06,
+      "loss": 0.9038,
+      "step": 22128
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5072015645851022,
+      "learning_rate": 9.794756367299383e-06,
+      "loss": 0.9169,
+      "step": 22129
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6886602803738975,
+      "learning_rate": 9.793993487569544e-06,
+      "loss": 0.9003,
+      "step": 22130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6231584346894543,
+      "learning_rate": 9.793230609039142e-06,
+      "loss": 0.9178,
+      "step": 22131
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4268668305135137,
+      "learning_rate": 9.792467731712618e-06,
+      "loss": 0.7981,
+      "step": 22132
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.649346346669236,
+      "learning_rate": 9.791704855594415e-06,
+      "loss": 0.849,
+      "step": 22133
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5659054989700383,
+      "learning_rate": 9.79094198068897e-06,
+      "loss": 0.922,
+      "step": 22134
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7982826470707804,
+      "learning_rate": 9.790179107000723e-06,
+      "loss": 0.6919,
+      "step": 22135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5375556422187409,
+      "learning_rate": 9.789416234534126e-06,
+      "loss": 0.6496,
+      "step": 22136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5852782136595345,
+      "learning_rate": 9.788653363293612e-06,
+      "loss": 0.7886,
+      "step": 22137
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7507947379509813,
+      "learning_rate": 9.787890493283626e-06,
+      "loss": 0.7506,
+      "step": 22138
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7599192912070849,
+      "learning_rate": 9.787127624508604e-06,
+      "loss": 0.722,
+      "step": 22139
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7910145430277855,
+      "learning_rate": 9.786364756972997e-06,
+      "loss": 0.7818,
+      "step": 22140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6818940570887233,
+      "learning_rate": 9.785601890681242e-06,
+      "loss": 0.8487,
+      "step": 22141
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.749099165658417,
+      "learning_rate": 9.78483902563778e-06,
+      "loss": 0.8508,
+      "step": 22142
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8140376576319922,
+      "learning_rate": 9.784076161847055e-06,
+      "loss": 0.7219,
+      "step": 22143
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6949599673138576,
+      "learning_rate": 9.783313299313502e-06,
+      "loss": 0.8991,
+      "step": 22144
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6534337780708486,
+      "learning_rate": 9.78255043804157e-06,
+      "loss": 0.8155,
+      "step": 22145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4524934168864518,
+      "learning_rate": 9.781787578035698e-06,
+      "loss": 0.8154,
+      "step": 22146
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7391656421673065,
+      "learning_rate": 9.781024719300332e-06,
+      "loss": 0.8232,
+      "step": 22147
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8724758939995232,
+      "learning_rate": 9.780261861839903e-06,
+      "loss": 1.0201,
+      "step": 22148
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7344469404416383,
+      "learning_rate": 9.779499005658862e-06,
+      "loss": 0.8896,
+      "step": 22149
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7757267706407603,
+      "learning_rate": 9.77873615076165e-06,
+      "loss": 0.7284,
+      "step": 22150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.380165250130941,
+      "learning_rate": 9.777973297152703e-06,
+      "loss": 0.8538,
+      "step": 22151
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7291077139815528,
+      "learning_rate": 9.777210444836463e-06,
+      "loss": 0.6752,
+      "step": 22152
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8815116785917896,
+      "learning_rate": 9.77644759381738e-06,
+      "loss": 0.8829,
+      "step": 22153
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.680555529240989,
+      "learning_rate": 9.775684744099889e-06,
+      "loss": 0.9072,
+      "step": 22154
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8018840876384792,
+      "learning_rate": 9.77492189568843e-06,
+      "loss": 0.8748,
+      "step": 22155
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.596480176721564,
+      "learning_rate": 9.774159048587445e-06,
+      "loss": 0.9053,
+      "step": 22156
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6426171654011927,
+      "learning_rate": 9.773396202801383e-06,
+      "loss": 0.8745,
+      "step": 22157
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8629810388301011,
+      "learning_rate": 9.772633358334678e-06,
+      "loss": 0.868,
+      "step": 22158
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.790283909764626,
+      "learning_rate": 9.771870515191773e-06,
+      "loss": 0.848,
+      "step": 22159
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4918115759270283,
+      "learning_rate": 9.771107673377106e-06,
+      "loss": 0.8625,
+      "step": 22160
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5911537796422262,
+      "learning_rate": 9.77034483289513e-06,
+      "loss": 0.8573,
+      "step": 22161
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8270134183859745,
+      "learning_rate": 9.769581993750277e-06,
+      "loss": 0.8024,
+      "step": 22162
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3423728867793425,
+      "learning_rate": 9.768819155946986e-06,
+      "loss": 0.7685,
+      "step": 22163
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5924118059837433,
+      "learning_rate": 9.768056319489711e-06,
+      "loss": 0.9186,
+      "step": 22164
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5888768493945316,
+      "learning_rate": 9.767293484382879e-06,
+      "loss": 0.82,
+      "step": 22165
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7852680629653007,
+      "learning_rate": 9.76653065063094e-06,
+      "loss": 0.7221,
+      "step": 22166
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7054151944738432,
+      "learning_rate": 9.76576781823833e-06,
+      "loss": 0.8906,
+      "step": 22167
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7300365122576569,
+      "learning_rate": 9.765004987209504e-06,
+      "loss": 0.8257,
+      "step": 22168
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5309153877816442,
+      "learning_rate": 9.764242157548884e-06,
+      "loss": 0.7619,
+      "step": 22169
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6952364317164763,
+      "learning_rate": 9.763479329260924e-06,
+      "loss": 0.8572,
+      "step": 22170
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6025335864546537,
+      "learning_rate": 9.762716502350058e-06,
+      "loss": 0.8978,
+      "step": 22171
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8445384091656958,
+      "learning_rate": 9.761953676820738e-06,
+      "loss": 0.7587,
+      "step": 22172
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6848131785472928,
+      "learning_rate": 9.761190852677397e-06,
+      "loss": 0.8365,
+      "step": 22173
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8303175271092351,
+      "learning_rate": 9.760428029924479e-06,
+      "loss": 1.0339,
+      "step": 22174
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3715592873591536,
+      "learning_rate": 9.759665208566425e-06,
+      "loss": 0.9048,
+      "step": 22175
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8009970242704723,
+      "learning_rate": 9.758902388607672e-06,
+      "loss": 0.7732,
+      "step": 22176
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.544165376581056,
+      "learning_rate": 9.75813957005267e-06,
+      "loss": 0.9439,
+      "step": 22177
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6865517039289595,
+      "learning_rate": 9.757376752905854e-06,
+      "loss": 0.8476,
+      "step": 22178
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6961345017812453,
+      "learning_rate": 9.756613937171669e-06,
+      "loss": 0.9244,
+      "step": 22179
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.816914637158217,
+      "learning_rate": 9.755851122854553e-06,
+      "loss": 0.7588,
+      "step": 22180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.248452267699518,
+      "learning_rate": 9.755088309958949e-06,
+      "loss": 0.7837,
+      "step": 22181
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.661157737601841,
+      "learning_rate": 9.7543254984893e-06,
+      "loss": 0.767,
+      "step": 22182
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8004116315639699,
+      "learning_rate": 9.753562688450047e-06,
+      "loss": 0.7637,
+      "step": 22183
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.609340272722338,
+      "learning_rate": 9.752799879845624e-06,
+      "loss": 0.8938,
+      "step": 22184
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.652523625512458,
+      "learning_rate": 9.752037072680484e-06,
+      "loss": 0.9007,
+      "step": 22185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6483492513192721,
+      "learning_rate": 9.751274266959064e-06,
+      "loss": 0.9156,
+      "step": 22186
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.50502009062703,
+      "learning_rate": 9.7505114626858e-06,
+      "loss": 0.7483,
+      "step": 22187
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.820409375307161,
+      "learning_rate": 9.749748659865136e-06,
+      "loss": 0.8237,
+      "step": 22188
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6386871539116588,
+      "learning_rate": 9.74898585850152e-06,
+      "loss": 0.9581,
+      "step": 22189
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.979557501968502,
+      "learning_rate": 9.748223058599383e-06,
+      "loss": 0.8545,
+      "step": 22190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5804877630815566,
+      "learning_rate": 9.747460260163173e-06,
+      "loss": 1.0682,
+      "step": 22191
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5822464280574526,
+      "learning_rate": 9.746697463197327e-06,
+      "loss": 0.8218,
+      "step": 22192
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.117486697777253,
+      "learning_rate": 9.745934667706291e-06,
+      "loss": 0.9296,
+      "step": 22193
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.778208724423302,
+      "learning_rate": 9.745171873694504e-06,
+      "loss": 0.887,
+      "step": 22194
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.880631312517983,
+      "learning_rate": 9.744409081166406e-06,
+      "loss": 0.8,
+      "step": 22195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.261324115928604,
+      "learning_rate": 9.743646290126444e-06,
+      "loss": 0.8148,
+      "step": 22196
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7260952495854411,
+      "learning_rate": 9.742883500579046e-06,
+      "loss": 0.7394,
+      "step": 22197
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6771891923515654,
+      "learning_rate": 9.742120712528665e-06,
+      "loss": 0.7988,
+      "step": 22198
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7257979929896436,
+      "learning_rate": 9.741357925979737e-06,
+      "loss": 1.0087,
+      "step": 22199
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6147759247250066,
+      "learning_rate": 9.740595140936714e-06,
+      "loss": 0.7838,
+      "step": 22200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.88487761203714,
+      "learning_rate": 9.73983235740402e-06,
+      "loss": 0.8792,
+      "step": 22201
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5665314386267353,
+      "learning_rate": 9.739069575386105e-06,
+      "loss": 0.9083,
+      "step": 22202
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0792303889790853,
+      "learning_rate": 9.738306794887413e-06,
+      "loss": 0.843,
+      "step": 22203
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6887979934574875,
+      "learning_rate": 9.73754401591238e-06,
+      "loss": 0.9197,
+      "step": 22204
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7667894912660032,
+      "learning_rate": 9.736781238465444e-06,
+      "loss": 0.7022,
+      "step": 22205
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.766734087781375,
+      "learning_rate": 9.736018462551057e-06,
+      "loss": 0.8685,
+      "step": 22206
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6559486748335326,
+      "learning_rate": 9.735255688173654e-06,
+      "loss": 0.8819,
+      "step": 22207
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5768575610100701,
+      "learning_rate": 9.734492915337675e-06,
+      "loss": 0.9377,
+      "step": 22208
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5505054808626604,
+      "learning_rate": 9.733730144047558e-06,
+      "loss": 0.8302,
+      "step": 22209
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5825033164598712,
+      "learning_rate": 9.732967374307755e-06,
+      "loss": 0.7424,
+      "step": 22210
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5946887172990167,
+      "learning_rate": 9.732204606122698e-06,
+      "loss": 0.784,
+      "step": 22211
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0822045140101633,
+      "learning_rate": 9.731441839496828e-06,
+      "loss": 0.8301,
+      "step": 22212
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.55570158221731,
+      "learning_rate": 9.73067907443459e-06,
+      "loss": 0.8734,
+      "step": 22213
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5457870187357017,
+      "learning_rate": 9.729916310940427e-06,
+      "loss": 0.8349,
+      "step": 22214
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5203299026311656,
+      "learning_rate": 9.729153549018775e-06,
+      "loss": 0.9749,
+      "step": 22215
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5856402405144354,
+      "learning_rate": 9.728390788674074e-06,
+      "loss": 0.819,
+      "step": 22216
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4692954616126053,
+      "learning_rate": 9.727628029910774e-06,
+      "loss": 0.7932,
+      "step": 22217
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6471212096177976,
+      "learning_rate": 9.726865272733304e-06,
+      "loss": 0.8301,
+      "step": 22218
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.339016400304775,
+      "learning_rate": 9.726102517146112e-06,
+      "loss": 0.8667,
+      "step": 22219
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.740189327270109,
+      "learning_rate": 9.725339763153636e-06,
+      "loss": 0.7519,
+      "step": 22220
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6422057318799692,
+      "learning_rate": 9.724577010760324e-06,
+      "loss": 0.8958,
+      "step": 22221
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.455594078458033,
+      "learning_rate": 9.723814259970608e-06,
+      "loss": 0.7857,
+      "step": 22222
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.575577051158319,
+      "learning_rate": 9.723051510788934e-06,
+      "loss": 1.0387,
+      "step": 22223
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7800182310269748,
+      "learning_rate": 9.722288763219744e-06,
+      "loss": 0.8658,
+      "step": 22224
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.895285033113373,
+      "learning_rate": 9.721526017267473e-06,
+      "loss": 0.9021,
+      "step": 22225
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.523078787926517,
+      "learning_rate": 9.720763272936568e-06,
+      "loss": 0.7956,
+      "step": 22226
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.883123041748739,
+      "learning_rate": 9.720000530231466e-06,
+      "loss": 0.92,
+      "step": 22227
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5207562094016596,
+      "learning_rate": 9.719237789156612e-06,
+      "loss": 0.8433,
+      "step": 22228
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4909930682649977,
+      "learning_rate": 9.718475049716441e-06,
+      "loss": 0.9145,
+      "step": 22229
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4673217519744246,
+      "learning_rate": 9.7177123119154e-06,
+      "loss": 0.8517,
+      "step": 22230
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7824270773140192,
+      "learning_rate": 9.71694957575793e-06,
+      "loss": 0.7931,
+      "step": 22231
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4955185715659618,
+      "learning_rate": 9.716186841248465e-06,
+      "loss": 0.796,
+      "step": 22232
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6854203899577425,
+      "learning_rate": 9.715424108391448e-06,
+      "loss": 0.8316,
+      "step": 22233
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7345457515184666,
+      "learning_rate": 9.714661377191324e-06,
+      "loss": 0.8151,
+      "step": 22234
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7648865492426682,
+      "learning_rate": 9.713898647652535e-06,
+      "loss": 0.7123,
+      "step": 22235
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.760187874030646,
+      "learning_rate": 9.713135919779516e-06,
+      "loss": 0.7471,
+      "step": 22236
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.649859587138505,
+      "learning_rate": 9.712373193576708e-06,
+      "loss": 0.8726,
+      "step": 22237
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.793577092536378,
+      "learning_rate": 9.71161046904856e-06,
+      "loss": 0.7898,
+      "step": 22238
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.052414993940084,
+      "learning_rate": 9.710847746199505e-06,
+      "loss": 0.7534,
+      "step": 22239
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9230796832184631,
+      "learning_rate": 9.710085025033984e-06,
+      "loss": 0.7701,
+      "step": 22240
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6353204633680547,
+      "learning_rate": 9.709322305556438e-06,
+      "loss": 0.9459,
+      "step": 22241
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.526205274506854,
+      "learning_rate": 9.708559587771316e-06,
+      "loss": 0.8448,
+      "step": 22242
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5462097511098907,
+      "learning_rate": 9.707796871683049e-06,
+      "loss": 0.9071,
+      "step": 22243
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6159657300604677,
+      "learning_rate": 9.70703415729608e-06,
+      "loss": 0.8675,
+      "step": 22244
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.629936322200382,
+      "learning_rate": 9.706271444614854e-06,
+      "loss": 0.9078,
+      "step": 22245
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7703832819501495,
+      "learning_rate": 9.705508733643805e-06,
+      "loss": 0.7471,
+      "step": 22246
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6760316494396286,
+      "learning_rate": 9.704746024387377e-06,
+      "loss": 0.8868,
+      "step": 22247
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.696468646879299,
+      "learning_rate": 9.703983316850012e-06,
+      "loss": 0.9301,
+      "step": 22248
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.75657388788681,
+      "learning_rate": 9.703220611036156e-06,
+      "loss": 0.9151,
+      "step": 22249
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8055128730990442,
+      "learning_rate": 9.702457906950235e-06,
+      "loss": 0.8326,
+      "step": 22250
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5704627420550183,
+      "learning_rate": 9.701695204596702e-06,
+      "loss": 0.9306,
+      "step": 22251
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5515488643447153,
+      "learning_rate": 9.700932503979996e-06,
+      "loss": 0.809,
+      "step": 22252
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4363271101491082,
+      "learning_rate": 9.700169805104553e-06,
+      "loss": 0.8639,
+      "step": 22253
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7850807222412011,
+      "learning_rate": 9.699407107974815e-06,
+      "loss": 0.6927,
+      "step": 22254
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.448791468852079,
+      "learning_rate": 9.698644412595225e-06,
+      "loss": 0.7975,
+      "step": 22255
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6021241462752174,
+      "learning_rate": 9.697881718970226e-06,
+      "loss": 0.9308,
+      "step": 22256
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5463910265770098,
+      "learning_rate": 9.697119027104252e-06,
+      "loss": 0.8927,
+      "step": 22257
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4312615331789997,
+      "learning_rate": 9.696356337001745e-06,
+      "loss": 0.8055,
+      "step": 22258
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.48368672804185,
+      "learning_rate": 9.695593648667152e-06,
+      "loss": 0.9296,
+      "step": 22259
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9820268697844368,
+      "learning_rate": 9.694830962104907e-06,
+      "loss": 0.8571,
+      "step": 22260
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5908847586161454,
+      "learning_rate": 9.694068277319451e-06,
+      "loss": 0.7899,
+      "step": 22261
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6767887661807976,
+      "learning_rate": 9.693305594315229e-06,
+      "loss": 0.9319,
+      "step": 22262
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.021937001502307,
+      "learning_rate": 9.69254291309668e-06,
+      "loss": 0.8313,
+      "step": 22263
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5820125014686501,
+      "learning_rate": 9.691780233668241e-06,
+      "loss": 0.8121,
+      "step": 22264
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5457425657172197,
+      "learning_rate": 9.691017556034354e-06,
+      "loss": 0.8587,
+      "step": 22265
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5904676760106606,
+      "learning_rate": 9.690254880199466e-06,
+      "loss": 0.9617,
+      "step": 22266
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.787779819976917,
+      "learning_rate": 9.689492206168006e-06,
+      "loss": 0.8674,
+      "step": 22267
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6323330095110342,
+      "learning_rate": 9.688729533944423e-06,
+      "loss": 0.9082,
+      "step": 22268
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8033992270350854,
+      "learning_rate": 9.687966863533154e-06,
+      "loss": 0.8122,
+      "step": 22269
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.996039505950336,
+      "learning_rate": 9.687204194938645e-06,
+      "loss": 0.7997,
+      "step": 22270
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.891905958561497,
+      "learning_rate": 9.686441528165327e-06,
+      "loss": 0.8911,
+      "step": 22271
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7871590498791214,
+      "learning_rate": 9.685678863217649e-06,
+      "loss": 0.8632,
+      "step": 22272
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7653506699701453,
+      "learning_rate": 9.68491620010005e-06,
+      "loss": 0.9456,
+      "step": 22273
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8446861095925469,
+      "learning_rate": 9.684153538816963e-06,
+      "loss": 0.9014,
+      "step": 22274
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2492501022185194,
+      "learning_rate": 9.683390879372837e-06,
+      "loss": 0.8234,
+      "step": 22275
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7749991298716639,
+      "learning_rate": 9.682628221772109e-06,
+      "loss": 0.8678,
+      "step": 22276
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.651965447128388,
+      "learning_rate": 9.681865566019224e-06,
+      "loss": 0.9452,
+      "step": 22277
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2974109088706833,
+      "learning_rate": 9.68110291211861e-06,
+      "loss": 0.8933,
+      "step": 22278
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6096607700390708,
+      "learning_rate": 9.680340260074721e-06,
+      "loss": 0.8966,
+      "step": 22279
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5763787053159135,
+      "learning_rate": 9.679577609891994e-06,
+      "loss": 0.9669,
+      "step": 22280
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4715063027406508,
+      "learning_rate": 9.678814961574866e-06,
+      "loss": 0.9243,
+      "step": 22281
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6479162330681723,
+      "learning_rate": 9.678052315127776e-06,
+      "loss": 0.8661,
+      "step": 22282
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8825132275719372,
+      "learning_rate": 9.67728967055517e-06,
+      "loss": 0.9002,
+      "step": 22283
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7186533611531294,
+      "learning_rate": 9.676527027861487e-06,
+      "loss": 0.8656,
+      "step": 22284
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6815903162831338,
+      "learning_rate": 9.675764387051164e-06,
+      "loss": 0.8128,
+      "step": 22285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.526180854848109,
+      "learning_rate": 9.675001748128641e-06,
+      "loss": 0.8128,
+      "step": 22286
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6174368209216095,
+      "learning_rate": 9.674239111098367e-06,
+      "loss": 0.786,
+      "step": 22287
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.616182662069702,
+      "learning_rate": 9.673476475964773e-06,
+      "loss": 0.8692,
+      "step": 22288
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.59851678327866,
+      "learning_rate": 9.6727138427323e-06,
+      "loss": 0.7897,
+      "step": 22289
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.474006207035954,
+      "learning_rate": 9.671951211405391e-06,
+      "loss": 0.7779,
+      "step": 22290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4143085630355405,
+      "learning_rate": 9.67118858198849e-06,
+      "loss": 0.7548,
+      "step": 22291
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4758583498113176,
+      "learning_rate": 9.67042595448603e-06,
+      "loss": 0.881,
+      "step": 22292
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.762327268505107,
+      "learning_rate": 9.669663328902455e-06,
+      "loss": 0.7539,
+      "step": 22293
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7650526474097377,
+      "learning_rate": 9.668900705242208e-06,
+      "loss": 0.9856,
+      "step": 22294
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9080859452593055,
+      "learning_rate": 9.66813808350972e-06,
+      "loss": 0.7894,
+      "step": 22295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7008265660890098,
+      "learning_rate": 9.66737546370944e-06,
+      "loss": 0.841,
+      "step": 22296
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8887091119258979,
+      "learning_rate": 9.666612845845803e-06,
+      "loss": 0.9179,
+      "step": 22297
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.714164831886229,
+      "learning_rate": 9.665850229923258e-06,
+      "loss": 0.9347,
+      "step": 22298
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5495849465725091,
+      "learning_rate": 9.665087615946231e-06,
+      "loss": 0.7697,
+      "step": 22299
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5075195702291282,
+      "learning_rate": 9.664325003919176e-06,
+      "loss": 0.9218,
+      "step": 22300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0438361673300127,
+      "learning_rate": 9.66356239384652e-06,
+      "loss": 0.8524,
+      "step": 22301
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.775793351005592,
+      "learning_rate": 9.66279978573272e-06,
+      "loss": 0.8803,
+      "step": 22302
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8852426071916437,
+      "learning_rate": 9.662037179582199e-06,
+      "loss": 0.8937,
+      "step": 22303
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.842237624416008,
+      "learning_rate": 9.661274575399406e-06,
+      "loss": 0.9354,
+      "step": 22304
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.655080125113747,
+      "learning_rate": 9.660511973188783e-06,
+      "loss": 0.7749,
+      "step": 22305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8699668920682677,
+      "learning_rate": 9.659749372954764e-06,
+      "loss": 0.936,
+      "step": 22306
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8088711460047762,
+      "learning_rate": 9.65898677470179e-06,
+      "loss": 0.791,
+      "step": 22307
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5448266737849572,
+      "learning_rate": 9.658224178434305e-06,
+      "loss": 0.8718,
+      "step": 22308
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.87399178899894,
+      "learning_rate": 9.657461584156748e-06,
+      "loss": 0.8107,
+      "step": 22309
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5599968521902816,
+      "learning_rate": 9.656698991873555e-06,
+      "loss": 0.8533,
+      "step": 22310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4890256350332696,
+      "learning_rate": 9.655936401589173e-06,
+      "loss": 0.8422,
+      "step": 22311
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.471973521662684,
+      "learning_rate": 9.655173813308037e-06,
+      "loss": 0.8264,
+      "step": 22312
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5898203163051392,
+      "learning_rate": 9.654411227034588e-06,
+      "loss": 0.9039,
+      "step": 22313
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6616129150022592,
+      "learning_rate": 9.653648642773264e-06,
+      "loss": 0.8801,
+      "step": 22314
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.608354632411386,
+      "learning_rate": 9.652886060528509e-06,
+      "loss": 0.8261,
+      "step": 22315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5384889144964284,
+      "learning_rate": 9.652123480304764e-06,
+      "loss": 0.8189,
+      "step": 22316
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6055350198245277,
+      "learning_rate": 9.651360902106463e-06,
+      "loss": 0.8445,
+      "step": 22317
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5753927607979108,
+      "learning_rate": 9.650598325938048e-06,
+      "loss": 1.0039,
+      "step": 22318
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5133296244496006,
+      "learning_rate": 9.649835751803965e-06,
+      "loss": 0.9052,
+      "step": 22319
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8455919350203012,
+      "learning_rate": 9.649073179708646e-06,
+      "loss": 0.8825,
+      "step": 22320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5198627278116459,
+      "learning_rate": 9.648310609656534e-06,
+      "loss": 0.9794,
+      "step": 22321
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6519944310549084,
+      "learning_rate": 9.647548041652066e-06,
+      "loss": 0.8737,
+      "step": 22322
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.361470403387042,
+      "learning_rate": 9.646785475699691e-06,
+      "loss": 0.8078,
+      "step": 22323
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.919550001381677,
+      "learning_rate": 9.64602291180384e-06,
+      "loss": 0.911,
+      "step": 22324
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8676974143539375,
+      "learning_rate": 9.645260349968955e-06,
+      "loss": 0.9612,
+      "step": 22325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6665066550295302,
+      "learning_rate": 9.644497790199478e-06,
+      "loss": 0.9023,
+      "step": 22326
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.99007702680479,
+      "learning_rate": 9.643735232499846e-06,
+      "loss": 0.8279,
+      "step": 22327
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.47513085665002,
+      "learning_rate": 9.6429726768745e-06,
+      "loss": 0.823,
+      "step": 22328
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3685930944807536,
+      "learning_rate": 9.642210123327882e-06,
+      "loss": 0.9569,
+      "step": 22329
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5540142773606564,
+      "learning_rate": 9.641447571864429e-06,
+      "loss": 0.7126,
+      "step": 22330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6005959907245693,
+      "learning_rate": 9.64068502248858e-06,
+      "loss": 0.933,
+      "step": 22331
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.69548961425907,
+      "learning_rate": 9.639922475204778e-06,
+      "loss": 0.801,
+      "step": 22332
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6057042898791956,
+      "learning_rate": 9.639159930017465e-06,
+      "loss": 0.9489,
+      "step": 22333
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9134515020378517,
+      "learning_rate": 9.638397386931072e-06,
+      "loss": 0.8731,
+      "step": 22334
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6327320248305446,
+      "learning_rate": 9.637634845950043e-06,
+      "loss": 0.8151,
+      "step": 22335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7796684127473049,
+      "learning_rate": 9.636872307078821e-06,
+      "loss": 0.7216,
+      "step": 22336
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5009390689844495,
+      "learning_rate": 9.636109770321846e-06,
+      "loss": 0.8601,
+      "step": 22337
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5530598263008262,
+      "learning_rate": 9.635347235683552e-06,
+      "loss": 1.0009,
+      "step": 22338
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6967282889842794,
+      "learning_rate": 9.634584703168381e-06,
+      "loss": 0.7928,
+      "step": 22339
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5138056113591296,
+      "learning_rate": 9.633822172780778e-06,
+      "loss": 0.7936,
+      "step": 22340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5341612308543575,
+      "learning_rate": 9.633059644525175e-06,
+      "loss": 0.8531,
+      "step": 22341
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5892100964628253,
+      "learning_rate": 9.632297118406015e-06,
+      "loss": 0.9048,
+      "step": 22342
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7988268872971527,
+      "learning_rate": 9.631534594427735e-06,
+      "loss": 0.9597,
+      "step": 22343
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6757157125084283,
+      "learning_rate": 9.630772072594784e-06,
+      "loss": 0.8487,
+      "step": 22344
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7730720219963516,
+      "learning_rate": 9.63000955291159e-06,
+      "loss": 0.9589,
+      "step": 22345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.743750007907229,
+      "learning_rate": 9.629247035382597e-06,
+      "loss": 0.8012,
+      "step": 22346
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.598726106667086,
+      "learning_rate": 9.62848452001225e-06,
+      "loss": 0.927,
+      "step": 22347
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.456231543248824,
+      "learning_rate": 9.627722006804977e-06,
+      "loss": 0.7941,
+      "step": 22348
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6050941613356056,
+      "learning_rate": 9.626959495765229e-06,
+      "loss": 0.8718,
+      "step": 22349
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7280239921770688,
+      "learning_rate": 9.626196986897437e-06,
+      "loss": 0.8087,
+      "step": 22350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7298372511809037,
+      "learning_rate": 9.62543448020605e-06,
+      "loss": 0.8621,
+      "step": 22351
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6411695805156803,
+      "learning_rate": 9.624671975695497e-06,
+      "loss": 0.8543,
+      "step": 22352
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7550350636221215,
+      "learning_rate": 9.623909473370225e-06,
+      "loss": 0.8768,
+      "step": 22353
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4583347496791472,
+      "learning_rate": 9.623146973234673e-06,
+      "loss": 0.8242,
+      "step": 22354
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7493476783783901,
+      "learning_rate": 9.622384475293278e-06,
+      "loss": 1.0209,
+      "step": 22355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9632454440533456,
+      "learning_rate": 9.621621979550475e-06,
+      "loss": 0.7947,
+      "step": 22356
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.533077095252002,
+      "learning_rate": 9.620859486010713e-06,
+      "loss": 0.9735,
+      "step": 22357
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.76332723984635,
+      "learning_rate": 9.620096994678429e-06,
+      "loss": 0.8742,
+      "step": 22358
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5816964352962677,
+      "learning_rate": 9.619334505558055e-06,
+      "loss": 0.9189,
+      "step": 22359
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3724895544726732,
+      "learning_rate": 9.618572018654039e-06,
+      "loss": 0.8603,
+      "step": 22360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5833152868622482,
+      "learning_rate": 9.61780953397082e-06,
+      "loss": 0.7919,
+      "step": 22361
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8494057523581957,
+      "learning_rate": 9.617047051512832e-06,
+      "loss": 0.8017,
+      "step": 22362
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5885017558519303,
+      "learning_rate": 9.616284571284515e-06,
+      "loss": 0.9651,
+      "step": 22363
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5767235115948852,
+      "learning_rate": 9.615522093290313e-06,
+      "loss": 0.8612,
+      "step": 22364
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7653445385607536,
+      "learning_rate": 9.614759617534665e-06,
+      "loss": 0.7092,
+      "step": 22365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6761793314390492,
+      "learning_rate": 9.613997144022006e-06,
+      "loss": 0.7886,
+      "step": 22366
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5482577741745038,
+      "learning_rate": 9.613234672756777e-06,
+      "loss": 0.9521,
+      "step": 22367
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6448684624714978,
+      "learning_rate": 9.612472203743423e-06,
+      "loss": 0.8561,
+      "step": 22368
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5524940647396532,
+      "learning_rate": 9.611709736986375e-06,
+      "loss": 0.8752,
+      "step": 22369
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6051929400587177,
+      "learning_rate": 9.610947272490077e-06,
+      "loss": 0.9832,
+      "step": 22370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7514772906575442,
+      "learning_rate": 9.610184810258964e-06,
+      "loss": 0.6954,
+      "step": 22371
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6713157264561371,
+      "learning_rate": 9.609422350297482e-06,
+      "loss": 0.9493,
+      "step": 22372
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4972857295334765,
+      "learning_rate": 9.608659892610067e-06,
+      "loss": 0.9012,
+      "step": 22373
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5527224503515102,
+      "learning_rate": 9.607897437201155e-06,
+      "loss": 0.7627,
+      "step": 22374
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4805862190371024,
+      "learning_rate": 9.607134984075193e-06,
+      "loss": 0.8767,
+      "step": 22375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7334734663877485,
+      "learning_rate": 9.60637253323661e-06,
+      "loss": 0.8288,
+      "step": 22376
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6959240882222124,
+      "learning_rate": 9.605610084689853e-06,
+      "loss": 0.8076,
+      "step": 22377
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5668596431058335,
+      "learning_rate": 9.604847638439358e-06,
+      "loss": 0.8318,
+      "step": 22378
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7753668495888347,
+      "learning_rate": 9.604085194489568e-06,
+      "loss": 0.7274,
+      "step": 22379
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.053758979075376,
+      "learning_rate": 9.603322752844917e-06,
+      "loss": 0.8015,
+      "step": 22380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6775990381148604,
+      "learning_rate": 9.602560313509846e-06,
+      "loss": 0.8773,
+      "step": 22381
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.493809416235298,
+      "learning_rate": 9.601797876488798e-06,
+      "loss": 0.8481,
+      "step": 22382
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7869374323426959,
+      "learning_rate": 9.601035441786206e-06,
+      "loss": 0.7786,
+      "step": 22383
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6917086032235973,
+      "learning_rate": 9.60027300940651e-06,
+      "loss": 1.0284,
+      "step": 22384
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6964625193785907,
+      "learning_rate": 9.599510579354153e-06,
+      "loss": 0.8301,
+      "step": 22385
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6491029991926653,
+      "learning_rate": 9.598748151633576e-06,
+      "loss": 0.8268,
+      "step": 22386
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7487197080301079,
+      "learning_rate": 9.597985726249211e-06,
+      "loss": 0.7462,
+      "step": 22387
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8947680495856851,
+      "learning_rate": 9.5972233032055e-06,
+      "loss": 0.8534,
+      "step": 22388
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7728911914721783,
+      "learning_rate": 9.596460882506885e-06,
+      "loss": 0.7806,
+      "step": 22389
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.540884100340958,
+      "learning_rate": 9.595698464157799e-06,
+      "loss": 0.9801,
+      "step": 22390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.630071045223335,
+      "learning_rate": 9.594936048162685e-06,
+      "loss": 0.7799,
+      "step": 22391
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.293536403708175,
+      "learning_rate": 9.59417363452598e-06,
+      "loss": 0.8083,
+      "step": 22392
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7314271758008268,
+      "learning_rate": 9.59341122325213e-06,
+      "loss": 0.8923,
+      "step": 22393
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7746242128346643,
+      "learning_rate": 9.592648814345565e-06,
+      "loss": 0.8011,
+      "step": 22394
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.741752766509117,
+      "learning_rate": 9.591886407810727e-06,
+      "loss": 0.8884,
+      "step": 22395
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5048817429037458,
+      "learning_rate": 9.591124003652062e-06,
+      "loss": 0.8582,
+      "step": 22396
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5435008798466245,
+      "learning_rate": 9.590361601873994e-06,
+      "loss": 0.8788,
+      "step": 22397
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5845412133701273,
+      "learning_rate": 9.589599202480974e-06,
+      "loss": 0.7993,
+      "step": 22398
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5470360448909874,
+      "learning_rate": 9.588836805477435e-06,
+      "loss": 0.9343,
+      "step": 22399
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4883233865434053,
+      "learning_rate": 9.588074410867825e-06,
+      "loss": 0.8528,
+      "step": 22400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7836267161706356,
+      "learning_rate": 9.587312018656568e-06,
+      "loss": 0.7625,
+      "step": 22401
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6677813896899731,
+      "learning_rate": 9.586549628848115e-06,
+      "loss": 0.9718,
+      "step": 22402
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6315775986354348,
+      "learning_rate": 9.585787241446903e-06,
+      "loss": 0.8604,
+      "step": 22403
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.568039811815096,
+      "learning_rate": 9.585024856457367e-06,
+      "loss": 0.9158,
+      "step": 22404
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5937616550028744,
+      "learning_rate": 9.584262473883945e-06,
+      "loss": 0.9201,
+      "step": 22405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8410354060135339,
+      "learning_rate": 9.58350009373108e-06,
+      "loss": 0.9069,
+      "step": 22406
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5620433852644362,
+      "learning_rate": 9.582737716003212e-06,
+      "loss": 0.7995,
+      "step": 22407
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7778373293103535,
+      "learning_rate": 9.581975340704772e-06,
+      "loss": 0.7542,
+      "step": 22408
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.48855950557866,
+      "learning_rate": 9.581212967840208e-06,
+      "loss": 0.8573,
+      "step": 22409
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9107998977147926,
+      "learning_rate": 9.580450597413955e-06,
+      "loss": 0.9466,
+      "step": 22410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5558140611517581,
+      "learning_rate": 9.57968822943045e-06,
+      "loss": 0.8702,
+      "step": 22411
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5772486141332138,
+      "learning_rate": 9.578925863894132e-06,
+      "loss": 0.8221,
+      "step": 22412
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5728937604710733,
+      "learning_rate": 9.578163500809443e-06,
+      "loss": 0.9269,
+      "step": 22413
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8209347465441426,
+      "learning_rate": 9.577401140180819e-06,
+      "loss": 0.9135,
+      "step": 22414
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5943060823645068,
+      "learning_rate": 9.5766387820127e-06,
+      "loss": 0.8315,
+      "step": 22415
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8057579637430867,
+      "learning_rate": 9.575876426309521e-06,
+      "loss": 0.85,
+      "step": 22416
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6013463526848308,
+      "learning_rate": 9.575114073075729e-06,
+      "loss": 0.8344,
+      "step": 22417
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6298236973697666,
+      "learning_rate": 9.574351722315754e-06,
+      "loss": 0.8573,
+      "step": 22418
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5712635179420544,
+      "learning_rate": 9.573589374034038e-06,
+      "loss": 0.9337,
+      "step": 22419
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6459938441372641,
+      "learning_rate": 9.572827028235018e-06,
+      "loss": 0.8792,
+      "step": 22420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7535494581893027,
+      "learning_rate": 9.572064684923138e-06,
+      "loss": 0.8439,
+      "step": 22421
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7699209395606954,
+      "learning_rate": 9.57130234410283e-06,
+      "loss": 0.8809,
+      "step": 22422
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.669954605123381,
+      "learning_rate": 9.570540005778539e-06,
+      "loss": 0.8496,
+      "step": 22423
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6280322925183102,
+      "learning_rate": 9.569777669954694e-06,
+      "loss": 0.9168,
+      "step": 22424
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7613970551199118,
+      "learning_rate": 9.569015336635746e-06,
+      "loss": 0.7189,
+      "step": 22425
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6725034119377278,
+      "learning_rate": 9.568253005826124e-06,
+      "loss": 0.7649,
+      "step": 22426
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5907880332753046,
+      "learning_rate": 9.56749067753027e-06,
+      "loss": 0.7672,
+      "step": 22427
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8882160427052863,
+      "learning_rate": 9.566728351752626e-06,
+      "loss": 0.8488,
+      "step": 22428
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.652200797699078,
+      "learning_rate": 9.56596602849762e-06,
+      "loss": 1.0362,
+      "step": 22429
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6621635311886316,
+      "learning_rate": 9.5652037077697e-06,
+      "loss": 0.9175,
+      "step": 22430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7712338155962726,
+      "learning_rate": 9.5644413895733e-06,
+      "loss": 0.7895,
+      "step": 22431
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4784695541830206,
+      "learning_rate": 9.563679073912866e-06,
+      "loss": 0.8653,
+      "step": 22432
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.721936814908092,
+      "learning_rate": 9.562916760792825e-06,
+      "loss": 0.8809,
+      "step": 22433
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6536229514133984,
+      "learning_rate": 9.562154450217623e-06,
+      "loss": 0.8021,
+      "step": 22434
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.656865584467021,
+      "learning_rate": 9.561392142191698e-06,
+      "loss": 0.7894,
+      "step": 22435
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7684626663297704,
+      "learning_rate": 9.560629836719485e-06,
+      "loss": 0.9495,
+      "step": 22436
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8834323308753007,
+      "learning_rate": 9.559867533805421e-06,
+      "loss": 0.7835,
+      "step": 22437
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8820223453975855,
+      "learning_rate": 9.55910523345395e-06,
+      "loss": 0.9594,
+      "step": 22438
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5952590465770504,
+      "learning_rate": 9.558342935669512e-06,
+      "loss": 0.9467,
+      "step": 22439
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5409028423125826,
+      "learning_rate": 9.557580640456538e-06,
+      "loss": 0.8227,
+      "step": 22440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8395178526092835,
+      "learning_rate": 9.556818347819467e-06,
+      "loss": 0.8285,
+      "step": 22441
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6460121850865488,
+      "learning_rate": 9.556056057762746e-06,
+      "loss": 0.9647,
+      "step": 22442
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.117719314567301,
+      "learning_rate": 9.555293770290802e-06,
+      "loss": 0.8891,
+      "step": 22443
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6553236656599948,
+      "learning_rate": 9.554531485408078e-06,
+      "loss": 0.9676,
+      "step": 22444
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.506862638869524,
+      "learning_rate": 9.553769203119014e-06,
+      "loss": 0.9081,
+      "step": 22445
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.582942492551136,
+      "learning_rate": 9.55300692342805e-06,
+      "loss": 0.8322,
+      "step": 22446
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7479690561936793,
+      "learning_rate": 9.552244646339619e-06,
+      "loss": 0.7369,
+      "step": 22447
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4469696768441787,
+      "learning_rate": 9.55148237185816e-06,
+      "loss": 0.8731,
+      "step": 22448
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4719207524272073,
+      "learning_rate": 9.550720099988117e-06,
+      "loss": 0.8382,
+      "step": 22449
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5746093304214623,
+      "learning_rate": 9.54995783073392e-06,
+      "loss": 0.9543,
+      "step": 22450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6708812125397792,
+      "learning_rate": 9.54919556410001e-06,
+      "loss": 0.8459,
+      "step": 22451
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.794398414030172,
+      "learning_rate": 9.548433300090826e-06,
+      "loss": 0.8257,
+      "step": 22452
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7243886289760837,
+      "learning_rate": 9.547671038710813e-06,
+      "loss": 0.8754,
+      "step": 22453
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9733573209838893,
+      "learning_rate": 9.546908779964394e-06,
+      "loss": 0.817,
+      "step": 22454
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8948587955248624,
+      "learning_rate": 9.546146523856021e-06,
+      "loss": 0.8924,
+      "step": 22455
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4718639937515945,
+      "learning_rate": 9.545384270390125e-06,
+      "loss": 0.8759,
+      "step": 22456
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.506783148994521,
+      "learning_rate": 9.544622019571144e-06,
+      "loss": 0.933,
+      "step": 22457
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5415150679675707,
+      "learning_rate": 9.543859771403518e-06,
+      "loss": 0.8117,
+      "step": 22458
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6324187287082426,
+      "learning_rate": 9.543097525891686e-06,
+      "loss": 0.96,
+      "step": 22459
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7547627933158496,
+      "learning_rate": 9.542335283040089e-06,
+      "loss": 0.8534,
+      "step": 22460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8547383069479377,
+      "learning_rate": 9.541573042853153e-06,
+      "loss": 0.9217,
+      "step": 22461
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8488762878951726,
+      "learning_rate": 9.540810805335327e-06,
+      "loss": 1.0077,
+      "step": 22462
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7115068302160497,
+      "learning_rate": 9.540048570491049e-06,
+      "loss": 0.7163,
+      "step": 22463
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7427937873080299,
+      "learning_rate": 9.539286338324753e-06,
+      "loss": 0.7157,
+      "step": 22464
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.73220389891557,
+      "learning_rate": 9.538524108840873e-06,
+      "loss": 0.7379,
+      "step": 22465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6867069524602667,
+      "learning_rate": 9.537761882043855e-06,
+      "loss": 0.8609,
+      "step": 22466
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9549457439760487,
+      "learning_rate": 9.536999657938135e-06,
+      "loss": 0.9341,
+      "step": 22467
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7300034457995332,
+      "learning_rate": 9.536237436528148e-06,
+      "loss": 0.7033,
+      "step": 22468
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7175761105468612,
+      "learning_rate": 9.535475217818331e-06,
+      "loss": 0.859,
+      "step": 22469
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8562534075437402,
+      "learning_rate": 9.534713001813131e-06,
+      "loss": 0.9155,
+      "step": 22470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.037467680184342,
+      "learning_rate": 9.533950788516974e-06,
+      "loss": 0.841,
+      "step": 22471
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5169081555151418,
+      "learning_rate": 9.533188577934304e-06,
+      "loss": 0.7017,
+      "step": 22472
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.694497874224283,
+      "learning_rate": 9.532426370069557e-06,
+      "loss": 0.8769,
+      "step": 22473
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5536182016874978,
+      "learning_rate": 9.531664164927175e-06,
+      "loss": 0.8661,
+      "step": 22474
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.880908440309415,
+      "learning_rate": 9.53090196251159e-06,
+      "loss": 0.8897,
+      "step": 22475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9104653960432556,
+      "learning_rate": 9.530139762827243e-06,
+      "loss": 0.847,
+      "step": 22476
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7538658016756241,
+      "learning_rate": 9.529377565878571e-06,
+      "loss": 0.8934,
+      "step": 22477
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.491432832069453,
+      "learning_rate": 9.52861537167001e-06,
+      "loss": 0.9934,
+      "step": 22478
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7532027368449006,
+      "learning_rate": 9.527853180206e-06,
+      "loss": 0.8435,
+      "step": 22479
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.619386242279798,
+      "learning_rate": 9.527090991490977e-06,
+      "loss": 1.0058,
+      "step": 22480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7468347193684108,
+      "learning_rate": 9.526328805529387e-06,
+      "loss": 0.7867,
+      "step": 22481
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5493121399191043,
+      "learning_rate": 9.525566622325653e-06,
+      "loss": 0.8039,
+      "step": 22482
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0557732928538597,
+      "learning_rate": 9.524804441884224e-06,
+      "loss": 0.7895,
+      "step": 22483
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5614374687783137,
+      "learning_rate": 9.524042264209535e-06,
+      "loss": 0.8456,
+      "step": 22484
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0723760242007687,
+      "learning_rate": 9.523280089306022e-06,
+      "loss": 0.7944,
+      "step": 22485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5721635642452945,
+      "learning_rate": 9.52251791717812e-06,
+      "loss": 0.8434,
+      "step": 22486
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7172852258876117,
+      "learning_rate": 9.52175574783027e-06,
+      "loss": 0.8878,
+      "step": 22487
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.662441556873438,
+      "learning_rate": 9.520993581266916e-06,
+      "loss": 0.8859,
+      "step": 22488
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5795599549898696,
+      "learning_rate": 9.520231417492484e-06,
+      "loss": 0.7543,
+      "step": 22489
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4804463516160231,
+      "learning_rate": 9.519469256511415e-06,
+      "loss": 0.8709,
+      "step": 22490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.626974449262683,
+      "learning_rate": 9.518707098328152e-06,
+      "loss": 0.8464,
+      "step": 22491
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7424052781652732,
+      "learning_rate": 9.517944942947126e-06,
+      "loss": 0.7043,
+      "step": 22492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6518612799724077,
+      "learning_rate": 9.517182790372776e-06,
+      "loss": 0.9535,
+      "step": 22493
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5812945893152661,
+      "learning_rate": 9.516420640609541e-06,
+      "loss": 0.9778,
+      "step": 22494
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7959574521619763,
+      "learning_rate": 9.515658493661863e-06,
+      "loss": 0.8483,
+      "step": 22495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3691958227625352,
+      "learning_rate": 9.51489634953417e-06,
+      "loss": 1.0328,
+      "step": 22496
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7883416256512213,
+      "learning_rate": 9.514134208230904e-06,
+      "loss": 0.7361,
+      "step": 22497
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5291790613296534,
+      "learning_rate": 9.513372069756508e-06,
+      "loss": 0.9866,
+      "step": 22498
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6175354028203592,
+      "learning_rate": 9.512609934115407e-06,
+      "loss": 0.8959,
+      "step": 22499
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0558129036851613,
+      "learning_rate": 9.511847801312048e-06,
+      "loss": 0.898,
+      "step": 22500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0121118269433573,
+      "learning_rate": 9.511085671350864e-06,
+      "loss": 0.8879,
+      "step": 22501
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5861789156711417,
+      "learning_rate": 9.510323544236298e-06,
+      "loss": 0.9296,
+      "step": 22502
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5229028458023415,
+      "learning_rate": 9.50956141997278e-06,
+      "loss": 0.8293,
+      "step": 22503
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7023486830506456,
+      "learning_rate": 9.508799298564753e-06,
+      "loss": 0.8987,
+      "step": 22504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4869492378703533,
+      "learning_rate": 9.508037180016652e-06,
+      "loss": 0.9793,
+      "step": 22505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.331644598221365,
+      "learning_rate": 9.50727506433291e-06,
+      "loss": 0.8541,
+      "step": 22506
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8364619774453503,
+      "learning_rate": 9.506512951517972e-06,
+      "loss": 0.8144,
+      "step": 22507
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.804272359479145,
+      "learning_rate": 9.50575084157627e-06,
+      "loss": 0.7798,
+      "step": 22508
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5857798941113872,
+      "learning_rate": 9.504988734512249e-06,
+      "loss": 0.8724,
+      "step": 22509
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6214076760291263,
+      "learning_rate": 9.504226630330333e-06,
+      "loss": 0.8237,
+      "step": 22510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.655352644973387,
+      "learning_rate": 9.503464529034967e-06,
+      "loss": 0.8427,
+      "step": 22511
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.758097716941435,
+      "learning_rate": 9.502702430630593e-06,
+      "loss": 0.8202,
+      "step": 22512
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5054901321824152,
+      "learning_rate": 9.50194033512164e-06,
+      "loss": 0.8587,
+      "step": 22513
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7544304243586455,
+      "learning_rate": 9.501178242512546e-06,
+      "loss": 0.9455,
+      "step": 22514
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7231682949634988,
+      "learning_rate": 9.500416152807752e-06,
+      "loss": 0.7685,
+      "step": 22515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.576831381239745,
+      "learning_rate": 9.499654066011696e-06,
+      "loss": 0.9845,
+      "step": 22516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7794363899078581,
+      "learning_rate": 9.498891982128809e-06,
+      "loss": 0.7662,
+      "step": 22517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7942903752491497,
+      "learning_rate": 9.49812990116353e-06,
+      "loss": 0.8683,
+      "step": 22518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4280982756659175,
+      "learning_rate": 9.497367823120302e-06,
+      "loss": 0.8645,
+      "step": 22519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.718981149204738,
+      "learning_rate": 9.496605748003556e-06,
+      "loss": 0.7266,
+      "step": 22520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7357387156477009,
+      "learning_rate": 9.49584367581773e-06,
+      "loss": 0.8109,
+      "step": 22521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7441103276238848,
+      "learning_rate": 9.495081606567261e-06,
+      "loss": 0.6335,
+      "step": 22522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6302615766673931,
+      "learning_rate": 9.49431954025659e-06,
+      "loss": 0.7768,
+      "step": 22523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4566605417694762,
+      "learning_rate": 9.493557476890146e-06,
+      "loss": 0.9186,
+      "step": 22524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6492130806320677,
+      "learning_rate": 9.492795416472374e-06,
+      "loss": 0.8882,
+      "step": 22525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5336053094878237,
+      "learning_rate": 9.49203335900771e-06,
+      "loss": 0.9319,
+      "step": 22526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9808662118172677,
+      "learning_rate": 9.491271304500581e-06,
+      "loss": 0.9058,
+      "step": 22527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7197070810464443,
+      "learning_rate": 9.490509252955437e-06,
+      "loss": 0.8229,
+      "step": 22528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6612827324759545,
+      "learning_rate": 9.489747204376707e-06,
+      "loss": 0.9482,
+      "step": 22529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4694536906105133,
+      "learning_rate": 9.488985158768836e-06,
+      "loss": 0.8595,
+      "step": 22530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6734929706753,
+      "learning_rate": 9.488223116136248e-06,
+      "loss": 0.7535,
+      "step": 22531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8477634365982827,
+      "learning_rate": 9.48746107648339e-06,
+      "loss": 0.8563,
+      "step": 22532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6433199208674711,
+      "learning_rate": 9.486699039814698e-06,
+      "loss": 0.7912,
+      "step": 22533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8350920823590754,
+      "learning_rate": 9.485937006134605e-06,
+      "loss": 0.7293,
+      "step": 22534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6184617673296724,
+      "learning_rate": 9.485174975447546e-06,
+      "loss": 0.7954,
+      "step": 22535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7679623592382813,
+      "learning_rate": 9.484412947757965e-06,
+      "loss": 0.6722,
+      "step": 22536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4730747047468022,
+      "learning_rate": 9.483650923070297e-06,
+      "loss": 0.9571,
+      "step": 22537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7587901767876875,
+      "learning_rate": 9.482888901388973e-06,
+      "loss": 0.7101,
+      "step": 22538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5336028993190813,
+      "learning_rate": 9.482126882718433e-06,
+      "loss": 0.8965,
+      "step": 22539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5575446523583187,
+      "learning_rate": 9.481364867063118e-06,
+      "loss": 0.8199,
+      "step": 22540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.532058391321206,
+      "learning_rate": 9.48060285442746e-06,
+      "loss": 0.947,
+      "step": 22541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7291235238033285,
+      "learning_rate": 9.479840844815892e-06,
+      "loss": 0.8644,
+      "step": 22542
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5491982113852318,
+      "learning_rate": 9.47907883823286e-06,
+      "loss": 0.7781,
+      "step": 22543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6483624899086728,
+      "learning_rate": 9.478316834682798e-06,
+      "loss": 0.9292,
+      "step": 22544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5931477398696492,
+      "learning_rate": 9.477554834170137e-06,
+      "loss": 0.8135,
+      "step": 22545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.855160364134732,
+      "learning_rate": 9.476792836699315e-06,
+      "loss": 0.8326,
+      "step": 22546
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4447098303100294,
+      "learning_rate": 9.476030842274773e-06,
+      "loss": 0.7761,
+      "step": 22547
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6709294715841394,
+      "learning_rate": 9.475268850900947e-06,
+      "loss": 0.7723,
+      "step": 22548
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7402839665657119,
+      "learning_rate": 9.47450686258227e-06,
+      "loss": 0.7127,
+      "step": 22549
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7339518360869042,
+      "learning_rate": 9.473744877323178e-06,
+      "loss": 0.658,
+      "step": 22550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.926178251954522,
+      "learning_rate": 9.472982895128116e-06,
+      "loss": 0.7588,
+      "step": 22551
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6283770921579586,
+      "learning_rate": 9.472220916001511e-06,
+      "loss": 0.7272,
+      "step": 22552
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6488785842506861,
+      "learning_rate": 9.471458939947804e-06,
+      "loss": 0.8217,
+      "step": 22553
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.975689033697665,
+      "learning_rate": 9.470696966971427e-06,
+      "loss": 0.8969,
+      "step": 22554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5235222821149876,
+      "learning_rate": 9.469934997076825e-06,
+      "loss": 0.8983,
+      "step": 22555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125140601171626,
+      "learning_rate": 9.469173030268424e-06,
+      "loss": 0.8032,
+      "step": 22556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7489142040234431,
+      "learning_rate": 9.46841106655067e-06,
+      "loss": 0.6906,
+      "step": 22557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5381426657138104,
+      "learning_rate": 9.467649105927995e-06,
+      "loss": 0.6658,
+      "step": 22558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4533044208952355,
+      "learning_rate": 9.466887148404831e-06,
+      "loss": 0.85,
+      "step": 22559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8648109982392058,
+      "learning_rate": 9.466125193985623e-06,
+      "loss": 0.8259,
+      "step": 22560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6686315913368597,
+      "learning_rate": 9.465363242674802e-06,
+      "loss": 0.9069,
+      "step": 22561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3097165448426362,
+      "learning_rate": 9.464601294476807e-06,
+      "loss": 0.799,
+      "step": 22562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.021259684056717,
+      "learning_rate": 9.463839349396069e-06,
+      "loss": 0.9359,
+      "step": 22563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9200801768641478,
+      "learning_rate": 9.463077407437028e-06,
+      "loss": 0.7716,
+      "step": 22564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4425763402174585,
+      "learning_rate": 9.462315468604126e-06,
+      "loss": 0.8142,
+      "step": 22565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0755444156175225,
+      "learning_rate": 9.461553532901789e-06,
+      "loss": 0.8549,
+      "step": 22566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4082537617694446,
+      "learning_rate": 9.460791600334457e-06,
+      "loss": 0.7876,
+      "step": 22567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6493630466284617,
+      "learning_rate": 9.460029670906568e-06,
+      "loss": 0.8424,
+      "step": 22568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4615458533693078,
+      "learning_rate": 9.45926774462256e-06,
+      "loss": 0.7981,
+      "step": 22569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7654390137235962,
+      "learning_rate": 9.458505821486864e-06,
+      "loss": 0.9462,
+      "step": 22570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6687316656989062,
+      "learning_rate": 9.457743901503917e-06,
+      "loss": 0.9706,
+      "step": 22571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.681902592294298,
+      "learning_rate": 9.45698198467816e-06,
+      "loss": 0.8889,
+      "step": 22572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.460855013847446,
+      "learning_rate": 9.456220071014025e-06,
+      "loss": 0.8765,
+      "step": 22573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5592608313475476,
+      "learning_rate": 9.455458160515948e-06,
+      "loss": 0.7999,
+      "step": 22574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6271000681598797,
+      "learning_rate": 9.454696253188365e-06,
+      "loss": 0.835,
+      "step": 22575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4368980737782178,
+      "learning_rate": 9.453934349035715e-06,
+      "loss": 0.899,
+      "step": 22576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.756503496790532,
+      "learning_rate": 9.453172448062433e-06,
+      "loss": 0.797,
+      "step": 22577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7563682395337683,
+      "learning_rate": 9.452410550272949e-06,
+      "loss": 0.8813,
+      "step": 22578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5732799229293435,
+      "learning_rate": 9.451648655671712e-06,
+      "loss": 0.967,
+      "step": 22579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7586218852737984,
+      "learning_rate": 9.450886764263144e-06,
+      "loss": 0.723,
+      "step": 22580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.775795572399488,
+      "learning_rate": 9.450124876051688e-06,
+      "loss": 0.7802,
+      "step": 22581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5628473040603525,
+      "learning_rate": 9.449362991041778e-06,
+      "loss": 0.8545,
+      "step": 22582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1985862992764367,
+      "learning_rate": 9.448601109237859e-06,
+      "loss": 0.8183,
+      "step": 22583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4228555486624224,
+      "learning_rate": 9.447839230644349e-06,
+      "loss": 0.8015,
+      "step": 22584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.627297139036074,
+      "learning_rate": 9.447077355265698e-06,
+      "loss": 0.8416,
+      "step": 22585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.611627213922212,
+      "learning_rate": 9.44631548310634e-06,
+      "loss": 0.8248,
+      "step": 22586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.63534035054456,
+      "learning_rate": 9.445553614170705e-06,
+      "loss": 0.9169,
+      "step": 22587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.592306359935706,
+      "learning_rate": 9.444791748463232e-06,
+      "loss": 1.013,
+      "step": 22588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6765985518789517,
+      "learning_rate": 9.444029885988359e-06,
+      "loss": 0.8753,
+      "step": 22589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7998056233797083,
+      "learning_rate": 9.443268026750521e-06,
+      "loss": 0.7126,
+      "step": 22590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.677535384871343,
+      "learning_rate": 9.442506170754149e-06,
+      "loss": 0.7938,
+      "step": 22591
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.988932580717588,
+      "learning_rate": 9.441744318003685e-06,
+      "loss": 0.8355,
+      "step": 22592
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7172935800587097,
+      "learning_rate": 9.440982468503563e-06,
+      "loss": 0.884,
+      "step": 22593
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8145763101616035,
+      "learning_rate": 9.440220622258217e-06,
+      "loss": 0.7877,
+      "step": 22594
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8202679486601685,
+      "learning_rate": 9.439458779272082e-06,
+      "loss": 0.7756,
+      "step": 22595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5371549792610513,
+      "learning_rate": 9.438696939549598e-06,
+      "loss": 0.8941,
+      "step": 22596
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5487979222127268,
+      "learning_rate": 9.437935103095201e-06,
+      "loss": 0.8343,
+      "step": 22597
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7643617988764224,
+      "learning_rate": 9.43717326991332e-06,
+      "loss": 0.6992,
+      "step": 22598
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6114448846793517,
+      "learning_rate": 9.436411440008392e-06,
+      "loss": 0.8855,
+      "step": 22599
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6827986621990607,
+      "learning_rate": 9.435649613384859e-06,
+      "loss": 0.8361,
+      "step": 22600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.791235067587359,
+      "learning_rate": 9.434887790047152e-06,
+      "loss": 0.9115,
+      "step": 22601
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.53465495210617,
+      "learning_rate": 9.434125969999707e-06,
+      "loss": 0.773,
+      "step": 22602
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5833129850828511,
+      "learning_rate": 9.433364153246958e-06,
+      "loss": 0.8255,
+      "step": 22603
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.885741475516056,
+      "learning_rate": 9.432602339793346e-06,
+      "loss": 0.902,
+      "step": 22604
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.863820599265244,
+      "learning_rate": 9.4318405296433e-06,
+      "loss": 0.8392,
+      "step": 22605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4416947201939512,
+      "learning_rate": 9.43107872280126e-06,
+      "loss": 0.7386,
+      "step": 22606
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8065876016536149,
+      "learning_rate": 9.43031691927166e-06,
+      "loss": 0.8753,
+      "step": 22607
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5803392652607664,
+      "learning_rate": 9.429555119058931e-06,
+      "loss": 0.8898,
+      "step": 22608
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6415139695774565,
+      "learning_rate": 9.428793322167516e-06,
+      "loss": 0.8525,
+      "step": 22609
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4736774253680653,
+      "learning_rate": 9.428031528601846e-06,
+      "loss": 1.0013,
+      "step": 22610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6100652770165569,
+      "learning_rate": 9.427269738366361e-06,
+      "loss": 0.9259,
+      "step": 22611
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7900662530999858,
+      "learning_rate": 9.426507951465487e-06,
+      "loss": 0.8214,
+      "step": 22612
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5265207935249856,
+      "learning_rate": 9.42574616790367e-06,
+      "loss": 0.8724,
+      "step": 22613
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7113600698044131,
+      "learning_rate": 9.42498438768534e-06,
+      "loss": 0.8863,
+      "step": 22614
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6041897244070005,
+      "learning_rate": 9.424222610814932e-06,
+      "loss": 0.7577,
+      "step": 22615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.485399507432066,
+      "learning_rate": 9.423460837296879e-06,
+      "loss": 0.939,
+      "step": 22616
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6319277123392677,
+      "learning_rate": 9.422699067135624e-06,
+      "loss": 0.9291,
+      "step": 22617
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6623273558343383,
+      "learning_rate": 9.4219373003356e-06,
+      "loss": 0.7875,
+      "step": 22618
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5323500163727624,
+      "learning_rate": 9.421175536901234e-06,
+      "loss": 0.9239,
+      "step": 22619
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6210265222865856,
+      "learning_rate": 9.420413776836968e-06,
+      "loss": 0.8821,
+      "step": 22620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6921329269097007,
+      "learning_rate": 9.41965202014724e-06,
+      "loss": 0.9781,
+      "step": 22621
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6100231394063207,
+      "learning_rate": 9.41889026683648e-06,
+      "loss": 0.9488,
+      "step": 22622
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7944017695076615,
+      "learning_rate": 9.418128516909126e-06,
+      "loss": 0.8062,
+      "step": 22623
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8074928023434169,
+      "learning_rate": 9.417366770369609e-06,
+      "loss": 0.7307,
+      "step": 22624
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.643903453954868,
+      "learning_rate": 9.41660502722237e-06,
+      "loss": 0.815,
+      "step": 22625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.614302278773522,
+      "learning_rate": 9.41584328747184e-06,
+      "loss": 0.9453,
+      "step": 22626
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4851682397080892,
+      "learning_rate": 9.415081551122453e-06,
+      "loss": 0.8604,
+      "step": 22627
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.816728369545545,
+      "learning_rate": 9.414319818178654e-06,
+      "loss": 1.003,
+      "step": 22628
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7770260011747802,
+      "learning_rate": 9.413558088644864e-06,
+      "loss": 0.669,
+      "step": 22629
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7242996917505822,
+      "learning_rate": 9.412796362525526e-06,
+      "loss": 0.8174,
+      "step": 22630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7815753285104787,
+      "learning_rate": 9.41203463982507e-06,
+      "loss": 0.713,
+      "step": 22631
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6705792062959819,
+      "learning_rate": 9.411272920547942e-06,
+      "loss": 0.8981,
+      "step": 22632
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5413470911218825,
+      "learning_rate": 9.410511204698562e-06,
+      "loss": 0.9084,
+      "step": 22633
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.405024950167107,
+      "learning_rate": 9.409749492281378e-06,
+      "loss": 0.9222,
+      "step": 22634
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0458995535717746,
+      "learning_rate": 9.408987783300818e-06,
+      "loss": 0.8149,
+      "step": 22635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8657163065116171,
+      "learning_rate": 9.408226077761316e-06,
+      "loss": 0.8947,
+      "step": 22636
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.714621104255463,
+      "learning_rate": 9.407464375667309e-06,
+      "loss": 0.8587,
+      "step": 22637
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6904206700782813,
+      "learning_rate": 9.406702677023233e-06,
+      "loss": 0.8883,
+      "step": 22638
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6830601652729562,
+      "learning_rate": 9.405940981833525e-06,
+      "loss": 0.975,
+      "step": 22639
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.762620947763751,
+      "learning_rate": 9.405179290102611e-06,
+      "loss": 0.8534,
+      "step": 22640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6663263238462027,
+      "learning_rate": 9.404417601834935e-06,
+      "loss": 0.9515,
+      "step": 22641
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.846945823486774,
+      "learning_rate": 9.40365591703493e-06,
+      "loss": 0.9533,
+      "step": 22642
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5817496463403602,
+      "learning_rate": 9.402894235707026e-06,
+      "loss": 0.8074,
+      "step": 22643
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8270425856480774,
+      "learning_rate": 9.402132557855659e-06,
+      "loss": 0.9323,
+      "step": 22644
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8998488835904281,
+      "learning_rate": 9.401370883485268e-06,
+      "loss": 0.8354,
+      "step": 22645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5753206672767237,
+      "learning_rate": 9.400609212600288e-06,
+      "loss": 0.8387,
+      "step": 22646
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6965856471967624,
+      "learning_rate": 9.399847545205148e-06,
+      "loss": 0.9841,
+      "step": 22647
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7255084972020713,
+      "learning_rate": 9.399085881304283e-06,
+      "loss": 0.9052,
+      "step": 22648
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5197686240417105,
+      "learning_rate": 9.398324220902135e-06,
+      "loss": 0.8249,
+      "step": 22649
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8130281237950263,
+      "learning_rate": 9.397562564003131e-06,
+      "loss": 0.7293,
+      "step": 22650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9698376192233091,
+      "learning_rate": 9.39680091061171e-06,
+      "loss": 0.8315,
+      "step": 22651
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5712691908933274,
+      "learning_rate": 9.396039260732302e-06,
+      "loss": 0.8418,
+      "step": 22652
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.560326645512282,
+      "learning_rate": 9.39527761436935e-06,
+      "loss": 0.7849,
+      "step": 22653
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.979758415331746,
+      "learning_rate": 9.39451597152728e-06,
+      "loss": 0.9329,
+      "step": 22654
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6353594255097472,
+      "learning_rate": 9.393754332210529e-06,
+      "loss": 0.8562,
+      "step": 22655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.519978176437285,
+      "learning_rate": 9.392992696423536e-06,
+      "loss": 0.7458,
+      "step": 22656
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.460629125654316,
+      "learning_rate": 9.392231064170728e-06,
+      "loss": 0.8821,
+      "step": 22657
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4726172937782447,
+      "learning_rate": 9.391469435456543e-06,
+      "loss": 0.8399,
+      "step": 22658
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7687411689835607,
+      "learning_rate": 9.390707810285416e-06,
+      "loss": 0.7216,
+      "step": 22659
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6345783042682436,
+      "learning_rate": 9.389946188661785e-06,
+      "loss": 0.9366,
+      "step": 22660
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9042310600950725,
+      "learning_rate": 9.389184570590075e-06,
+      "loss": 0.9114,
+      "step": 22661
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5700485139706866,
+      "learning_rate": 9.388422956074727e-06,
+      "loss": 0.7822,
+      "step": 22662
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9908280406274639,
+      "learning_rate": 9.387661345120178e-06,
+      "loss": 0.8348,
+      "step": 22663
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.13536405437362,
+      "learning_rate": 9.386899737730854e-06,
+      "loss": 0.7884,
+      "step": 22664
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.571552098637546,
+      "learning_rate": 9.386138133911193e-06,
+      "loss": 0.8755,
+      "step": 22665
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4705971609251474,
+      "learning_rate": 9.385376533665633e-06,
+      "loss": 0.7493,
+      "step": 22666
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7197302478129903,
+      "learning_rate": 9.384614936998607e-06,
+      "loss": 0.8588,
+      "step": 22667
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.788226089155665,
+      "learning_rate": 9.383853343914546e-06,
+      "loss": 0.9025,
+      "step": 22668
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5827327569616414,
+      "learning_rate": 9.383091754417884e-06,
+      "loss": 0.9539,
+      "step": 22669
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7690384988566434,
+      "learning_rate": 9.382330168513058e-06,
+      "loss": 0.9206,
+      "step": 22670
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4863663355701968,
+      "learning_rate": 9.381568586204504e-06,
+      "loss": 0.9174,
+      "step": 22671
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7550234685358763,
+      "learning_rate": 9.380807007496652e-06,
+      "loss": 0.6699,
+      "step": 22672
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7000856255626546,
+      "learning_rate": 9.380045432393934e-06,
+      "loss": 0.8501,
+      "step": 22673
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8137620761714583,
+      "learning_rate": 9.379283860900793e-06,
+      "loss": 0.7018,
+      "step": 22674
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.720967957436107,
+      "learning_rate": 9.378522293021656e-06,
+      "loss": 0.9333,
+      "step": 22675
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8236805400882348,
+      "learning_rate": 9.377760728760957e-06,
+      "loss": 0.9306,
+      "step": 22676
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.57386173766932,
+      "learning_rate": 9.376999168123134e-06,
+      "loss": 0.9097,
+      "step": 22677
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.761049534753989,
+      "learning_rate": 9.37623761111262e-06,
+      "loss": 0.878,
+      "step": 22678
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7323146874752696,
+      "learning_rate": 9.375476057733847e-06,
+      "loss": 0.7554,
+      "step": 22679
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6158165795784736,
+      "learning_rate": 9.374714507991248e-06,
+      "loss": 0.8929,
+      "step": 22680
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.597134169570352,
+      "learning_rate": 9.373952961889265e-06,
+      "loss": 0.8153,
+      "step": 22681
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.729857105179001,
+      "learning_rate": 9.373191419432319e-06,
+      "loss": 0.9214,
+      "step": 22682
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7334105125866313,
+      "learning_rate": 9.372429880624856e-06,
+      "loss": 0.8679,
+      "step": 22683
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4145360926305615,
+      "learning_rate": 9.3716683454713e-06,
+      "loss": 0.8071,
+      "step": 22684
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6318325453643938,
+      "learning_rate": 9.370906813976097e-06,
+      "loss": 0.8413,
+      "step": 22685
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4902390660851752,
+      "learning_rate": 9.370145286143667e-06,
+      "loss": 0.8533,
+      "step": 22686
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6343736325286975,
+      "learning_rate": 9.369383761978453e-06,
+      "loss": 0.8065,
+      "step": 22687
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.462920097648161,
+      "learning_rate": 9.36862224148489e-06,
+      "loss": 0.7851,
+      "step": 22688
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6620332887524256,
+      "learning_rate": 9.367860724667401e-06,
+      "loss": 0.928,
+      "step": 22689
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5272341333914694,
+      "learning_rate": 9.367099211530432e-06,
+      "loss": 0.9339,
+      "step": 22690
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7518877359062444,
+      "learning_rate": 9.36633770207841e-06,
+      "loss": 0.8603,
+      "step": 22691
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.516650941131488,
+      "learning_rate": 9.365576196315772e-06,
+      "loss": 0.8151,
+      "step": 22692
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.122994112499247,
+      "learning_rate": 9.364814694246947e-06,
+      "loss": 0.8757,
+      "step": 22693
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6085904526070247,
+      "learning_rate": 9.364053195876375e-06,
+      "loss": 0.8945,
+      "step": 22694
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.494149030906431,
+      "learning_rate": 9.363291701208488e-06,
+      "loss": 0.8965,
+      "step": 22695
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5304440597298916,
+      "learning_rate": 9.362530210247715e-06,
+      "loss": 0.8104,
+      "step": 22696
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4670080533405743,
+      "learning_rate": 9.361768722998494e-06,
+      "loss": 0.9299,
+      "step": 22697
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.550441997005649,
+      "learning_rate": 9.361007239465257e-06,
+      "loss": 0.9247,
+      "step": 22698
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7183905214060067,
+      "learning_rate": 9.36024575965244e-06,
+      "loss": 0.952,
+      "step": 22699
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6731772151747824,
+      "learning_rate": 9.359484283564474e-06,
+      "loss": 0.9092,
+      "step": 22700
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.876229397881864,
+      "learning_rate": 9.35872281120579e-06,
+      "loss": 0.7542,
+      "step": 22701
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.253494449265299,
+      "learning_rate": 9.35796134258083e-06,
+      "loss": 0.9068,
+      "step": 22702
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6199386059682894,
+      "learning_rate": 9.35719987769402e-06,
+      "loss": 0.9578,
+      "step": 22703
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5472140632219045,
+      "learning_rate": 9.356438416549795e-06,
+      "loss": 0.992,
+      "step": 22704
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.541798779814976,
+      "learning_rate": 9.355676959152588e-06,
+      "loss": 0.9524,
+      "step": 22705
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.541668744682956,
+      "learning_rate": 9.354915505506839e-06,
+      "loss": 0.9055,
+      "step": 22706
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7653733978567492,
+      "learning_rate": 9.354154055616971e-06,
+      "loss": 0.8961,
+      "step": 22707
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6345324671905503,
+      "learning_rate": 9.353392609487427e-06,
+      "loss": 0.8466,
+      "step": 22708
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9336002513054924,
+      "learning_rate": 9.352631167122635e-06,
+      "loss": 0.9793,
+      "step": 22709
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6088277311575074,
+      "learning_rate": 9.351869728527024e-06,
+      "loss": 0.8018,
+      "step": 22710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6307969197717833,
+      "learning_rate": 9.351108293705037e-06,
+      "loss": 0.8009,
+      "step": 22711
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5284685392451782,
+      "learning_rate": 9.3503468626611e-06,
+      "loss": 0.9343,
+      "step": 22712
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5492483335827416,
+      "learning_rate": 9.349585435399656e-06,
+      "loss": 0.9458,
+      "step": 22713
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3150231277430477,
+      "learning_rate": 9.348824011925124e-06,
+      "loss": 0.8887,
+      "step": 22714
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5507825872566539,
+      "learning_rate": 9.34806259224195e-06,
+      "loss": 0.8311,
+      "step": 22715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6959478170628597,
+      "learning_rate": 9.34730117635456e-06,
+      "loss": 0.8929,
+      "step": 22716
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.552164124755709,
+      "learning_rate": 9.346539764267389e-06,
+      "loss": 0.7631,
+      "step": 22717
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6358572165766994,
+      "learning_rate": 9.345778355984869e-06,
+      "loss": 0.8444,
+      "step": 22718
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6931417013463563,
+      "learning_rate": 9.345016951511436e-06,
+      "loss": 0.8666,
+      "step": 22719
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7643106355445731,
+      "learning_rate": 9.344255550851523e-06,
+      "loss": 0.7287,
+      "step": 22720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.186177807571048,
+      "learning_rate": 9.34349415400956e-06,
+      "loss": 0.8233,
+      "step": 22721
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0105460816373553,
+      "learning_rate": 9.34273276098998e-06,
+      "loss": 0.9593,
+      "step": 22722
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8911355257942155,
+      "learning_rate": 9.341971371797223e-06,
+      "loss": 0.9624,
+      "step": 22723
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4920474140833877,
+      "learning_rate": 9.341209986435715e-06,
+      "loss": 0.8911,
+      "step": 22724
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7315599731607617,
+      "learning_rate": 9.340448604909888e-06,
+      "loss": 0.6644,
+      "step": 22725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.592937276906333,
+      "learning_rate": 9.339687227224178e-06,
+      "loss": 0.9114,
+      "step": 22726
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.620300863513551,
+      "learning_rate": 9.338925853383024e-06,
+      "loss": 0.8325,
+      "step": 22727
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8114792548205383,
+      "learning_rate": 9.338164483390849e-06,
+      "loss": 0.6971,
+      "step": 22728
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4731497760542784,
+      "learning_rate": 9.337403117252089e-06,
+      "loss": 0.7093,
+      "step": 22729
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9638769906760591,
+      "learning_rate": 9.336641754971183e-06,
+      "loss": 0.8606,
+      "step": 22730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.723596695269634,
+      "learning_rate": 9.335880396552552e-06,
+      "loss": 0.8347,
+      "step": 22731
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7813130338846155,
+      "learning_rate": 9.335119042000637e-06,
+      "loss": 0.8151,
+      "step": 22732
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9329707763194726,
+      "learning_rate": 9.33435769131987e-06,
+      "loss": 0.7987,
+      "step": 22733
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7493901637228266,
+      "learning_rate": 9.333596344514689e-06,
+      "loss": 0.7813,
+      "step": 22734
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5949324508449372,
+      "learning_rate": 9.332835001589514e-06,
+      "loss": 0.9221,
+      "step": 22735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7799928281285319,
+      "learning_rate": 9.332073662548785e-06,
+      "loss": 0.7494,
+      "step": 22736
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.599094043368221,
+      "learning_rate": 9.33131232739694e-06,
+      "loss": 0.9462,
+      "step": 22737
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.840179710676346,
+      "learning_rate": 9.3305509961384e-06,
+      "loss": 0.7699,
+      "step": 22738
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4486128731171393,
+      "learning_rate": 9.329789668777606e-06,
+      "loss": 0.9346,
+      "step": 22739
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5750979166079744,
+      "learning_rate": 9.32902834531899e-06,
+      "loss": 0.8889,
+      "step": 22740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4953952437855949,
+      "learning_rate": 9.328267025766984e-06,
+      "loss": 0.9338,
+      "step": 22741
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4333602366275209,
+      "learning_rate": 9.327505710126018e-06,
+      "loss": 0.8313,
+      "step": 22742
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7483416464614716,
+      "learning_rate": 9.326744398400527e-06,
+      "loss": 0.8946,
+      "step": 22743
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5700976031377816,
+      "learning_rate": 9.325983090594946e-06,
+      "loss": 0.7534,
+      "step": 22744
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6502151891720125,
+      "learning_rate": 9.325221786713704e-06,
+      "loss": 0.8606,
+      "step": 22745
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5779623125762228,
+      "learning_rate": 9.324460486761231e-06,
+      "loss": 0.723,
+      "step": 22746
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6609932240504115,
+      "learning_rate": 9.323699190741964e-06,
+      "loss": 0.8435,
+      "step": 22747
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4014352519007456,
+      "learning_rate": 9.322937898660339e-06,
+      "loss": 0.7752,
+      "step": 22748
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.785498312674068,
+      "learning_rate": 9.32217661052078e-06,
+      "loss": 0.8594,
+      "step": 22749
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.751698266006535,
+      "learning_rate": 9.321415326327721e-06,
+      "loss": 0.7116,
+      "step": 22750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.013703956950767,
+      "learning_rate": 9.320654046085602e-06,
+      "loss": 0.7196,
+      "step": 22751
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7254899503371217,
+      "learning_rate": 9.319892769798848e-06,
+      "loss": 0.8766,
+      "step": 22752
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.708428459400726,
+      "learning_rate": 9.319131497471894e-06,
+      "loss": 0.8636,
+      "step": 22753
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5149719677985836,
+      "learning_rate": 9.318370229109168e-06,
+      "loss": 0.8213,
+      "step": 22754
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5475054323249606,
+      "learning_rate": 9.317608964715113e-06,
+      "loss": 0.8788,
+      "step": 22755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5761180793020704,
+      "learning_rate": 9.31684770429415e-06,
+      "loss": 0.8825,
+      "step": 22756
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9663991123095212,
+      "learning_rate": 9.316086447850718e-06,
+      "loss": 0.8392,
+      "step": 22757
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7235679719969983,
+      "learning_rate": 9.31532519538925e-06,
+      "loss": 0.7738,
+      "step": 22758
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9847950946525705,
+      "learning_rate": 9.314563946914169e-06,
+      "loss": 0.9602,
+      "step": 22759
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5280737209498492,
+      "learning_rate": 9.313802702429918e-06,
+      "loss": 0.8964,
+      "step": 22760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4460936141756275,
+      "learning_rate": 9.31304146194092e-06,
+      "loss": 0.7961,
+      "step": 22761
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7479166841696506,
+      "learning_rate": 9.31228022545162e-06,
+      "loss": 0.7082,
+      "step": 22762
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6684898164304738,
+      "learning_rate": 9.311518992966436e-06,
+      "loss": 0.8238,
+      "step": 22763
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0158320114331083,
+      "learning_rate": 9.310757764489808e-06,
+      "loss": 0.8953,
+      "step": 22764
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8311897452931207,
+      "learning_rate": 9.309996540026169e-06,
+      "loss": 0.8737,
+      "step": 22765
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7189330741627131,
+      "learning_rate": 9.309235319579947e-06,
+      "loss": 0.8353,
+      "step": 22766
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6420921532112462,
+      "learning_rate": 9.308474103155573e-06,
+      "loss": 0.8154,
+      "step": 22767
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8267401108250658,
+      "learning_rate": 9.307712890757483e-06,
+      "loss": 0.8403,
+      "step": 22768
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7084032379238321,
+      "learning_rate": 9.306951682390112e-06,
+      "loss": 0.73,
+      "step": 22769
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0030096983371837,
+      "learning_rate": 9.306190478057884e-06,
+      "loss": 0.8754,
+      "step": 22770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5316466642520095,
+      "learning_rate": 9.305429277765233e-06,
+      "loss": 0.9016,
+      "step": 22771
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6289826413879076,
+      "learning_rate": 9.304668081516598e-06,
+      "loss": 0.8635,
+      "step": 22772
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6779702104651284,
+      "learning_rate": 9.3039068893164e-06,
+      "loss": 0.7757,
+      "step": 22773
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6778093428364405,
+      "learning_rate": 9.303145701169077e-06,
+      "loss": 0.9954,
+      "step": 22774
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4471118995336867,
+      "learning_rate": 9.302384517079061e-06,
+      "loss": 0.8229,
+      "step": 22775
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3672270201806547,
+      "learning_rate": 9.301623337050788e-06,
+      "loss": 0.7202,
+      "step": 22776
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5695959999848126,
+      "learning_rate": 9.30086216108868e-06,
+      "loss": 0.8765,
+      "step": 22777
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5368137165211058,
+      "learning_rate": 9.300100989197174e-06,
+      "loss": 0.8486,
+      "step": 22778
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6020645160560072,
+      "learning_rate": 9.299339821380708e-06,
+      "loss": 0.9541,
+      "step": 22779
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.901090747190357,
+      "learning_rate": 9.2985786576437e-06,
+      "loss": 0.6823,
+      "step": 22780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.599615175998798,
+      "learning_rate": 9.29781749799059e-06,
+      "loss": 0.8241,
+      "step": 22781
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5525486698602013,
+      "learning_rate": 9.297056342425807e-06,
+      "loss": 0.9251,
+      "step": 22782
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.800494734394771,
+      "learning_rate": 9.296295190953792e-06,
+      "loss": 0.8841,
+      "step": 22783
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4948095509111594,
+      "learning_rate": 9.295534043578963e-06,
+      "loss": 0.7446,
+      "step": 22784
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5872963923805141,
+      "learning_rate": 9.29477290030576e-06,
+      "loss": 0.9134,
+      "step": 22785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4685438128630115,
+      "learning_rate": 9.294011761138614e-06,
+      "loss": 0.9336,
+      "step": 22786
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.819410465486956,
+      "learning_rate": 9.293250626081951e-06,
+      "loss": 0.8411,
+      "step": 22787
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5364265531445418,
+      "learning_rate": 9.292489495140209e-06,
+      "loss": 0.9093,
+      "step": 22788
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0668235612747705,
+      "learning_rate": 9.291728368317816e-06,
+      "loss": 0.9026,
+      "step": 22789
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8183435725619965,
+      "learning_rate": 9.290967245619208e-06,
+      "loss": 0.989,
+      "step": 22790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7720457335710689,
+      "learning_rate": 9.290206127048809e-06,
+      "loss": 0.736,
+      "step": 22791
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6893349504897952,
+      "learning_rate": 9.289445012611056e-06,
+      "loss": 0.7769,
+      "step": 22792
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8453274727288556,
+      "learning_rate": 9.28868390231038e-06,
+      "loss": 0.819,
+      "step": 22793
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9558521646258065,
+      "learning_rate": 9.287922796151214e-06,
+      "loss": 0.8765,
+      "step": 22794
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.822338722146411,
+      "learning_rate": 9.28716169413798e-06,
+      "loss": 0.8046,
+      "step": 22795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6105114662684157,
+      "learning_rate": 9.28640059627512e-06,
+      "loss": 0.8781,
+      "step": 22796
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.955726460852354,
+      "learning_rate": 9.285639502567065e-06,
+      "loss": 0.8871,
+      "step": 22797
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0699933500801015,
+      "learning_rate": 9.284878413018238e-06,
+      "loss": 0.9427,
+      "step": 22798
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8138263023900234,
+      "learning_rate": 9.284117327633077e-06,
+      "loss": 1.0398,
+      "step": 22799
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9258751882863265,
+      "learning_rate": 9.28335624641601e-06,
+      "loss": 0.8165,
+      "step": 22800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1995058146674085,
+      "learning_rate": 9.282595169371474e-06,
+      "loss": 0.9279,
+      "step": 22801
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4768437817838287,
+      "learning_rate": 9.281834096503895e-06,
+      "loss": 0.8332,
+      "step": 22802
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.591401797541437,
+      "learning_rate": 9.281073027817702e-06,
+      "loss": 0.8098,
+      "step": 22803
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7092314086401095,
+      "learning_rate": 9.280311963317332e-06,
+      "loss": 0.7963,
+      "step": 22804
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6151112360471787,
+      "learning_rate": 9.279550903007212e-06,
+      "loss": 0.7676,
+      "step": 22805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.696525874371138,
+      "learning_rate": 9.278789846891777e-06,
+      "loss": 0.884,
+      "step": 22806
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5007577471095719,
+      "learning_rate": 9.278028794975452e-06,
+      "loss": 0.7777,
+      "step": 22807
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4361434583251202,
+      "learning_rate": 9.277267747262676e-06,
+      "loss": 0.8927,
+      "step": 22808
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7031391661264268,
+      "learning_rate": 9.276506703757875e-06,
+      "loss": 0.924,
+      "step": 22809
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2513489444407955,
+      "learning_rate": 9.275745664465478e-06,
+      "loss": 0.9274,
+      "step": 22810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7307005194345486,
+      "learning_rate": 9.274984629389927e-06,
+      "loss": 0.9704,
+      "step": 22811
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7981613699776104,
+      "learning_rate": 9.274223598535635e-06,
+      "loss": 0.8276,
+      "step": 22812
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6207009241691868,
+      "learning_rate": 9.273462571907049e-06,
+      "loss": 0.8305,
+      "step": 22813
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7036065377287541,
+      "learning_rate": 9.272701549508589e-06,
+      "loss": 0.8731,
+      "step": 22814
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.583343894236727,
+      "learning_rate": 9.271940531344699e-06,
+      "loss": 0.9334,
+      "step": 22815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6176381206791939,
+      "learning_rate": 9.271179517419794e-06,
+      "loss": 0.8408,
+      "step": 22816
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5578538287758996,
+      "learning_rate": 9.270418507738316e-06,
+      "loss": 0.9481,
+      "step": 22817
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6064320181815395,
+      "learning_rate": 9.269657502304694e-06,
+      "loss": 0.8701,
+      "step": 22818
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6080459073471578,
+      "learning_rate": 9.268896501123354e-06,
+      "loss": 0.8138,
+      "step": 22819
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.434278860935168,
+      "learning_rate": 9.26813550419873e-06,
+      "loss": 0.8276,
+      "step": 22820
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.817349061400161,
+      "learning_rate": 9.267374511535254e-06,
+      "loss": 0.8956,
+      "step": 22821
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4813475105245149,
+      "learning_rate": 9.266613523137358e-06,
+      "loss": 0.936,
+      "step": 22822
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6752273147627186,
+      "learning_rate": 9.265852539009466e-06,
+      "loss": 0.9193,
+      "step": 22823
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4912277667003726,
+      "learning_rate": 9.265091559156014e-06,
+      "loss": 0.8154,
+      "step": 22824
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8834550374514174,
+      "learning_rate": 9.264330583581434e-06,
+      "loss": 0.7282,
+      "step": 22825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8230335213433377,
+      "learning_rate": 9.263569612290154e-06,
+      "loss": 0.8333,
+      "step": 22826
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5566222342783236,
+      "learning_rate": 9.2628086452866e-06,
+      "loss": 0.8501,
+      "step": 22827
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5512195767253236,
+      "learning_rate": 9.262047682575211e-06,
+      "loss": 0.8085,
+      "step": 22828
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6402339223325408,
+      "learning_rate": 9.261286724160417e-06,
+      "loss": 0.8902,
+      "step": 22829
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.852280482540295,
+      "learning_rate": 9.260525770046642e-06,
+      "loss": 0.6906,
+      "step": 22830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4989477278069043,
+      "learning_rate": 9.259764820238318e-06,
+      "loss": 0.9366,
+      "step": 22831
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8534685514613063,
+      "learning_rate": 9.259003874739885e-06,
+      "loss": 0.9434,
+      "step": 22832
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.158077468304837,
+      "learning_rate": 9.258242933555759e-06,
+      "loss": 0.8105,
+      "step": 22833
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6056205421135317,
+      "learning_rate": 9.25748199669038e-06,
+      "loss": 0.9386,
+      "step": 22834
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.799052732272174,
+      "learning_rate": 9.256721064148173e-06,
+      "loss": 0.697,
+      "step": 22835
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8824568560114274,
+      "learning_rate": 9.255960135933577e-06,
+      "loss": 0.8304,
+      "step": 22836
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4858149567317815,
+      "learning_rate": 9.255199212051012e-06,
+      "loss": 0.8367,
+      "step": 22837
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6680867839238183,
+      "learning_rate": 9.254438292504915e-06,
+      "loss": 0.9124,
+      "step": 22838
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7690916273262371,
+      "learning_rate": 9.253677377299716e-06,
+      "loss": 0.6854,
+      "step": 22839
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6476824249188986,
+      "learning_rate": 9.252916466439839e-06,
+      "loss": 0.8317,
+      "step": 22840
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5583030935305713,
+      "learning_rate": 9.252155559929722e-06,
+      "loss": 0.852,
+      "step": 22841
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.419454130343862,
+      "learning_rate": 9.25139465777379e-06,
+      "loss": 0.8417,
+      "step": 22842
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0896955646748454,
+      "learning_rate": 9.25063375997648e-06,
+      "loss": 0.9513,
+      "step": 22843
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6657993682130472,
+      "learning_rate": 9.249872866542211e-06,
+      "loss": 0.894,
+      "step": 22844
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3668852592499776,
+      "learning_rate": 9.249111977475422e-06,
+      "loss": 0.7922,
+      "step": 22845
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4465333685291373,
+      "learning_rate": 9.248351092780544e-06,
+      "loss": 0.9022,
+      "step": 22846
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6294954957466077,
+      "learning_rate": 9.247590212462001e-06,
+      "loss": 0.9607,
+      "step": 22847
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6812649617337603,
+      "learning_rate": 9.246829336524225e-06,
+      "loss": 0.9205,
+      "step": 22848
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8336256018056969,
+      "learning_rate": 9.246068464971649e-06,
+      "loss": 0.7571,
+      "step": 22849
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.400921379402302,
+      "learning_rate": 9.245307597808702e-06,
+      "loss": 0.7723,
+      "step": 22850
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4883673669509663,
+      "learning_rate": 9.244546735039811e-06,
+      "loss": 0.7959,
+      "step": 22851
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.619276476692481,
+      "learning_rate": 9.243785876669407e-06,
+      "loss": 0.8265,
+      "step": 22852
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7608837890849298,
+      "learning_rate": 9.243025022701926e-06,
+      "loss": 0.7746,
+      "step": 22853
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7543173860676499,
+      "learning_rate": 9.242264173141789e-06,
+      "loss": 0.7162,
+      "step": 22854
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6331970523139625,
+      "learning_rate": 9.24150332799343e-06,
+      "loss": 0.9574,
+      "step": 22855
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8124827438914746,
+      "learning_rate": 9.240742487261279e-06,
+      "loss": 0.8969,
+      "step": 22856
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6279277197589093,
+      "learning_rate": 9.239981650949768e-06,
+      "loss": 0.865,
+      "step": 22857
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6193609935402251,
+      "learning_rate": 9.239220819063322e-06,
+      "loss": 0.7286,
+      "step": 22858
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5901943387427975,
+      "learning_rate": 9.238459991606371e-06,
+      "loss": 0.8782,
+      "step": 22859
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7982398475242418,
+      "learning_rate": 9.237699168583355e-06,
+      "loss": 0.7673,
+      "step": 22860
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8462799151814544,
+      "learning_rate": 9.236938349998688e-06,
+      "loss": 0.8239,
+      "step": 22861
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8025835212790755,
+      "learning_rate": 9.23617753585681e-06,
+      "loss": 0.9321,
+      "step": 22862
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.150557925617014,
+      "learning_rate": 9.235416726162147e-06,
+      "loss": 0.9081,
+      "step": 22863
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6142567184857621,
+      "learning_rate": 9.234655920919135e-06,
+      "loss": 0.9158,
+      "step": 22864
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.729301219055645,
+      "learning_rate": 9.233895120132192e-06,
+      "loss": 0.7915,
+      "step": 22865
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6514658577681127,
+      "learning_rate": 9.233134323805757e-06,
+      "loss": 0.9319,
+      "step": 22866
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8330534298248746,
+      "learning_rate": 9.23237353194426e-06,
+      "loss": 0.8617,
+      "step": 22867
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.473459311117769,
+      "learning_rate": 9.231612744552123e-06,
+      "loss": 0.8108,
+      "step": 22868
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7208289843151763,
+      "learning_rate": 9.230851961633779e-06,
+      "loss": 0.8338,
+      "step": 22869
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4295975241610133,
+      "learning_rate": 9.230091183193659e-06,
+      "loss": 0.7172,
+      "step": 22870
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.65840570881269,
+      "learning_rate": 9.229330409236196e-06,
+      "loss": 0.853,
+      "step": 22871
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6240495920799167,
+      "learning_rate": 9.22856963976581e-06,
+      "loss": 0.8669,
+      "step": 22872
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7622950963407549,
+      "learning_rate": 9.227808874786937e-06,
+      "loss": 0.9689,
+      "step": 22873
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7783428977708355,
+      "learning_rate": 9.22704811430401e-06,
+      "loss": 0.8647,
+      "step": 22874
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8090865658627304,
+      "learning_rate": 9.22628735832145e-06,
+      "loss": 0.9123,
+      "step": 22875
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6173861840176345,
+      "learning_rate": 9.225526606843687e-06,
+      "loss": 0.9606,
+      "step": 22876
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.066316010681888,
+      "learning_rate": 9.224765859875157e-06,
+      "loss": 0.9412,
+      "step": 22877
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1289970981446964,
+      "learning_rate": 9.224005117420287e-06,
+      "loss": 0.7694,
+      "step": 22878
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6862163429775416,
+      "learning_rate": 9.223244379483502e-06,
+      "loss": 0.7321,
+      "step": 22879
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7371001633813115,
+      "learning_rate": 9.222483646069233e-06,
+      "loss": 0.8879,
+      "step": 22880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6908459268695097,
+      "learning_rate": 9.221722917181917e-06,
+      "loss": 0.8209,
+      "step": 22881
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6958469190967758,
+      "learning_rate": 9.22096219282597e-06,
+      "loss": 0.7978,
+      "step": 22882
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7941920254581578,
+      "learning_rate": 9.220201473005828e-06,
+      "loss": 0.8945,
+      "step": 22883
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.534187479374231,
+      "learning_rate": 9.21944075772592e-06,
+      "loss": 0.8915,
+      "step": 22884
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2689013257231463,
+      "learning_rate": 9.218680046990678e-06,
+      "loss": 0.8947,
+      "step": 22885
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7351270248907393,
+      "learning_rate": 9.217919340804527e-06,
+      "loss": 0.9327,
+      "step": 22886
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7343637761099444,
+      "learning_rate": 9.217158639171898e-06,
+      "loss": 0.6904,
+      "step": 22887
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.680663992470602,
+      "learning_rate": 9.216397942097219e-06,
+      "loss": 0.8137,
+      "step": 22888
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5616091879267424,
+      "learning_rate": 9.215637249584915e-06,
+      "loss": 0.8112,
+      "step": 22889
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5973571951811076,
+      "learning_rate": 9.214876561639424e-06,
+      "loss": 0.9476,
+      "step": 22890
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6480202838621734,
+      "learning_rate": 9.214115878265168e-06,
+      "loss": 0.8173,
+      "step": 22891
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5581899379163606,
+      "learning_rate": 9.213355199466582e-06,
+      "loss": 0.7652,
+      "step": 22892
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2855703784173875,
+      "learning_rate": 9.212594525248086e-06,
+      "loss": 0.888,
+      "step": 22893
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8736954036688864,
+      "learning_rate": 9.211833855614115e-06,
+      "loss": 0.819,
+      "step": 22894
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8020990404920434,
+      "learning_rate": 9.2110731905691e-06,
+      "loss": 0.8616,
+      "step": 22895
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.539948017436802,
+      "learning_rate": 9.210312530117464e-06,
+      "loss": 0.8748,
+      "step": 22896
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4419037219420563,
+      "learning_rate": 9.209551874263637e-06,
+      "loss": 0.8186,
+      "step": 22897
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6473495098263498,
+      "learning_rate": 9.208791223012051e-06,
+      "loss": 0.8315,
+      "step": 22898
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4303954284392013,
+      "learning_rate": 9.208030576367136e-06,
+      "loss": 0.8779,
+      "step": 22899
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7385195462797318,
+      "learning_rate": 9.207269934333315e-06,
+      "loss": 0.8473,
+      "step": 22900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6997957835536253,
+      "learning_rate": 9.206509296915016e-06,
+      "loss": 0.8535,
+      "step": 22901
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6884307088928874,
+      "learning_rate": 9.205748664116677e-06,
+      "loss": 0.7865,
+      "step": 22902
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8154163712674855,
+      "learning_rate": 9.20498803594272e-06,
+      "loss": 0.7734,
+      "step": 22903
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7786131494122313,
+      "learning_rate": 9.20422741239757e-06,
+      "loss": 0.7837,
+      "step": 22904
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6953346843190908,
+      "learning_rate": 9.203466793485662e-06,
+      "loss": 0.8656,
+      "step": 22905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7427221503938681,
+      "learning_rate": 9.202706179211425e-06,
+      "loss": 0.758,
+      "step": 22906
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1654646636198414,
+      "learning_rate": 9.201945569579283e-06,
+      "loss": 0.842,
+      "step": 22907
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6197197520703261,
+      "learning_rate": 9.201184964593664e-06,
+      "loss": 0.8092,
+      "step": 22908
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8161750856099552,
+      "learning_rate": 9.200424364259006e-06,
+      "loss": 0.8839,
+      "step": 22909
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7255592284081192,
+      "learning_rate": 9.199663768579724e-06,
+      "loss": 0.836,
+      "step": 22910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6647712958561882,
+      "learning_rate": 9.198903177560255e-06,
+      "loss": 0.8234,
+      "step": 22911
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7526253413803365,
+      "learning_rate": 9.198142591205022e-06,
+      "loss": 0.8731,
+      "step": 22912
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6159193331767205,
+      "learning_rate": 9.197382009518465e-06,
+      "loss": 0.8558,
+      "step": 22913
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.523378808200742,
+      "learning_rate": 9.196621432504996e-06,
+      "loss": 0.9902,
+      "step": 22914
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7428988359087093,
+      "learning_rate": 9.195860860169054e-06,
+      "loss": 0.7223,
+      "step": 22915
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.669810733282531,
+      "learning_rate": 9.195100292515064e-06,
+      "loss": 0.8289,
+      "step": 22916
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6071578390127792,
+      "learning_rate": 9.194339729547461e-06,
+      "loss": 0.762,
+      "step": 22917
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5912033555172505,
+      "learning_rate": 9.193579171270659e-06,
+      "loss": 1.0128,
+      "step": 22918
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.789172446303598,
+      "learning_rate": 9.192818617689097e-06,
+      "loss": 0.9007,
+      "step": 22919
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8348154469134157,
+      "learning_rate": 9.192058068807203e-06,
+      "loss": 0.8611,
+      "step": 22920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7530688771657175,
+      "learning_rate": 9.191297524629398e-06,
+      "loss": 0.8457,
+      "step": 22921
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.136489652441868,
+      "learning_rate": 9.190536985160119e-06,
+      "loss": 0.8632,
+      "step": 22922
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.619811118971683,
+      "learning_rate": 9.189776450403789e-06,
+      "loss": 0.8481,
+      "step": 22923
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8085922239981511,
+      "learning_rate": 9.189015920364839e-06,
+      "loss": 0.6943,
+      "step": 22924
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4904158401745804,
+      "learning_rate": 9.18825539504769e-06,
+      "loss": 0.7692,
+      "step": 22925
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7374584220674871,
+      "learning_rate": 9.187494874456777e-06,
+      "loss": 0.8006,
+      "step": 22926
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.794199673212808,
+      "learning_rate": 9.186734358596529e-06,
+      "loss": 0.6343,
+      "step": 22927
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.710573048693814,
+      "learning_rate": 9.185973847471369e-06,
+      "loss": 1.0163,
+      "step": 22928
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4418022968802897,
+      "learning_rate": 9.185213341085725e-06,
+      "loss": 0.9412,
+      "step": 22929
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6065651234049374,
+      "learning_rate": 9.184452839444029e-06,
+      "loss": 0.922,
+      "step": 22930
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6435831215382712,
+      "learning_rate": 9.183692342550708e-06,
+      "loss": 0.8428,
+      "step": 22931
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8878529637175205,
+      "learning_rate": 9.182931850410187e-06,
+      "loss": 0.7463,
+      "step": 22932
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1139610942524354,
+      "learning_rate": 9.182171363026894e-06,
+      "loss": 0.941,
+      "step": 22933
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6221920699206027,
+      "learning_rate": 9.181410880405262e-06,
+      "loss": 0.8163,
+      "step": 22934
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3808355933850116,
+      "learning_rate": 9.180650402549713e-06,
+      "loss": 0.8745,
+      "step": 22935
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6679718727674722,
+      "learning_rate": 9.179889929464676e-06,
+      "loss": 0.9489,
+      "step": 22936
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8021248136000398,
+      "learning_rate": 9.179129461154578e-06,
+      "loss": 0.7496,
+      "step": 22937
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.69591516521016,
+      "learning_rate": 9.178368997623853e-06,
+      "loss": 0.8285,
+      "step": 22938
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7984899007681596,
+      "learning_rate": 9.17760853887692e-06,
+      "loss": 0.8417,
+      "step": 22939
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.440719674266089,
+      "learning_rate": 9.176848084918212e-06,
+      "loss": 0.8407,
+      "step": 22940
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2404860668978968,
+      "learning_rate": 9.176087635752157e-06,
+      "loss": 0.8881,
+      "step": 22941
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7745273849618561,
+      "learning_rate": 9.175327191383175e-06,
+      "loss": 0.6847,
+      "step": 22942
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.413433300043305,
+      "learning_rate": 9.174566751815704e-06,
+      "loss": 0.8151,
+      "step": 22943
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8300129386537465,
+      "learning_rate": 9.173806317054161e-06,
+      "loss": 0.9338,
+      "step": 22944
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.686014265339895,
+      "learning_rate": 9.173045887102988e-06,
+      "loss": 0.8191,
+      "step": 22945
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5651737230018228,
+      "learning_rate": 9.172285461966597e-06,
+      "loss": 0.8584,
+      "step": 22946
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.637494889587607,
+      "learning_rate": 9.171525041649423e-06,
+      "loss": 0.8988,
+      "step": 22947
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0793007775334713,
+      "learning_rate": 9.170764626155896e-06,
+      "loss": 0.792,
+      "step": 22948
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2836712508701793,
+      "learning_rate": 9.170004215490438e-06,
+      "loss": 0.9507,
+      "step": 22949
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6892007271258442,
+      "learning_rate": 9.169243809657476e-06,
+      "loss": 0.8473,
+      "step": 22950
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.696114922726681,
+      "learning_rate": 9.16848340866144e-06,
+      "loss": 0.7609,
+      "step": 22951
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5943194520383488,
+      "learning_rate": 9.16772301250676e-06,
+      "loss": 0.9711,
+      "step": 22952
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.824147846643854,
+      "learning_rate": 9.16696262119786e-06,
+      "loss": 0.8703,
+      "step": 22953
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5793767164006707,
+      "learning_rate": 9.166202234739161e-06,
+      "loss": 0.8883,
+      "step": 22954
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6339695476867067,
+      "learning_rate": 9.165441853135104e-06,
+      "loss": 0.899,
+      "step": 22955
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.90243974283206,
+      "learning_rate": 9.164681476390107e-06,
+      "loss": 0.9184,
+      "step": 22956
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4771010019103437,
+      "learning_rate": 9.163921104508593e-06,
+      "loss": 0.8108,
+      "step": 22957
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5919858002370726,
+      "learning_rate": 9.163160737495e-06,
+      "loss": 0.903,
+      "step": 22958
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.623887506913462,
+      "learning_rate": 9.162400375353752e-06,
+      "loss": 0.8903,
+      "step": 22959
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9280543069547416,
+      "learning_rate": 9.161640018089273e-06,
+      "loss": 0.8362,
+      "step": 22960
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6922199013814736,
+      "learning_rate": 9.160879665705988e-06,
+      "loss": 0.988,
+      "step": 22961
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6507533746243153,
+      "learning_rate": 9.160119318208334e-06,
+      "loss": 0.8692,
+      "step": 22962
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4941146604608035,
+      "learning_rate": 9.159358975600723e-06,
+      "loss": 0.8089,
+      "step": 22963
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8210393878961615,
+      "learning_rate": 9.158598637887594e-06,
+      "loss": 0.8394,
+      "step": 22964
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5475205831157448,
+      "learning_rate": 9.15783830507337e-06,
+      "loss": 0.9175,
+      "step": 22965
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.562668274259771,
+      "learning_rate": 9.157077977162481e-06,
+      "loss": 0.8109,
+      "step": 22966
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7528602732294039,
+      "learning_rate": 9.156317654159346e-06,
+      "loss": 0.8654,
+      "step": 22967
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4714541582408738,
+      "learning_rate": 9.155557336068399e-06,
+      "loss": 0.838,
+      "step": 22968
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9003467566390335,
+      "learning_rate": 9.154797022894067e-06,
+      "loss": 0.8834,
+      "step": 22969
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7190978625300362,
+      "learning_rate": 9.154036714640768e-06,
+      "loss": 0.7711,
+      "step": 22970
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8582922743982426,
+      "learning_rate": 9.15327641131294e-06,
+      "loss": 0.7979,
+      "step": 22971
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4713057723168441,
+      "learning_rate": 9.152516112915002e-06,
+      "loss": 0.9018,
+      "step": 22972
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.704058959658619,
+      "learning_rate": 9.151755819451387e-06,
+      "loss": 0.8379,
+      "step": 22973
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8003022357207095,
+      "learning_rate": 9.150995530926515e-06,
+      "loss": 0.8061,
+      "step": 22974
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6463381363049436,
+      "learning_rate": 9.150235247344817e-06,
+      "loss": 0.8228,
+      "step": 22975
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5660142079275494,
+      "learning_rate": 9.14947496871072e-06,
+      "loss": 0.7593,
+      "step": 22976
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7533799780707218,
+      "learning_rate": 9.148714695028647e-06,
+      "loss": 0.856,
+      "step": 22977
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5418874953897848,
+      "learning_rate": 9.147954426303024e-06,
+      "loss": 0.7748,
+      "step": 22978
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7662179394017231,
+      "learning_rate": 9.147194162538284e-06,
+      "loss": 0.7353,
+      "step": 22979
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5674591875111088,
+      "learning_rate": 9.14643390373885e-06,
+      "loss": 0.8109,
+      "step": 22980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5319848661289246,
+      "learning_rate": 9.145673649909147e-06,
+      "loss": 0.9304,
+      "step": 22981
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7591064256343214,
+      "learning_rate": 9.144913401053599e-06,
+      "loss": 0.6925,
+      "step": 22982
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7481180466582822,
+      "learning_rate": 9.144153157176641e-06,
+      "loss": 0.9862,
+      "step": 22983
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5911732959892566,
+      "learning_rate": 9.143392918282693e-06,
+      "loss": 0.7909,
+      "step": 22984
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7893907039922818,
+      "learning_rate": 9.14263268437618e-06,
+      "loss": 0.7328,
+      "step": 22985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5565139168960493,
+      "learning_rate": 9.141872455461532e-06,
+      "loss": 0.7925,
+      "step": 22986
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6776801173641243,
+      "learning_rate": 9.141112231543177e-06,
+      "loss": 1.0165,
+      "step": 22987
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6898144113629687,
+      "learning_rate": 9.140352012625538e-06,
+      "loss": 0.9389,
+      "step": 22988
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9288329545942011,
+      "learning_rate": 9.139591798713038e-06,
+      "loss": 0.8713,
+      "step": 22989
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8638595293111169,
+      "learning_rate": 9.138831589810111e-06,
+      "loss": 0.889,
+      "step": 22990
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.711081718674951,
+      "learning_rate": 9.138071385921176e-06,
+      "loss": 0.7152,
+      "step": 22991
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5928288753963862,
+      "learning_rate": 9.137311187050663e-06,
+      "loss": 0.9491,
+      "step": 22992
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0374996966110785,
+      "learning_rate": 9.136550993202997e-06,
+      "loss": 0.8487,
+      "step": 22993
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7895238264415674,
+      "learning_rate": 9.13579080438261e-06,
+      "loss": 0.679,
+      "step": 22994
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9851371100734108,
+      "learning_rate": 9.135030620593915e-06,
+      "loss": 0.7964,
+      "step": 22995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4522418077596861,
+      "learning_rate": 9.13427044184135e-06,
+      "loss": 0.9423,
+      "step": 22996
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6171211351831265,
+      "learning_rate": 9.133510268129337e-06,
+      "loss": 0.8195,
+      "step": 22997
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7198651867736872,
+      "learning_rate": 9.1327500994623e-06,
+      "loss": 0.6546,
+      "step": 22998
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9541195288092679,
+      "learning_rate": 9.131989935844663e-06,
+      "loss": 0.9093,
+      "step": 22999
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5607358365051411,
+      "learning_rate": 9.131229777280859e-06,
+      "loss": 0.6619,
+      "step": 23000
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4714298041486222,
+      "learning_rate": 9.130469623775312e-06,
+      "loss": 0.8802,
+      "step": 23001
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6777519268454482,
+      "learning_rate": 9.129709475332445e-06,
+      "loss": 0.8054,
+      "step": 23002
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7967354290894243,
+      "learning_rate": 9.12894933195668e-06,
+      "loss": 0.6999,
+      "step": 23003
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5150917208777535,
+      "learning_rate": 9.128189193652455e-06,
+      "loss": 0.8652,
+      "step": 23004
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4843981182461146,
+      "learning_rate": 9.127429060424184e-06,
+      "loss": 0.7922,
+      "step": 23005
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5306419442817718,
+      "learning_rate": 9.126668932276298e-06,
+      "loss": 0.9201,
+      "step": 23006
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7498223789924323,
+      "learning_rate": 9.125908809213221e-06,
+      "loss": 0.6954,
+      "step": 23007
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7347642236145625,
+      "learning_rate": 9.125148691239383e-06,
+      "loss": 0.6935,
+      "step": 23008
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8364307664958481,
+      "learning_rate": 9.124388578359206e-06,
+      "loss": 0.9461,
+      "step": 23009
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6881953807892267,
+      "learning_rate": 9.123628470577111e-06,
+      "loss": 0.9029,
+      "step": 23010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9183849093816003,
+      "learning_rate": 9.122868367897537e-06,
+      "loss": 0.8227,
+      "step": 23011
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4560220947477907,
+      "learning_rate": 9.122108270324894e-06,
+      "loss": 0.7168,
+      "step": 23012
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9308530616579866,
+      "learning_rate": 9.121348177863615e-06,
+      "loss": 0.814,
+      "step": 23013
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.004188341963954,
+      "learning_rate": 9.120588090518126e-06,
+      "loss": 0.8863,
+      "step": 23014
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1231865824842657,
+      "learning_rate": 9.119828008292856e-06,
+      "loss": 0.7842,
+      "step": 23015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6262114338201887,
+      "learning_rate": 9.11906793119222e-06,
+      "loss": 0.8929,
+      "step": 23016
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8003665611993602,
+      "learning_rate": 9.118307859220651e-06,
+      "loss": 0.9806,
+      "step": 23017
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7321181331475778,
+      "learning_rate": 9.117547792382577e-06,
+      "loss": 0.7063,
+      "step": 23018
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.086727470994316,
+      "learning_rate": 9.116787730682413e-06,
+      "loss": 0.9176,
+      "step": 23019
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0081846974508384,
+      "learning_rate": 9.116027674124592e-06,
+      "loss": 0.9093,
+      "step": 23020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.706369607691953,
+      "learning_rate": 9.115267622713539e-06,
+      "loss": 0.6817,
+      "step": 23021
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6331038716462731,
+      "learning_rate": 9.11450757645368e-06,
+      "loss": 0.9594,
+      "step": 23022
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5933445269403572,
+      "learning_rate": 9.113747535349433e-06,
+      "loss": 0.9052,
+      "step": 23023
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.917936113842475,
+      "learning_rate": 9.112987499405232e-06,
+      "loss": 0.8025,
+      "step": 23024
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5000159277239504,
+      "learning_rate": 9.112227468625499e-06,
+      "loss": 0.7799,
+      "step": 23025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5263340217501002,
+      "learning_rate": 9.11146744301466e-06,
+      "loss": 0.8023,
+      "step": 23026
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5935784828063917,
+      "learning_rate": 9.110707422577133e-06,
+      "loss": 0.8025,
+      "step": 23027
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6419460737470446,
+      "learning_rate": 9.109947407317352e-06,
+      "loss": 0.8664,
+      "step": 23028
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5318425164418261,
+      "learning_rate": 9.109187397239742e-06,
+      "loss": 0.8624,
+      "step": 23029
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7583590786844187,
+      "learning_rate": 9.108427392348722e-06,
+      "loss": 0.7708,
+      "step": 23030
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6116552534885447,
+      "learning_rate": 9.107667392648718e-06,
+      "loss": 1.0255,
+      "step": 23031
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7438585602220411,
+      "learning_rate": 9.10690739814416e-06,
+      "loss": 0.7336,
+      "step": 23032
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4742149888449867,
+      "learning_rate": 9.10614740883947e-06,
+      "loss": 0.8642,
+      "step": 23033
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.469732538794093,
+      "learning_rate": 9.105387424739073e-06,
+      "loss": 0.9014,
+      "step": 23034
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6165292752115203,
+      "learning_rate": 9.10462744584739e-06,
+      "loss": 0.8341,
+      "step": 23035
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.495150580366537,
+      "learning_rate": 9.103867472168854e-06,
+      "loss": 0.791,
+      "step": 23036
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6077613115488845,
+      "learning_rate": 9.103107503707885e-06,
+      "loss": 0.886,
+      "step": 23037
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.187784502488066,
+      "learning_rate": 9.102347540468906e-06,
+      "loss": 0.9362,
+      "step": 23038
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7572644585277174,
+      "learning_rate": 9.101587582456342e-06,
+      "loss": 0.7787,
+      "step": 23039
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6567451450804118,
+      "learning_rate": 9.100827629674623e-06,
+      "loss": 0.9775,
+      "step": 23040
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.568763244266842,
+      "learning_rate": 9.10006768212817e-06,
+      "loss": 0.9457,
+      "step": 23041
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7126771404475454,
+      "learning_rate": 9.099307739821404e-06,
+      "loss": 0.7336,
+      "step": 23042
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7021548953586303,
+      "learning_rate": 9.098547802758762e-06,
+      "loss": 0.8088,
+      "step": 23043
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4895424285935965,
+      "learning_rate": 9.09778787094465e-06,
+      "loss": 0.8274,
+      "step": 23044
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.638986207350254,
+      "learning_rate": 9.097027944383508e-06,
+      "loss": 0.8707,
+      "step": 23045
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.715848581355866,
+      "learning_rate": 9.096268023079753e-06,
+      "loss": 0.8899,
+      "step": 23046
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.439638992996208,
+      "learning_rate": 9.095508107037817e-06,
+      "loss": 0.8971,
+      "step": 23047
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5617514218563178,
+      "learning_rate": 9.094748196262114e-06,
+      "loss": 0.9411,
+      "step": 23048
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4938018940010165,
+      "learning_rate": 9.093988290757075e-06,
+      "loss": 0.829,
+      "step": 23049
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.513097545199607,
+      "learning_rate": 9.093228390527123e-06,
+      "loss": 0.8366,
+      "step": 23050
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6205069656059226,
+      "learning_rate": 9.092468495576682e-06,
+      "loss": 0.8174,
+      "step": 23051
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6551727978176836,
+      "learning_rate": 9.091708605910175e-06,
+      "loss": 0.8752,
+      "step": 23052
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.119331549084122,
+      "learning_rate": 9.09094872153203e-06,
+      "loss": 0.9254,
+      "step": 23053
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.994071497709122,
+      "learning_rate": 9.090188842446672e-06,
+      "loss": 0.9622,
+      "step": 23054
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.894369482585837,
+      "learning_rate": 9.089428968658518e-06,
+      "loss": 0.8841,
+      "step": 23055
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6230252283458297,
+      "learning_rate": 9.088669100171999e-06,
+      "loss": 0.8281,
+      "step": 23056
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5516281803743035,
+      "learning_rate": 9.08790923699154e-06,
+      "loss": 0.8999,
+      "step": 23057
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5993437341590935,
+      "learning_rate": 9.087149379121556e-06,
+      "loss": 0.9067,
+      "step": 23058
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7602924031546514,
+      "learning_rate": 9.08638952656648e-06,
+      "loss": 0.7613,
+      "step": 23059
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.674521682075243,
+      "learning_rate": 9.085629679330733e-06,
+      "loss": 0.819,
+      "step": 23060
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6770704509214909,
+      "learning_rate": 9.084869837418742e-06,
+      "loss": 0.8432,
+      "step": 23061
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7626271318129159,
+      "learning_rate": 9.084110000834928e-06,
+      "loss": 0.757,
+      "step": 23062
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.683266716766806,
+      "learning_rate": 9.083350169583711e-06,
+      "loss": 0.7647,
+      "step": 23063
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.761118778167235,
+      "learning_rate": 9.082590343669526e-06,
+      "loss": 0.8596,
+      "step": 23064
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6468651745238898,
+      "learning_rate": 9.081830523096786e-06,
+      "loss": 0.7161,
+      "step": 23065
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6568448159010758,
+      "learning_rate": 9.081070707869921e-06,
+      "loss": 0.9354,
+      "step": 23066
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.123141129675091,
+      "learning_rate": 9.08031089799335e-06,
+      "loss": 1.0196,
+      "step": 23067
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6667418136161627,
+      "learning_rate": 9.079551093471507e-06,
+      "loss": 0.855,
+      "step": 23068
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5885013104730623,
+      "learning_rate": 9.078791294308803e-06,
+      "loss": 0.806,
+      "step": 23069
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2459588522026173,
+      "learning_rate": 9.07803150050967e-06,
+      "loss": 0.7377,
+      "step": 23070
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9214338232007808,
+      "learning_rate": 9.077271712078531e-06,
+      "loss": 0.7914,
+      "step": 23071
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7709586108222651,
+      "learning_rate": 9.076511929019804e-06,
+      "loss": 0.7539,
+      "step": 23072
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.690727559192968,
+      "learning_rate": 9.07575215133792e-06,
+      "loss": 0.8793,
+      "step": 23073
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.649305375231017,
+      "learning_rate": 9.074992379037298e-06,
+      "loss": 0.8915,
+      "step": 23074
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.564753838663814,
+      "learning_rate": 9.074232612122367e-06,
+      "loss": 0.795,
+      "step": 23075
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3482023211292518,
+      "learning_rate": 9.073472850597542e-06,
+      "loss": 0.8412,
+      "step": 23076
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5868077080373364,
+      "learning_rate": 9.072713094467253e-06,
+      "loss": 0.8749,
+      "step": 23077
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5340378176744838,
+      "learning_rate": 9.071953343735925e-06,
+      "loss": 0.9429,
+      "step": 23078
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6508738353166774,
+      "learning_rate": 9.071193598407975e-06,
+      "loss": 0.8257,
+      "step": 23079
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6135783820087608,
+      "learning_rate": 9.070433858487829e-06,
+      "loss": 0.8771,
+      "step": 23080
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5139080859468579,
+      "learning_rate": 9.069674123979912e-06,
+      "loss": 0.866,
+      "step": 23081
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7659797780638922,
+      "learning_rate": 9.068914394888651e-06,
+      "loss": 0.6851,
+      "step": 23082
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1931647183452707,
+      "learning_rate": 9.068154671218462e-06,
+      "loss": 0.7678,
+      "step": 23083
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0322574916332394,
+      "learning_rate": 9.06739495297377e-06,
+      "loss": 0.8927,
+      "step": 23084
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5529608363050929,
+      "learning_rate": 9.066635240159002e-06,
+      "loss": 0.8257,
+      "step": 23085
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7236552407207653,
+      "learning_rate": 9.065875532778578e-06,
+      "loss": 0.9593,
+      "step": 23086
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6042655056526052,
+      "learning_rate": 9.065115830836923e-06,
+      "loss": 0.9847,
+      "step": 23087
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7687989383997595,
+      "learning_rate": 9.064356134338459e-06,
+      "loss": 0.7743,
+      "step": 23088
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6583320807425295,
+      "learning_rate": 9.063596443287612e-06,
+      "loss": 0.8083,
+      "step": 23089
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7277268387622924,
+      "learning_rate": 9.0628367576888e-06,
+      "loss": 0.6869,
+      "step": 23090
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5034735618009338,
+      "learning_rate": 9.062077077546449e-06,
+      "loss": 0.9248,
+      "step": 23091
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7165545787524668,
+      "learning_rate": 9.061317402864988e-06,
+      "loss": 0.808,
+      "step": 23092
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4792380859697867,
+      "learning_rate": 9.060557733648827e-06,
+      "loss": 0.7898,
+      "step": 23093
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6462163944305896,
+      "learning_rate": 9.059798069902399e-06,
+      "loss": 0.8697,
+      "step": 23094
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.367368846780734,
+      "learning_rate": 9.059038411630123e-06,
+      "loss": 0.8428,
+      "step": 23095
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9147378459177034,
+      "learning_rate": 9.058278758836428e-06,
+      "loss": 0.8381,
+      "step": 23096
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.799322902238758,
+      "learning_rate": 9.057519111525728e-06,
+      "loss": 0.8575,
+      "step": 23097
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8493561977222348,
+      "learning_rate": 9.05675946970245e-06,
+      "loss": 0.8693,
+      "step": 23098
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.935449175409041,
+      "learning_rate": 9.05599983337102e-06,
+      "loss": 0.9502,
+      "step": 23099
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5731674429798423,
+      "learning_rate": 9.055240202535856e-06,
+      "loss": 0.7501,
+      "step": 23100
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7779636265377956,
+      "learning_rate": 9.054480577201381e-06,
+      "loss": 0.7683,
+      "step": 23101
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5203286286380482,
+      "learning_rate": 9.05372095737202e-06,
+      "loss": 0.745,
+      "step": 23102
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.73169623331864,
+      "learning_rate": 9.052961343052199e-06,
+      "loss": 0.9755,
+      "step": 23103
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7024994152222088,
+      "learning_rate": 9.052201734246332e-06,
+      "loss": 0.9196,
+      "step": 23104
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6010551644549964,
+      "learning_rate": 9.051442130958848e-06,
+      "loss": 0.897,
+      "step": 23105
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.47620977066172,
+      "learning_rate": 9.050682533194171e-06,
+      "loss": 0.8043,
+      "step": 23106
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6257824347285261,
+      "learning_rate": 9.049922940956719e-06,
+      "loss": 0.8882,
+      "step": 23107
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7653527097339772,
+      "learning_rate": 9.049163354250914e-06,
+      "loss": 0.9195,
+      "step": 23108
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.005296543933828,
+      "learning_rate": 9.048403773081182e-06,
+      "loss": 0.8837,
+      "step": 23109
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6986568076335538,
+      "learning_rate": 9.047644197451949e-06,
+      "loss": 0.8407,
+      "step": 23110
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6669489019716555,
+      "learning_rate": 9.046884627367628e-06,
+      "loss": 0.8485,
+      "step": 23111
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.563934468179961,
+      "learning_rate": 9.046125062832646e-06,
+      "loss": 0.859,
+      "step": 23112
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7505127258335831,
+      "learning_rate": 9.045365503851432e-06,
+      "loss": 0.7332,
+      "step": 23113
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.641985028889278,
+      "learning_rate": 9.044605950428397e-06,
+      "loss": 0.8145,
+      "step": 23114
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4456626371921688,
+      "learning_rate": 9.043846402567969e-06,
+      "loss": 0.7837,
+      "step": 23115
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.090963546616983,
+      "learning_rate": 9.043086860274567e-06,
+      "loss": 0.811,
+      "step": 23116
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5384627699929894,
+      "learning_rate": 9.042327323552624e-06,
+      "loss": 0.8572,
+      "step": 23117
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7910557070043127,
+      "learning_rate": 9.04156779240655e-06,
+      "loss": 0.7847,
+      "step": 23118
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.335269999885331,
+      "learning_rate": 9.040808266840773e-06,
+      "loss": 0.8837,
+      "step": 23119
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5486497346736134,
+      "learning_rate": 9.040048746859715e-06,
+      "loss": 0.8835,
+      "step": 23120
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.758474137704637,
+      "learning_rate": 9.039289232467792e-06,
+      "loss": 0.7944,
+      "step": 23121
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5048029616824867,
+      "learning_rate": 9.038529723669435e-06,
+      "loss": 0.8378,
+      "step": 23122
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0076993622972235,
+      "learning_rate": 9.037770220469062e-06,
+      "loss": 0.7752,
+      "step": 23123
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3898948572757595,
+      "learning_rate": 9.037010722871098e-06,
+      "loss": 0.8597,
+      "step": 23124
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4908526169413425,
+      "learning_rate": 9.036251230879956e-06,
+      "loss": 0.8222,
+      "step": 23125
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.571594528746516,
+      "learning_rate": 9.035491744500069e-06,
+      "loss": 0.8969,
+      "step": 23126
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.536220944912178,
+      "learning_rate": 9.034732263735856e-06,
+      "loss": 0.7866,
+      "step": 23127
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6495262630336471,
+      "learning_rate": 9.033972788591735e-06,
+      "loss": 0.9558,
+      "step": 23128
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6726384953469016,
+      "learning_rate": 9.033213319072126e-06,
+      "loss": 0.825,
+      "step": 23129
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8638522596458753,
+      "learning_rate": 9.032453855181462e-06,
+      "loss": 0.8723,
+      "step": 23130
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9746136984249554,
+      "learning_rate": 9.031694396924156e-06,
+      "loss": 0.8708,
+      "step": 23131
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.696803289479229,
+      "learning_rate": 9.030934944304632e-06,
+      "loss": 0.8623,
+      "step": 23132
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5104846325430203,
+      "learning_rate": 9.030175497327308e-06,
+      "loss": 0.9478,
+      "step": 23133
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.868539093829379,
+      "learning_rate": 9.029416055996614e-06,
+      "loss": 0.8226,
+      "step": 23134
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0639869585904425,
+      "learning_rate": 9.028656620316964e-06,
+      "loss": 0.814,
+      "step": 23135
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.58350393562863,
+      "learning_rate": 9.027897190292784e-06,
+      "loss": 0.8819,
+      "step": 23136
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.822481122661794,
+      "learning_rate": 9.027137765928492e-06,
+      "loss": 0.8001,
+      "step": 23137
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.755853302828208,
+      "learning_rate": 9.026378347228516e-06,
+      "loss": 0.7972,
+      "step": 23138
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6343558472543447,
+      "learning_rate": 9.025618934197272e-06,
+      "loss": 0.8644,
+      "step": 23139
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6705408444529009,
+      "learning_rate": 9.02485952683918e-06,
+      "loss": 0.6846,
+      "step": 23140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.558677483556314,
+      "learning_rate": 9.024100125158671e-06,
+      "loss": 0.8434,
+      "step": 23141
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5323974419952886,
+      "learning_rate": 9.023340729160154e-06,
+      "loss": 0.8604,
+      "step": 23142
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6418403511421016,
+      "learning_rate": 9.022581338848058e-06,
+      "loss": 0.8725,
+      "step": 23143
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2889557964133007,
+      "learning_rate": 9.021821954226801e-06,
+      "loss": 0.9772,
+      "step": 23144
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.491341549126808,
+      "learning_rate": 9.021062575300814e-06,
+      "loss": 0.9962,
+      "step": 23145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7843330334823566,
+      "learning_rate": 9.020303202074503e-06,
+      "loss": 0.7962,
+      "step": 23146
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0021133272107487,
+      "learning_rate": 9.0195438345523e-06,
+      "loss": 0.8008,
+      "step": 23147
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.697371465196622,
+      "learning_rate": 9.018784472738624e-06,
+      "loss": 0.8029,
+      "step": 23148
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34133686532403,
+      "learning_rate": 9.018025116637895e-06,
+      "loss": 0.8509,
+      "step": 23149
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4415705917368282,
+      "learning_rate": 9.01726576625453e-06,
+      "loss": 0.7744,
+      "step": 23150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5899649241695697,
+      "learning_rate": 9.016506421592959e-06,
+      "loss": 0.7662,
+      "step": 23151
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4883750829393503,
+      "learning_rate": 9.0157470826576e-06,
+      "loss": 0.8707,
+      "step": 23152
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7751313186273288,
+      "learning_rate": 9.01498774945287e-06,
+      "loss": 0.6997,
+      "step": 23153
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9667823479863131,
+      "learning_rate": 9.014228421983194e-06,
+      "loss": 0.766,
+      "step": 23154
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7631848047949151,
+      "learning_rate": 9.013469100252995e-06,
+      "loss": 0.6397,
+      "step": 23155
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2754761496753955,
+      "learning_rate": 9.012709784266688e-06,
+      "loss": 0.8403,
+      "step": 23156
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6597323648218616,
+      "learning_rate": 9.011950474028697e-06,
+      "loss": 0.8393,
+      "step": 23157
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7794098722098571,
+      "learning_rate": 9.011191169543444e-06,
+      "loss": 0.716,
+      "step": 23158
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.530102387828581,
+      "learning_rate": 9.010431870815351e-06,
+      "loss": 0.8974,
+      "step": 23159
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7715531498472739,
+      "learning_rate": 9.009672577848836e-06,
+      "loss": 0.7389,
+      "step": 23160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4064600493761288,
+      "learning_rate": 9.008913290648317e-06,
+      "loss": 0.8516,
+      "step": 23161
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.582779530970621,
+      "learning_rate": 9.008154009218228e-06,
+      "loss": 0.9045,
+      "step": 23162
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7789627510397492,
+      "learning_rate": 9.00739473356297e-06,
+      "loss": 0.8077,
+      "step": 23163
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.618208438646683,
+      "learning_rate": 9.00663546368698e-06,
+      "loss": 0.8223,
+      "step": 23164
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8029631798536863,
+      "learning_rate": 9.005876199594668e-06,
+      "loss": 0.8372,
+      "step": 23165
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9956336925674774,
+      "learning_rate": 9.005116941290465e-06,
+      "loss": 0.8709,
+      "step": 23166
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4381293872810004,
+      "learning_rate": 9.004357688778784e-06,
+      "loss": 0.8491,
+      "step": 23167
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7376776359508606,
+      "learning_rate": 9.003598442064048e-06,
+      "loss": 0.7209,
+      "step": 23168
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7486105682292585,
+      "learning_rate": 9.002839201150673e-06,
+      "loss": 0.7191,
+      "step": 23169
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8894116065052389,
+      "learning_rate": 9.00207996604309e-06,
+      "loss": 0.7558,
+      "step": 23170
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9645352809922372,
+      "learning_rate": 9.001320736745711e-06,
+      "loss": 0.7951,
+      "step": 23171
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.129650747770493,
+      "learning_rate": 9.00056151326296e-06,
+      "loss": 0.8584,
+      "step": 23172
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3733990880827383,
+      "learning_rate": 8.999802295599257e-06,
+      "loss": 0.8043,
+      "step": 23173
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1625242916062657,
+      "learning_rate": 8.999043083759016e-06,
+      "loss": 0.8224,
+      "step": 23174
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5471188691167426,
+      "learning_rate": 8.998283877746669e-06,
+      "loss": 0.8416,
+      "step": 23175
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6431482805687887,
+      "learning_rate": 8.997524677566627e-06,
+      "loss": 0.8337,
+      "step": 23176
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6406224976851194,
+      "learning_rate": 8.996765483223317e-06,
+      "loss": 0.9785,
+      "step": 23177
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.557094126373792,
+      "learning_rate": 8.996006294721153e-06,
+      "loss": 0.8123,
+      "step": 23178
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9336563638461723,
+      "learning_rate": 8.99524711206456e-06,
+      "loss": 0.716,
+      "step": 23179
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8787029920374274,
+      "learning_rate": 8.994487935257957e-06,
+      "loss": 0.9298,
+      "step": 23180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.186972749447638,
+      "learning_rate": 8.993728764305762e-06,
+      "loss": 0.9104,
+      "step": 23181
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8535319383237965,
+      "learning_rate": 8.992969599212395e-06,
+      "loss": 0.77,
+      "step": 23182
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6949913741802891,
+      "learning_rate": 8.99221043998228e-06,
+      "loss": 0.8297,
+      "step": 23183
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4929933518218463,
+      "learning_rate": 8.991451286619837e-06,
+      "loss": 0.8169,
+      "step": 23184
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3920414683567945,
+      "learning_rate": 8.990692139129481e-06,
+      "loss": 0.867,
+      "step": 23185
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9105313422260053,
+      "learning_rate": 8.989932997515634e-06,
+      "loss": 0.9016,
+      "step": 23186
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7638722862229484,
+      "learning_rate": 8.98917386178272e-06,
+      "loss": 0.7219,
+      "step": 23187
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5147936903210735,
+      "learning_rate": 8.988414731935153e-06,
+      "loss": 0.8624,
+      "step": 23188
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8407181064478244,
+      "learning_rate": 8.987655607977357e-06,
+      "loss": 0.9024,
+      "step": 23189
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7564209010399969,
+      "learning_rate": 8.98689648991375e-06,
+      "loss": 0.6494,
+      "step": 23190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3945145835777115,
+      "learning_rate": 8.986137377748754e-06,
+      "loss": 0.8439,
+      "step": 23191
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6882811510534297,
+      "learning_rate": 8.985378271486786e-06,
+      "loss": 0.8206,
+      "step": 23192
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8267191435497068,
+      "learning_rate": 8.984619171132264e-06,
+      "loss": 0.8475,
+      "step": 23193
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5894004186962811,
+      "learning_rate": 8.983860076689617e-06,
+      "loss": 0.7561,
+      "step": 23194
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4984979611693878,
+      "learning_rate": 8.983100988163253e-06,
+      "loss": 0.8092,
+      "step": 23195
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3400976775655167,
+      "learning_rate": 8.982341905557599e-06,
+      "loss": 0.7406,
+      "step": 23196
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1682128777523832,
+      "learning_rate": 8.981582828877072e-06,
+      "loss": 0.9151,
+      "step": 23197
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5870794932994938,
+      "learning_rate": 8.980823758126096e-06,
+      "loss": 0.7137,
+      "step": 23198
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6173908504524857,
+      "learning_rate": 8.98006469330908e-06,
+      "loss": 0.9663,
+      "step": 23199
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4724796549203474,
+      "learning_rate": 8.979305634430453e-06,
+      "loss": 0.8976,
+      "step": 23200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4319826103644384,
+      "learning_rate": 8.978546581494637e-06,
+      "loss": 0.9545,
+      "step": 23201
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7347143363862317,
+      "learning_rate": 8.977787534506038e-06,
+      "loss": 0.6755,
+      "step": 23202
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6136955775406772,
+      "learning_rate": 8.977028493469088e-06,
+      "loss": 0.8687,
+      "step": 23203
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7917005201638172,
+      "learning_rate": 8.976269458388201e-06,
+      "loss": 0.7367,
+      "step": 23204
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6569423634512366,
+      "learning_rate": 8.9755104292678e-06,
+      "loss": 0.8245,
+      "step": 23205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6373200562249373,
+      "learning_rate": 8.974751406112296e-06,
+      "loss": 0.8453,
+      "step": 23206
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9057430489200056,
+      "learning_rate": 8.973992388926117e-06,
+      "loss": 0.8467,
+      "step": 23207
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7692852961507656,
+      "learning_rate": 8.973233377713681e-06,
+      "loss": 0.6686,
+      "step": 23208
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.693882452805859,
+      "learning_rate": 8.972474372479404e-06,
+      "loss": 0.8572,
+      "step": 23209
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.337968253009703,
+      "learning_rate": 8.971715373227704e-06,
+      "loss": 0.9282,
+      "step": 23210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1315734641796653,
+      "learning_rate": 8.970956379963006e-06,
+      "loss": 0.7982,
+      "step": 23211
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5096959396247118,
+      "learning_rate": 8.970197392689725e-06,
+      "loss": 0.8716,
+      "step": 23212
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5167432261070677,
+      "learning_rate": 8.96943841141228e-06,
+      "loss": 0.8882,
+      "step": 23213
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.586560662643379,
+      "learning_rate": 8.96867943613509e-06,
+      "loss": 0.986,
+      "step": 23214
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5034910474951537,
+      "learning_rate": 8.967920466862577e-06,
+      "loss": 0.9392,
+      "step": 23215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0700663288794616,
+      "learning_rate": 8.967161503599157e-06,
+      "loss": 0.8946,
+      "step": 23216
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6397717353136358,
+      "learning_rate": 8.96640254634925e-06,
+      "loss": 0.8055,
+      "step": 23217
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6481939989306837,
+      "learning_rate": 8.96564359511727e-06,
+      "loss": 0.826,
+      "step": 23218
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.604900516892501,
+      "learning_rate": 8.964884649907648e-06,
+      "loss": 0.7503,
+      "step": 23219
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5062422774255906,
+      "learning_rate": 8.964125710724791e-06,
+      "loss": 0.7995,
+      "step": 23220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5020689653378239,
+      "learning_rate": 8.963366777573123e-06,
+      "loss": 0.9549,
+      "step": 23221
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6216552738151047,
+      "learning_rate": 8.962607850457063e-06,
+      "loss": 0.8062,
+      "step": 23222
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9442856268167095,
+      "learning_rate": 8.961848929381027e-06,
+      "loss": 0.8854,
+      "step": 23223
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.645894101826824,
+      "learning_rate": 8.961090014349433e-06,
+      "loss": 0.9167,
+      "step": 23224
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6490334121295855,
+      "learning_rate": 8.960331105366705e-06,
+      "loss": 0.754,
+      "step": 23225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5952505018165186,
+      "learning_rate": 8.95957220243726e-06,
+      "loss": 0.7279,
+      "step": 23226
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7626212416416698,
+      "learning_rate": 8.95881330556551e-06,
+      "loss": 0.6568,
+      "step": 23227
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5711201256867853,
+      "learning_rate": 8.958054414755882e-06,
+      "loss": 0.861,
+      "step": 23228
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7875830930387797,
+      "learning_rate": 8.957295530012792e-06,
+      "loss": 0.9151,
+      "step": 23229
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8731328214031566,
+      "learning_rate": 8.956536651340655e-06,
+      "loss": 0.7925,
+      "step": 23230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8581902151905307,
+      "learning_rate": 8.955777778743891e-06,
+      "loss": 0.8371,
+      "step": 23231
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.232319904194559,
+      "learning_rate": 8.95501891222692e-06,
+      "loss": 0.7726,
+      "step": 23232
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4553473618738033,
+      "learning_rate": 8.954260051794165e-06,
+      "loss": 0.752,
+      "step": 23233
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8412000405658173,
+      "learning_rate": 8.953501197450035e-06,
+      "loss": 0.877,
+      "step": 23234
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5268115611109763,
+      "learning_rate": 8.952742349198951e-06,
+      "loss": 0.8518,
+      "step": 23235
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5073814416562152,
+      "learning_rate": 8.951983507045335e-06,
+      "loss": 0.8821,
+      "step": 23236
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8728563440034882,
+      "learning_rate": 8.951224670993603e-06,
+      "loss": 0.7817,
+      "step": 23237
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.731122908459255,
+      "learning_rate": 8.950465841048168e-06,
+      "loss": 0.6685,
+      "step": 23238
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4033464675375942,
+      "learning_rate": 8.949707017213457e-06,
+      "loss": 0.7777,
+      "step": 23239
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5535757281270164,
+      "learning_rate": 8.948948199493887e-06,
+      "loss": 0.9623,
+      "step": 23240
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6699714454256505,
+      "learning_rate": 8.948189387893871e-06,
+      "loss": 0.8276,
+      "step": 23241
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5600967465617166,
+      "learning_rate": 8.947430582417827e-06,
+      "loss": 0.9298,
+      "step": 23242
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8502863949949948,
+      "learning_rate": 8.946671783070182e-06,
+      "loss": 0.9593,
+      "step": 23243
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5189194933326757,
+      "learning_rate": 8.94591298985534e-06,
+      "loss": 0.8905,
+      "step": 23244
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5811886982256003,
+      "learning_rate": 8.945154202777729e-06,
+      "loss": 0.8231,
+      "step": 23245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7052328238689312,
+      "learning_rate": 8.94439542184176e-06,
+      "loss": 0.7739,
+      "step": 23246
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7180840263049192,
+      "learning_rate": 8.943636647051864e-06,
+      "loss": 0.7207,
+      "step": 23247
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4163210640331467,
+      "learning_rate": 8.942877878412444e-06,
+      "loss": 0.8114,
+      "step": 23248
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.664630839989858,
+      "learning_rate": 8.942119115927925e-06,
+      "loss": 0.8931,
+      "step": 23249
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5553358556858605,
+      "learning_rate": 8.941360359602724e-06,
+      "loss": 0.8162,
+      "step": 23250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7662684439695961,
+      "learning_rate": 8.940601609441256e-06,
+      "loss": 0.6684,
+      "step": 23251
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.04648151753552,
+      "learning_rate": 8.93984286544794e-06,
+      "loss": 0.8573,
+      "step": 23252
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7671867425004284,
+      "learning_rate": 8.939084127627198e-06,
+      "loss": 0.8213,
+      "step": 23253
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.662612031357752,
+      "learning_rate": 8.938325395983445e-06,
+      "loss": 0.8284,
+      "step": 23254
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5495652385526149,
+      "learning_rate": 8.937566670521092e-06,
+      "loss": 0.7584,
+      "step": 23255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8850614383996738,
+      "learning_rate": 8.936807951244566e-06,
+      "loss": 0.8162,
+      "step": 23256
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5610838253964603,
+      "learning_rate": 8.936049238158283e-06,
+      "loss": 0.8957,
+      "step": 23257
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5447773315523783,
+      "learning_rate": 8.935290531266656e-06,
+      "loss": 0.8579,
+      "step": 23258
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6049827706445516,
+      "learning_rate": 8.934531830574101e-06,
+      "loss": 0.8467,
+      "step": 23259
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5838305217663111,
+      "learning_rate": 8.933773136085042e-06,
+      "loss": 0.9583,
+      "step": 23260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4418764357476104,
+      "learning_rate": 8.933014447803896e-06,
+      "loss": 0.8749,
+      "step": 23261
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5328474434998816,
+      "learning_rate": 8.932255765735075e-06,
+      "loss": 0.8114,
+      "step": 23262
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7950366988647048,
+      "learning_rate": 8.931497089882997e-06,
+      "loss": 0.999,
+      "step": 23263
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6711591559884238,
+      "learning_rate": 8.930738420252085e-06,
+      "loss": 0.8002,
+      "step": 23264
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7036505935711113,
+      "learning_rate": 8.92997975684675e-06,
+      "loss": 0.8591,
+      "step": 23265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8140683340754331,
+      "learning_rate": 8.929221099671414e-06,
+      "loss": 0.7384,
+      "step": 23266
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7369016275070142,
+      "learning_rate": 8.928462448730487e-06,
+      "loss": 0.8739,
+      "step": 23267
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5568117746102021,
+      "learning_rate": 8.927703804028396e-06,
+      "loss": 0.9708,
+      "step": 23268
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8967386279491463,
+      "learning_rate": 8.926945165569551e-06,
+      "loss": 0.9069,
+      "step": 23269
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7513564189302988,
+      "learning_rate": 8.92618653335837e-06,
+      "loss": 0.8953,
+      "step": 23270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6053668335543017,
+      "learning_rate": 8.925427907399276e-06,
+      "loss": 0.859,
+      "step": 23271
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7187350122112581,
+      "learning_rate": 8.924669287696673e-06,
+      "loss": 0.6951,
+      "step": 23272
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5714769567607119,
+      "learning_rate": 8.92391067425499e-06,
+      "loss": 0.8789,
+      "step": 23273
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7623579796503686,
+      "learning_rate": 8.92315206707864e-06,
+      "loss": 0.7748,
+      "step": 23274
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5149744863137633,
+      "learning_rate": 8.922393466172042e-06,
+      "loss": 0.743,
+      "step": 23275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5830386503227847,
+      "learning_rate": 8.921634871539603e-06,
+      "loss": 0.9214,
+      "step": 23276
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5856297879988208,
+      "learning_rate": 8.920876283185753e-06,
+      "loss": 0.8857,
+      "step": 23277
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6114117429812556,
+      "learning_rate": 8.920117701114903e-06,
+      "loss": 0.8461,
+      "step": 23278
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7784934593170848,
+      "learning_rate": 8.919359125331468e-06,
+      "loss": 0.7553,
+      "step": 23279
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4120473733902381,
+      "learning_rate": 8.918600555839865e-06,
+      "loss": 0.7513,
+      "step": 23280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4299399187570692,
+      "learning_rate": 8.917841992644513e-06,
+      "loss": 0.7844,
+      "step": 23281
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.672131058987703,
+      "learning_rate": 8.91708343574983e-06,
+      "loss": 0.8078,
+      "step": 23282
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.613828602032894,
+      "learning_rate": 8.916324885160228e-06,
+      "loss": 0.798,
+      "step": 23283
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7340377844803485,
+      "learning_rate": 8.915566340880122e-06,
+      "loss": 0.8222,
+      "step": 23284
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5684933653982773,
+      "learning_rate": 8.914807802913939e-06,
+      "loss": 0.8508,
+      "step": 23285
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4614675060564912,
+      "learning_rate": 8.914049271266085e-06,
+      "loss": 0.8214,
+      "step": 23286
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3968320886128454,
+      "learning_rate": 8.913290745940979e-06,
+      "loss": 0.8087,
+      "step": 23287
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.704859454325803,
+      "learning_rate": 8.912532226943038e-06,
+      "loss": 1.0112,
+      "step": 23288
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6748918083701816,
+      "learning_rate": 8.911773714276682e-06,
+      "loss": 0.9102,
+      "step": 23289
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5163025966640427,
+      "learning_rate": 8.911015207946324e-06,
+      "loss": 0.769,
+      "step": 23290
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7000606813663173,
+      "learning_rate": 8.910256707956375e-06,
+      "loss": 0.7982,
+      "step": 23291
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.724743681260214,
+      "learning_rate": 8.90949821431126e-06,
+      "loss": 0.9445,
+      "step": 23292
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.464670218838483,
+      "learning_rate": 8.908739727015394e-06,
+      "loss": 0.8367,
+      "step": 23293
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.531821348510884,
+      "learning_rate": 8.907981246073188e-06,
+      "loss": 0.9126,
+      "step": 23294
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9390434787689612,
+      "learning_rate": 8.907222771489059e-06,
+      "loss": 0.8946,
+      "step": 23295
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5020546276408047,
+      "learning_rate": 8.90646430326743e-06,
+      "loss": 0.9346,
+      "step": 23296
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6135226136540248,
+      "learning_rate": 8.905705841412708e-06,
+      "loss": 0.85,
+      "step": 23297
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7478748585867108,
+      "learning_rate": 8.904947385929314e-06,
+      "loss": 0.8018,
+      "step": 23298
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5898250769596736,
+      "learning_rate": 8.90418893682166e-06,
+      "loss": 0.8917,
+      "step": 23299
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5133042238766654,
+      "learning_rate": 8.903430494094171e-06,
+      "loss": 0.9597,
+      "step": 23300
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2060339291019346,
+      "learning_rate": 8.902672057751254e-06,
+      "loss": 0.838,
+      "step": 23301
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4342607072134947,
+      "learning_rate": 8.901913627797328e-06,
+      "loss": 0.8579,
+      "step": 23302
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6337501095630358,
+      "learning_rate": 8.90115520423681e-06,
+      "loss": 0.8474,
+      "step": 23303
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.050275177791525,
+      "learning_rate": 8.90039678707411e-06,
+      "loss": 0.7948,
+      "step": 23304
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.637134221944142,
+      "learning_rate": 8.899638376313651e-06,
+      "loss": 0.7528,
+      "step": 23305
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.646578388761976,
+      "learning_rate": 8.898879971959844e-06,
+      "loss": 0.8267,
+      "step": 23306
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5037911180605088,
+      "learning_rate": 8.89812157401711e-06,
+      "loss": 0.9227,
+      "step": 23307
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9539114900694226,
+      "learning_rate": 8.897363182489856e-06,
+      "loss": 0.83,
+      "step": 23308
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7896506658614828,
+      "learning_rate": 8.896604797382507e-06,
+      "loss": 0.8758,
+      "step": 23309
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5468804228466642,
+      "learning_rate": 8.895846418699475e-06,
+      "loss": 0.898,
+      "step": 23310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.539774671850541,
+      "learning_rate": 8.89508804644517e-06,
+      "loss": 0.8002,
+      "step": 23311
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.650942522184648,
+      "learning_rate": 8.894329680624012e-06,
+      "loss": 0.9713,
+      "step": 23312
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.723840676125585,
+      "learning_rate": 8.89357132124042e-06,
+      "loss": 0.734,
+      "step": 23313
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5204019682511092,
+      "learning_rate": 8.892812968298807e-06,
+      "loss": 1.0139,
+      "step": 23314
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7974969017879011,
+      "learning_rate": 8.892054621803586e-06,
+      "loss": 0.7962,
+      "step": 23315
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0950794912037365,
+      "learning_rate": 8.89129628175917e-06,
+      "loss": 0.8954,
+      "step": 23316
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6051844951563463,
+      "learning_rate": 8.890537948169984e-06,
+      "loss": 0.9124,
+      "step": 23317
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.570795135158842,
+      "learning_rate": 8.889779621040435e-06,
+      "loss": 0.795,
+      "step": 23318
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6264217728741357,
+      "learning_rate": 8.88902130037494e-06,
+      "loss": 0.9108,
+      "step": 23319
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5518575996954183,
+      "learning_rate": 8.888262986177913e-06,
+      "loss": 0.9075,
+      "step": 23320
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.909269056277327,
+      "learning_rate": 8.887504678453775e-06,
+      "loss": 0.8343,
+      "step": 23321
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2834204082628684,
+      "learning_rate": 8.886746377206934e-06,
+      "loss": 0.8389,
+      "step": 23322
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5892402477147576,
+      "learning_rate": 8.88598808244181e-06,
+      "loss": 0.8156,
+      "step": 23323
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.644542512867851,
+      "learning_rate": 8.885229794162818e-06,
+      "loss": 0.8445,
+      "step": 23324
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6284980536010396,
+      "learning_rate": 8.884471512374363e-06,
+      "loss": 0.7957,
+      "step": 23325
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.159614053673555,
+      "learning_rate": 8.883713237080875e-06,
+      "loss": 0.8475,
+      "step": 23326
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3496879764682768,
+      "learning_rate": 8.882954968286756e-06,
+      "loss": 0.9955,
+      "step": 23327
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6286149679946957,
+      "learning_rate": 8.882196705996436e-06,
+      "loss": 0.7884,
+      "step": 23328
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8024623149976267,
+      "learning_rate": 8.881438450214313e-06,
+      "loss": 0.7834,
+      "step": 23329
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7770762559759734,
+      "learning_rate": 8.880680200944812e-06,
+      "loss": 0.8574,
+      "step": 23330
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7383388055189495,
+      "learning_rate": 8.879921958192347e-06,
+      "loss": 0.848,
+      "step": 23331
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5430486593876038,
+      "learning_rate": 8.87916372196133e-06,
+      "loss": 0.8452,
+      "step": 23332
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6093168393914272,
+      "learning_rate": 8.878405492256174e-06,
+      "loss": 0.9218,
+      "step": 23333
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.626778540229173,
+      "learning_rate": 8.877647269081298e-06,
+      "loss": 0.8525,
+      "step": 23334
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8010938951211453,
+      "learning_rate": 8.876889052441117e-06,
+      "loss": 0.8619,
+      "step": 23335
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6374576334252078,
+      "learning_rate": 8.876130842340039e-06,
+      "loss": 0.9816,
+      "step": 23336
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5381916558740885,
+      "learning_rate": 8.875372638782485e-06,
+      "loss": 0.9459,
+      "step": 23337
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7769785454111156,
+      "learning_rate": 8.874614441772871e-06,
+      "loss": 0.7781,
+      "step": 23338
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.686791097234361,
+      "learning_rate": 8.873856251315605e-06,
+      "loss": 0.8699,
+      "step": 23339
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7883708038998247,
+      "learning_rate": 8.873098067415103e-06,
+      "loss": 0.786,
+      "step": 23340
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.607491326725573,
+      "learning_rate": 8.872339890075782e-06,
+      "loss": 0.9218,
+      "step": 23341
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5989477577978377,
+      "learning_rate": 8.871581719302057e-06,
+      "loss": 0.8397,
+      "step": 23342
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.568340242979973,
+      "learning_rate": 8.870823555098339e-06,
+      "loss": 0.8506,
+      "step": 23343
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7201834865992145,
+      "learning_rate": 8.870065397469043e-06,
+      "loss": 0.9194,
+      "step": 23344
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8032691581193367,
+      "learning_rate": 8.869307246418589e-06,
+      "loss": 0.9054,
+      "step": 23345
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5892943874098209,
+      "learning_rate": 8.86854910195138e-06,
+      "loss": 0.8386,
+      "step": 23346
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.471596445185681,
+      "learning_rate": 8.867790964071838e-06,
+      "loss": 0.7613,
+      "step": 23347
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7959253084840686,
+      "learning_rate": 8.867032832784372e-06,
+      "loss": 0.7063,
+      "step": 23348
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.442608322745331,
+      "learning_rate": 8.866274708093407e-06,
+      "loss": 0.9218,
+      "step": 23349
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6378805614334446,
+      "learning_rate": 8.865516590003344e-06,
+      "loss": 0.9706,
+      "step": 23350
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5414336128538362,
+      "learning_rate": 8.864758478518606e-06,
+      "loss": 0.7493,
+      "step": 23351
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9543497454549619,
+      "learning_rate": 8.864000373643604e-06,
+      "loss": 0.9076,
+      "step": 23352
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.501769896102029,
+      "learning_rate": 8.863242275382747e-06,
+      "loss": 0.8764,
+      "step": 23353
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.761293905253321,
+      "learning_rate": 8.862484183740457e-06,
+      "loss": 0.7368,
+      "step": 23354
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.115591901490947,
+      "learning_rate": 8.861726098721142e-06,
+      "loss": 0.8395,
+      "step": 23355
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8193345649580434,
+      "learning_rate": 8.860968020329222e-06,
+      "loss": 0.7233,
+      "step": 23356
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4353159177385257,
+      "learning_rate": 8.8602099485691e-06,
+      "loss": 0.8045,
+      "step": 23357
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6922379579879356,
+      "learning_rate": 8.859451883445201e-06,
+      "loss": 0.9064,
+      "step": 23358
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7893220488757593,
+      "learning_rate": 8.858693824961936e-06,
+      "loss": 0.8462,
+      "step": 23359
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3808547775731246,
+      "learning_rate": 8.857935773123715e-06,
+      "loss": 0.7807,
+      "step": 23360
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.161681441782409,
+      "learning_rate": 8.85717772793495e-06,
+      "loss": 0.8316,
+      "step": 23361
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.782574523392062,
+      "learning_rate": 8.856419689400062e-06,
+      "loss": 0.9054,
+      "step": 23362
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5816522457020232,
+      "learning_rate": 8.855661657523461e-06,
+      "loss": 0.863,
+      "step": 23363
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8706302185320667,
+      "learning_rate": 8.85490363230956e-06,
+      "loss": 0.8111,
+      "step": 23364
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.723497096459199,
+      "learning_rate": 8.854145613762767e-06,
+      "loss": 0.841,
+      "step": 23365
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.567536041867134,
+      "learning_rate": 8.85338760188751e-06,
+      "loss": 0.9093,
+      "step": 23366
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7577280335523432,
+      "learning_rate": 8.852629596688189e-06,
+      "loss": 0.697,
+      "step": 23367
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.509940766559824,
+      "learning_rate": 8.85187159816922e-06,
+      "loss": 0.9145,
+      "step": 23368
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6276097883147271,
+      "learning_rate": 8.851113606335018e-06,
+      "loss": 0.9701,
+      "step": 23369
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7221596077375891,
+      "learning_rate": 8.85035562119e-06,
+      "loss": 0.898,
+      "step": 23370
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8042479053321405,
+      "learning_rate": 8.849597642738573e-06,
+      "loss": 0.7252,
+      "step": 23371
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.008499364340877,
+      "learning_rate": 8.848839670985151e-06,
+      "loss": 0.9159,
+      "step": 23372
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.614513490823274,
+      "learning_rate": 8.848081705934155e-06,
+      "loss": 0.8737,
+      "step": 23373
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.801415914513901,
+      "learning_rate": 8.847323747589987e-06,
+      "loss": 0.6988,
+      "step": 23374
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6723653733804245,
+      "learning_rate": 8.846565795957066e-06,
+      "loss": 0.9107,
+      "step": 23375
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6304313108577175,
+      "learning_rate": 8.845807851039801e-06,
+      "loss": 0.7498,
+      "step": 23376
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1434517823492243,
+      "learning_rate": 8.845049912842616e-06,
+      "loss": 0.7425,
+      "step": 23377
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6616895276860348,
+      "learning_rate": 8.844291981369911e-06,
+      "loss": 0.8394,
+      "step": 23378
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5972186092633365,
+      "learning_rate": 8.843534056626103e-06,
+      "loss": 0.8207,
+      "step": 23379
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.566002668748403,
+      "learning_rate": 8.842776138615609e-06,
+      "loss": 0.7377,
+      "step": 23380
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8333012858818463,
+      "learning_rate": 8.842018227342839e-06,
+      "loss": 0.786,
+      "step": 23381
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6743412360850256,
+      "learning_rate": 8.8412603228122e-06,
+      "loss": 0.8266,
+      "step": 23382
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6528798270745608,
+      "learning_rate": 8.840502425028114e-06,
+      "loss": 0.8144,
+      "step": 23383
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6241739800991986,
+      "learning_rate": 8.839744533994991e-06,
+      "loss": 0.8569,
+      "step": 23384
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6969207942480984,
+      "learning_rate": 8.83898664971724e-06,
+      "loss": 0.9656,
+      "step": 23385
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7408868925868632,
+      "learning_rate": 8.838228772199278e-06,
+      "loss": 0.7803,
+      "step": 23386
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5707529373904636,
+      "learning_rate": 8.837470901445518e-06,
+      "loss": 0.8208,
+      "step": 23387
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6820718284274407,
+      "learning_rate": 8.836713037460368e-06,
+      "loss": 0.8357,
+      "step": 23388
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6517772441088423,
+      "learning_rate": 8.83595518024824e-06,
+      "loss": 0.8903,
+      "step": 23389
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3993476139514127,
+      "learning_rate": 8.835197329813552e-06,
+      "loss": 0.8292,
+      "step": 23390
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6050257162882953,
+      "learning_rate": 8.834439486160716e-06,
+      "loss": 0.8748,
+      "step": 23391
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7590343988595663,
+      "learning_rate": 8.833681649294142e-06,
+      "loss": 0.7543,
+      "step": 23392
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4354539095533507,
+      "learning_rate": 8.832923819218238e-06,
+      "loss": 0.8253,
+      "step": 23393
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.758975369520572,
+      "learning_rate": 8.83216599593743e-06,
+      "loss": 0.8296,
+      "step": 23394
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8079096375691052,
+      "learning_rate": 8.831408179456112e-06,
+      "loss": 0.7186,
+      "step": 23395
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.759551925436874,
+      "learning_rate": 8.830650369778711e-06,
+      "loss": 0.9085,
+      "step": 23396
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5793424820937143,
+      "learning_rate": 8.829892566909629e-06,
+      "loss": 0.8856,
+      "step": 23397
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7446600060339321,
+      "learning_rate": 8.829134770853289e-06,
+      "loss": 0.6588,
+      "step": 23398
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.657917223119895,
+      "learning_rate": 8.828376981614093e-06,
+      "loss": 0.9364,
+      "step": 23399
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6362786406713243,
+      "learning_rate": 8.827619199196457e-06,
+      "loss": 0.8996,
+      "step": 23400
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6121637560319597,
+      "learning_rate": 8.826861423604797e-06,
+      "loss": 0.7866,
+      "step": 23401
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6912445003648422,
+      "learning_rate": 8.826103654843516e-06,
+      "loss": 0.8439,
+      "step": 23402
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.663016048097996,
+      "learning_rate": 8.825345892917035e-06,
+      "loss": 0.7543,
+      "step": 23403
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7936296241005031,
+      "learning_rate": 8.824588137829761e-06,
+      "loss": 0.7725,
+      "step": 23404
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6099324966636748,
+      "learning_rate": 8.823830389586108e-06,
+      "loss": 0.7922,
+      "step": 23405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7903614040765568,
+      "learning_rate": 8.823072648190484e-06,
+      "loss": 0.9435,
+      "step": 23406
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6304179408818051,
+      "learning_rate": 8.822314913647305e-06,
+      "loss": 0.9596,
+      "step": 23407
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.672711156040143,
+      "learning_rate": 8.821557185960984e-06,
+      "loss": 0.7766,
+      "step": 23408
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4571570772028477,
+      "learning_rate": 8.820799465135928e-06,
+      "loss": 0.7537,
+      "step": 23409
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.414292693089927,
+      "learning_rate": 8.820041751176549e-06,
+      "loss": 0.8676,
+      "step": 23410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6528395621608534,
+      "learning_rate": 8.81928404408726e-06,
+      "loss": 0.832,
+      "step": 23411
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6233878530815553,
+      "learning_rate": 8.81852634387248e-06,
+      "loss": 0.8213,
+      "step": 23412
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5427376977033607,
+      "learning_rate": 8.817768650536607e-06,
+      "loss": 0.8675,
+      "step": 23413
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.459366887016306,
+      "learning_rate": 8.817010964084059e-06,
+      "loss": 0.8969,
+      "step": 23414
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.642380788833524,
+      "learning_rate": 8.81625328451925e-06,
+      "loss": 0.8417,
+      "step": 23415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.712133130132426,
+      "learning_rate": 8.815495611846592e-06,
+      "loss": 0.7565,
+      "step": 23416
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8243855611885833,
+      "learning_rate": 8.814737946070489e-06,
+      "loss": 0.7449,
+      "step": 23417
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.763351894362198,
+      "learning_rate": 8.813980287195358e-06,
+      "loss": 0.8265,
+      "step": 23418
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7391685300642623,
+      "learning_rate": 8.81322263522561e-06,
+      "loss": 0.9076,
+      "step": 23419
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5532338314717178,
+      "learning_rate": 8.812464990165656e-06,
+      "loss": 0.8369,
+      "step": 23420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7204215726862306,
+      "learning_rate": 8.811707352019905e-06,
+      "loss": 0.8765,
+      "step": 23421
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5577800817276664,
+      "learning_rate": 8.810949720792771e-06,
+      "loss": 0.8871,
+      "step": 23422
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6179099972169972,
+      "learning_rate": 8.810192096488666e-06,
+      "loss": 0.9492,
+      "step": 23423
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4800130859557694,
+      "learning_rate": 8.809434479111998e-06,
+      "loss": 0.8818,
+      "step": 23424
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.3811418843401775,
+      "learning_rate": 8.808676868667177e-06,
+      "loss": 0.8576,
+      "step": 23425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7445233312353736,
+      "learning_rate": 8.807919265158622e-06,
+      "loss": 0.9616,
+      "step": 23426
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.563203991085246,
+      "learning_rate": 8.807161668590733e-06,
+      "loss": 0.8091,
+      "step": 23427
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9471579537353407,
+      "learning_rate": 8.806404078967928e-06,
+      "loss": 0.9487,
+      "step": 23428
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6642867309982101,
+      "learning_rate": 8.805646496294613e-06,
+      "loss": 0.7565,
+      "step": 23429
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.503319953450312,
+      "learning_rate": 8.80488892057521e-06,
+      "loss": 0.8359,
+      "step": 23430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.388135518658166,
+      "learning_rate": 8.804131351814117e-06,
+      "loss": 0.825,
+      "step": 23431
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3941433561008532,
+      "learning_rate": 8.80337379001575e-06,
+      "loss": 0.8783,
+      "step": 23432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5703319924697725,
+      "learning_rate": 8.802616235184524e-06,
+      "loss": 0.8432,
+      "step": 23433
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4808386013809196,
+      "learning_rate": 8.801858687324839e-06,
+      "loss": 0.8494,
+      "step": 23434
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7922377785609385,
+      "learning_rate": 8.801101146441114e-06,
+      "loss": 0.9704,
+      "step": 23435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7718141902243352,
+      "learning_rate": 8.800343612537758e-06,
+      "loss": 0.8017,
+      "step": 23436
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5336971511516135,
+      "learning_rate": 8.799586085619184e-06,
+      "loss": 0.8006,
+      "step": 23437
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6989432587727507,
+      "learning_rate": 8.798828565689796e-06,
+      "loss": 0.9628,
+      "step": 23438
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5873819806119662,
+      "learning_rate": 8.79807105275401e-06,
+      "loss": 0.7862,
+      "step": 23439
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3922191433424103,
+      "learning_rate": 8.797313546816238e-06,
+      "loss": 0.9191,
+      "step": 23440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5337729788153636,
+      "learning_rate": 8.796556047880883e-06,
+      "loss": 0.9494,
+      "step": 23441
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.605928732041886,
+      "learning_rate": 8.79579855595236e-06,
+      "loss": 0.8153,
+      "step": 23442
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.585095959552823,
+      "learning_rate": 8.795041071035078e-06,
+      "loss": 0.8669,
+      "step": 23443
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6255944312750414,
+      "learning_rate": 8.794283593133454e-06,
+      "loss": 0.738,
+      "step": 23444
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7586598690393678,
+      "learning_rate": 8.793526122251888e-06,
+      "loss": 0.7155,
+      "step": 23445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7208555859291013,
+      "learning_rate": 8.792768658394794e-06,
+      "loss": 0.8471,
+      "step": 23446
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5970693835244936,
+      "learning_rate": 8.792011201566587e-06,
+      "loss": 0.8589,
+      "step": 23447
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5976041936576533,
+      "learning_rate": 8.791253751771671e-06,
+      "loss": 1.0917,
+      "step": 23448
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5644924562042817,
+      "learning_rate": 8.790496309014457e-06,
+      "loss": 0.8689,
+      "step": 23449
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5955683908732632,
+      "learning_rate": 8.789738873299356e-06,
+      "loss": 0.9612,
+      "step": 23450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8272898600707508,
+      "learning_rate": 8.788981444630784e-06,
+      "loss": 0.9292,
+      "step": 23451
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6398860954659424,
+      "learning_rate": 8.78822402301314e-06,
+      "loss": 0.9597,
+      "step": 23452
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9397556542514456,
+      "learning_rate": 8.787466608450843e-06,
+      "loss": 0.7595,
+      "step": 23453
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.800371151823851,
+      "learning_rate": 8.786709200948299e-06,
+      "loss": 0.6881,
+      "step": 23454
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7480635136538574,
+      "learning_rate": 8.785951800509915e-06,
+      "loss": 0.7051,
+      "step": 23455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6736951404854474,
+      "learning_rate": 8.785194407140105e-06,
+      "loss": 0.8321,
+      "step": 23456
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.592732650745439,
+      "learning_rate": 8.784437020843278e-06,
+      "loss": 0.7631,
+      "step": 23457
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4683227407450132,
+      "learning_rate": 8.783679641623845e-06,
+      "loss": 0.8526,
+      "step": 23458
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.412295620922528,
+      "learning_rate": 8.782922269486211e-06,
+      "loss": 0.8429,
+      "step": 23459
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5628321745502651,
+      "learning_rate": 8.782164904434791e-06,
+      "loss": 0.824,
+      "step": 23460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5523176960076805,
+      "learning_rate": 8.781407546473995e-06,
+      "loss": 0.7855,
+      "step": 23461
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7126044222328882,
+      "learning_rate": 8.780650195608228e-06,
+      "loss": 0.8412,
+      "step": 23462
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.567746027116616,
+      "learning_rate": 8.779892851841897e-06,
+      "loss": 0.9374,
+      "step": 23463
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7757469177925155,
+      "learning_rate": 8.77913551517942e-06,
+      "loss": 0.7543,
+      "step": 23464
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8368234555793582,
+      "learning_rate": 8.778378185625206e-06,
+      "loss": 0.7667,
+      "step": 23465
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7930876540447511,
+      "learning_rate": 8.777620863183658e-06,
+      "loss": 0.7465,
+      "step": 23466
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4595192761088798,
+      "learning_rate": 8.776863547859185e-06,
+      "loss": 0.8877,
+      "step": 23467
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5432843843301454,
+      "learning_rate": 8.776106239656206e-06,
+      "loss": 0.8163,
+      "step": 23468
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7212004127501804,
+      "learning_rate": 8.77534893857912e-06,
+      "loss": 0.7093,
+      "step": 23469
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7701079610584022,
+      "learning_rate": 8.774591644632338e-06,
+      "loss": 0.6865,
+      "step": 23470
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3492674378883114,
+      "learning_rate": 8.773834357820275e-06,
+      "loss": 0.911,
+      "step": 23471
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7628814464642961,
+      "learning_rate": 8.773077078147337e-06,
+      "loss": 0.7481,
+      "step": 23472
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4609272631452934,
+      "learning_rate": 8.772319805617931e-06,
+      "loss": 0.8673,
+      "step": 23473
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5629762270997987,
+      "learning_rate": 8.771562540236467e-06,
+      "loss": 0.8171,
+      "step": 23474
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5495786670130516,
+      "learning_rate": 8.77080528200736e-06,
+      "loss": 0.8962,
+      "step": 23475
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2219265313095904,
+      "learning_rate": 8.770048030935007e-06,
+      "loss": 0.8256,
+      "step": 23476
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.05074346479166,
+      "learning_rate": 8.769290787023828e-06,
+      "loss": 0.889,
+      "step": 23477
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6894809398125459,
+      "learning_rate": 8.768533550278223e-06,
+      "loss": 0.9174,
+      "step": 23478
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7178872752109269,
+      "learning_rate": 8.767776320702614e-06,
+      "loss": 0.7673,
+      "step": 23479
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5283493460085125,
+      "learning_rate": 8.767019098301394e-06,
+      "loss": 0.836,
+      "step": 23480
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.823740198133042,
+      "learning_rate": 8.766261883078981e-06,
+      "loss": 0.8687,
+      "step": 23481
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3805988489644747,
+      "learning_rate": 8.765504675039784e-06,
+      "loss": 0.8488,
+      "step": 23482
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.743815663694007,
+      "learning_rate": 8.764747474188205e-06,
+      "loss": 0.9743,
+      "step": 23483
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6074513355750135,
+      "learning_rate": 8.763990280528661e-06,
+      "loss": 0.9351,
+      "step": 23484
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6277000952783256,
+      "learning_rate": 8.763233094065556e-06,
+      "loss": 0.7521,
+      "step": 23485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8288906844744224,
+      "learning_rate": 8.7624759148033e-06,
+      "loss": 0.7785,
+      "step": 23486
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7108136920634334,
+      "learning_rate": 8.7617187427463e-06,
+      "loss": 0.7825,
+      "step": 23487
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.456536969509168,
+      "learning_rate": 8.760961577898965e-06,
+      "loss": 0.8702,
+      "step": 23488
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6946932999131845,
+      "learning_rate": 8.760204420265707e-06,
+      "loss": 0.9479,
+      "step": 23489
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2097541351717793,
+      "learning_rate": 8.75944726985093e-06,
+      "loss": 0.897,
+      "step": 23490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8369017481906413,
+      "learning_rate": 8.758690126659041e-06,
+      "loss": 0.7507,
+      "step": 23491
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7814148982373927,
+      "learning_rate": 8.757932990694454e-06,
+      "loss": 0.7324,
+      "step": 23492
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8587628942195626,
+      "learning_rate": 8.757175861961576e-06,
+      "loss": 0.9226,
+      "step": 23493
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.657924831184941,
+      "learning_rate": 8.75641874046481e-06,
+      "loss": 0.8925,
+      "step": 23494
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.827710552734634,
+      "learning_rate": 8.755661626208568e-06,
+      "loss": 0.7609,
+      "step": 23495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5408486243978043,
+      "learning_rate": 8.75490451919726e-06,
+      "loss": 0.8438,
+      "step": 23496
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5789110350615947,
+      "learning_rate": 8.754147419435291e-06,
+      "loss": 0.8136,
+      "step": 23497
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4952202383160451,
+      "learning_rate": 8.75339032692707e-06,
+      "loss": 0.8815,
+      "step": 23498
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4750581192351708,
+      "learning_rate": 8.752633241677003e-06,
+      "loss": 0.8216,
+      "step": 23499
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8858416075410813,
+      "learning_rate": 8.751876163689506e-06,
+      "loss": 0.8808,
+      "step": 23500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8073601577720413,
+      "learning_rate": 8.751119092968978e-06,
+      "loss": 0.8963,
+      "step": 23501
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6396992609787717,
+      "learning_rate": 8.75036202951983e-06,
+      "loss": 0.7994,
+      "step": 23502
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7463649742389347,
+      "learning_rate": 8.74960497334647e-06,
+      "loss": 0.8696,
+      "step": 23503
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.646379660697829,
+      "learning_rate": 8.748847924453304e-06,
+      "loss": 0.7899,
+      "step": 23504
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5916391777071397,
+      "learning_rate": 8.748090882844742e-06,
+      "loss": 0.9067,
+      "step": 23505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7949279526002512,
+      "learning_rate": 8.747333848525193e-06,
+      "loss": 0.7028,
+      "step": 23506
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6834118895527652,
+      "learning_rate": 8.746576821499062e-06,
+      "loss": 0.8861,
+      "step": 23507
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8692439218375676,
+      "learning_rate": 8.745819801770756e-06,
+      "loss": 1.0177,
+      "step": 23508
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5499614122424883,
+      "learning_rate": 8.745062789344686e-06,
+      "loss": 0.748,
+      "step": 23509
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.497998271053261,
+      "learning_rate": 8.74430578422526e-06,
+      "loss": 0.8249,
+      "step": 23510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.651660584422252,
+      "learning_rate": 8.74354878641688e-06,
+      "loss": 0.8757,
+      "step": 23511
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1296981317174697,
+      "learning_rate": 8.742791795923956e-06,
+      "loss": 0.958,
+      "step": 23512
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.56104905592789,
+      "learning_rate": 8.742034812750896e-06,
+      "loss": 0.9043,
+      "step": 23513
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6122968784964273,
+      "learning_rate": 8.741277836902114e-06,
+      "loss": 0.7491,
+      "step": 23514
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7946481520483445,
+      "learning_rate": 8.740520868382005e-06,
+      "loss": 0.73,
+      "step": 23515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.458322015569223,
+      "learning_rate": 8.739763907194981e-06,
+      "loss": 0.7967,
+      "step": 23516
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7004341940959382,
+      "learning_rate": 8.739006953345455e-06,
+      "loss": 0.8245,
+      "step": 23517
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4285341271031735,
+      "learning_rate": 8.738250006837828e-06,
+      "loss": 0.7499,
+      "step": 23518
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7502811302331582,
+      "learning_rate": 8.737493067676508e-06,
+      "loss": 0.8632,
+      "step": 23519
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7431353543292418,
+      "learning_rate": 8.736736135865901e-06,
+      "loss": 0.8639,
+      "step": 23520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6353489350058386,
+      "learning_rate": 8.735979211410423e-06,
+      "loss": 0.8302,
+      "step": 23521
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.846742549441083,
+      "learning_rate": 8.73522229431447e-06,
+      "loss": 0.8285,
+      "step": 23522
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4555682474145182,
+      "learning_rate": 8.73446538458245e-06,
+      "loss": 0.8386,
+      "step": 23523
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5413378575899033,
+      "learning_rate": 8.733708482218782e-06,
+      "loss": 0.856,
+      "step": 23524
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5559148940944059,
+      "learning_rate": 8.732951587227856e-06,
+      "loss": 0.9694,
+      "step": 23525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6935567060594823,
+      "learning_rate": 8.732194699614089e-06,
+      "loss": 0.8208,
+      "step": 23526
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7218114147943175,
+      "learning_rate": 8.731437819381884e-06,
+      "loss": 0.8524,
+      "step": 23527
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8759094971708354,
+      "learning_rate": 8.730680946535657e-06,
+      "loss": 0.8533,
+      "step": 23528
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6655882679488658,
+      "learning_rate": 8.7299240810798e-06,
+      "loss": 0.8243,
+      "step": 23529
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5876702262040918,
+      "learning_rate": 8.72916722301873e-06,
+      "loss": 0.9907,
+      "step": 23530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7307312798844854,
+      "learning_rate": 8.72841037235685e-06,
+      "loss": 0.94,
+      "step": 23531
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.434396944755178,
+      "learning_rate": 8.727653529098567e-06,
+      "loss": 0.7411,
+      "step": 23532
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6719519761818098,
+      "learning_rate": 8.726896693248288e-06,
+      "loss": 0.8893,
+      "step": 23533
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2671510996159245,
+      "learning_rate": 8.726139864810422e-06,
+      "loss": 0.8526,
+      "step": 23534
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7525866546487989,
+      "learning_rate": 8.725383043789372e-06,
+      "loss": 0.6907,
+      "step": 23535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3938879198305547,
+      "learning_rate": 8.724626230189544e-06,
+      "loss": 0.9142,
+      "step": 23536
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8896995368721696,
+      "learning_rate": 8.723869424015345e-06,
+      "loss": 0.78,
+      "step": 23537
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7231702673942066,
+      "learning_rate": 8.723112625271184e-06,
+      "loss": 0.8618,
+      "step": 23538
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8192772022533146,
+      "learning_rate": 8.722355833961467e-06,
+      "loss": 1.0312,
+      "step": 23539
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5266120552576452,
+      "learning_rate": 8.721599050090596e-06,
+      "loss": 0.8965,
+      "step": 23540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.571859883702076,
+      "learning_rate": 8.72084227366298e-06,
+      "loss": 0.9055,
+      "step": 23541
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4087140281280615,
+      "learning_rate": 8.720085504683028e-06,
+      "loss": 0.8802,
+      "step": 23542
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7911802115529274,
+      "learning_rate": 8.719328743155144e-06,
+      "loss": 0.916,
+      "step": 23543
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6282349336530253,
+      "learning_rate": 8.718571989083729e-06,
+      "loss": 0.8953,
+      "step": 23544
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7737962417126208,
+      "learning_rate": 8.717815242473196e-06,
+      "loss": 0.7366,
+      "step": 23545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6238006571177936,
+      "learning_rate": 8.71705850332795e-06,
+      "loss": 0.8609,
+      "step": 23546
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4865387382656192,
+      "learning_rate": 8.716301771652393e-06,
+      "loss": 0.7953,
+      "step": 23547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7815795566192357,
+      "learning_rate": 8.715545047450933e-06,
+      "loss": 0.6965,
+      "step": 23548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.565193459815535,
+      "learning_rate": 8.71478833072798e-06,
+      "loss": 0.6449,
+      "step": 23549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.745430535657735,
+      "learning_rate": 8.714031621487934e-06,
+      "loss": 0.9502,
+      "step": 23550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7850618393144404,
+      "learning_rate": 8.713274919735204e-06,
+      "loss": 0.7477,
+      "step": 23551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7981286695731737,
+      "learning_rate": 8.712518225474191e-06,
+      "loss": 0.8811,
+      "step": 23552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7190221238679277,
+      "learning_rate": 8.71176153870931e-06,
+      "loss": 0.8807,
+      "step": 23553
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.503834750245392,
+      "learning_rate": 8.711004859444957e-06,
+      "loss": 0.9034,
+      "step": 23554
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8434270836534405,
+      "learning_rate": 8.710248187685545e-06,
+      "loss": 0.9011,
+      "step": 23555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5903574178862268,
+      "learning_rate": 8.709491523435477e-06,
+      "loss": 0.7098,
+      "step": 23556
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9178897007457152,
+      "learning_rate": 8.708734866699153e-06,
+      "loss": 0.8202,
+      "step": 23557
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4820691287107193,
+      "learning_rate": 8.707978217480986e-06,
+      "loss": 0.8206,
+      "step": 23558
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.570590582272844,
+      "learning_rate": 8.707221575785379e-06,
+      "loss": 0.8714,
+      "step": 23559
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8629151151608472,
+      "learning_rate": 8.706464941616741e-06,
+      "loss": 0.8196,
+      "step": 23560
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.058029827340547,
+      "learning_rate": 8.705708314979468e-06,
+      "loss": 0.8632,
+      "step": 23561
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4107004813856596,
+      "learning_rate": 8.704951695877973e-06,
+      "loss": 0.796,
+      "step": 23562
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8376083904426417,
+      "learning_rate": 8.70419508431666e-06,
+      "loss": 0.7366,
+      "step": 23563
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1699071425434386,
+      "learning_rate": 8.703438480299932e-06,
+      "loss": 0.9749,
+      "step": 23564
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8158693241051695,
+      "learning_rate": 8.702681883832196e-06,
+      "loss": 0.9214,
+      "step": 23565
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7752795325258665,
+      "learning_rate": 8.701925294917857e-06,
+      "loss": 0.6675,
+      "step": 23566
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8771894699256353,
+      "learning_rate": 8.70116871356132e-06,
+      "loss": 0.7837,
+      "step": 23567
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6398564601343606,
+      "learning_rate": 8.700412139766989e-06,
+      "loss": 0.834,
+      "step": 23568
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7994757597569742,
+      "learning_rate": 8.69965557353927e-06,
+      "loss": 0.844,
+      "step": 23569
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6833143302652207,
+      "learning_rate": 8.698899014882572e-06,
+      "loss": 0.722,
+      "step": 23570
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.428985359020794,
+      "learning_rate": 8.698142463801292e-06,
+      "loss": 0.9621,
+      "step": 23571
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8141444410940806,
+      "learning_rate": 8.697385920299838e-06,
+      "loss": 0.8708,
+      "step": 23572
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8700558662943476,
+      "learning_rate": 8.696629384382616e-06,
+      "loss": 0.7287,
+      "step": 23573
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1007384836620915,
+      "learning_rate": 8.695872856054034e-06,
+      "loss": 0.9175,
+      "step": 23574
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.133964203820027,
+      "learning_rate": 8.695116335318492e-06,
+      "loss": 0.8043,
+      "step": 23575
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.820230393981212,
+      "learning_rate": 8.694359822180391e-06,
+      "loss": 0.85,
+      "step": 23576
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5098239399247462,
+      "learning_rate": 8.693603316644147e-06,
+      "loss": 0.8031,
+      "step": 23577
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0384113656270575,
+      "learning_rate": 8.692846818714154e-06,
+      "loss": 1.0299,
+      "step": 23578
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7875649416185538,
+      "learning_rate": 8.692090328394821e-06,
+      "loss": 0.8055,
+      "step": 23579
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.794673584067406,
+      "learning_rate": 8.69133384569055e-06,
+      "loss": 0.8144,
+      "step": 23580
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.718370238290883,
+      "learning_rate": 8.690577370605752e-06,
+      "loss": 0.9275,
+      "step": 23581
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6979095193374631,
+      "learning_rate": 8.689820903144827e-06,
+      "loss": 0.953,
+      "step": 23582
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2386802522159015,
+      "learning_rate": 8.689064443312177e-06,
+      "loss": 0.8153,
+      "step": 23583
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.53057143150011,
+      "learning_rate": 8.688307991112212e-06,
+      "loss": 0.9299,
+      "step": 23584
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7473819672185504,
+      "learning_rate": 8.68755154654933e-06,
+      "loss": 0.7949,
+      "step": 23585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8107857213554326,
+      "learning_rate": 8.68679510962794e-06,
+      "loss": 0.9971,
+      "step": 23586
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5036554924222705,
+      "learning_rate": 8.686038680352442e-06,
+      "loss": 0.8494,
+      "step": 23587
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4006304946760824,
+      "learning_rate": 8.685282258727248e-06,
+      "loss": 0.8125,
+      "step": 23588
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8407436019847028,
+      "learning_rate": 8.684525844756751e-06,
+      "loss": 0.8477,
+      "step": 23589
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.967943357060203,
+      "learning_rate": 8.683769438445365e-06,
+      "loss": 0.8605,
+      "step": 23590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7399642538486397,
+      "learning_rate": 8.68301303979749e-06,
+      "loss": 0.9629,
+      "step": 23591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5173311272622643,
+      "learning_rate": 8.682256648817528e-06,
+      "loss": 0.8073,
+      "step": 23592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4183508955885102,
+      "learning_rate": 8.681500265509884e-06,
+      "loss": 0.9224,
+      "step": 23593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.710634884390193,
+      "learning_rate": 8.680743889878964e-06,
+      "loss": 0.9574,
+      "step": 23594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7548893306554871,
+      "learning_rate": 8.679987521929174e-06,
+      "loss": 0.8742,
+      "step": 23595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.604739596229927,
+      "learning_rate": 8.679231161664911e-06,
+      "loss": 0.801,
+      "step": 23596
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.704624606084419,
+      "learning_rate": 8.678474809090583e-06,
+      "loss": 0.8744,
+      "step": 23597
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5923107542925603,
+      "learning_rate": 8.677718464210594e-06,
+      "loss": 0.7702,
+      "step": 23598
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.424008985933511,
+      "learning_rate": 8.676962127029348e-06,
+      "loss": 0.859,
+      "step": 23599
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7707757237780203,
+      "learning_rate": 8.676205797551243e-06,
+      "loss": 0.7205,
+      "step": 23600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7674110237305917,
+      "learning_rate": 8.675449475780689e-06,
+      "loss": 0.7078,
+      "step": 23601
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8771214840724806,
+      "learning_rate": 8.674693161722089e-06,
+      "loss": 0.8676,
+      "step": 23602
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7927466787200964,
+      "learning_rate": 8.673936855379844e-06,
+      "loss": 0.6884,
+      "step": 23603
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6156564886598068,
+      "learning_rate": 8.673180556758358e-06,
+      "loss": 0.8215,
+      "step": 23604
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.412092598780023,
+      "learning_rate": 8.672424265862037e-06,
+      "loss": 0.8484,
+      "step": 23605
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.561476935138716,
+      "learning_rate": 8.671667982695279e-06,
+      "loss": 0.815,
+      "step": 23606
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6895977011977872,
+      "learning_rate": 8.670911707262491e-06,
+      "loss": 0.853,
+      "step": 23607
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.756025368376842,
+      "learning_rate": 8.670155439568076e-06,
+      "loss": 0.8812,
+      "step": 23608
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.815023567364621,
+      "learning_rate": 8.669399179616441e-06,
+      "loss": 0.891,
+      "step": 23609
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4074624844424446,
+      "learning_rate": 8.66864292741198e-06,
+      "loss": 0.8426,
+      "step": 23610
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8343195821531157,
+      "learning_rate": 8.667886682959103e-06,
+      "loss": 0.919,
+      "step": 23611
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8957978328402285,
+      "learning_rate": 8.667130446262214e-06,
+      "loss": 0.8655,
+      "step": 23612
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5640414995297325,
+      "learning_rate": 8.666374217325713e-06,
+      "loss": 0.8703,
+      "step": 23613
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7578103165811647,
+      "learning_rate": 8.665617996154e-06,
+      "loss": 0.758,
+      "step": 23614
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7163652257407378,
+      "learning_rate": 8.664861782751482e-06,
+      "loss": 0.9601,
+      "step": 23615
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5936894907140937,
+      "learning_rate": 8.664105577122566e-06,
+      "loss": 0.9296,
+      "step": 23616
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6659208331345379,
+      "learning_rate": 8.663349379271645e-06,
+      "loss": 0.8518,
+      "step": 23617
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6881380205109524,
+      "learning_rate": 8.662593189203128e-06,
+      "loss": 0.7921,
+      "step": 23618
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7657692698023695,
+      "learning_rate": 8.661837006921422e-06,
+      "loss": 0.8672,
+      "step": 23619
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5835984829119492,
+      "learning_rate": 8.661080832430919e-06,
+      "loss": 0.7948,
+      "step": 23620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6891396862610164,
+      "learning_rate": 8.660324665736026e-06,
+      "loss": 0.9673,
+      "step": 23621
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5054310203391958,
+      "learning_rate": 8.659568506841151e-06,
+      "loss": 0.9635,
+      "step": 23622
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6705995050317608,
+      "learning_rate": 8.658812355750692e-06,
+      "loss": 0.8512,
+      "step": 23623
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4205972975275094,
+      "learning_rate": 8.658056212469052e-06,
+      "loss": 0.7516,
+      "step": 23624
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4835324381886363,
+      "learning_rate": 8.657300077000632e-06,
+      "loss": 0.8666,
+      "step": 23625
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5202042945535432,
+      "learning_rate": 8.65654394934984e-06,
+      "loss": 0.8914,
+      "step": 23626
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.254138611998555,
+      "learning_rate": 8.65578782952107e-06,
+      "loss": 1.0321,
+      "step": 23627
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8346839324434125,
+      "learning_rate": 8.655031717518732e-06,
+      "loss": 0.8763,
+      "step": 23628
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7945696328362756,
+      "learning_rate": 8.65427561334722e-06,
+      "loss": 0.7205,
+      "step": 23629
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3482796714351275,
+      "learning_rate": 8.653519517010947e-06,
+      "loss": 0.8677,
+      "step": 23630
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1914662729942793,
+      "learning_rate": 8.652763428514309e-06,
+      "loss": 0.9131,
+      "step": 23631
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.508382437769598,
+      "learning_rate": 8.652007347861707e-06,
+      "loss": 0.794,
+      "step": 23632
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.076595825122285,
+      "learning_rate": 8.651251275057547e-06,
+      "loss": 0.9257,
+      "step": 23633
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5256437070776523,
+      "learning_rate": 8.650495210106225e-06,
+      "loss": 0.8399,
+      "step": 23634
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5620506394460145,
+      "learning_rate": 8.64973915301215e-06,
+      "loss": 0.7835,
+      "step": 23635
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5967207547361668,
+      "learning_rate": 8.648983103779722e-06,
+      "loss": 0.8761,
+      "step": 23636
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1181720769325194,
+      "learning_rate": 8.648227062413342e-06,
+      "loss": 0.7548,
+      "step": 23637
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.566467895243868,
+      "learning_rate": 8.647471028917409e-06,
+      "loss": 0.8869,
+      "step": 23638
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.08140560013311,
+      "learning_rate": 8.64671500329633e-06,
+      "loss": 0.871,
+      "step": 23639
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.521812290291339,
+      "learning_rate": 8.645958985554506e-06,
+      "loss": 0.7825,
+      "step": 23640
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6624893708473003,
+      "learning_rate": 8.645202975696336e-06,
+      "loss": 0.8048,
+      "step": 23641
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4598252835053651,
+      "learning_rate": 8.644446973726222e-06,
+      "loss": 0.858,
+      "step": 23642
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6540623503226008,
+      "learning_rate": 8.643690979648567e-06,
+      "loss": 0.9939,
+      "step": 23643
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7628756162448613,
+      "learning_rate": 8.642934993467776e-06,
+      "loss": 0.8349,
+      "step": 23644
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5246792723671283,
+      "learning_rate": 8.642179015188245e-06,
+      "loss": 0.9118,
+      "step": 23645
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.604985941506053,
+      "learning_rate": 8.641423044814375e-06,
+      "loss": 0.8447,
+      "step": 23646
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7626631922466709,
+      "learning_rate": 8.640667082350574e-06,
+      "loss": 0.7297,
+      "step": 23647
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6603891555715642,
+      "learning_rate": 8.639911127801238e-06,
+      "loss": 0.9604,
+      "step": 23648
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9711929685348102,
+      "learning_rate": 8.639155181170771e-06,
+      "loss": 0.8325,
+      "step": 23649
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8498019074620098,
+      "learning_rate": 8.63839924246357e-06,
+      "loss": 0.8671,
+      "step": 23650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7073831429102413,
+      "learning_rate": 8.637643311684045e-06,
+      "loss": 0.8412,
+      "step": 23651
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9081723553029903,
+      "learning_rate": 8.636887388836588e-06,
+      "loss": 0.9031,
+      "step": 23652
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.709405371667164,
+      "learning_rate": 8.636131473925605e-06,
+      "loss": 0.8804,
+      "step": 23653
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5412841691954358,
+      "learning_rate": 8.6353755669555e-06,
+      "loss": 0.7268,
+      "step": 23654
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.484547175559415,
+      "learning_rate": 8.634619667930664e-06,
+      "loss": 0.889,
+      "step": 23655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7693964275783531,
+      "learning_rate": 8.633863776855508e-06,
+      "loss": 0.876,
+      "step": 23656
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7726118040135351,
+      "learning_rate": 8.633107893734427e-06,
+      "loss": 0.7015,
+      "step": 23657
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.645630010661164,
+      "learning_rate": 8.63235201857183e-06,
+      "loss": 0.8911,
+      "step": 23658
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3531348303807793,
+      "learning_rate": 8.631596151372107e-06,
+      "loss": 0.8176,
+      "step": 23659
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9788260789575325,
+      "learning_rate": 8.630840292139668e-06,
+      "loss": 0.8688,
+      "step": 23660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6692272477783177,
+      "learning_rate": 8.630084440878906e-06,
+      "loss": 0.8648,
+      "step": 23661
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.572314948365646,
+      "learning_rate": 8.629328597594233e-06,
+      "loss": 0.8714,
+      "step": 23662
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6130968361468594,
+      "learning_rate": 8.628572762290038e-06,
+      "loss": 0.9448,
+      "step": 23663
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5607297054915747,
+      "learning_rate": 8.627816934970728e-06,
+      "loss": 0.761,
+      "step": 23664
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5266840039122356,
+      "learning_rate": 8.627061115640703e-06,
+      "loss": 0.8666,
+      "step": 23665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6092054759862606,
+      "learning_rate": 8.62630530430436e-06,
+      "loss": 1.0002,
+      "step": 23666
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9915157287893304,
+      "learning_rate": 8.625549500966103e-06,
+      "loss": 0.8079,
+      "step": 23667
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6991753378411572,
+      "learning_rate": 8.624793705630334e-06,
+      "loss": 1.0045,
+      "step": 23668
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8699483050379995,
+      "learning_rate": 8.624037918301453e-06,
+      "loss": 0.7985,
+      "step": 23669
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7665730811442503,
+      "learning_rate": 8.623282138983854e-06,
+      "loss": 0.6999,
+      "step": 23670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.707367204300003,
+      "learning_rate": 8.622526367681944e-06,
+      "loss": 0.7417,
+      "step": 23671
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5961735148920841,
+      "learning_rate": 8.621770604400126e-06,
+      "loss": 0.9286,
+      "step": 23672
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6099902980580079,
+      "learning_rate": 8.621014849142792e-06,
+      "loss": 0.9464,
+      "step": 23673
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7825192280744901,
+      "learning_rate": 8.620259101914344e-06,
+      "loss": 0.8962,
+      "step": 23674
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7648004795291239,
+      "learning_rate": 8.619503362719186e-06,
+      "loss": 0.8867,
+      "step": 23675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6223117634956907,
+      "learning_rate": 8.618747631561719e-06,
+      "loss": 0.9506,
+      "step": 23676
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5849006747914787,
+      "learning_rate": 8.617991908446339e-06,
+      "loss": 0.8311,
+      "step": 23677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8736711920701705,
+      "learning_rate": 8.617236193377445e-06,
+      "loss": 0.8721,
+      "step": 23678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6408890073775697,
+      "learning_rate": 8.616480486359444e-06,
+      "loss": 0.9648,
+      "step": 23679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5538308323460286,
+      "learning_rate": 8.61572478739673e-06,
+      "loss": 0.8402,
+      "step": 23680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6946789026297566,
+      "learning_rate": 8.614969096493704e-06,
+      "loss": 1.007,
+      "step": 23681
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5597681173215228,
+      "learning_rate": 8.614213413654764e-06,
+      "loss": 0.8268,
+      "step": 23682
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7468259182580702,
+      "learning_rate": 8.613457738884316e-06,
+      "loss": 0.7574,
+      "step": 23683
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.482486779032277,
+      "learning_rate": 8.612702072186757e-06,
+      "loss": 0.7276,
+      "step": 23684
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7996781928928955,
+      "learning_rate": 8.611946413566481e-06,
+      "loss": 0.6782,
+      "step": 23685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5206984533742944,
+      "learning_rate": 8.611190763027898e-06,
+      "loss": 0.9485,
+      "step": 23686
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.737278083186934,
+      "learning_rate": 8.610435120575396e-06,
+      "loss": 0.8204,
+      "step": 23687
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5075096506973218,
+      "learning_rate": 8.609679486213384e-06,
+      "loss": 0.776,
+      "step": 23688
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.798322053010691,
+      "learning_rate": 8.608923859946259e-06,
+      "loss": 0.9545,
+      "step": 23689
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5044286578785686,
+      "learning_rate": 8.60816824177842e-06,
+      "loss": 0.8012,
+      "step": 23690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5836308387720197,
+      "learning_rate": 8.607412631714263e-06,
+      "loss": 0.9724,
+      "step": 23691
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3071435241639797,
+      "learning_rate": 8.606657029758193e-06,
+      "loss": 0.8616,
+      "step": 23692
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6917982087094827,
+      "learning_rate": 8.605901435914608e-06,
+      "loss": 0.9055,
+      "step": 23693
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.644013600870287,
+      "learning_rate": 8.605145850187904e-06,
+      "loss": 0.8483,
+      "step": 23694
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7810910562736801,
+      "learning_rate": 8.60439027258248e-06,
+      "loss": 0.8934,
+      "step": 23695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1449150697686448,
+      "learning_rate": 8.60363470310274e-06,
+      "loss": 0.8368,
+      "step": 23696
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7407662865865794,
+      "learning_rate": 8.602879141753084e-06,
+      "loss": 0.7175,
+      "step": 23697
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4838552100155498,
+      "learning_rate": 8.602123588537906e-06,
+      "loss": 0.7792,
+      "step": 23698
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.693983335746834,
+      "learning_rate": 8.601368043461604e-06,
+      "loss": 0.8897,
+      "step": 23699
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4293575596652377,
+      "learning_rate": 8.600612506528584e-06,
+      "loss": 0.8775,
+      "step": 23700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6690041994725178,
+      "learning_rate": 8.59985697774324e-06,
+      "loss": 0.7451,
+      "step": 23701
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5404819129677376,
+      "learning_rate": 8.599101457109971e-06,
+      "loss": 1.0042,
+      "step": 23702
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7344197754984272,
+      "learning_rate": 8.598345944633174e-06,
+      "loss": 0.9303,
+      "step": 23703
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7281708594086451,
+      "learning_rate": 8.597590440317257e-06,
+      "loss": 0.8499,
+      "step": 23704
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9232148763355745,
+      "learning_rate": 8.596834944166608e-06,
+      "loss": 0.8294,
+      "step": 23705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5709957340738867,
+      "learning_rate": 8.59607945618563e-06,
+      "loss": 0.9251,
+      "step": 23706
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9770632746473198,
+      "learning_rate": 8.595323976378727e-06,
+      "loss": 0.9097,
+      "step": 23707
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5790428354432888,
+      "learning_rate": 8.594568504750287e-06,
+      "loss": 0.7078,
+      "step": 23708
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7310407922612199,
+      "learning_rate": 8.593813041304716e-06,
+      "loss": 0.8293,
+      "step": 23709
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8018640445519414,
+      "learning_rate": 8.593057586046409e-06,
+      "loss": 0.8567,
+      "step": 23710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.541331950210689,
+      "learning_rate": 8.592302138979773e-06,
+      "loss": 0.8497,
+      "step": 23711
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7309666907904515,
+      "learning_rate": 8.59154670010919e-06,
+      "loss": 0.8943,
+      "step": 23712
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7212818111187724,
+      "learning_rate": 8.590791269439074e-06,
+      "loss": 0.7574,
+      "step": 23713
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.7365466471162545,
+      "learning_rate": 8.590035846973817e-06,
+      "loss": 0.8218,
+      "step": 23714
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7282257860962578,
+      "learning_rate": 8.589280432717814e-06,
+      "loss": 0.8919,
+      "step": 23715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6940174116516753,
+      "learning_rate": 8.58852502667547e-06,
+      "loss": 0.8706,
+      "step": 23716
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.783310114254053,
+      "learning_rate": 8.58776962885118e-06,
+      "loss": 0.7528,
+      "step": 23717
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6759647269845748,
+      "learning_rate": 8.587014239249344e-06,
+      "loss": 0.8925,
+      "step": 23718
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.865517717174713,
+      "learning_rate": 8.586258857874355e-06,
+      "loss": 0.9156,
+      "step": 23719
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6658731415679557,
+      "learning_rate": 8.585503484730616e-06,
+      "loss": 0.7746,
+      "step": 23720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5609194737598502,
+      "learning_rate": 8.584748119822525e-06,
+      "loss": 0.8233,
+      "step": 23721
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6495502537906528,
+      "learning_rate": 8.583992763154476e-06,
+      "loss": 0.8362,
+      "step": 23722
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4420087411148448,
+      "learning_rate": 8.583237414730869e-06,
+      "loss": 0.925,
+      "step": 23723
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6847588944613006,
+      "learning_rate": 8.582482074556103e-06,
+      "loss": 0.8387,
+      "step": 23724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8064931659028021,
+      "learning_rate": 8.581726742634579e-06,
+      "loss": 0.7351,
+      "step": 23725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6774038229202213,
+      "learning_rate": 8.580971418970687e-06,
+      "loss": 0.7881,
+      "step": 23726
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.693088598271128,
+      "learning_rate": 8.580216103568828e-06,
+      "loss": 0.7133,
+      "step": 23727
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6612652594980664,
+      "learning_rate": 8.579460796433403e-06,
+      "loss": 0.8509,
+      "step": 23728
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7302568760700274,
+      "learning_rate": 8.578705497568807e-06,
+      "loss": 0.8293,
+      "step": 23729
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6711275036461437,
+      "learning_rate": 8.577950206979436e-06,
+      "loss": 0.7619,
+      "step": 23730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9185927138240368,
+      "learning_rate": 8.577194924669687e-06,
+      "loss": 0.9489,
+      "step": 23731
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.44779793405611,
+      "learning_rate": 8.576439650643964e-06,
+      "loss": 0.9189,
+      "step": 23732
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.792873140991813,
+      "learning_rate": 8.575684384906658e-06,
+      "loss": 0.7111,
+      "step": 23733
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.842315926144465,
+      "learning_rate": 8.574929127462168e-06,
+      "loss": 0.9108,
+      "step": 23734
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0845530207505716,
+      "learning_rate": 8.574173878314895e-06,
+      "loss": 0.92,
+      "step": 23735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5286061693525497,
+      "learning_rate": 8.573418637469228e-06,
+      "loss": 0.8869,
+      "step": 23736
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3637406993483956,
+      "learning_rate": 8.572663404929573e-06,
+      "loss": 1.0068,
+      "step": 23737
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5072604509349334,
+      "learning_rate": 8.57190818070032e-06,
+      "loss": 0.8925,
+      "step": 23738
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7754515021112693,
+      "learning_rate": 8.571152964785876e-06,
+      "loss": 0.7417,
+      "step": 23739
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9522402084266264,
+      "learning_rate": 8.570397757190624e-06,
+      "loss": 0.9668,
+      "step": 23740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7129749476611578,
+      "learning_rate": 8.569642557918974e-06,
+      "loss": 0.8708,
+      "step": 23741
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7693267202193819,
+      "learning_rate": 8.568887366975318e-06,
+      "loss": 0.8307,
+      "step": 23742
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.816969310938273,
+      "learning_rate": 8.56813218436405e-06,
+      "loss": 0.8796,
+      "step": 23743
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.498222820221204,
+      "learning_rate": 8.567377010089571e-06,
+      "loss": 0.9272,
+      "step": 23744
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.452085618136701,
+      "learning_rate": 8.566621844156275e-06,
+      "loss": 0.8144,
+      "step": 23745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4953968338741002,
+      "learning_rate": 8.565866686568565e-06,
+      "loss": 0.8665,
+      "step": 23746
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5723623597147594,
+      "learning_rate": 8.565111537330831e-06,
+      "loss": 0.9199,
+      "step": 23747
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.471455248136451,
+      "learning_rate": 8.56435639644747e-06,
+      "loss": 0.9157,
+      "step": 23748
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5333651599029297,
+      "learning_rate": 8.563601263922883e-06,
+      "loss": 0.9743,
+      "step": 23749
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8267156609216944,
+      "learning_rate": 8.562846139761462e-06,
+      "loss": 0.9462,
+      "step": 23750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.517506205267313,
+      "learning_rate": 8.562091023967606e-06,
+      "loss": 0.8602,
+      "step": 23751
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7175145450221512,
+      "learning_rate": 8.56133591654571e-06,
+      "loss": 0.7336,
+      "step": 23752
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.863385858963122,
+      "learning_rate": 8.560580817500177e-06,
+      "loss": 0.8066,
+      "step": 23753
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7002592243219954,
+      "learning_rate": 8.559825726835395e-06,
+      "loss": 0.8188,
+      "step": 23754
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.808288419178667,
+      "learning_rate": 8.559070644555761e-06,
+      "loss": 0.946,
+      "step": 23755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7134062039576654,
+      "learning_rate": 8.55831557066568e-06,
+      "loss": 0.8486,
+      "step": 23756
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.213294556115209,
+      "learning_rate": 8.557560505169538e-06,
+      "loss": 0.8165,
+      "step": 23757
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.589321503371513,
+      "learning_rate": 8.556805448071736e-06,
+      "loss": 0.9596,
+      "step": 23758
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5999982426908346,
+      "learning_rate": 8.556050399376667e-06,
+      "loss": 0.7358,
+      "step": 23759
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8351339374885298,
+      "learning_rate": 8.555295359088736e-06,
+      "loss": 0.7859,
+      "step": 23760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4467844119021613,
+      "learning_rate": 8.554540327212328e-06,
+      "loss": 0.9196,
+      "step": 23761
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.810008429499622,
+      "learning_rate": 8.553785303751844e-06,
+      "loss": 0.9291,
+      "step": 23762
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.680673455608165,
+      "learning_rate": 8.553030288711683e-06,
+      "loss": 0.8392,
+      "step": 23763
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.652508767188381,
+      "learning_rate": 8.552275282096233e-06,
+      "loss": 0.9975,
+      "step": 23764
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6455312068914714,
+      "learning_rate": 8.551520283909898e-06,
+      "loss": 0.927,
+      "step": 23765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5572316944596822,
+      "learning_rate": 8.550765294157069e-06,
+      "loss": 0.8953,
+      "step": 23766
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8928655893798996,
+      "learning_rate": 8.550010312842148e-06,
+      "loss": 0.8356,
+      "step": 23767
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8481914631624288,
+      "learning_rate": 8.54925533996952e-06,
+      "loss": 0.7973,
+      "step": 23768
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8212910431649187,
+      "learning_rate": 8.548500375543587e-06,
+      "loss": 0.7427,
+      "step": 23769
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4980015192779945,
+      "learning_rate": 8.547745419568749e-06,
+      "loss": 0.8801,
+      "step": 23770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.794290203979569,
+      "learning_rate": 8.546990472049394e-06,
+      "loss": 0.8357,
+      "step": 23771
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.652212751552555,
+      "learning_rate": 8.546235532989919e-06,
+      "loss": 0.9053,
+      "step": 23772
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8152417654100106,
+      "learning_rate": 8.545480602394721e-06,
+      "loss": 0.864,
+      "step": 23773
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6218987689766966,
+      "learning_rate": 8.544725680268199e-06,
+      "loss": 0.992,
+      "step": 23774
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.681010312398231,
+      "learning_rate": 8.543970766614743e-06,
+      "loss": 1.0234,
+      "step": 23775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5797562563112653,
+      "learning_rate": 8.543215861438749e-06,
+      "loss": 0.9025,
+      "step": 23776
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5378984255817851,
+      "learning_rate": 8.542460964744615e-06,
+      "loss": 0.7614,
+      "step": 23777
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7379176102702487,
+      "learning_rate": 8.541706076536735e-06,
+      "loss": 0.8243,
+      "step": 23778
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8467425845292137,
+      "learning_rate": 8.540951196819502e-06,
+      "loss": 0.7712,
+      "step": 23779
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4914916803915987,
+      "learning_rate": 8.540196325597312e-06,
+      "loss": 0.883,
+      "step": 23780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7947761830093458,
+      "learning_rate": 8.539441462874566e-06,
+      "loss": 0.7209,
+      "step": 23781
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8634171546924783,
+      "learning_rate": 8.538686608655652e-06,
+      "loss": 0.8755,
+      "step": 23782
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6101990688507515,
+      "learning_rate": 8.537931762944966e-06,
+      "loss": 0.7486,
+      "step": 23783
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.230845437618248,
+      "learning_rate": 8.537176925746902e-06,
+      "loss": 0.7845,
+      "step": 23784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5822647732188782,
+      "learning_rate": 8.536422097065863e-06,
+      "loss": 0.7812,
+      "step": 23785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.605800346065234,
+      "learning_rate": 8.535667276906234e-06,
+      "loss": 0.856,
+      "step": 23786
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7786388347135853,
+      "learning_rate": 8.534912465272414e-06,
+      "loss": 0.8944,
+      "step": 23787
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5189623705369477,
+      "learning_rate": 8.5341576621688e-06,
+      "loss": 0.9102,
+      "step": 23788
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8505429888197835,
+      "learning_rate": 8.53340286759978e-06,
+      "loss": 0.9065,
+      "step": 23789
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6139763185255107,
+      "learning_rate": 8.532648081569756e-06,
+      "loss": 0.8036,
+      "step": 23790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2720242518426805,
+      "learning_rate": 8.531893304083117e-06,
+      "loss": 0.7353,
+      "step": 23791
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5593842940273621,
+      "learning_rate": 8.531138535144265e-06,
+      "loss": 0.8107,
+      "step": 23792
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5334337764680244,
+      "learning_rate": 8.530383774757583e-06,
+      "loss": 0.8606,
+      "step": 23793
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4791646034884014,
+      "learning_rate": 8.529629022927474e-06,
+      "loss": 0.8511,
+      "step": 23794
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6664543334316075,
+      "learning_rate": 8.528874279658335e-06,
+      "loss": 0.8999,
+      "step": 23795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.340070339377043,
+      "learning_rate": 8.528119544954551e-06,
+      "loss": 0.8241,
+      "step": 23796
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.607273099013546,
+      "learning_rate": 8.52736481882052e-06,
+      "loss": 0.9,
+      "step": 23797
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.040859076765264,
+      "learning_rate": 8.526610101260639e-06,
+      "loss": 0.8981,
+      "step": 23798
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5343193675375164,
+      "learning_rate": 8.525855392279302e-06,
+      "loss": 0.8688,
+      "step": 23799
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8128922885135231,
+      "learning_rate": 8.525100691880898e-06,
+      "loss": 0.7563,
+      "step": 23800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3848917714340996,
+      "learning_rate": 8.524346000069828e-06,
+      "loss": 0.8913,
+      "step": 23801
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4741905776831143,
+      "learning_rate": 8.523591316850483e-06,
+      "loss": 0.8756,
+      "step": 23802
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4306870646222496,
+      "learning_rate": 8.522836642227254e-06,
+      "loss": 0.9465,
+      "step": 23803
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.629909819644344,
+      "learning_rate": 8.522081976204539e-06,
+      "loss": 0.7512,
+      "step": 23804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7151308304835855,
+      "learning_rate": 8.52132731878673e-06,
+      "loss": 0.6647,
+      "step": 23805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4881399033882665,
+      "learning_rate": 8.520572669978224e-06,
+      "loss": 0.8286,
+      "step": 23806
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.471601744314138,
+      "learning_rate": 8.519818029783412e-06,
+      "loss": 0.7736,
+      "step": 23807
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5057496655818476,
+      "learning_rate": 8.519063398206684e-06,
+      "loss": 0.7622,
+      "step": 23808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8640689962179762,
+      "learning_rate": 8.518308775252444e-06,
+      "loss": 0.767,
+      "step": 23809
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7805106860488343,
+      "learning_rate": 8.517554160925073e-06,
+      "loss": 0.7075,
+      "step": 23810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7374690333730807,
+      "learning_rate": 8.516799555228974e-06,
+      "loss": 0.7483,
+      "step": 23811
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7662296695674796,
+      "learning_rate": 8.516044958168535e-06,
+      "loss": 0.745,
+      "step": 23812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6910239589066791,
+      "learning_rate": 8.515290369748158e-06,
+      "loss": 0.9699,
+      "step": 23813
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7282878031218212,
+      "learning_rate": 8.514535789972226e-06,
+      "loss": 0.8022,
+      "step": 23814
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4749822773884624,
+      "learning_rate": 8.513781218845137e-06,
+      "loss": 0.8154,
+      "step": 23815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8448000601259984,
+      "learning_rate": 8.513026656371288e-06,
+      "loss": 0.913,
+      "step": 23816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4907463364230658,
+      "learning_rate": 8.512272102555063e-06,
+      "loss": 0.8607,
+      "step": 23817
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8944830362646894,
+      "learning_rate": 8.511517557400864e-06,
+      "loss": 0.9237,
+      "step": 23818
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5471947407938889,
+      "learning_rate": 8.51076302091308e-06,
+      "loss": 0.7799,
+      "step": 23819
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.860362760252289,
+      "learning_rate": 8.510008493096109e-06,
+      "loss": 0.7155,
+      "step": 23820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4398074883753398,
+      "learning_rate": 8.509253973954334e-06,
+      "loss": 0.7864,
+      "step": 23821
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.614713798740135,
+      "learning_rate": 8.508499463492157e-06,
+      "loss": 0.9443,
+      "step": 23822
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.885084316028589,
+      "learning_rate": 8.507744961713972e-06,
+      "loss": 0.8024,
+      "step": 23823
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0142140705781135,
+      "learning_rate": 8.506990468624164e-06,
+      "loss": 0.8268,
+      "step": 23824
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7684890477705509,
+      "learning_rate": 8.506235984227129e-06,
+      "loss": 0.7111,
+      "step": 23825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6418536581676302,
+      "learning_rate": 8.505481508527263e-06,
+      "loss": 0.908,
+      "step": 23826
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4937190127364892,
+      "learning_rate": 8.504727041528959e-06,
+      "loss": 0.8799,
+      "step": 23827
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2876198444898512,
+      "learning_rate": 8.503972583236605e-06,
+      "loss": 0.8597,
+      "step": 23828
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.584821939745262,
+      "learning_rate": 8.503218133654595e-06,
+      "loss": 1.0312,
+      "step": 23829
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4908587604351935,
+      "learning_rate": 8.502463692787327e-06,
+      "loss": 0.9197,
+      "step": 23830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9208255094842637,
+      "learning_rate": 8.501709260639187e-06,
+      "loss": 0.9371,
+      "step": 23831
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5810412239904288,
+      "learning_rate": 8.50095483721457e-06,
+      "loss": 0.9466,
+      "step": 23832
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2243796917810084,
+      "learning_rate": 8.500200422517866e-06,
+      "loss": 0.7429,
+      "step": 23833
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.543057256926123,
+      "learning_rate": 8.499446016553475e-06,
+      "loss": 0.9939,
+      "step": 23834
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6952877246814169,
+      "learning_rate": 8.498691619325782e-06,
+      "loss": 0.8656,
+      "step": 23835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.697132479685649,
+      "learning_rate": 8.497937230839182e-06,
+      "loss": 0.949,
+      "step": 23836
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.670737475416145,
+      "learning_rate": 8.497182851098068e-06,
+      "loss": 0.9578,
+      "step": 23837
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9004048782058671,
+      "learning_rate": 8.496428480106828e-06,
+      "loss": 0.758,
+      "step": 23838
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6970717466300895,
+      "learning_rate": 8.495674117869859e-06,
+      "loss": 0.9042,
+      "step": 23839
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4385260408688183,
+      "learning_rate": 8.494919764391548e-06,
+      "loss": 0.8009,
+      "step": 23840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5772352943552006,
+      "learning_rate": 8.4941654196763e-06,
+      "loss": 0.9131,
+      "step": 23841
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.346553645021218,
+      "learning_rate": 8.493411083728491e-06,
+      "loss": 0.81,
+      "step": 23842
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9459286557248547,
+      "learning_rate": 8.49265675655252e-06,
+      "loss": 0.8297,
+      "step": 23843
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7482774471901037,
+      "learning_rate": 8.49190243815278e-06,
+      "loss": 0.6818,
+      "step": 23844
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5988338041679417,
+      "learning_rate": 8.491148128533662e-06,
+      "loss": 0.8623,
+      "step": 23845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6391360565140587,
+      "learning_rate": 8.490393827699554e-06,
+      "loss": 0.958,
+      "step": 23846
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7202933032784462,
+      "learning_rate": 8.489639535654855e-06,
+      "loss": 0.8296,
+      "step": 23847
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7416581912885893,
+      "learning_rate": 8.488885252403953e-06,
+      "loss": 0.7978,
+      "step": 23848
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7380883228670039,
+      "learning_rate": 8.488130977951235e-06,
+      "loss": 0.83,
+      "step": 23849
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7236894136441165,
+      "learning_rate": 8.487376712301099e-06,
+      "loss": 0.8758,
+      "step": 23850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4484536246673516,
+      "learning_rate": 8.486622455457938e-06,
+      "loss": 0.7552,
+      "step": 23851
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5443227173291971,
+      "learning_rate": 8.485868207426136e-06,
+      "loss": 1.0245,
+      "step": 23852
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8849046449891687,
+      "learning_rate": 8.485113968210089e-06,
+      "loss": 0.9571,
+      "step": 23853
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4994030198110728,
+      "learning_rate": 8.484359737814189e-06,
+      "loss": 0.7768,
+      "step": 23854
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5696909679402098,
+      "learning_rate": 8.483605516242829e-06,
+      "loss": 0.8091,
+      "step": 23855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5993287923296495,
+      "learning_rate": 8.482851303500395e-06,
+      "loss": 0.9672,
+      "step": 23856
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.780382182673302,
+      "learning_rate": 8.48209709959128e-06,
+      "loss": 0.9729,
+      "step": 23857
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.022571198856448,
+      "learning_rate": 8.481342904519882e-06,
+      "loss": 0.8017,
+      "step": 23858
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4990941103427449,
+      "learning_rate": 8.480588718290579e-06,
+      "loss": 0.9816,
+      "step": 23859
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5658625418615193,
+      "learning_rate": 8.479834540907774e-06,
+      "loss": 0.8283,
+      "step": 23860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7669582837653235,
+      "learning_rate": 8.479080372375848e-06,
+      "loss": 0.7349,
+      "step": 23861
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6175688228551874,
+      "learning_rate": 8.478326212699205e-06,
+      "loss": 0.9186,
+      "step": 23862
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.633727529913108,
+      "learning_rate": 8.477572061882225e-06,
+      "loss": 0.8958,
+      "step": 23863
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7930765304449419,
+      "learning_rate": 8.476817919929304e-06,
+      "loss": 0.7058,
+      "step": 23864
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.71142632452797,
+      "learning_rate": 8.47606378684483e-06,
+      "loss": 0.9869,
+      "step": 23865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8298871677427316,
+      "learning_rate": 8.475309662633195e-06,
+      "loss": 0.7149,
+      "step": 23866
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.696461197539184,
+      "learning_rate": 8.474555547298789e-06,
+      "loss": 0.9147,
+      "step": 23867
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5655387187251173,
+      "learning_rate": 8.473801440846005e-06,
+      "loss": 0.9276,
+      "step": 23868
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6387091598386638,
+      "learning_rate": 8.473047343279233e-06,
+      "loss": 0.8255,
+      "step": 23869
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5582109622805742,
+      "learning_rate": 8.472293254602858e-06,
+      "loss": 0.8427,
+      "step": 23870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4726780867190812,
+      "learning_rate": 8.47153917482128e-06,
+      "loss": 0.8428,
+      "step": 23871
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.679990705857954,
+      "learning_rate": 8.470785103938887e-06,
+      "loss": 0.8494,
+      "step": 23872
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9281453511438378,
+      "learning_rate": 8.470031041960062e-06,
+      "loss": 0.7682,
+      "step": 23873
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.586834808736407,
+      "learning_rate": 8.469276988889202e-06,
+      "loss": 0.8682,
+      "step": 23874
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6810280993792912,
+      "learning_rate": 8.468522944730697e-06,
+      "loss": 0.8607,
+      "step": 23875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6533032234818938,
+      "learning_rate": 8.467768909488938e-06,
+      "loss": 0.9047,
+      "step": 23876
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.627986355924687,
+      "learning_rate": 8.46701488316831e-06,
+      "loss": 0.9652,
+      "step": 23877
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6050583545724175,
+      "learning_rate": 8.466260865773207e-06,
+      "loss": 0.8323,
+      "step": 23878
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3697957228620283,
+      "learning_rate": 8.465506857308021e-06,
+      "loss": 0.7951,
+      "step": 23879
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6603765783241677,
+      "learning_rate": 8.46475285777714e-06,
+      "loss": 0.9038,
+      "step": 23880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.649061025367383,
+      "learning_rate": 8.463998867184952e-06,
+      "loss": 0.8291,
+      "step": 23881
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6587837881395509,
+      "learning_rate": 8.463244885535849e-06,
+      "loss": 0.9267,
+      "step": 23882
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7590212589754497,
+      "learning_rate": 8.462490912834222e-06,
+      "loss": 0.9362,
+      "step": 23883
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.680034885551862,
+      "learning_rate": 8.461736949084458e-06,
+      "loss": 0.9562,
+      "step": 23884
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5104356829057464,
+      "learning_rate": 8.460982994290951e-06,
+      "loss": 0.8877,
+      "step": 23885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7532061647866783,
+      "learning_rate": 8.46022904845809e-06,
+      "loss": 0.9332,
+      "step": 23886
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4952895797916543,
+      "learning_rate": 8.459475111590256e-06,
+      "loss": 0.7758,
+      "step": 23887
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7837115124027644,
+      "learning_rate": 8.45872118369185e-06,
+      "loss": 0.7098,
+      "step": 23888
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.659592282628042,
+      "learning_rate": 8.457967264767253e-06,
+      "loss": 0.8486,
+      "step": 23889
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.589946349495192,
+      "learning_rate": 8.457213354820868e-06,
+      "loss": 0.8739,
+      "step": 23890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3619387464795722,
+      "learning_rate": 8.456459453857067e-06,
+      "loss": 0.8451,
+      "step": 23891
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0113256141218057,
+      "learning_rate": 8.45570556188025e-06,
+      "loss": 0.8568,
+      "step": 23892
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5182810602600385,
+      "learning_rate": 8.454951678894807e-06,
+      "loss": 0.824,
+      "step": 23893
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.777144976946202,
+      "learning_rate": 8.454197804905123e-06,
+      "loss": 0.8411,
+      "step": 23894
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.029074534095529,
+      "learning_rate": 8.453443939915586e-06,
+      "loss": 0.751,
+      "step": 23895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6552100668440244,
+      "learning_rate": 8.45269008393059e-06,
+      "loss": 0.8273,
+      "step": 23896
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6271277333761802,
+      "learning_rate": 8.451936236954524e-06,
+      "loss": 0.7005,
+      "step": 23897
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6227129469406307,
+      "learning_rate": 8.45118239899177e-06,
+      "loss": 0.8144,
+      "step": 23898
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6760686014992618,
+      "learning_rate": 8.450428570046725e-06,
+      "loss": 0.9166,
+      "step": 23899
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7449126658419187,
+      "learning_rate": 8.44967475012378e-06,
+      "loss": 0.8826,
+      "step": 23900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.721387947998397,
+      "learning_rate": 8.448920939227316e-06,
+      "loss": 0.9211,
+      "step": 23901
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5620814364806863,
+      "learning_rate": 8.448167137361723e-06,
+      "loss": 0.8485,
+      "step": 23902
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.760587889399215,
+      "learning_rate": 8.447413344531394e-06,
+      "loss": 0.7027,
+      "step": 23903
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6161840048358207,
+      "learning_rate": 8.446659560740718e-06,
+      "loss": 0.8833,
+      "step": 23904
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.485337957315034,
+      "learning_rate": 8.44590578599408e-06,
+      "loss": 0.7126,
+      "step": 23905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5391644285662924,
+      "learning_rate": 8.445152020295869e-06,
+      "loss": 0.9256,
+      "step": 23906
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6227742750283027,
+      "learning_rate": 8.444398263650477e-06,
+      "loss": 0.8704,
+      "step": 23907
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5639891429215447,
+      "learning_rate": 8.443644516062293e-06,
+      "loss": 0.8069,
+      "step": 23908
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8350501795976213,
+      "learning_rate": 8.4428907775357e-06,
+      "loss": 0.8931,
+      "step": 23909
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7903835717097721,
+      "learning_rate": 8.44213704807509e-06,
+      "loss": 0.7271,
+      "step": 23910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7112111946470474,
+      "learning_rate": 8.441383327684854e-06,
+      "loss": 0.9069,
+      "step": 23911
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4981284591033253,
+      "learning_rate": 8.440629616369375e-06,
+      "loss": 0.9279,
+      "step": 23912
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6673971677208337,
+      "learning_rate": 8.439875914133044e-06,
+      "loss": 0.9782,
+      "step": 23913
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.617864990895652,
+      "learning_rate": 8.439122220980249e-06,
+      "loss": 0.8577,
+      "step": 23914
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5583316829065068,
+      "learning_rate": 8.438368536915382e-06,
+      "loss": 0.8392,
+      "step": 23915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.780117846916719,
+      "learning_rate": 8.437614861942826e-06,
+      "loss": 0.873,
+      "step": 23916
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.858764770347725,
+      "learning_rate": 8.43686119606697e-06,
+      "loss": 0.9046,
+      "step": 23917
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8361482831820868,
+      "learning_rate": 8.436107539292206e-06,
+      "loss": 0.9439,
+      "step": 23918
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7904816635878444,
+      "learning_rate": 8.435353891622915e-06,
+      "loss": 0.7382,
+      "step": 23919
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5767573967987092,
+      "learning_rate": 8.434600253063489e-06,
+      "loss": 0.7638,
+      "step": 23920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6061058266911015,
+      "learning_rate": 8.433846623618319e-06,
+      "loss": 0.8379,
+      "step": 23921
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5089978315148147,
+      "learning_rate": 8.43309300329179e-06,
+      "loss": 0.787,
+      "step": 23922
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6767006939963793,
+      "learning_rate": 8.432339392088286e-06,
+      "loss": 0.9343,
+      "step": 23923
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5116425603680768,
+      "learning_rate": 8.4315857900122e-06,
+      "loss": 0.8615,
+      "step": 23924
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7929208387665605,
+      "learning_rate": 8.43083219706792e-06,
+      "loss": 0.8732,
+      "step": 23925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0358987139927525,
+      "learning_rate": 8.430078613259832e-06,
+      "loss": 0.8048,
+      "step": 23926
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5625275393011153,
+      "learning_rate": 8.42932503859232e-06,
+      "loss": 0.9789,
+      "step": 23927
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5915121390803084,
+      "learning_rate": 8.428571473069775e-06,
+      "loss": 0.9118,
+      "step": 23928
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.508479538744773,
+      "learning_rate": 8.427817916696588e-06,
+      "loss": 0.6839,
+      "step": 23929
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5296127026062485,
+      "learning_rate": 8.42706436947714e-06,
+      "loss": 0.8481,
+      "step": 23930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7321174306933432,
+      "learning_rate": 8.42631083141582e-06,
+      "loss": 0.7246,
+      "step": 23931
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.705348295973627,
+      "learning_rate": 8.42555730251702e-06,
+      "loss": 0.8513,
+      "step": 23932
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6674863373100202,
+      "learning_rate": 8.424803782785123e-06,
+      "loss": 0.9197,
+      "step": 23933
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.668286854764303,
+      "learning_rate": 8.424050272224515e-06,
+      "loss": 0.8174,
+      "step": 23934
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9857043529631992,
+      "learning_rate": 8.423296770839584e-06,
+      "loss": 0.7314,
+      "step": 23935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.507840460129069,
+      "learning_rate": 8.422543278634726e-06,
+      "loss": 0.9701,
+      "step": 23936
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.78510444824719,
+      "learning_rate": 8.421789795614315e-06,
+      "loss": 0.8315,
+      "step": 23937
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.64800625961581,
+      "learning_rate": 8.421036321782741e-06,
+      "loss": 0.8013,
+      "step": 23938
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.490146094527612,
+      "learning_rate": 8.4202828571444e-06,
+      "loss": 0.8639,
+      "step": 23939
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5277537763860423,
+      "learning_rate": 8.419529401703667e-06,
+      "loss": 0.9792,
+      "step": 23940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5670106002821669,
+      "learning_rate": 8.418775955464935e-06,
+      "loss": 0.9672,
+      "step": 23941
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6218310024370999,
+      "learning_rate": 8.418022518432588e-06,
+      "loss": 0.9799,
+      "step": 23942
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7613404707512547,
+      "learning_rate": 8.417269090611023e-06,
+      "loss": 0.7398,
+      "step": 23943
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7455251444125963,
+      "learning_rate": 8.416515672004611e-06,
+      "loss": 0.7726,
+      "step": 23944
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6843707271184007,
+      "learning_rate": 8.415762262617748e-06,
+      "loss": 0.9051,
+      "step": 23945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6286158154129302,
+      "learning_rate": 8.41500886245482e-06,
+      "loss": 0.8244,
+      "step": 23946
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7442547645970624,
+      "learning_rate": 8.41425547152021e-06,
+      "loss": 0.7087,
+      "step": 23947
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.706444360751978,
+      "learning_rate": 8.413502089818307e-06,
+      "loss": 0.7617,
+      "step": 23948
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6981583745490227,
+      "learning_rate": 8.412748717353499e-06,
+      "loss": 0.946,
+      "step": 23949
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6593457587959626,
+      "learning_rate": 8.41199535413017e-06,
+      "loss": 0.866,
+      "step": 23950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5372756373680432,
+      "learning_rate": 8.411242000152705e-06,
+      "loss": 0.84,
+      "step": 23951
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.719246353182875,
+      "learning_rate": 8.410488655425493e-06,
+      "loss": 0.7861,
+      "step": 23952
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7358255492841796,
+      "learning_rate": 8.40973531995292e-06,
+      "loss": 0.7281,
+      "step": 23953
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5115547526318796,
+      "learning_rate": 8.40898199373937e-06,
+      "loss": 0.7145,
+      "step": 23954
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7605344869806996,
+      "learning_rate": 8.40822867678923e-06,
+      "loss": 0.8551,
+      "step": 23955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.578002265796653,
+      "learning_rate": 8.407475369106889e-06,
+      "loss": 0.8681,
+      "step": 23956
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8349076329332178,
+      "learning_rate": 8.406722070696731e-06,
+      "loss": 0.7572,
+      "step": 23957
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4753674758672402,
+      "learning_rate": 8.405968781563141e-06,
+      "loss": 0.8392,
+      "step": 23958
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6015083431945367,
+      "learning_rate": 8.405215501710501e-06,
+      "loss": 0.8551,
+      "step": 23959
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.576955740076691,
+      "learning_rate": 8.404462231143208e-06,
+      "loss": 0.8783,
+      "step": 23960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5018668113900662,
+      "learning_rate": 8.403708969865638e-06,
+      "loss": 0.7401,
+      "step": 23961
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8596981835779065,
+      "learning_rate": 8.402955717882179e-06,
+      "loss": 0.8913,
+      "step": 23962
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.497928888807153,
+      "learning_rate": 8.402202475197216e-06,
+      "loss": 0.7983,
+      "step": 23963
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6288642561104003,
+      "learning_rate": 8.40144924181514e-06,
+      "loss": 0.8143,
+      "step": 23964
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6310746758932873,
+      "learning_rate": 8.400696017740329e-06,
+      "loss": 0.9402,
+      "step": 23965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7502846276313522,
+      "learning_rate": 8.399942802977175e-06,
+      "loss": 0.6605,
+      "step": 23966
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.456768612119286,
+      "learning_rate": 8.39918959753006e-06,
+      "loss": 0.9347,
+      "step": 23967
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5247387737436566,
+      "learning_rate": 8.398436401403368e-06,
+      "loss": 0.9584,
+      "step": 23968
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6486619033815495,
+      "learning_rate": 8.397683214601487e-06,
+      "loss": 0.8321,
+      "step": 23969
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9083805649937702,
+      "learning_rate": 8.396930037128803e-06,
+      "loss": 0.8576,
+      "step": 23970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7310746205215337,
+      "learning_rate": 8.3961768689897e-06,
+      "loss": 0.7179,
+      "step": 23971
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5446717730699995,
+      "learning_rate": 8.395423710188561e-06,
+      "loss": 0.8818,
+      "step": 23972
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5924572234969907,
+      "learning_rate": 8.394670560729774e-06,
+      "loss": 0.8408,
+      "step": 23973
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7406159736593421,
+      "learning_rate": 8.393917420617726e-06,
+      "loss": 0.7283,
+      "step": 23974
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4354792192753203,
+      "learning_rate": 8.393164289856797e-06,
+      "loss": 0.8773,
+      "step": 23975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6436613348468538,
+      "learning_rate": 8.392411168451371e-06,
+      "loss": 0.8975,
+      "step": 23976
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7853944631921453,
+      "learning_rate": 8.39165805640584e-06,
+      "loss": 0.6755,
+      "step": 23977
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7736155808490649,
+      "learning_rate": 8.390904953724586e-06,
+      "loss": 0.7387,
+      "step": 23978
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7805111549835124,
+      "learning_rate": 8.390151860411991e-06,
+      "loss": 0.8498,
+      "step": 23979
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6713645833572637,
+      "learning_rate": 8.389398776472439e-06,
+      "loss": 0.9181,
+      "step": 23980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7834651796156831,
+      "learning_rate": 8.38864570191032e-06,
+      "loss": 0.779,
+      "step": 23981
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6225966550037552,
+      "learning_rate": 8.387892636730016e-06,
+      "loss": 0.9719,
+      "step": 23982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6500553893139263,
+      "learning_rate": 8.387139580935912e-06,
+      "loss": 0.8093,
+      "step": 23983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4674012163432804,
+      "learning_rate": 8.386386534532388e-06,
+      "loss": 0.8688,
+      "step": 23984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4903800390353628,
+      "learning_rate": 8.385633497523838e-06,
+      "loss": 0.8225,
+      "step": 23985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5455394275739467,
+      "learning_rate": 8.384880469914638e-06,
+      "loss": 0.7972,
+      "step": 23986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7171313769228433,
+      "learning_rate": 8.384127451709173e-06,
+      "loss": 0.8349,
+      "step": 23987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5630331388645746,
+      "learning_rate": 8.383374442911836e-06,
+      "loss": 0.786,
+      "step": 23988
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.47359256732089,
+      "learning_rate": 8.382621443527e-06,
+      "loss": 0.9032,
+      "step": 23989
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.564923857093584,
+      "learning_rate": 8.381868453559053e-06,
+      "loss": 0.8599,
+      "step": 23990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6411743891846968,
+      "learning_rate": 8.381115473012381e-06,
+      "loss": 0.7671,
+      "step": 23991
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8707692947830494,
+      "learning_rate": 8.380362501891373e-06,
+      "loss": 1.079,
+      "step": 23992
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9321929973219398,
+      "learning_rate": 8.3796095402004e-06,
+      "loss": 0.9634,
+      "step": 23993
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7784755037034283,
+      "learning_rate": 8.378856587943856e-06,
+      "loss": 0.9721,
+      "step": 23994
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8130902072717425,
+      "learning_rate": 8.378103645126124e-06,
+      "loss": 0.8775,
+      "step": 23995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7789310656749766,
+      "learning_rate": 8.377350711751583e-06,
+      "loss": 0.9293,
+      "step": 23996
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6088353237834498,
+      "learning_rate": 8.376597787824618e-06,
+      "loss": 0.8414,
+      "step": 23997
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0034696174337783,
+      "learning_rate": 8.37584487334962e-06,
+      "loss": 0.8806,
+      "step": 23998
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1487798441476125,
+      "learning_rate": 8.375091968330966e-06,
+      "loss": 0.9593,
+      "step": 23999
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6757109647852289,
+      "learning_rate": 8.374339072773036e-06,
+      "loss": 0.9534,
+      "step": 24000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.731682022099757,
+      "learning_rate": 8.373586186680224e-06,
+      "loss": 0.8419,
+      "step": 24001
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2636704511993857,
+      "learning_rate": 8.37283331005691e-06,
+      "loss": 0.7872,
+      "step": 24002
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7483101960955277,
+      "learning_rate": 8.372080442907471e-06,
+      "loss": 0.7746,
+      "step": 24003
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7479342764836341,
+      "learning_rate": 8.371327585236294e-06,
+      "loss": 0.6898,
+      "step": 24004
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6422350801461396,
+      "learning_rate": 8.370574737047765e-06,
+      "loss": 0.7711,
+      "step": 24005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7541507540432335,
+      "learning_rate": 8.369821898346269e-06,
+      "loss": 0.9025,
+      "step": 24006
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.59132987405706,
+      "learning_rate": 8.369069069136184e-06,
+      "loss": 0.8091,
+      "step": 24007
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6201595388848768,
+      "learning_rate": 8.368316249421894e-06,
+      "loss": 0.9094,
+      "step": 24008
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.472655587333738,
+      "learning_rate": 8.367563439207784e-06,
+      "loss": 0.8253,
+      "step": 24009
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.392547762748454,
+      "learning_rate": 8.366810638498237e-06,
+      "loss": 0.7858,
+      "step": 24010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.546579295821558,
+      "learning_rate": 8.366057847297635e-06,
+      "loss": 0.8944,
+      "step": 24011
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7800627353143745,
+      "learning_rate": 8.36530506561036e-06,
+      "loss": 0.8752,
+      "step": 24012
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4073207186398027,
+      "learning_rate": 8.3645522934408e-06,
+      "loss": 0.8715,
+      "step": 24013
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3913389503367695,
+      "learning_rate": 8.363799530793332e-06,
+      "loss": 0.7949,
+      "step": 24014
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7721912831937072,
+      "learning_rate": 8.363046777672342e-06,
+      "loss": 0.9351,
+      "step": 24015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5783120364430772,
+      "learning_rate": 8.362294034082214e-06,
+      "loss": 0.8794,
+      "step": 24016
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3999242607905122,
+      "learning_rate": 8.361541300027323e-06,
+      "loss": 0.8963,
+      "step": 24017
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5582364659029404,
+      "learning_rate": 8.36078857551206e-06,
+      "loss": 0.9107,
+      "step": 24018
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9208453770275273,
+      "learning_rate": 8.360035860540806e-06,
+      "loss": 0.841,
+      "step": 24019
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.733443844596743,
+      "learning_rate": 8.359283155117943e-06,
+      "loss": 0.8414,
+      "step": 24020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5824275168057234,
+      "learning_rate": 8.35853045924785e-06,
+      "loss": 0.8294,
+      "step": 24021
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6472912980278753,
+      "learning_rate": 8.357777772934914e-06,
+      "loss": 0.8853,
+      "step": 24022
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1272097675684414,
+      "learning_rate": 8.357025096183518e-06,
+      "loss": 0.8929,
+      "step": 24023
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6431936110290266,
+      "learning_rate": 8.356272428998039e-06,
+      "loss": 0.8444,
+      "step": 24024
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5806985055035419,
+      "learning_rate": 8.355519771382861e-06,
+      "loss": 0.7736,
+      "step": 24025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7281033802349793,
+      "learning_rate": 8.35476712334237e-06,
+      "loss": 0.7402,
+      "step": 24026
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7220251924360084,
+      "learning_rate": 8.354014484880946e-06,
+      "loss": 0.754,
+      "step": 24027
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8799974890509596,
+      "learning_rate": 8.35326185600297e-06,
+      "loss": 0.8957,
+      "step": 24028
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3559018699582457,
+      "learning_rate": 8.352509236712823e-06,
+      "loss": 0.7571,
+      "step": 24029
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.902777097458223,
+      "learning_rate": 8.351756627014894e-06,
+      "loss": 0.7657,
+      "step": 24030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.588670926206474,
+      "learning_rate": 8.351004026913555e-06,
+      "loss": 0.9571,
+      "step": 24031
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5940935912456493,
+      "learning_rate": 8.350251436413194e-06,
+      "loss": 0.8651,
+      "step": 24032
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3828278738087574,
+      "learning_rate": 8.34949885551819e-06,
+      "loss": 0.8556,
+      "step": 24033
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.678640506407953,
+      "learning_rate": 8.34874628423293e-06,
+      "loss": 0.8887,
+      "step": 24034
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.798837593010554,
+      "learning_rate": 8.347993722561789e-06,
+      "loss": 0.7183,
+      "step": 24035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8130607873177456,
+      "learning_rate": 8.34724117050915e-06,
+      "loss": 0.738,
+      "step": 24036
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7413242015121226,
+      "learning_rate": 8.346488628079397e-06,
+      "loss": 0.8738,
+      "step": 24037
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5139294880795209,
+      "learning_rate": 8.345736095276914e-06,
+      "loss": 0.9307,
+      "step": 24038
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5565508864427444,
+      "learning_rate": 8.344983572106077e-06,
+      "loss": 0.9243,
+      "step": 24039
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6241858407108514,
+      "learning_rate": 8.344231058571268e-06,
+      "loss": 0.9005,
+      "step": 24040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6983318068368625,
+      "learning_rate": 8.343478554676876e-06,
+      "loss": 0.8043,
+      "step": 24041
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.569181576448147,
+      "learning_rate": 8.342726060427271e-06,
+      "loss": 0.7921,
+      "step": 24042
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7543830099365296,
+      "learning_rate": 8.34197357582684e-06,
+      "loss": 0.9353,
+      "step": 24043
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5790993034749747,
+      "learning_rate": 8.341221100879962e-06,
+      "loss": 0.8542,
+      "step": 24044
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7870229419294853,
+      "learning_rate": 8.340468635591023e-06,
+      "loss": 0.7518,
+      "step": 24045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6881642761944093,
+      "learning_rate": 8.3397161799644e-06,
+      "loss": 0.855,
+      "step": 24046
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.657519335657696,
+      "learning_rate": 8.338963734004477e-06,
+      "loss": 0.8639,
+      "step": 24047
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9281082079206315,
+      "learning_rate": 8.338211297715633e-06,
+      "loss": 0.8888,
+      "step": 24048
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.061272867441909,
+      "learning_rate": 8.337458871102245e-06,
+      "loss": 0.8155,
+      "step": 24049
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5036912537799345,
+      "learning_rate": 8.336706454168701e-06,
+      "loss": 0.9408,
+      "step": 24050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.489010640547188,
+      "learning_rate": 8.335954046919376e-06,
+      "loss": 0.8751,
+      "step": 24051
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9603541780203797,
+      "learning_rate": 8.335201649358658e-06,
+      "loss": 0.8238,
+      "step": 24052
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6198949091003054,
+      "learning_rate": 8.334449261490917e-06,
+      "loss": 0.8752,
+      "step": 24053
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4239497892344026,
+      "learning_rate": 8.333696883320544e-06,
+      "loss": 0.7159,
+      "step": 24054
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.624831208664363,
+      "learning_rate": 8.332944514851916e-06,
+      "loss": 0.9623,
+      "step": 24055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7777884502782906,
+      "learning_rate": 8.332192156089411e-06,
+      "loss": 0.7083,
+      "step": 24056
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6175145370904989,
+      "learning_rate": 8.33143980703741e-06,
+      "loss": 0.822,
+      "step": 24057
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.702287258443282,
+      "learning_rate": 8.330687467700295e-06,
+      "loss": 0.8475,
+      "step": 24058
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.957076356365862,
+      "learning_rate": 8.32993513808245e-06,
+      "loss": 0.8791,
+      "step": 24059
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5663009807768344,
+      "learning_rate": 8.329182818188247e-06,
+      "loss": 0.9206,
+      "step": 24060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4951386455242495,
+      "learning_rate": 8.32843050802207e-06,
+      "loss": 0.9015,
+      "step": 24061
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5004777489190733,
+      "learning_rate": 8.327678207588304e-06,
+      "loss": 0.8649,
+      "step": 24062
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7616037790981336,
+      "learning_rate": 8.326925916891322e-06,
+      "loss": 0.812,
+      "step": 24063
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7888938027215338,
+      "learning_rate": 8.326173635935509e-06,
+      "loss": 0.6955,
+      "step": 24064
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4352907409862796,
+      "learning_rate": 8.325421364725237e-06,
+      "loss": 0.8605,
+      "step": 24065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8086772535952732,
+      "learning_rate": 8.324669103264899e-06,
+      "loss": 0.7201,
+      "step": 24066
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7397903758233121,
+      "learning_rate": 8.323916851558867e-06,
+      "loss": 0.7468,
+      "step": 24067
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6611656542119764,
+      "learning_rate": 8.32316460961152e-06,
+      "loss": 0.8089,
+      "step": 24068
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9339979086826442,
+      "learning_rate": 8.322412377427242e-06,
+      "loss": 0.7975,
+      "step": 24069
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.532086660858693,
+      "learning_rate": 8.321660155010407e-06,
+      "loss": 0.8381,
+      "step": 24070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.570400639822382,
+      "learning_rate": 8.320907942365397e-06,
+      "loss": 0.8977,
+      "step": 24071
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5059882668087676,
+      "learning_rate": 8.320155739496593e-06,
+      "loss": 0.7284,
+      "step": 24072
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.42839781357529,
+      "learning_rate": 8.31940354640838e-06,
+      "loss": 0.8129,
+      "step": 24073
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6787026172222872,
+      "learning_rate": 8.318651363105125e-06,
+      "loss": 0.8211,
+      "step": 24074
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6020501176599617,
+      "learning_rate": 8.317899189591216e-06,
+      "loss": 1.028,
+      "step": 24075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5081324542063828,
+      "learning_rate": 8.317147025871033e-06,
+      "loss": 1.0031,
+      "step": 24076
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.80009566647381,
+      "learning_rate": 8.31639487194895e-06,
+      "loss": 0.7734,
+      "step": 24077
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6426630928380552,
+      "learning_rate": 8.315642727829347e-06,
+      "loss": 0.8694,
+      "step": 24078
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4892148556923908,
+      "learning_rate": 8.31489059351661e-06,
+      "loss": 0.9179,
+      "step": 24079
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.791152073440371,
+      "learning_rate": 8.314138469015112e-06,
+      "loss": 0.7598,
+      "step": 24080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5847222298230896,
+      "learning_rate": 8.313386354329234e-06,
+      "loss": 0.9206,
+      "step": 24081
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5899920536080991,
+      "learning_rate": 8.31263424946335e-06,
+      "loss": 0.8187,
+      "step": 24082
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7259344205196112,
+      "learning_rate": 8.31188215442185e-06,
+      "loss": 0.907,
+      "step": 24083
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5182047545195376,
+      "learning_rate": 8.311130069209103e-06,
+      "loss": 0.755,
+      "step": 24084
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.160771314543529,
+      "learning_rate": 8.310377993829491e-06,
+      "loss": 0.8384,
+      "step": 24085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6209033518707336,
+      "learning_rate": 8.309625928287394e-06,
+      "loss": 0.8283,
+      "step": 24086
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1351389317647866,
+      "learning_rate": 8.308873872587193e-06,
+      "loss": 0.9016,
+      "step": 24087
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6370840680098109,
+      "learning_rate": 8.30812182673326e-06,
+      "loss": 0.8751,
+      "step": 24088
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6999064269361701,
+      "learning_rate": 8.307369790729977e-06,
+      "loss": 0.9211,
+      "step": 24089
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6928601875261944,
+      "learning_rate": 8.306617764581728e-06,
+      "loss": 0.9496,
+      "step": 24090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.731827356328373,
+      "learning_rate": 8.305865748292882e-06,
+      "loss": 0.7783,
+      "step": 24091
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5495959853048662,
+      "learning_rate": 8.305113741867822e-06,
+      "loss": 0.8278,
+      "step": 24092
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.64438970043263,
+      "learning_rate": 8.304361745310926e-06,
+      "loss": 0.6827,
+      "step": 24093
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5614055640646107,
+      "learning_rate": 8.303609758626574e-06,
+      "loss": 0.931,
+      "step": 24094
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4130817744339956,
+      "learning_rate": 8.302857781819144e-06,
+      "loss": 0.7626,
+      "step": 24095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5680230318070598,
+      "learning_rate": 8.302105814893014e-06,
+      "loss": 0.8423,
+      "step": 24096
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7328101949058836,
+      "learning_rate": 8.301353857852561e-06,
+      "loss": 0.7272,
+      "step": 24097
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0476321356721385,
+      "learning_rate": 8.30060191070216e-06,
+      "loss": 0.8364,
+      "step": 24098
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5759661187447451,
+      "learning_rate": 8.299849973446196e-06,
+      "loss": 0.7994,
+      "step": 24099
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7552029826789308,
+      "learning_rate": 8.299098046089042e-06,
+      "loss": 0.8663,
+      "step": 24100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7006487711210314,
+      "learning_rate": 8.298346128635081e-06,
+      "loss": 0.801,
+      "step": 24101
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6867392033492628,
+      "learning_rate": 8.297594221088683e-06,
+      "loss": 0.8996,
+      "step": 24102
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.806707839392089,
+      "learning_rate": 8.296842323454234e-06,
+      "loss": 0.9509,
+      "step": 24103
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7894579196224387,
+      "learning_rate": 8.296090435736107e-06,
+      "loss": 0.6945,
+      "step": 24104
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5739685218124393,
+      "learning_rate": 8.295338557938682e-06,
+      "loss": 0.8927,
+      "step": 24105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6355016291467077,
+      "learning_rate": 8.294586690066331e-06,
+      "loss": 0.7428,
+      "step": 24106
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7959309977248683,
+      "learning_rate": 8.29383483212344e-06,
+      "loss": 0.8151,
+      "step": 24107
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8191001452179032,
+      "learning_rate": 8.293082984114384e-06,
+      "loss": 0.8007,
+      "step": 24108
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5798953398235749,
+      "learning_rate": 8.292331146043537e-06,
+      "loss": 0.8314,
+      "step": 24109
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3800669328843502,
+      "learning_rate": 8.291579317915278e-06,
+      "loss": 0.591,
+      "step": 24110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.49381477984584,
+      "learning_rate": 8.290827499733988e-06,
+      "loss": 0.8772,
+      "step": 24111
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7962676654709973,
+      "learning_rate": 8.290075691504041e-06,
+      "loss": 0.8968,
+      "step": 24112
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5911463783820285,
+      "learning_rate": 8.289323893229812e-06,
+      "loss": 0.8198,
+      "step": 24113
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7793050129291623,
+      "learning_rate": 8.28857210491568e-06,
+      "loss": 1.0017,
+      "step": 24114
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.895871336156866,
+      "learning_rate": 8.287820326566028e-06,
+      "loss": 0.8602,
+      "step": 24115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3867748601317367,
+      "learning_rate": 8.287068558185225e-06,
+      "loss": 0.9247,
+      "step": 24116
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4725245067157298,
+      "learning_rate": 8.286316799777651e-06,
+      "loss": 0.8555,
+      "step": 24117
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7662187821846989,
+      "learning_rate": 8.285565051347688e-06,
+      "loss": 0.7305,
+      "step": 24118
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8180187255466271,
+      "learning_rate": 8.284813312899701e-06,
+      "loss": 0.8933,
+      "step": 24119
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7267921707877203,
+      "learning_rate": 8.284061584438077e-06,
+      "loss": 0.721,
+      "step": 24120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.450848797222273,
+      "learning_rate": 8.283309865967187e-06,
+      "loss": 0.7492,
+      "step": 24121
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5911337185618688,
+      "learning_rate": 8.282558157491417e-06,
+      "loss": 0.7984,
+      "step": 24122
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7409733011329915,
+      "learning_rate": 8.281806459015132e-06,
+      "loss": 0.9016,
+      "step": 24123
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7409329481657655,
+      "learning_rate": 8.281054770542713e-06,
+      "loss": 0.7134,
+      "step": 24124
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5173457509479962,
+      "learning_rate": 8.280303092078542e-06,
+      "loss": 0.9185,
+      "step": 24125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7211745455098664,
+      "learning_rate": 8.279551423626988e-06,
+      "loss": 0.9213,
+      "step": 24126
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515229013443092,
+      "learning_rate": 8.278799765192427e-06,
+      "loss": 0.954,
+      "step": 24127
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5077740023778952,
+      "learning_rate": 8.278048116779242e-06,
+      "loss": 0.8619,
+      "step": 24128
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6891338707906645,
+      "learning_rate": 8.27729647839181e-06,
+      "loss": 0.8685,
+      "step": 24129
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9145526347240143,
+      "learning_rate": 8.276544850034498e-06,
+      "loss": 0.8017,
+      "step": 24130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5459423015285017,
+      "learning_rate": 8.275793231711687e-06,
+      "loss": 0.9337,
+      "step": 24131
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7617240759238479,
+      "learning_rate": 8.275041623427757e-06,
+      "loss": 0.7673,
+      "step": 24132
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5759635370517922,
+      "learning_rate": 8.27429002518708e-06,
+      "loss": 0.8893,
+      "step": 24133
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6832016672058703,
+      "learning_rate": 8.27353843699403e-06,
+      "loss": 0.9983,
+      "step": 24134
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6904454118965844,
+      "learning_rate": 8.272786858852989e-06,
+      "loss": 0.8551,
+      "step": 24135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4993808796043981,
+      "learning_rate": 8.27203529076833e-06,
+      "loss": 0.9562,
+      "step": 24136
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4706950929553477,
+      "learning_rate": 8.271283732744427e-06,
+      "loss": 0.866,
+      "step": 24137
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5652192851390108,
+      "learning_rate": 8.270532184785656e-06,
+      "loss": 0.864,
+      "step": 24138
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8883039779003292,
+      "learning_rate": 8.2697806468964e-06,
+      "loss": 0.7374,
+      "step": 24139
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7380060615497883,
+      "learning_rate": 8.269029119081022e-06,
+      "loss": 0.9648,
+      "step": 24140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6163858696102493,
+      "learning_rate": 8.268277601343908e-06,
+      "loss": 0.9647,
+      "step": 24141
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5090521068303908,
+      "learning_rate": 8.267526093689426e-06,
+      "loss": 0.791,
+      "step": 24142
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.573581404584281,
+      "learning_rate": 8.266774596121962e-06,
+      "loss": 0.8694,
+      "step": 24143
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9739643860934921,
+      "learning_rate": 8.26602310864588e-06,
+      "loss": 0.7816,
+      "step": 24144
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8044681483387077,
+      "learning_rate": 8.265271631265564e-06,
+      "loss": 0.7268,
+      "step": 24145
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7756703565860634,
+      "learning_rate": 8.264520163985385e-06,
+      "loss": 0.7105,
+      "step": 24146
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.488832042679129,
+      "learning_rate": 8.263768706809715e-06,
+      "loss": 0.7467,
+      "step": 24147
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.934659783080685,
+      "learning_rate": 8.263017259742938e-06,
+      "loss": 0.8645,
+      "step": 24148
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5088275197175656,
+      "learning_rate": 8.262265822789422e-06,
+      "loss": 0.7898,
+      "step": 24149
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5901049422208318,
+      "learning_rate": 8.261514395953548e-06,
+      "loss": 0.8979,
+      "step": 24150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.575963607135288,
+      "learning_rate": 8.260762979239683e-06,
+      "loss": 0.9366,
+      "step": 24151
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.634336349324208,
+      "learning_rate": 8.260011572652208e-06,
+      "loss": 0.8206,
+      "step": 24152
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7735507308282752,
+      "learning_rate": 8.259260176195499e-06,
+      "loss": 0.7136,
+      "step": 24153
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6642146107123819,
+      "learning_rate": 8.258508789873926e-06,
+      "loss": 0.9122,
+      "step": 24154
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5074778752837015,
+      "learning_rate": 8.257757413691865e-06,
+      "loss": 0.7567,
+      "step": 24155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9025005887360498,
+      "learning_rate": 8.257006047653693e-06,
+      "loss": 0.8868,
+      "step": 24156
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7050614132402202,
+      "learning_rate": 8.256254691763787e-06,
+      "loss": 0.8965,
+      "step": 24157
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9341587843180008,
+      "learning_rate": 8.255503346026514e-06,
+      "loss": 0.9551,
+      "step": 24158
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8497211202786308,
+      "learning_rate": 8.254752010446252e-06,
+      "loss": 0.8343,
+      "step": 24159
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.600989505002579,
+      "learning_rate": 8.254000685027379e-06,
+      "loss": 0.9062,
+      "step": 24160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5788990513225694,
+      "learning_rate": 8.253249369774268e-06,
+      "loss": 0.8612,
+      "step": 24161
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5998716731360332,
+      "learning_rate": 8.252498064691291e-06,
+      "loss": 0.7725,
+      "step": 24162
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.526247284882419,
+      "learning_rate": 8.251746769782821e-06,
+      "loss": 0.8232,
+      "step": 24163
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1346365522158646,
+      "learning_rate": 8.250995485053238e-06,
+      "loss": 0.8173,
+      "step": 24164
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0633688261397114,
+      "learning_rate": 8.250244210506913e-06,
+      "loss": 0.9772,
+      "step": 24165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7191737082485579,
+      "learning_rate": 8.249492946148217e-06,
+      "loss": 0.9252,
+      "step": 24166
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6872480851726908,
+      "learning_rate": 8.248741691981528e-06,
+      "loss": 0.9137,
+      "step": 24167
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2901015858962643,
+      "learning_rate": 8.247990448011222e-06,
+      "loss": 0.7778,
+      "step": 24168
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6851748000087488,
+      "learning_rate": 8.24723921424167e-06,
+      "loss": 0.9231,
+      "step": 24169
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5559394693891897,
+      "learning_rate": 8.246487990677242e-06,
+      "loss": 0.8514,
+      "step": 24170
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0074510165153363,
+      "learning_rate": 8.245736777322324e-06,
+      "loss": 0.8073,
+      "step": 24171
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.469363119410042,
+      "learning_rate": 8.244985574181274e-06,
+      "loss": 0.9883,
+      "step": 24172
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.509784075647699,
+      "learning_rate": 8.244234381258477e-06,
+      "loss": 0.7715,
+      "step": 24173
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.327986863750292,
+      "learning_rate": 8.243483198558299e-06,
+      "loss": 0.8268,
+      "step": 24174
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7160269305690714,
+      "learning_rate": 8.242732026085127e-06,
+      "loss": 0.8125,
+      "step": 24175
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.525044867006575,
+      "learning_rate": 8.24198086384332e-06,
+      "loss": 0.7739,
+      "step": 24176
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4647392160548693,
+      "learning_rate": 8.241229711837256e-06,
+      "loss": 0.8206,
+      "step": 24177
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.867328244673453,
+      "learning_rate": 8.240478570071312e-06,
+      "loss": 0.7779,
+      "step": 24178
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5449935135066661,
+      "learning_rate": 8.239727438549859e-06,
+      "loss": 0.76,
+      "step": 24179
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5761926688885686,
+      "learning_rate": 8.238976317277266e-06,
+      "loss": 0.7344,
+      "step": 24180
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6588732764400431,
+      "learning_rate": 8.238225206257913e-06,
+      "loss": 0.8208,
+      "step": 24181
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7610448917019492,
+      "learning_rate": 8.237474105496172e-06,
+      "loss": 1.0103,
+      "step": 24182
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6008183128321067,
+      "learning_rate": 8.23672301499641e-06,
+      "loss": 0.7883,
+      "step": 24183
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4779593765289991,
+      "learning_rate": 8.23597193476301e-06,
+      "loss": 0.8305,
+      "step": 24184
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6288629850473935,
+      "learning_rate": 8.23522086480034e-06,
+      "loss": 0.8435,
+      "step": 24185
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5867354240740517,
+      "learning_rate": 8.23446980511277e-06,
+      "loss": 1.0144,
+      "step": 24186
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.68594863970948,
+      "learning_rate": 8.233718755704676e-06,
+      "loss": 0.8867,
+      "step": 24187
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1047430983886843,
+      "learning_rate": 8.232967716580431e-06,
+      "loss": 0.8812,
+      "step": 24188
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8854995340346117,
+      "learning_rate": 8.23221668774441e-06,
+      "loss": 0.8289,
+      "step": 24189
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7968483626649767,
+      "learning_rate": 8.23146566920098e-06,
+      "loss": 0.9053,
+      "step": 24190
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5181187654872377,
+      "learning_rate": 8.230714660954516e-06,
+      "loss": 0.8244,
+      "step": 24191
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6097653545473367,
+      "learning_rate": 8.229963663009394e-06,
+      "loss": 0.8798,
+      "step": 24192
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.559166231858763,
+      "learning_rate": 8.229212675369984e-06,
+      "loss": 0.809,
+      "step": 24193
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1162248929449934,
+      "learning_rate": 8.228461698040657e-06,
+      "loss": 0.8073,
+      "step": 24194
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5144271030566003,
+      "learning_rate": 8.227710731025785e-06,
+      "loss": 0.8936,
+      "step": 24195
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5560088953345357,
+      "learning_rate": 8.226959774329747e-06,
+      "loss": 0.8499,
+      "step": 24196
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6305125473759816,
+      "learning_rate": 8.226208827956908e-06,
+      "loss": 0.7767,
+      "step": 24197
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7599545194167772,
+      "learning_rate": 8.225457891911643e-06,
+      "loss": 0.7495,
+      "step": 24198
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5379172257392761,
+      "learning_rate": 8.224706966198325e-06,
+      "loss": 0.7817,
+      "step": 24199
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.761687940043448,
+      "learning_rate": 8.223956050821322e-06,
+      "loss": 0.8626,
+      "step": 24200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.573671205526512,
+      "learning_rate": 8.223205145785012e-06,
+      "loss": 0.8612,
+      "step": 24201
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5340136878357171,
+      "learning_rate": 8.222454251093763e-06,
+      "loss": 0.8822,
+      "step": 24202
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5186948382308074,
+      "learning_rate": 8.22170336675195e-06,
+      "loss": 0.9215,
+      "step": 24203
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7673080019273402,
+      "learning_rate": 8.220952492763938e-06,
+      "loss": 0.8011,
+      "step": 24204
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.872040368901797,
+      "learning_rate": 8.220201629134109e-06,
+      "loss": 0.8404,
+      "step": 24205
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7588030732754858,
+      "learning_rate": 8.219450775866828e-06,
+      "loss": 0.8459,
+      "step": 24206
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6883940123409824,
+      "learning_rate": 8.218699932966468e-06,
+      "loss": 0.8326,
+      "step": 24207
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.57598552895212,
+      "learning_rate": 8.2179491004374e-06,
+      "loss": 0.8005,
+      "step": 24208
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7519453150796804,
+      "learning_rate": 8.217198278283997e-06,
+      "loss": 0.7323,
+      "step": 24209
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1128723456034058,
+      "learning_rate": 8.216447466510633e-06,
+      "loss": 0.8823,
+      "step": 24210
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7776419191017736,
+      "learning_rate": 8.215696665121674e-06,
+      "loss": 0.7369,
+      "step": 24211
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5058957439273752,
+      "learning_rate": 8.214945874121491e-06,
+      "loss": 0.8425,
+      "step": 24212
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.572524190160496,
+      "learning_rate": 8.214195093514463e-06,
+      "loss": 0.8119,
+      "step": 24213
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5762557205514285,
+      "learning_rate": 8.213444323304955e-06,
+      "loss": 0.813,
+      "step": 24214
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6146977810024115,
+      "learning_rate": 8.21269356349734e-06,
+      "loss": 0.843,
+      "step": 24215
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.593434710186875,
+      "learning_rate": 8.211942814095986e-06,
+      "loss": 0.7617,
+      "step": 24216
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.271129707029268,
+      "learning_rate": 8.211192075105271e-06,
+      "loss": 0.8625,
+      "step": 24217
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5134340958570038,
+      "learning_rate": 8.21044134652956e-06,
+      "loss": 0.7833,
+      "step": 24218
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.557662275350856,
+      "learning_rate": 8.209690628373225e-06,
+      "loss": 0.7277,
+      "step": 24219
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8744344251011487,
+      "learning_rate": 8.208939920640644e-06,
+      "loss": 0.899,
+      "step": 24220
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.742455885658203,
+      "learning_rate": 8.208189223336175e-06,
+      "loss": 0.8409,
+      "step": 24221
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6276862865553314,
+      "learning_rate": 8.207438536464197e-06,
+      "loss": 0.8257,
+      "step": 24222
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4399997854008206,
+      "learning_rate": 8.206687860029079e-06,
+      "loss": 0.7704,
+      "step": 24223
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.617632195688384,
+      "learning_rate": 8.205937194035197e-06,
+      "loss": 0.8392,
+      "step": 24224
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.651244838759274,
+      "learning_rate": 8.205186538486911e-06,
+      "loss": 0.782,
+      "step": 24225
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7616334810313758,
+      "learning_rate": 8.204435893388599e-06,
+      "loss": 0.696,
+      "step": 24226
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9454977012922179,
+      "learning_rate": 8.203685258744633e-06,
+      "loss": 0.9771,
+      "step": 24227
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.520088558200056,
+      "learning_rate": 8.202934634559375e-06,
+      "loss": 0.7465,
+      "step": 24228
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.79983783753231,
+      "learning_rate": 8.202184020837202e-06,
+      "loss": 0.8066,
+      "step": 24229
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3417448593155057,
+      "learning_rate": 8.201433417582482e-06,
+      "loss": 0.9334,
+      "step": 24230
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5612460470275409,
+      "learning_rate": 8.200682824799591e-06,
+      "loss": 0.8087,
+      "step": 24231
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7949703803807608,
+      "learning_rate": 8.199932242492887e-06,
+      "loss": 0.7257,
+      "step": 24232
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6157593049453525,
+      "learning_rate": 8.199181670666751e-06,
+      "loss": 0.7539,
+      "step": 24233
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5774353925169604,
+      "learning_rate": 8.198431109325552e-06,
+      "loss": 0.8985,
+      "step": 24234
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8679016043764929,
+      "learning_rate": 8.197680558473655e-06,
+      "loss": 0.83,
+      "step": 24235
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4780405879136183,
+      "learning_rate": 8.196930018115429e-06,
+      "loss": 0.8667,
+      "step": 24236
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5344518903610727,
+      "learning_rate": 8.196179488255251e-06,
+      "loss": 0.9371,
+      "step": 24237
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.591393487118158,
+      "learning_rate": 8.19542896889749e-06,
+      "loss": 0.7009,
+      "step": 24238
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7686278116389844,
+      "learning_rate": 8.194678460046509e-06,
+      "loss": 0.8809,
+      "step": 24239
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8010259723377982,
+      "learning_rate": 8.19392796170668e-06,
+      "loss": 0.8691,
+      "step": 24240
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6210202314972755,
+      "learning_rate": 8.19317747388238e-06,
+      "loss": 0.9015,
+      "step": 24241
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7371780663034566,
+      "learning_rate": 8.192426996577967e-06,
+      "loss": 0.8159,
+      "step": 24242
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.509737585033754,
+      "learning_rate": 8.191676529797818e-06,
+      "loss": 0.8832,
+      "step": 24243
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6635203612765295,
+      "learning_rate": 8.1909260735463e-06,
+      "loss": 0.8571,
+      "step": 24244
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4757157504695648,
+      "learning_rate": 8.190175627827785e-06,
+      "loss": 0.8053,
+      "step": 24245
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4860261932825227,
+      "learning_rate": 8.18942519264664e-06,
+      "loss": 0.9881,
+      "step": 24246
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6534910313154736,
+      "learning_rate": 8.188674768007234e-06,
+      "loss": 0.833,
+      "step": 24247
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.680992433294381,
+      "learning_rate": 8.18792435391394e-06,
+      "loss": 0.6661,
+      "step": 24248
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4547961823561002,
+      "learning_rate": 8.18717395037112e-06,
+      "loss": 0.8822,
+      "step": 24249
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.186266123536559,
+      "learning_rate": 8.18642355738315e-06,
+      "loss": 0.9679,
+      "step": 24250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5104361576622511,
+      "learning_rate": 8.185673174954394e-06,
+      "loss": 0.8598,
+      "step": 24251
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8138258967605105,
+      "learning_rate": 8.184922803089226e-06,
+      "loss": 0.9643,
+      "step": 24252
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7753551484271936,
+      "learning_rate": 8.184172441792008e-06,
+      "loss": 0.7386,
+      "step": 24253
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7380987063138842,
+      "learning_rate": 8.183422091067115e-06,
+      "loss": 0.9241,
+      "step": 24254
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5469629573981545,
+      "learning_rate": 8.182671750918915e-06,
+      "loss": 0.8738,
+      "step": 24255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5945280499238714,
+      "learning_rate": 8.181921421351775e-06,
+      "loss": 0.7933,
+      "step": 24256
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7471576055527545,
+      "learning_rate": 8.181171102370062e-06,
+      "loss": 0.8446,
+      "step": 24257
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4959691945486855,
+      "learning_rate": 8.180420793978148e-06,
+      "loss": 0.9185,
+      "step": 24258
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5677734332407223,
+      "learning_rate": 8.179670496180401e-06,
+      "loss": 0.7692,
+      "step": 24259
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3647148670851856,
+      "learning_rate": 8.178920208981189e-06,
+      "loss": 0.922,
+      "step": 24260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4385323968058747,
+      "learning_rate": 8.178169932384877e-06,
+      "loss": 0.8215,
+      "step": 24261
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2708919614714866,
+      "learning_rate": 8.177419666395839e-06,
+      "loss": 0.9011,
+      "step": 24262
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6080428083633855,
+      "learning_rate": 8.176669411018441e-06,
+      "loss": 0.7771,
+      "step": 24263
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8040819029684125,
+      "learning_rate": 8.17591916625705e-06,
+      "loss": 0.71,
+      "step": 24264
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6834322550169933,
+      "learning_rate": 8.175168932116033e-06,
+      "loss": 0.9175,
+      "step": 24265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6066296107586244,
+      "learning_rate": 8.174418708599764e-06,
+      "loss": 0.7692,
+      "step": 24266
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5080626035672464,
+      "learning_rate": 8.173668495712605e-06,
+      "loss": 0.9301,
+      "step": 24267
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.909423605070551,
+      "learning_rate": 8.172918293458925e-06,
+      "loss": 0.7886,
+      "step": 24268
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9685632332913339,
+      "learning_rate": 8.1721681018431e-06,
+      "loss": 0.8119,
+      "step": 24269
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5966521910026357,
+      "learning_rate": 8.171417920869483e-06,
+      "loss": 0.8546,
+      "step": 24270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5664732734317497,
+      "learning_rate": 8.170667750542453e-06,
+      "loss": 0.6915,
+      "step": 24271
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4823668615424872,
+      "learning_rate": 8.169917590866372e-06,
+      "loss": 0.8615,
+      "step": 24272
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0553432358511916,
+      "learning_rate": 8.169167441845617e-06,
+      "loss": 0.8402,
+      "step": 24273
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5281558951930825,
+      "learning_rate": 8.168417303484542e-06,
+      "loss": 0.8514,
+      "step": 24274
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6176000538183124,
+      "learning_rate": 8.167667175787524e-06,
+      "loss": 0.8855,
+      "step": 24275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5676862551252762,
+      "learning_rate": 8.16691705875893e-06,
+      "loss": 0.8425,
+      "step": 24276
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.541116969960667,
+      "learning_rate": 8.16616695240312e-06,
+      "loss": 0.8263,
+      "step": 24277
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7060244665174105,
+      "learning_rate": 8.165416856724468e-06,
+      "loss": 0.8651,
+      "step": 24278
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7837938018715715,
+      "learning_rate": 8.164666771727343e-06,
+      "loss": 0.8577,
+      "step": 24279
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6694491838569354,
+      "learning_rate": 8.163916697416109e-06,
+      "loss": 0.8969,
+      "step": 24280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.574815546314811,
+      "learning_rate": 8.163166633795129e-06,
+      "loss": 0.8678,
+      "step": 24281
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6854896820434837,
+      "learning_rate": 8.162416580868777e-06,
+      "loss": 0.7846,
+      "step": 24282
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5461934315800947,
+      "learning_rate": 8.161666538641418e-06,
+      "loss": 0.8556,
+      "step": 24283
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3879276712549964,
+      "learning_rate": 8.160916507117418e-06,
+      "loss": 0.8435,
+      "step": 24284
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4876222264261274,
+      "learning_rate": 8.160166486301141e-06,
+      "loss": 0.8376,
+      "step": 24285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6773920783981124,
+      "learning_rate": 8.159416476196962e-06,
+      "loss": 0.8046,
+      "step": 24286
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.535068189194927,
+      "learning_rate": 8.158666476809243e-06,
+      "loss": 0.7815,
+      "step": 24287
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9034552560114208,
+      "learning_rate": 8.157916488142349e-06,
+      "loss": 0.8337,
+      "step": 24288
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2316090164690214,
+      "learning_rate": 8.157166510200647e-06,
+      "loss": 0.9289,
+      "step": 24289
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6853264853331709,
+      "learning_rate": 8.156416542988505e-06,
+      "loss": 0.861,
+      "step": 24290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.726493479762524,
+      "learning_rate": 8.155666586510293e-06,
+      "loss": 0.7782,
+      "step": 24291
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7110734147949165,
+      "learning_rate": 8.154916640770373e-06,
+      "loss": 0.9022,
+      "step": 24292
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6977292299020141,
+      "learning_rate": 8.15416670577311e-06,
+      "loss": 0.9213,
+      "step": 24293
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4442007847576221,
+      "learning_rate": 8.153416781522876e-06,
+      "loss": 0.7915,
+      "step": 24294
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.64309763281781,
+      "learning_rate": 8.152666868024032e-06,
+      "loss": 0.84,
+      "step": 24295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3656165061616703,
+      "learning_rate": 8.151916965280947e-06,
+      "loss": 0.8947,
+      "step": 24296
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5902858540542364,
+      "learning_rate": 8.151167073297983e-06,
+      "loss": 0.8575,
+      "step": 24297
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6259488919985208,
+      "learning_rate": 8.150417192079516e-06,
+      "loss": 0.9066,
+      "step": 24298
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5529479374384132,
+      "learning_rate": 8.149667321629902e-06,
+      "loss": 0.9452,
+      "step": 24299
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6586827012468963,
+      "learning_rate": 8.148917461953511e-06,
+      "loss": 0.8895,
+      "step": 24300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5658129942406718,
+      "learning_rate": 8.148167613054711e-06,
+      "loss": 0.8008,
+      "step": 24301
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5883552609865452,
+      "learning_rate": 8.14741777493786e-06,
+      "loss": 0.8727,
+      "step": 24302
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.59325711188701,
+      "learning_rate": 8.146667947607335e-06,
+      "loss": 1.0053,
+      "step": 24303
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6002211133872475,
+      "learning_rate": 8.14591813106749e-06,
+      "loss": 0.8712,
+      "step": 24304
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.471876971820935,
+      "learning_rate": 8.145168325322704e-06,
+      "loss": 0.9206,
+      "step": 24305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6184752090835386,
+      "learning_rate": 8.144418530377329e-06,
+      "loss": 0.7898,
+      "step": 24306
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.609230672726084,
+      "learning_rate": 8.14366874623574e-06,
+      "loss": 0.8883,
+      "step": 24307
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6841582666993467,
+      "learning_rate": 8.142918972902301e-06,
+      "loss": 0.8347,
+      "step": 24308
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6072877797777658,
+      "learning_rate": 8.142169210381373e-06,
+      "loss": 0.8164,
+      "step": 24309
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4710275935957633,
+      "learning_rate": 8.141419458677321e-06,
+      "loss": 0.8651,
+      "step": 24310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5875839284574473,
+      "learning_rate": 8.140669717794515e-06,
+      "loss": 0.8863,
+      "step": 24311
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.870062810123447,
+      "learning_rate": 8.139919987737324e-06,
+      "loss": 0.9343,
+      "step": 24312
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5683710068626755,
+      "learning_rate": 8.139170268510104e-06,
+      "loss": 0.8875,
+      "step": 24313
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.571200430894107,
+      "learning_rate": 8.13842056011722e-06,
+      "loss": 1.0255,
+      "step": 24314
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.69604600266894,
+      "learning_rate": 8.137670862563047e-06,
+      "loss": 0.8448,
+      "step": 24315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.621736563205837,
+      "learning_rate": 8.136921175851942e-06,
+      "loss": 0.7351,
+      "step": 24316
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6513825710840282,
+      "learning_rate": 8.136171499988267e-06,
+      "loss": 0.7947,
+      "step": 24317
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7397396232129736,
+      "learning_rate": 8.135421834976396e-06,
+      "loss": 0.7438,
+      "step": 24318
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9031876684294788,
+      "learning_rate": 8.134672180820692e-06,
+      "loss": 0.8501,
+      "step": 24319
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0108733501760225,
+      "learning_rate": 8.133922537525515e-06,
+      "loss": 0.8921,
+      "step": 24320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7638974176238191,
+      "learning_rate": 8.133172905095227e-06,
+      "loss": 0.7477,
+      "step": 24321
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0143614362945077,
+      "learning_rate": 8.132423283534206e-06,
+      "loss": 0.9253,
+      "step": 24322
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4305046367430976,
+      "learning_rate": 8.1316736728468e-06,
+      "loss": 0.9892,
+      "step": 24323
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6604521677179427,
+      "learning_rate": 8.130924073037385e-06,
+      "loss": 0.7868,
+      "step": 24324
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6064431200258895,
+      "learning_rate": 8.13017448411032e-06,
+      "loss": 0.9552,
+      "step": 24325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6463586573582643,
+      "learning_rate": 8.129424906069974e-06,
+      "loss": 0.8691,
+      "step": 24326
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5232047877178756,
+      "learning_rate": 8.128675338920707e-06,
+      "loss": 0.8746,
+      "step": 24327
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7355433989885947,
+      "learning_rate": 8.127925782666884e-06,
+      "loss": 0.9193,
+      "step": 24328
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.672314255354309,
+      "learning_rate": 8.127176237312872e-06,
+      "loss": 0.9229,
+      "step": 24329
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.757719812101903,
+      "learning_rate": 8.126426702863029e-06,
+      "loss": 0.7729,
+      "step": 24330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0501525000779837,
+      "learning_rate": 8.125677179321725e-06,
+      "loss": 0.919,
+      "step": 24331
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6708469501727703,
+      "learning_rate": 8.124927666693324e-06,
+      "loss": 0.8425,
+      "step": 24332
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7448859901002876,
+      "learning_rate": 8.124178164982186e-06,
+      "loss": 0.7896,
+      "step": 24333
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1033022678904376,
+      "learning_rate": 8.123428674192675e-06,
+      "loss": 0.7891,
+      "step": 24334
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6148040717169856,
+      "learning_rate": 8.122679194329157e-06,
+      "loss": 0.8431,
+      "step": 24335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.513535451509474,
+      "learning_rate": 8.121929725396e-06,
+      "loss": 0.9537,
+      "step": 24336
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.132180528661629,
+      "learning_rate": 8.121180267397556e-06,
+      "loss": 0.7559,
+      "step": 24337
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.935214730323073,
+      "learning_rate": 8.120430820338195e-06,
+      "loss": 0.8609,
+      "step": 24338
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.474491770918599,
+      "learning_rate": 8.119681384222284e-06,
+      "loss": 0.8946,
+      "step": 24339
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8603966763488031,
+      "learning_rate": 8.118931959054185e-06,
+      "loss": 0.7512,
+      "step": 24340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.436824537159386,
+      "learning_rate": 8.118182544838258e-06,
+      "loss": 0.9203,
+      "step": 24341
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.674264028627007,
+      "learning_rate": 8.117433141578865e-06,
+      "loss": 0.6982,
+      "step": 24342
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7629593373377758,
+      "learning_rate": 8.116683749280376e-06,
+      "loss": 0.7046,
+      "step": 24343
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6129131498291804,
+      "learning_rate": 8.115934367947151e-06,
+      "loss": 0.8639,
+      "step": 24344
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7750286988820211,
+      "learning_rate": 8.11518499758355e-06,
+      "loss": 0.727,
+      "step": 24345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7446561616458979,
+      "learning_rate": 8.114435638193937e-06,
+      "loss": 0.6952,
+      "step": 24346
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9477893706608918,
+      "learning_rate": 8.113686289782682e-06,
+      "loss": 0.9165,
+      "step": 24347
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6198632379599962,
+      "learning_rate": 8.112936952354139e-06,
+      "loss": 0.8332,
+      "step": 24348
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7904555363978798,
+      "learning_rate": 8.112187625912675e-06,
+      "loss": 0.7343,
+      "step": 24349
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6519851971206856,
+      "learning_rate": 8.111438310462656e-06,
+      "loss": 0.8409,
+      "step": 24350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6190063947370383,
+      "learning_rate": 8.110689006008435e-06,
+      "loss": 0.9687,
+      "step": 24351
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5473220745188665,
+      "learning_rate": 8.109939712554383e-06,
+      "loss": 0.8146,
+      "step": 24352
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6877911500062113,
+      "learning_rate": 8.109190430104858e-06,
+      "loss": 0.8856,
+      "step": 24353
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7469239409229924,
+      "learning_rate": 8.108441158664234e-06,
+      "loss": 0.7002,
+      "step": 24354
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.568718210008863,
+      "learning_rate": 8.107691898236855e-06,
+      "loss": 0.9281,
+      "step": 24355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7641712604465438,
+      "learning_rate": 8.106942648827097e-06,
+      "loss": 0.8574,
+      "step": 24356
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5901892063979488,
+      "learning_rate": 8.106193410439318e-06,
+      "loss": 0.7924,
+      "step": 24357
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.323990504355945,
+      "learning_rate": 8.10544418307788e-06,
+      "loss": 0.8696,
+      "step": 24358
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5169842043350787,
+      "learning_rate": 8.104694966747143e-06,
+      "loss": 0.7739,
+      "step": 24359
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.76115289414948,
+      "learning_rate": 8.103945761451474e-06,
+      "loss": 0.7533,
+      "step": 24360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8495581940744463,
+      "learning_rate": 8.103196567195237e-06,
+      "loss": 0.9538,
+      "step": 24361
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.770367331913125,
+      "learning_rate": 8.102447383982785e-06,
+      "loss": 0.9428,
+      "step": 24362
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5053275661280217,
+      "learning_rate": 8.101698211818485e-06,
+      "loss": 0.7786,
+      "step": 24363
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5400889768959767,
+      "learning_rate": 8.100949050706704e-06,
+      "loss": 0.8192,
+      "step": 24364
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6519700447424988,
+      "learning_rate": 8.100199900651794e-06,
+      "loss": 0.8548,
+      "step": 24365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5177310541062943,
+      "learning_rate": 8.099450761658122e-06,
+      "loss": 0.803,
+      "step": 24366
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2300188667704486,
+      "learning_rate": 8.098701633730051e-06,
+      "loss": 0.9388,
+      "step": 24367
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7340553993270154,
+      "learning_rate": 8.097952516871944e-06,
+      "loss": 0.8956,
+      "step": 24368
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.407018091684441,
+      "learning_rate": 8.097203411088155e-06,
+      "loss": 0.8744,
+      "step": 24369
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9796812529464634,
+      "learning_rate": 8.096454316383049e-06,
+      "loss": 0.8764,
+      "step": 24370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4071751142744544,
+      "learning_rate": 8.095705232760994e-06,
+      "loss": 0.9235,
+      "step": 24371
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.477143361267971,
+      "learning_rate": 8.094956160226342e-06,
+      "loss": 0.8146,
+      "step": 24372
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.428936203298,
+      "learning_rate": 8.09420709878346e-06,
+      "loss": 0.7879,
+      "step": 24373
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9448737542629333,
+      "learning_rate": 8.093458048436705e-06,
+      "loss": 0.8178,
+      "step": 24374
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6374788493116108,
+      "learning_rate": 8.092709009190446e-06,
+      "loss": 0.9583,
+      "step": 24375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3534675179071576,
+      "learning_rate": 8.091959981049034e-06,
+      "loss": 0.7466,
+      "step": 24376
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.712954370675592,
+      "learning_rate": 8.091210964016837e-06,
+      "loss": 0.8818,
+      "step": 24377
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3995512724663506,
+      "learning_rate": 8.090461958098216e-06,
+      "loss": 0.7677,
+      "step": 24378
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.584676276217695,
+      "learning_rate": 8.089712963297527e-06,
+      "loss": 0.8866,
+      "step": 24379
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7761983037369989,
+      "learning_rate": 8.088963979619135e-06,
+      "loss": 0.7421,
+      "step": 24380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7417674083872237,
+      "learning_rate": 8.0882150070674e-06,
+      "loss": 0.7921,
+      "step": 24381
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.523674452771141,
+      "learning_rate": 8.087466045646684e-06,
+      "loss": 0.82,
+      "step": 24382
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7726946453396588,
+      "learning_rate": 8.086717095361342e-06,
+      "loss": 0.6806,
+      "step": 24383
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4524057366348504,
+      "learning_rate": 8.085968156215741e-06,
+      "loss": 0.8732,
+      "step": 24384
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.675425945725925,
+      "learning_rate": 8.085219228214241e-06,
+      "loss": 0.7999,
+      "step": 24385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7788791799396872,
+      "learning_rate": 8.0844703113612e-06,
+      "loss": 0.6951,
+      "step": 24386
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6232318648843573,
+      "learning_rate": 8.083721405660978e-06,
+      "loss": 0.8826,
+      "step": 24387
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7815797255296568,
+      "learning_rate": 8.082972511117936e-06,
+      "loss": 0.8648,
+      "step": 24388
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6519148822891336,
+      "learning_rate": 8.082223627736439e-06,
+      "loss": 0.7438,
+      "step": 24389
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8054100863077904,
+      "learning_rate": 8.08147475552084e-06,
+      "loss": 0.724,
+      "step": 24390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4886281413020759,
+      "learning_rate": 8.0807258944755e-06,
+      "loss": 0.9208,
+      "step": 24391
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7812241200569006,
+      "learning_rate": 8.079977044604788e-06,
+      "loss": 0.8579,
+      "step": 24392
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6456950448779954,
+      "learning_rate": 8.079228205913053e-06,
+      "loss": 0.7497,
+      "step": 24393
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5173344336299583,
+      "learning_rate": 8.07847937840466e-06,
+      "loss": 0.8141,
+      "step": 24394
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.692375208436697,
+      "learning_rate": 8.077730562083967e-06,
+      "loss": 0.8412,
+      "step": 24395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1545351693368544,
+      "learning_rate": 8.07698175695534e-06,
+      "loss": 0.8807,
+      "step": 24396
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6132939728715494,
+      "learning_rate": 8.07623296302313e-06,
+      "loss": 0.7705,
+      "step": 24397
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4727375361008181,
+      "learning_rate": 8.075484180291702e-06,
+      "loss": 0.8675,
+      "step": 24398
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5116977131296856,
+      "learning_rate": 8.074735408765417e-06,
+      "loss": 0.7807,
+      "step": 24399
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7237248573770554,
+      "learning_rate": 8.073986648448627e-06,
+      "loss": 0.8265,
+      "step": 24400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7876179737258443,
+      "learning_rate": 8.073237899345699e-06,
+      "loss": 0.7056,
+      "step": 24401
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5603863544198684,
+      "learning_rate": 8.072489161460986e-06,
+      "loss": 0.8558,
+      "step": 24402
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5067700818903704,
+      "learning_rate": 8.07174043479886e-06,
+      "loss": 1.0259,
+      "step": 24403
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6680238197171413,
+      "learning_rate": 8.070991719363664e-06,
+      "loss": 0.9386,
+      "step": 24404
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.65535140013937,
+      "learning_rate": 8.070243015159767e-06,
+      "loss": 0.8975,
+      "step": 24405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4229725536392412,
+      "learning_rate": 8.069494322191526e-06,
+      "loss": 0.7729,
+      "step": 24406
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5569120803705254,
+      "learning_rate": 8.068745640463305e-06,
+      "loss": 0.843,
+      "step": 24407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5298998761835614,
+      "learning_rate": 8.067996969979451e-06,
+      "loss": 0.8261,
+      "step": 24408
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4643178109893147,
+      "learning_rate": 8.067248310744334e-06,
+      "loss": 0.9123,
+      "step": 24409
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7318550901834369,
+      "learning_rate": 8.066499662762312e-06,
+      "loss": 0.8958,
+      "step": 24410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8124400775544387,
+      "learning_rate": 8.065751026037738e-06,
+      "loss": 0.9482,
+      "step": 24411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7641073283507285,
+      "learning_rate": 8.065002400574971e-06,
+      "loss": 0.7478,
+      "step": 24412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.377205823942053,
+      "learning_rate": 8.064253786378375e-06,
+      "loss": 0.899,
+      "step": 24413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4735539178935626,
+      "learning_rate": 8.063505183452309e-06,
+      "loss": 0.8798,
+      "step": 24414
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7625310392137447,
+      "learning_rate": 8.062756591801124e-06,
+      "loss": 0.6961,
+      "step": 24415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8504353233641938,
+      "learning_rate": 8.062008011429187e-06,
+      "loss": 0.7381,
+      "step": 24416
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7667373204827712,
+      "learning_rate": 8.061259442340852e-06,
+      "loss": 0.7027,
+      "step": 24417
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.564910473402532,
+      "learning_rate": 8.06051088454048e-06,
+      "loss": 0.8552,
+      "step": 24418
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7560398816537655,
+      "learning_rate": 8.059762338032423e-06,
+      "loss": 0.8934,
+      "step": 24419
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.44186400767697,
+      "learning_rate": 8.059013802821047e-06,
+      "loss": 0.8464,
+      "step": 24420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.53394976893052,
+      "learning_rate": 8.05826527891071e-06,
+      "loss": 0.9724,
+      "step": 24421
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6652323968572675,
+      "learning_rate": 8.057516766305765e-06,
+      "loss": 0.8101,
+      "step": 24422
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6550539253591092,
+      "learning_rate": 8.056768265010568e-06,
+      "loss": 0.8578,
+      "step": 24423
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7942037879483064,
+      "learning_rate": 8.056019775029488e-06,
+      "loss": 0.7702,
+      "step": 24424
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6682654270570703,
+      "learning_rate": 8.055271296366874e-06,
+      "loss": 0.8807,
+      "step": 24425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.674472921705052,
+      "learning_rate": 8.054522829027084e-06,
+      "loss": 0.9005,
+      "step": 24426
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.825270689160014,
+      "learning_rate": 8.053774373014478e-06,
+      "loss": 0.9443,
+      "step": 24427
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6365061921514144,
+      "learning_rate": 8.053025928333418e-06,
+      "loss": 0.7929,
+      "step": 24428
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7786135957843339,
+      "learning_rate": 8.052277494988254e-06,
+      "loss": 0.7724,
+      "step": 24429
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6853972320189692,
+      "learning_rate": 8.051529072983349e-06,
+      "loss": 0.8459,
+      "step": 24430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3884653236763622,
+      "learning_rate": 8.050780662323059e-06,
+      "loss": 0.8586,
+      "step": 24431
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.573204004452894,
+      "learning_rate": 8.050032263011738e-06,
+      "loss": 0.8112,
+      "step": 24432
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6583773400494255,
+      "learning_rate": 8.049283875053747e-06,
+      "loss": 0.9499,
+      "step": 24433
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.123087935908738,
+      "learning_rate": 8.048535498453445e-06,
+      "loss": 0.8013,
+      "step": 24434
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3961168342169408,
+      "learning_rate": 8.047787133215188e-06,
+      "loss": 0.8213,
+      "step": 24435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6612609709281176,
+      "learning_rate": 8.047038779343327e-06,
+      "loss": 0.787,
+      "step": 24436
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7021332008534236,
+      "learning_rate": 8.046290436842229e-06,
+      "loss": 0.7956,
+      "step": 24437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8684659066954197,
+      "learning_rate": 8.045542105716247e-06,
+      "loss": 0.8813,
+      "step": 24438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5750459933671561,
+      "learning_rate": 8.044793785969736e-06,
+      "loss": 0.9384,
+      "step": 24439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.548528901430367,
+      "learning_rate": 8.044045477607053e-06,
+      "loss": 0.8977,
+      "step": 24440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6837623329516715,
+      "learning_rate": 8.04329718063256e-06,
+      "loss": 0.7882,
+      "step": 24441
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.64274497689376,
+      "learning_rate": 8.042548895050609e-06,
+      "loss": 0.8295,
+      "step": 24442
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.559007479958853,
+      "learning_rate": 8.041800620865558e-06,
+      "loss": 0.8724,
+      "step": 24443
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7390837582950536,
+      "learning_rate": 8.041052358081761e-06,
+      "loss": 0.8837,
+      "step": 24444
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6179064569615236,
+      "learning_rate": 8.040304106703582e-06,
+      "loss": 0.7726,
+      "step": 24445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1562922670264086,
+      "learning_rate": 8.039555866735371e-06,
+      "loss": 0.7761,
+      "step": 24446
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.333451854174602,
+      "learning_rate": 8.038807638181486e-06,
+      "loss": 0.923,
+      "step": 24447
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3733156734795988,
+      "learning_rate": 8.038059421046281e-06,
+      "loss": 0.9127,
+      "step": 24448
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.531808779040094,
+      "learning_rate": 8.037311215334122e-06,
+      "loss": 0.79,
+      "step": 24449
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.56730104477116,
+      "learning_rate": 8.036563021049354e-06,
+      "loss": 0.8624,
+      "step": 24450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8608368205048689,
+      "learning_rate": 8.035814838196335e-06,
+      "loss": 0.8037,
+      "step": 24451
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.552284020627727,
+      "learning_rate": 8.035066666779432e-06,
+      "loss": 0.9809,
+      "step": 24452
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6600792177820085,
+      "learning_rate": 8.034318506802987e-06,
+      "loss": 0.9202,
+      "step": 24453
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8586683869997074,
+      "learning_rate": 8.033570358271362e-06,
+      "loss": 0.9057,
+      "step": 24454
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4629163823753664,
+      "learning_rate": 8.032822221188913e-06,
+      "loss": 0.9169,
+      "step": 24455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6625391016814584,
+      "learning_rate": 8.03207409556e-06,
+      "loss": 0.8057,
+      "step": 24456
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7011312035479178,
+      "learning_rate": 8.03132598138897e-06,
+      "loss": 0.7558,
+      "step": 24457
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9307707374558536,
+      "learning_rate": 8.030577878680185e-06,
+      "loss": 0.7125,
+      "step": 24458
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.651095671485328,
+      "learning_rate": 8.029829787438001e-06,
+      "loss": 0.7855,
+      "step": 24459
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.682915882274788,
+      "learning_rate": 8.02908170766677e-06,
+      "loss": 0.9991,
+      "step": 24460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.586924683315242,
+      "learning_rate": 8.028333639370846e-06,
+      "loss": 0.9349,
+      "step": 24461
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4170966571208274,
+      "learning_rate": 8.027585582554592e-06,
+      "loss": 0.8494,
+      "step": 24462
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8602322148258639,
+      "learning_rate": 8.02683753722236e-06,
+      "loss": 0.8448,
+      "step": 24463
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.023210885697124,
+      "learning_rate": 8.0260895033785e-06,
+      "loss": 0.8482,
+      "step": 24464
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9186963023536399,
+      "learning_rate": 8.025341481027374e-06,
+      "loss": 0.8959,
+      "step": 24465
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4694966624142392,
+      "learning_rate": 8.024593470173337e-06,
+      "loss": 0.7851,
+      "step": 24466
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4550771109806278,
+      "learning_rate": 8.023845470820742e-06,
+      "loss": 0.9404,
+      "step": 24467
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.563150484723514,
+      "learning_rate": 8.02309748297394e-06,
+      "loss": 0.7991,
+      "step": 24468
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.392149076335609,
+      "learning_rate": 8.022349506637294e-06,
+      "loss": 0.8676,
+      "step": 24469
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.592805060867067,
+      "learning_rate": 8.021601541815156e-06,
+      "loss": 0.8082,
+      "step": 24470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6355693388113821,
+      "learning_rate": 8.02085358851188e-06,
+      "loss": 0.9655,
+      "step": 24471
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8689236321897522,
+      "learning_rate": 8.020105646731815e-06,
+      "loss": 0.8702,
+      "step": 24472
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5868945086448234,
+      "learning_rate": 8.01935771647933e-06,
+      "loss": 0.7416,
+      "step": 24473
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.52969514583215,
+      "learning_rate": 8.018609797758768e-06,
+      "loss": 0.7544,
+      "step": 24474
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6905786464766592,
+      "learning_rate": 8.017861890574488e-06,
+      "loss": 0.9207,
+      "step": 24475
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5466019801986597,
+      "learning_rate": 8.01711399493084e-06,
+      "loss": 0.9224,
+      "step": 24476
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.851775604851119,
+      "learning_rate": 8.016366110832186e-06,
+      "loss": 1.0983,
+      "step": 24477
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3689224064485996,
+      "learning_rate": 8.015618238282876e-06,
+      "loss": 0.9898,
+      "step": 24478
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5350237456776683,
+      "learning_rate": 8.014870377287263e-06,
+      "loss": 0.868,
+      "step": 24479
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7624306466639134,
+      "learning_rate": 8.014122527849706e-06,
+      "loss": 0.7116,
+      "step": 24480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0979189769536832,
+      "learning_rate": 8.013374689974554e-06,
+      "loss": 0.7882,
+      "step": 24481
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.572346509976468,
+      "learning_rate": 8.012626863666163e-06,
+      "loss": 0.9202,
+      "step": 24482
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5822551392448374,
+      "learning_rate": 8.011879048928888e-06,
+      "loss": 0.8718,
+      "step": 24483
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.200711436864692,
+      "learning_rate": 8.011131245767086e-06,
+      "loss": 0.8244,
+      "step": 24484
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.804656254324243,
+      "learning_rate": 8.010383454185103e-06,
+      "loss": 0.7777,
+      "step": 24485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4950445265628434,
+      "learning_rate": 8.0096356741873e-06,
+      "loss": 0.7804,
+      "step": 24486
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7407812847045867,
+      "learning_rate": 8.00888790577803e-06,
+      "loss": 0.6938,
+      "step": 24487
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.40892664692632,
+      "learning_rate": 8.008140148961642e-06,
+      "loss": 0.7927,
+      "step": 24488
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.865282363642515,
+      "learning_rate": 8.007392403742492e-06,
+      "loss": 0.9161,
+      "step": 24489
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6031519638121996,
+      "learning_rate": 8.006644670124935e-06,
+      "loss": 0.9406,
+      "step": 24490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4505979449444315,
+      "learning_rate": 8.005896948113326e-06,
+      "loss": 0.9679,
+      "step": 24491
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9532632040956335,
+      "learning_rate": 8.005149237712014e-06,
+      "loss": 0.8331,
+      "step": 24492
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5123180327917582,
+      "learning_rate": 8.004401538925353e-06,
+      "loss": 0.8513,
+      "step": 24493
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.637691551005423,
+      "learning_rate": 8.003653851757703e-06,
+      "loss": 0.7665,
+      "step": 24494
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.070226888851795,
+      "learning_rate": 8.00290617621341e-06,
+      "loss": 0.7626,
+      "step": 24495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5167256104285627,
+      "learning_rate": 8.002158512296829e-06,
+      "loss": 0.7529,
+      "step": 24496
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.542210779269812,
+      "learning_rate": 8.001410860012311e-06,
+      "loss": 0.8054,
+      "step": 24497
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8750409944468558,
+      "learning_rate": 8.000663219364217e-06,
+      "loss": 1.0481,
+      "step": 24498
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9578662046841548,
+      "learning_rate": 7.99991559035689e-06,
+      "loss": 0.8816,
+      "step": 24499
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.508521340495785,
+      "learning_rate": 7.99916797299469e-06,
+      "loss": 0.7231,
+      "step": 24500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6538789791959667,
+      "learning_rate": 7.99842036728197e-06,
+      "loss": 0.8184,
+      "step": 24501
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5490560690090955,
+      "learning_rate": 7.997672773223077e-06,
+      "loss": 0.7675,
+      "step": 24502
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6506853529454368,
+      "learning_rate": 7.996925190822367e-06,
+      "loss": 0.9153,
+      "step": 24503
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6135246220504746,
+      "learning_rate": 7.996177620084191e-06,
+      "loss": 0.848,
+      "step": 24504
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9006488806721666,
+      "learning_rate": 7.99543006101291e-06,
+      "loss": 0.924,
+      "step": 24505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7069471686216433,
+      "learning_rate": 7.994682513612863e-06,
+      "loss": 0.9319,
+      "step": 24506
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6605772513584767,
+      "learning_rate": 7.993934977888412e-06,
+      "loss": 0.7583,
+      "step": 24507
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7436993456602328,
+      "learning_rate": 7.993187453843908e-06,
+      "loss": 0.8838,
+      "step": 24508
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7108050638697889,
+      "learning_rate": 7.992439941483702e-06,
+      "loss": 0.6744,
+      "step": 24509
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6914854854663277,
+      "learning_rate": 7.991692440812143e-06,
+      "loss": 0.74,
+      "step": 24510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7922301144312153,
+      "learning_rate": 7.990944951833586e-06,
+      "loss": 0.7756,
+      "step": 24511
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.77281372919489,
+      "learning_rate": 7.99019747455239e-06,
+      "loss": 0.8895,
+      "step": 24512
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7601957131989878,
+      "learning_rate": 7.989450008972893e-06,
+      "loss": 0.8884,
+      "step": 24513
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5908716273198358,
+      "learning_rate": 7.988702555099458e-06,
+      "loss": 0.9437,
+      "step": 24514
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7636884773237067,
+      "learning_rate": 7.987955112936435e-06,
+      "loss": 0.8343,
+      "step": 24515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4712888384575726,
+      "learning_rate": 7.987207682488172e-06,
+      "loss": 0.9244,
+      "step": 24516
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5948506702361904,
+      "learning_rate": 7.986460263759022e-06,
+      "loss": 0.748,
+      "step": 24517
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5334398788636512,
+      "learning_rate": 7.98571285675334e-06,
+      "loss": 0.7585,
+      "step": 24518
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4048699116223893,
+      "learning_rate": 7.984965461475477e-06,
+      "loss": 0.8516,
+      "step": 24519
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.604829932294646,
+      "learning_rate": 7.984218077929782e-06,
+      "loss": 0.8759,
+      "step": 24520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.670315308193254,
+      "learning_rate": 7.983470706120604e-06,
+      "loss": 0.9394,
+      "step": 24521
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7559801098696153,
+      "learning_rate": 7.982723346052302e-06,
+      "loss": 0.8464,
+      "step": 24522
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7596898494677851,
+      "learning_rate": 7.981975997729222e-06,
+      "loss": 0.6938,
+      "step": 24523
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7661836514063438,
+      "learning_rate": 7.981228661155718e-06,
+      "loss": 0.732,
+      "step": 24524
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5379793602106817,
+      "learning_rate": 7.980481336336135e-06,
+      "loss": 0.7885,
+      "step": 24525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.64826583466341,
+      "learning_rate": 7.979734023274835e-06,
+      "loss": 0.7725,
+      "step": 24526
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8115258569007202,
+      "learning_rate": 7.97898672197616e-06,
+      "loss": 0.9142,
+      "step": 24527
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4706753742245582,
+      "learning_rate": 7.978239432444466e-06,
+      "loss": 0.8969,
+      "step": 24528
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7256179433963452,
+      "learning_rate": 7.977492154684097e-06,
+      "loss": 0.9344,
+      "step": 24529
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.639538542856582,
+      "learning_rate": 7.976744888699416e-06,
+      "loss": 0.8407,
+      "step": 24530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4856411213855432,
+      "learning_rate": 7.975997634494764e-06,
+      "loss": 0.8992,
+      "step": 24531
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7845385628863757,
+      "learning_rate": 7.975250392074493e-06,
+      "loss": 0.7781,
+      "step": 24532
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5727530179906515,
+      "learning_rate": 7.97450316144296e-06,
+      "loss": 0.8078,
+      "step": 24533
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2395165223943456,
+      "learning_rate": 7.973755942604505e-06,
+      "loss": 0.8099,
+      "step": 24534
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.550118200295701,
+      "learning_rate": 7.973008735563486e-06,
+      "loss": 0.9437,
+      "step": 24535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7926964951442967,
+      "learning_rate": 7.972261540324252e-06,
+      "loss": 0.733,
+      "step": 24536
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7012417265142454,
+      "learning_rate": 7.971514356891157e-06,
+      "loss": 0.736,
+      "step": 24537
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5932526992977452,
+      "learning_rate": 7.970767185268541e-06,
+      "loss": 0.8963,
+      "step": 24538
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.047385402570625,
+      "learning_rate": 7.970020025460765e-06,
+      "loss": 0.8094,
+      "step": 24539
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7521401171515416,
+      "learning_rate": 7.969272877472178e-06,
+      "loss": 0.8645,
+      "step": 24540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5296210730806665,
+      "learning_rate": 7.968525741307123e-06,
+      "loss": 0.7851,
+      "step": 24541
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7792590145245422,
+      "learning_rate": 7.967778616969952e-06,
+      "loss": 0.7154,
+      "step": 24542
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5101517650051501,
+      "learning_rate": 7.96703150446502e-06,
+      "loss": 0.9416,
+      "step": 24543
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6346759071151078,
+      "learning_rate": 7.966284403796677e-06,
+      "loss": 0.9206,
+      "step": 24544
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7676958372708443,
+      "learning_rate": 7.965537314969267e-06,
+      "loss": 0.749,
+      "step": 24545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.770130257990734,
+      "learning_rate": 7.964790237987142e-06,
+      "loss": 0.9831,
+      "step": 24546
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6174508772371874,
+      "learning_rate": 7.964043172854657e-06,
+      "loss": 0.8386,
+      "step": 24547
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.58971433393241,
+      "learning_rate": 7.963296119576154e-06,
+      "loss": 0.912,
+      "step": 24548
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5131060891091621,
+      "learning_rate": 7.962549078155984e-06,
+      "loss": 0.9487,
+      "step": 24549
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7344316810101943,
+      "learning_rate": 7.9618020485985e-06,
+      "loss": 0.7705,
+      "step": 24550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9981943685326649,
+      "learning_rate": 7.961055030908052e-06,
+      "loss": 0.813,
+      "step": 24551
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5797320767467389,
+      "learning_rate": 7.960308025088985e-06,
+      "loss": 0.7654,
+      "step": 24552
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7376287242086269,
+      "learning_rate": 7.959561031145648e-06,
+      "loss": 0.6602,
+      "step": 24553
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8757811531399673,
+      "learning_rate": 7.958814049082399e-06,
+      "loss": 0.9988,
+      "step": 24554
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6320427936162654,
+      "learning_rate": 7.958067078903574e-06,
+      "loss": 0.8422,
+      "step": 24555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0225554789345157,
+      "learning_rate": 7.957320120613535e-06,
+      "loss": 0.8745,
+      "step": 24556
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6273964099689555,
+      "learning_rate": 7.956573174216619e-06,
+      "loss": 0.9018,
+      "step": 24557
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7003112708713182,
+      "learning_rate": 7.955826239717189e-06,
+      "loss": 0.8458,
+      "step": 24558
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6352849293813292,
+      "learning_rate": 7.955079317119578e-06,
+      "loss": 0.9108,
+      "step": 24559
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4905350624014708,
+      "learning_rate": 7.954332406428145e-06,
+      "loss": 0.9201,
+      "step": 24560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8579413825868492,
+      "learning_rate": 7.953585507647239e-06,
+      "loss": 0.8075,
+      "step": 24561
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.561508708709047,
+      "learning_rate": 7.952838620781201e-06,
+      "loss": 0.7907,
+      "step": 24562
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.602342334847503,
+      "learning_rate": 7.952091745834387e-06,
+      "loss": 0.9928,
+      "step": 24563
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7528387331344317,
+      "learning_rate": 7.951344882811145e-06,
+      "loss": 0.8363,
+      "step": 24564
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6149116485099335,
+      "learning_rate": 7.950598031715822e-06,
+      "loss": 0.8491,
+      "step": 24565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.608226481212312,
+      "learning_rate": 7.949851192552762e-06,
+      "loss": 0.8788,
+      "step": 24566
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.455336205825761,
+      "learning_rate": 7.94910436532632e-06,
+      "loss": 0.902,
+      "step": 24567
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8311696044623729,
+      "learning_rate": 7.948357550040843e-06,
+      "loss": 0.8362,
+      "step": 24568
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.498035414816963,
+      "learning_rate": 7.947610746700677e-06,
+      "loss": 0.8475,
+      "step": 24569
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.528229287593426,
+      "learning_rate": 7.946863955310169e-06,
+      "loss": 0.8302,
+      "step": 24570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4622036348728409,
+      "learning_rate": 7.94611717587367e-06,
+      "loss": 0.9535,
+      "step": 24571
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5092288374643124,
+      "learning_rate": 7.94537040839553e-06,
+      "loss": 0.8415,
+      "step": 24572
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4421834419673591,
+      "learning_rate": 7.944623652880092e-06,
+      "loss": 0.9119,
+      "step": 24573
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5087670465054412,
+      "learning_rate": 7.943876909331702e-06,
+      "loss": 0.931,
+      "step": 24574
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.683351519004729,
+      "learning_rate": 7.943130177754718e-06,
+      "loss": 0.9309,
+      "step": 24575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0947278472367374,
+      "learning_rate": 7.942383458153478e-06,
+      "loss": 0.8471,
+      "step": 24576
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4743218412194568,
+      "learning_rate": 7.941636750532334e-06,
+      "loss": 0.8257,
+      "step": 24577
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5104820434777804,
+      "learning_rate": 7.940890054895627e-06,
+      "loss": 0.7926,
+      "step": 24578
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7271539646771157,
+      "learning_rate": 7.940143371247718e-06,
+      "loss": 0.8119,
+      "step": 24579
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5197430379710128,
+      "learning_rate": 7.939396699592942e-06,
+      "loss": 0.8856,
+      "step": 24580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7846532444013827,
+      "learning_rate": 7.938650039935654e-06,
+      "loss": 0.8059,
+      "step": 24581
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8857590347038264,
+      "learning_rate": 7.937903392280198e-06,
+      "loss": 0.8168,
+      "step": 24582
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.613863511800793,
+      "learning_rate": 7.937156756630915e-06,
+      "loss": 0.8603,
+      "step": 24583
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8200449308127514,
+      "learning_rate": 7.936410132992163e-06,
+      "loss": 0.7245,
+      "step": 24584
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6655109051761572,
+      "learning_rate": 7.935663521368283e-06,
+      "loss": 0.7933,
+      "step": 24585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5240035724547554,
+      "learning_rate": 7.93491692176363e-06,
+      "loss": 0.8135,
+      "step": 24586
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.726542267307289,
+      "learning_rate": 7.934170334182537e-06,
+      "loss": 0.7471,
+      "step": 24587
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4986897220974094,
+      "learning_rate": 7.93342375862936e-06,
+      "loss": 0.7423,
+      "step": 24588
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7206398804652471,
+      "learning_rate": 7.932677195108447e-06,
+      "loss": 0.7422,
+      "step": 24589
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7685312252809384,
+      "learning_rate": 7.93193064362414e-06,
+      "loss": 0.8991,
+      "step": 24590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6365822311235043,
+      "learning_rate": 7.931184104180784e-06,
+      "loss": 0.7915,
+      "step": 24591
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7306414238109684,
+      "learning_rate": 7.930437576782732e-06,
+      "loss": 0.88,
+      "step": 24592
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5498294657449647,
+      "learning_rate": 7.92969106143433e-06,
+      "loss": 0.7984,
+      "step": 24593
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.143940452486402,
+      "learning_rate": 7.928944558139921e-06,
+      "loss": 0.7923,
+      "step": 24594
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5099373646666154,
+      "learning_rate": 7.928198066903849e-06,
+      "loss": 0.9309,
+      "step": 24595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6062982141656963,
+      "learning_rate": 7.927451587730467e-06,
+      "loss": 0.9756,
+      "step": 24596
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.830039234432617,
+      "learning_rate": 7.926705120624119e-06,
+      "loss": 0.9048,
+      "step": 24597
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8362881122925214,
+      "learning_rate": 7.925958665589146e-06,
+      "loss": 0.8396,
+      "step": 24598
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7108329928214854,
+      "learning_rate": 7.9252122226299e-06,
+      "loss": 0.7756,
+      "step": 24599
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6297032334901123,
+      "learning_rate": 7.924465791750728e-06,
+      "loss": 0.9485,
+      "step": 24600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7942757059235233,
+      "learning_rate": 7.923719372955973e-06,
+      "loss": 0.7894,
+      "step": 24601
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6041695492566181,
+      "learning_rate": 7.922972966249977e-06,
+      "loss": 0.7922,
+      "step": 24602
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.81865622517933,
+      "learning_rate": 7.922226571637096e-06,
+      "loss": 1.0136,
+      "step": 24603
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7544829129575925,
+      "learning_rate": 7.921480189121665e-06,
+      "loss": 0.7603,
+      "step": 24604
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5574307447468652,
+      "learning_rate": 7.920733818708036e-06,
+      "loss": 0.882,
+      "step": 24605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0544047790192557,
+      "learning_rate": 7.919987460400549e-06,
+      "loss": 0.863,
+      "step": 24606
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4445138653867784,
+      "learning_rate": 7.919241114203563e-06,
+      "loss": 0.8407,
+      "step": 24607
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.480839513819328,
+      "learning_rate": 7.918494780121404e-06,
+      "loss": 0.8708,
+      "step": 24608
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7851997673572833,
+      "learning_rate": 7.917748458158432e-06,
+      "loss": 0.7716,
+      "step": 24609
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6017799136102617,
+      "learning_rate": 7.917002148318989e-06,
+      "loss": 0.9182,
+      "step": 24610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4170375831039341,
+      "learning_rate": 7.916255850607415e-06,
+      "loss": 0.7593,
+      "step": 24611
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.732129522745466,
+      "learning_rate": 7.915509565028061e-06,
+      "loss": 0.8677,
+      "step": 24612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4868335999863334,
+      "learning_rate": 7.914763291585271e-06,
+      "loss": 0.903,
+      "step": 24613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5271775722613652,
+      "learning_rate": 7.91401703028339e-06,
+      "loss": 0.9557,
+      "step": 24614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6556778340118765,
+      "learning_rate": 7.91327078112676e-06,
+      "loss": 0.8932,
+      "step": 24615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9083144079279386,
+      "learning_rate": 7.912524544119728e-06,
+      "loss": 0.8657,
+      "step": 24616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7533678746535772,
+      "learning_rate": 7.911778319266643e-06,
+      "loss": 0.6975,
+      "step": 24617
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6946587975876108,
+      "learning_rate": 7.911032106571842e-06,
+      "loss": 0.9009,
+      "step": 24618
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9829336978754364,
+      "learning_rate": 7.910285906039673e-06,
+      "loss": 0.8441,
+      "step": 24619
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6173365846448375,
+      "learning_rate": 7.909539717674483e-06,
+      "loss": 0.8442,
+      "step": 24620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6436346416623338,
+      "learning_rate": 7.908793541480615e-06,
+      "loss": 0.9328,
+      "step": 24621
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6982783708185,
+      "learning_rate": 7.908047377462413e-06,
+      "loss": 0.985,
+      "step": 24622
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5035925387416453,
+      "learning_rate": 7.907301225624218e-06,
+      "loss": 0.8519,
+      "step": 24623
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7526330326356825,
+      "learning_rate": 7.906555085970382e-06,
+      "loss": 0.7451,
+      "step": 24624
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1385063936931545,
+      "learning_rate": 7.905808958505243e-06,
+      "loss": 0.894,
+      "step": 24625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7737845569013804,
+      "learning_rate": 7.905062843233148e-06,
+      "loss": 0.7941,
+      "step": 24626
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5006578143401237,
+      "learning_rate": 7.904316740158437e-06,
+      "loss": 0.9569,
+      "step": 24627
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4533385922276325,
+      "learning_rate": 7.903570649285463e-06,
+      "loss": 0.8733,
+      "step": 24628
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7280579420009703,
+      "learning_rate": 7.902824570618562e-06,
+      "loss": 0.9482,
+      "step": 24629
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7851540010970763,
+      "learning_rate": 7.90207850416208e-06,
+      "loss": 0.8883,
+      "step": 24630
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5489365962809405,
+      "learning_rate": 7.901332449920365e-06,
+      "loss": 0.8566,
+      "step": 24631
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6109186925500416,
+      "learning_rate": 7.90058640789775e-06,
+      "loss": 0.9917,
+      "step": 24632
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6721191284412502,
+      "learning_rate": 7.899840378098589e-06,
+      "loss": 0.8804,
+      "step": 24633
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6680618419356417,
+      "learning_rate": 7.89909436052722e-06,
+      "loss": 0.865,
+      "step": 24634
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7668146519490767,
+      "learning_rate": 7.898348355187994e-06,
+      "loss": 0.6511,
+      "step": 24635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8445668764017091,
+      "learning_rate": 7.897602362085246e-06,
+      "loss": 0.8687,
+      "step": 24636
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6626452869434944,
+      "learning_rate": 7.896856381223321e-06,
+      "loss": 1.0034,
+      "step": 24637
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5587634273075985,
+      "learning_rate": 7.896110412606569e-06,
+      "loss": 0.8631,
+      "step": 24638
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5859593233711673,
+      "learning_rate": 7.895364456239324e-06,
+      "loss": 0.7775,
+      "step": 24639
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7370595161457716,
+      "learning_rate": 7.894618512125932e-06,
+      "loss": 0.8608,
+      "step": 24640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6885871199463085,
+      "learning_rate": 7.893872580270741e-06,
+      "loss": 0.9644,
+      "step": 24641
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7788350818839657,
+      "learning_rate": 7.893126660678091e-06,
+      "loss": 0.738,
+      "step": 24642
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.623418376476411,
+      "learning_rate": 7.892380753352324e-06,
+      "loss": 0.8194,
+      "step": 24643
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5776002613165745,
+      "learning_rate": 7.89163485829778e-06,
+      "loss": 0.8136,
+      "step": 24644
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6609302111230626,
+      "learning_rate": 7.89088897551881e-06,
+      "loss": 0.763,
+      "step": 24645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6643074186940756,
+      "learning_rate": 7.89014310501975e-06,
+      "loss": 0.7854,
+      "step": 24646
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6377820424788236,
+      "learning_rate": 7.889397246804943e-06,
+      "loss": 0.8071,
+      "step": 24647
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8275670311116275,
+      "learning_rate": 7.888651400878736e-06,
+      "loss": 0.7021,
+      "step": 24648
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.573655435939081,
+      "learning_rate": 7.88790556724547e-06,
+      "loss": 0.8699,
+      "step": 24649
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7967287077158028,
+      "learning_rate": 7.887159745909484e-06,
+      "loss": 0.7512,
+      "step": 24650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8156097466392633,
+      "learning_rate": 7.886413936875122e-06,
+      "loss": 0.869,
+      "step": 24651
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6007354844588266,
+      "learning_rate": 7.88566814014673e-06,
+      "loss": 0.8973,
+      "step": 24652
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5053525871577063,
+      "learning_rate": 7.884922355728649e-06,
+      "loss": 0.8164,
+      "step": 24653
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1444069682715887,
+      "learning_rate": 7.884176583625215e-06,
+      "loss": 0.9532,
+      "step": 24654
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6869139475997421,
+      "learning_rate": 7.883430823840775e-06,
+      "loss": 0.8283,
+      "step": 24655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6305681222331572,
+      "learning_rate": 7.882685076379675e-06,
+      "loss": 0.7916,
+      "step": 24656
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8462774015953451,
+      "learning_rate": 7.88193934124625e-06,
+      "loss": 0.8783,
+      "step": 24657
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5345154705881712,
+      "learning_rate": 7.881193618444846e-06,
+      "loss": 0.8063,
+      "step": 24658
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.087087874817856,
+      "learning_rate": 7.8804479079798e-06,
+      "loss": 0.9845,
+      "step": 24659
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6107428267259,
+      "learning_rate": 7.879702209855463e-06,
+      "loss": 0.9719,
+      "step": 24660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6695576951428122,
+      "learning_rate": 7.878956524076167e-06,
+      "loss": 0.7969,
+      "step": 24661
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5186786892681412,
+      "learning_rate": 7.87821085064626e-06,
+      "loss": 0.7598,
+      "step": 24662
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7403554791793447,
+      "learning_rate": 7.877465189570081e-06,
+      "loss": 0.8264,
+      "step": 24663
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5622967239489407,
+      "learning_rate": 7.876719540851969e-06,
+      "loss": 0.8982,
+      "step": 24664
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6198658347957442,
+      "learning_rate": 7.87597390449627e-06,
+      "loss": 0.849,
+      "step": 24665
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7279043998028842,
+      "learning_rate": 7.875228280507322e-06,
+      "loss": 0.881,
+      "step": 24666
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7817192796465391,
+      "learning_rate": 7.874482668889472e-06,
+      "loss": 0.7475,
+      "step": 24667
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6237977382018218,
+      "learning_rate": 7.87373706964705e-06,
+      "loss": 0.9068,
+      "step": 24668
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5460051633123806,
+      "learning_rate": 7.872991482784408e-06,
+      "loss": 0.7909,
+      "step": 24669
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7528170502125433,
+      "learning_rate": 7.872245908305885e-06,
+      "loss": 0.9176,
+      "step": 24670
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.691656321625147,
+      "learning_rate": 7.871500346215817e-06,
+      "loss": 0.913,
+      "step": 24671
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6759332978160393,
+      "learning_rate": 7.870754796518547e-06,
+      "loss": 0.9163,
+      "step": 24672
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.147703238151797,
+      "learning_rate": 7.870009259218417e-06,
+      "loss": 0.8774,
+      "step": 24673
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6089919697295079,
+      "learning_rate": 7.86926373431977e-06,
+      "loss": 0.7683,
+      "step": 24674
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5521326904864934,
+      "learning_rate": 7.868518221826943e-06,
+      "loss": 0.8136,
+      "step": 24675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.750546794941546,
+      "learning_rate": 7.867772721744275e-06,
+      "loss": 0.912,
+      "step": 24676
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4784532384075302,
+      "learning_rate": 7.867027234076115e-06,
+      "loss": 0.8773,
+      "step": 24677
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.544996937728891,
+      "learning_rate": 7.866281758826793e-06,
+      "loss": 0.9967,
+      "step": 24678
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.652826143497985,
+      "learning_rate": 7.865536296000655e-06,
+      "loss": 0.8975,
+      "step": 24679
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8131516874923301,
+      "learning_rate": 7.86479084560204e-06,
+      "loss": 0.9082,
+      "step": 24680
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6344357367623534,
+      "learning_rate": 7.86404540763529e-06,
+      "loss": 0.7298,
+      "step": 24681
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.925830494126416,
+      "learning_rate": 7.863299982104745e-06,
+      "loss": 0.7665,
+      "step": 24682
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7267730230590086,
+      "learning_rate": 7.862554569014739e-06,
+      "loss": 0.6279,
+      "step": 24683
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5613833605256096,
+      "learning_rate": 7.861809168369626e-06,
+      "loss": 0.9491,
+      "step": 24684
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0378157762072955,
+      "learning_rate": 7.861063780173728e-06,
+      "loss": 0.872,
+      "step": 24685
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4667760514020607,
+      "learning_rate": 7.860318404431397e-06,
+      "loss": 0.9459,
+      "step": 24686
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7258947808611567,
+      "learning_rate": 7.859573041146968e-06,
+      "loss": 0.7733,
+      "step": 24687
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.610003427851894,
+      "learning_rate": 7.858827690324788e-06,
+      "loss": 1.0159,
+      "step": 24688
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7495504293944908,
+      "learning_rate": 7.858082351969185e-06,
+      "loss": 0.7729,
+      "step": 24689
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5817572548847532,
+      "learning_rate": 7.857337026084508e-06,
+      "loss": 0.7778,
+      "step": 24690
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203806927096585,
+      "learning_rate": 7.856591712675093e-06,
+      "loss": 0.8901,
+      "step": 24691
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.759852988358218,
+      "learning_rate": 7.855846411745278e-06,
+      "loss": 0.6859,
+      "step": 24692
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5744147977132503,
+      "learning_rate": 7.855101123299403e-06,
+      "loss": 0.7575,
+      "step": 24693
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9660140051729424,
+      "learning_rate": 7.85435584734181e-06,
+      "loss": 0.7881,
+      "step": 24694
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.659256100929761,
+      "learning_rate": 7.853610583876839e-06,
+      "loss": 0.7771,
+      "step": 24695
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.807854117695289,
+      "learning_rate": 7.852865332908822e-06,
+      "loss": 0.8248,
+      "step": 24696
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.661941330085071,
+      "learning_rate": 7.852120094442105e-06,
+      "loss": 0.9268,
+      "step": 24697
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.715142837896559,
+      "learning_rate": 7.851374868481027e-06,
+      "loss": 0.9817,
+      "step": 24698
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.604233688903576,
+      "learning_rate": 7.850629655029923e-06,
+      "loss": 0.9482,
+      "step": 24699
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7826157167229468,
+      "learning_rate": 7.849884454093132e-06,
+      "loss": 0.9926,
+      "step": 24700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4237172296738483,
+      "learning_rate": 7.849139265674995e-06,
+      "loss": 0.9044,
+      "step": 24701
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8426658562832594,
+      "learning_rate": 7.848394089779854e-06,
+      "loss": 1.0358,
+      "step": 24702
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7693284350915458,
+      "learning_rate": 7.84764892641204e-06,
+      "loss": 0.7409,
+      "step": 24703
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6577278103166813,
+      "learning_rate": 7.846903775575893e-06,
+      "loss": 0.7361,
+      "step": 24704
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0771259256536534,
+      "learning_rate": 7.846158637275761e-06,
+      "loss": 0.8587,
+      "step": 24705
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.897252912804103,
+      "learning_rate": 7.84541351151597e-06,
+      "loss": 0.8718,
+      "step": 24706
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.563633150077389,
+      "learning_rate": 7.844668398300866e-06,
+      "loss": 0.9221,
+      "step": 24707
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7091231793165491,
+      "learning_rate": 7.843923297634781e-06,
+      "loss": 0.8839,
+      "step": 24708
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7978303955227883,
+      "learning_rate": 7.843178209522063e-06,
+      "loss": 0.8305,
+      "step": 24709
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7502512174842647,
+      "learning_rate": 7.842433133967041e-06,
+      "loss": 0.7756,
+      "step": 24710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.24611712722171,
+      "learning_rate": 7.841688070974056e-06,
+      "loss": 0.8873,
+      "step": 24711
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5336024438608693,
+      "learning_rate": 7.84094302054745e-06,
+      "loss": 0.8025,
+      "step": 24712
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.577327193048389,
+      "learning_rate": 7.840197982691552e-06,
+      "loss": 0.8636,
+      "step": 24713
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.667502549864424,
+      "learning_rate": 7.839452957410707e-06,
+      "loss": 0.8009,
+      "step": 24714
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7072604427920204,
+      "learning_rate": 7.838707944709252e-06,
+      "loss": 0.8998,
+      "step": 24715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9432861024075936,
+      "learning_rate": 7.837962944591525e-06,
+      "loss": 0.7994,
+      "step": 24716
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4181600680646014,
+      "learning_rate": 7.837217957061858e-06,
+      "loss": 0.8551,
+      "step": 24717
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6321766799243251,
+      "learning_rate": 7.836472982124596e-06,
+      "loss": 0.9893,
+      "step": 24718
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6616454170487258,
+      "learning_rate": 7.835728019784074e-06,
+      "loss": 0.9834,
+      "step": 24719
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6493214110061016,
+      "learning_rate": 7.834983070044626e-06,
+      "loss": 0.7683,
+      "step": 24720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7782210781449346,
+      "learning_rate": 7.83423813291059e-06,
+      "loss": 0.7476,
+      "step": 24721
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7754219582098762,
+      "learning_rate": 7.833493208386308e-06,
+      "loss": 0.8673,
+      "step": 24722
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7580393549849549,
+      "learning_rate": 7.832748296476117e-06,
+      "loss": 0.7282,
+      "step": 24723
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.66050088587018,
+      "learning_rate": 7.832003397184351e-06,
+      "loss": 0.84,
+      "step": 24724
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4131353934134951,
+      "learning_rate": 7.831258510515342e-06,
+      "loss": 0.7992,
+      "step": 24725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5119638603719519,
+      "learning_rate": 7.83051363647344e-06,
+      "loss": 0.7783,
+      "step": 24726
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5794556446495691,
+      "learning_rate": 7.82976877506297e-06,
+      "loss": 0.8179,
+      "step": 24727
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.65979565799517,
+      "learning_rate": 7.829023926288275e-06,
+      "loss": 0.8425,
+      "step": 24728
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6492309674958676,
+      "learning_rate": 7.828279090153686e-06,
+      "loss": 0.8694,
+      "step": 24729
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5654213219676125,
+      "learning_rate": 7.827534266663548e-06,
+      "loss": 0.8687,
+      "step": 24730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5502926832108723,
+      "learning_rate": 7.826789455822194e-06,
+      "loss": 0.8565,
+      "step": 24731
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6118021603519255,
+      "learning_rate": 7.826044657633956e-06,
+      "loss": 0.9423,
+      "step": 24732
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5157291511216215,
+      "learning_rate": 7.825299872103181e-06,
+      "loss": 0.853,
+      "step": 24733
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6357645392314166,
+      "learning_rate": 7.82455509923419e-06,
+      "loss": 0.9064,
+      "step": 24734
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9859976921930418,
+      "learning_rate": 7.823810339031332e-06,
+      "loss": 0.8972,
+      "step": 24735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4926298035622785,
+      "learning_rate": 7.823065591498937e-06,
+      "loss": 0.7319,
+      "step": 24736
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.838520664566594,
+      "learning_rate": 7.82232085664135e-06,
+      "loss": 0.7959,
+      "step": 24737
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5801907885342559,
+      "learning_rate": 7.821576134462893e-06,
+      "loss": 0.8304,
+      "step": 24738
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6999989097057906,
+      "learning_rate": 7.820831424967912e-06,
+      "loss": 0.7698,
+      "step": 24739
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6519158470635946,
+      "learning_rate": 7.820086728160743e-06,
+      "loss": 0.8677,
+      "step": 24740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7738310010338529,
+      "learning_rate": 7.819342044045716e-06,
+      "loss": 0.687,
+      "step": 24741
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8277635844053356,
+      "learning_rate": 7.81859737262717e-06,
+      "loss": 0.7433,
+      "step": 24742
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9500022791694807,
+      "learning_rate": 7.817852713909442e-06,
+      "loss": 0.9088,
+      "step": 24743
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7677950993752126,
+      "learning_rate": 7.817108067896865e-06,
+      "loss": 0.6847,
+      "step": 24744
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5392687906713696,
+      "learning_rate": 7.816363434593776e-06,
+      "loss": 0.9384,
+      "step": 24745
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7327105989393699,
+      "learning_rate": 7.815618814004511e-06,
+      "loss": 1.0465,
+      "step": 24746
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.695276644268223,
+      "learning_rate": 7.814874206133407e-06,
+      "loss": 0.9348,
+      "step": 24747
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.510675749245645,
+      "learning_rate": 7.814129610984795e-06,
+      "loss": 0.9842,
+      "step": 24748
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.726740151977503,
+      "learning_rate": 7.81338502856301e-06,
+      "loss": 0.8541,
+      "step": 24749
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.540711907006517,
+      "learning_rate": 7.812640458872393e-06,
+      "loss": 0.9714,
+      "step": 24750
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5885932183678693,
+      "learning_rate": 7.811895901917277e-06,
+      "loss": 0.75,
+      "step": 24751
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.590767103469473,
+      "learning_rate": 7.811151357701993e-06,
+      "loss": 0.9628,
+      "step": 24752
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5648228526029404,
+      "learning_rate": 7.810406826230877e-06,
+      "loss": 0.9251,
+      "step": 24753
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3610537411875026,
+      "learning_rate": 7.809662307508271e-06,
+      "loss": 0.8689,
+      "step": 24754
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5950898569444871,
+      "learning_rate": 7.808917801538502e-06,
+      "loss": 0.818,
+      "step": 24755
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4768234667755165,
+      "learning_rate": 7.808173308325905e-06,
+      "loss": 0.8141,
+      "step": 24756
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7380475506641042,
+      "learning_rate": 7.807428827874818e-06,
+      "loss": 0.9381,
+      "step": 24757
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6743666647092625,
+      "learning_rate": 7.806684360189579e-06,
+      "loss": 0.8658,
+      "step": 24758
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.744345614218831,
+      "learning_rate": 7.805939905274514e-06,
+      "loss": 0.9574,
+      "step": 24759
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5344138186658627,
+      "learning_rate": 7.80519546313396e-06,
+      "loss": 0.9623,
+      "step": 24760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6732724307215419,
+      "learning_rate": 7.804451033772258e-06,
+      "loss": 0.7522,
+      "step": 24761
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.752094493407002,
+      "learning_rate": 7.803706617193731e-06,
+      "loss": 0.7547,
+      "step": 24762
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7726862158213232,
+      "learning_rate": 7.802962213402723e-06,
+      "loss": 0.7575,
+      "step": 24763
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.468357510002145,
+      "learning_rate": 7.802217822403563e-06,
+      "loss": 0.965,
+      "step": 24764
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4323558892902724,
+      "learning_rate": 7.80147344420059e-06,
+      "loss": 0.9048,
+      "step": 24765
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.722465539991634,
+      "learning_rate": 7.80072907879813e-06,
+      "loss": 0.8798,
+      "step": 24766
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6648274956214413,
+      "learning_rate": 7.799984726200524e-06,
+      "loss": 1.0145,
+      "step": 24767
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.46277969706762,
+      "learning_rate": 7.799240386412106e-06,
+      "loss": 0.7837,
+      "step": 24768
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5933779572672622,
+      "learning_rate": 7.798496059437204e-06,
+      "loss": 0.8311,
+      "step": 24769
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7938241418016583,
+      "learning_rate": 7.797751745280153e-06,
+      "loss": 0.8362,
+      "step": 24770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5691032803582514,
+      "learning_rate": 7.797007443945291e-06,
+      "loss": 0.8416,
+      "step": 24771
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7991729915651307,
+      "learning_rate": 7.796263155436952e-06,
+      "loss": 0.8471,
+      "step": 24772
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4736805116781138,
+      "learning_rate": 7.795518879759462e-06,
+      "loss": 0.9042,
+      "step": 24773
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4657880530073304,
+      "learning_rate": 7.79477461691716e-06,
+      "loss": 0.9158,
+      "step": 24774
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6056356585481748,
+      "learning_rate": 7.794030366914379e-06,
+      "loss": 0.7793,
+      "step": 24775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6948513890974555,
+      "learning_rate": 7.793286129755453e-06,
+      "loss": 0.9044,
+      "step": 24776
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7544399802231754,
+      "learning_rate": 7.792541905444713e-06,
+      "loss": 1.0003,
+      "step": 24777
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7123569891814525,
+      "learning_rate": 7.79179769398649e-06,
+      "loss": 0.8945,
+      "step": 24778
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.934174965780157,
+      "learning_rate": 7.791053495385124e-06,
+      "loss": 0.9156,
+      "step": 24779
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.840908437549605,
+      "learning_rate": 7.790309309644942e-06,
+      "loss": 0.8027,
+      "step": 24780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6844712895162195,
+      "learning_rate": 7.789565136770278e-06,
+      "loss": 0.8846,
+      "step": 24781
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.81479317125429,
+      "learning_rate": 7.788820976765466e-06,
+      "loss": 0.8102,
+      "step": 24782
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5356246646375347,
+      "learning_rate": 7.788076829634842e-06,
+      "loss": 0.8508,
+      "step": 24783
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9056068152848618,
+      "learning_rate": 7.78733269538273e-06,
+      "loss": 0.8302,
+      "step": 24784
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7634929656790872,
+      "learning_rate": 7.786588574013467e-06,
+      "loss": 0.6976,
+      "step": 24785
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9008337397694324,
+      "learning_rate": 7.785844465531393e-06,
+      "loss": 0.7946,
+      "step": 24786
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6309424189843533,
+      "learning_rate": 7.785100369940826e-06,
+      "loss": 0.9087,
+      "step": 24787
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7216334296340958,
+      "learning_rate": 7.78435628724611e-06,
+      "loss": 0.939,
+      "step": 24788
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6816969450291717,
+      "learning_rate": 7.78361221745157e-06,
+      "loss": 0.8615,
+      "step": 24789
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7187642288059297,
+      "learning_rate": 7.782868160561548e-06,
+      "loss": 0.8946,
+      "step": 24790
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9060994280486858,
+      "learning_rate": 7.782124116580362e-06,
+      "loss": 0.7385,
+      "step": 24791
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.564593053965685,
+      "learning_rate": 7.781380085512355e-06,
+      "loss": 0.9486,
+      "step": 24792
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7591557179637642,
+      "learning_rate": 7.780636067361857e-06,
+      "loss": 0.8265,
+      "step": 24793
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6992313384886018,
+      "learning_rate": 7.779892062133193e-06,
+      "loss": 0.821,
+      "step": 24794
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6987767192174053,
+      "learning_rate": 7.779148069830705e-06,
+      "loss": 0.7623,
+      "step": 24795
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9555956993702734,
+      "learning_rate": 7.778404090458718e-06,
+      "loss": 0.8464,
+      "step": 24796
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6013607677135797,
+      "learning_rate": 7.777660124021569e-06,
+      "loss": 0.7964,
+      "step": 24797
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2225454258085557,
+      "learning_rate": 7.77691617052358e-06,
+      "loss": 0.8983,
+      "step": 24798
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.664388743302945,
+      "learning_rate": 7.776172229969095e-06,
+      "loss": 0.926,
+      "step": 24799
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6926793797608017,
+      "learning_rate": 7.775428302362437e-06,
+      "loss": 0.81,
+      "step": 24800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4619227981577134,
+      "learning_rate": 7.77468438770794e-06,
+      "loss": 0.8242,
+      "step": 24801
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5187963990972508,
+      "learning_rate": 7.773940486009934e-06,
+      "loss": 0.8473,
+      "step": 24802
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4325073952345757,
+      "learning_rate": 7.773196597272751e-06,
+      "loss": 0.9122,
+      "step": 24803
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9318858322049812,
+      "learning_rate": 7.772452721500726e-06,
+      "loss": 0.7842,
+      "step": 24804
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4795776238260065,
+      "learning_rate": 7.771708858698184e-06,
+      "loss": 0.779,
+      "step": 24805
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.461238515176431,
+      "learning_rate": 7.770965008869455e-06,
+      "loss": 0.6707,
+      "step": 24806
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.700334331791672,
+      "learning_rate": 7.77022117201888e-06,
+      "loss": 0.896,
+      "step": 24807
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5797859922998905,
+      "learning_rate": 7.76947734815078e-06,
+      "loss": 0.8275,
+      "step": 24808
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7860586880563183,
+      "learning_rate": 7.768733537269488e-06,
+      "loss": 0.9606,
+      "step": 24809
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5690369520941885,
+      "learning_rate": 7.767989739379336e-06,
+      "loss": 0.8791,
+      "step": 24810
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.829524952834316,
+      "learning_rate": 7.767245954484657e-06,
+      "loss": 0.9432,
+      "step": 24811
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7562237033627088,
+      "learning_rate": 7.766502182589776e-06,
+      "loss": 0.965,
+      "step": 24812
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7445870426010367,
+      "learning_rate": 7.765758423699027e-06,
+      "loss": 0.6752,
+      "step": 24813
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.764101360476551,
+      "learning_rate": 7.765014677816745e-06,
+      "loss": 0.9077,
+      "step": 24814
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5509462902890048,
+      "learning_rate": 7.764270944947248e-06,
+      "loss": 0.8468,
+      "step": 24815
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.636067444877857,
+      "learning_rate": 7.763527225094876e-06,
+      "loss": 0.9368,
+      "step": 24816
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.752639094709337,
+      "learning_rate": 7.762783518263953e-06,
+      "loss": 0.8153,
+      "step": 24817
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7583726347951797,
+      "learning_rate": 7.76203982445882e-06,
+      "loss": 0.7122,
+      "step": 24818
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.757786125624338,
+      "learning_rate": 7.761296143683793e-06,
+      "loss": 0.8829,
+      "step": 24819
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7929148256137526,
+      "learning_rate": 7.760552475943212e-06,
+      "loss": 0.7952,
+      "step": 24820
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5711722437596174,
+      "learning_rate": 7.759808821241406e-06,
+      "loss": 0.8076,
+      "step": 24821
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.578340473895687,
+      "learning_rate": 7.759065179582698e-06,
+      "loss": 0.7594,
+      "step": 24822
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.688273915282926,
+      "learning_rate": 7.758321550971421e-06,
+      "loss": 0.8227,
+      "step": 24823
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.644571084940423,
+      "learning_rate": 7.757577935411908e-06,
+      "loss": 0.9302,
+      "step": 24824
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7633771678354624,
+      "learning_rate": 7.75683433290849e-06,
+      "loss": 1.0136,
+      "step": 24825
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8497542414371035,
+      "learning_rate": 7.756090743465488e-06,
+      "loss": 0.6328,
+      "step": 24826
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6301438185902166,
+      "learning_rate": 7.755347167087234e-06,
+      "loss": 0.9033,
+      "step": 24827
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6048285080641935,
+      "learning_rate": 7.754603603778063e-06,
+      "loss": 0.7915,
+      "step": 24828
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.530887518110084,
+      "learning_rate": 7.753860053542302e-06,
+      "loss": 0.864,
+      "step": 24829
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7414698119163934,
+      "learning_rate": 7.753116516384274e-06,
+      "loss": 0.8017,
+      "step": 24830
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0003459605470386,
+      "learning_rate": 7.752372992308316e-06,
+      "loss": 0.7712,
+      "step": 24831
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5999748033233658,
+      "learning_rate": 7.751629481318756e-06,
+      "loss": 0.867,
+      "step": 24832
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7747577189359585,
+      "learning_rate": 7.750885983419919e-06,
+      "loss": 0.8168,
+      "step": 24833
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8698286585302584,
+      "learning_rate": 7.750142498616133e-06,
+      "loss": 0.7695,
+      "step": 24834
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7536778983195747,
+      "learning_rate": 7.749399026911736e-06,
+      "loss": 0.7677,
+      "step": 24835
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7204714428297287,
+      "learning_rate": 7.748655568311045e-06,
+      "loss": 0.7058,
+      "step": 24836
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4941271181758284,
+      "learning_rate": 7.747912122818395e-06,
+      "loss": 0.7433,
+      "step": 24837
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.741742308242689,
+      "learning_rate": 7.747168690438112e-06,
+      "loss": 0.7008,
+      "step": 24838
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8241022954306452,
+      "learning_rate": 7.746425271174533e-06,
+      "loss": 0.6888,
+      "step": 24839
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9912319829878884,
+      "learning_rate": 7.745681865031973e-06,
+      "loss": 0.8913,
+      "step": 24840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5695516958405926,
+      "learning_rate": 7.74493847201477e-06,
+      "loss": 0.9988,
+      "step": 24841
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.502121425368308,
+      "learning_rate": 7.74419509212725e-06,
+      "loss": 0.884,
+      "step": 24842
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.763015899818654,
+      "learning_rate": 7.743451725373734e-06,
+      "loss": 0.8332,
+      "step": 24843
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5703917521426431,
+      "learning_rate": 7.742708371758562e-06,
+      "loss": 0.7188,
+      "step": 24844
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5363285915816938,
+      "learning_rate": 7.741965031286055e-06,
+      "loss": 0.8784,
+      "step": 24845
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7696220706975695,
+      "learning_rate": 7.741221703960547e-06,
+      "loss": 0.6689,
+      "step": 24846
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4412849317144676,
+      "learning_rate": 7.740478389786354e-06,
+      "loss": 0.8565,
+      "step": 24847
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7959725039324541,
+      "learning_rate": 7.739735088767815e-06,
+      "loss": 0.7738,
+      "step": 24848
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8032793670710651,
+      "learning_rate": 7.738991800909254e-06,
+      "loss": 0.7127,
+      "step": 24849
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5786629408908244,
+      "learning_rate": 7.738248526214998e-06,
+      "loss": 0.823,
+      "step": 24850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.762098369524249,
+      "learning_rate": 7.737505264689374e-06,
+      "loss": 0.9018,
+      "step": 24851
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.561195374485899,
+      "learning_rate": 7.73676201633671e-06,
+      "loss": 0.8895,
+      "step": 24852
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.503450391468733,
+      "learning_rate": 7.736018781161339e-06,
+      "loss": 0.805,
+      "step": 24853
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8459191397439014,
+      "learning_rate": 7.73527555916758e-06,
+      "loss": 0.7942,
+      "step": 24854
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8066004392294089,
+      "learning_rate": 7.734532350359762e-06,
+      "loss": 0.6957,
+      "step": 24855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6589618581828454,
+      "learning_rate": 7.733789154742217e-06,
+      "loss": 0.8925,
+      "step": 24856
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6299400059565996,
+      "learning_rate": 7.733045972319268e-06,
+      "loss": 0.8562,
+      "step": 24857
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.609541885618568,
+      "learning_rate": 7.732302803095242e-06,
+      "loss": 0.8477,
+      "step": 24858
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.430299486266131,
+      "learning_rate": 7.731559647074466e-06,
+      "loss": 0.832,
+      "step": 24859
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.691360365691107,
+      "learning_rate": 7.730816504261271e-06,
+      "loss": 0.8902,
+      "step": 24860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5040971303923931,
+      "learning_rate": 7.73007337465998e-06,
+      "loss": 0.8576,
+      "step": 24861
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6749315228758295,
+      "learning_rate": 7.729330258274919e-06,
+      "loss": 0.7819,
+      "step": 24862
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7532390953585236,
+      "learning_rate": 7.72858715511042e-06,
+      "loss": 0.7395,
+      "step": 24863
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6758636960794029,
+      "learning_rate": 7.7278440651708e-06,
+      "loss": 0.8376,
+      "step": 24864
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.749873085816927,
+      "learning_rate": 7.727100988460394e-06,
+      "loss": 0.9157,
+      "step": 24865
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4505010593621053,
+      "learning_rate": 7.726357924983524e-06,
+      "loss": 0.8767,
+      "step": 24866
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7750701793524585,
+      "learning_rate": 7.725614874744524e-06,
+      "loss": 0.7185,
+      "step": 24867
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4980067427600177,
+      "learning_rate": 7.724871837747708e-06,
+      "loss": 0.8062,
+      "step": 24868
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.536326699402633,
+      "learning_rate": 7.72412881399741e-06,
+      "loss": 0.761,
+      "step": 24869
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4661027869962717,
+      "learning_rate": 7.723385803497956e-06,
+      "loss": 0.8733,
+      "step": 24870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7371117232806819,
+      "learning_rate": 7.72264280625367e-06,
+      "loss": 0.8366,
+      "step": 24871
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5089165111757896,
+      "learning_rate": 7.721899822268875e-06,
+      "loss": 0.859,
+      "step": 24872
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9151198542382777,
+      "learning_rate": 7.721156851547903e-06,
+      "loss": 0.8258,
+      "step": 24873
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8120373668364805,
+      "learning_rate": 7.72041389409508e-06,
+      "loss": 0.7482,
+      "step": 24874
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5351110539990498,
+      "learning_rate": 7.719670949914729e-06,
+      "loss": 0.8732,
+      "step": 24875
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8483549250876627,
+      "learning_rate": 7.71892801901117e-06,
+      "loss": 0.7995,
+      "step": 24876
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7159800365747166,
+      "learning_rate": 7.71818510138874e-06,
+      "loss": 1.0061,
+      "step": 24877
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5438827057691131,
+      "learning_rate": 7.717442197051756e-06,
+      "loss": 0.9104,
+      "step": 24878
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4486333739897888,
+      "learning_rate": 7.716699306004544e-06,
+      "loss": 0.8928,
+      "step": 24879
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7283043806728446,
+      "learning_rate": 7.715956428251433e-06,
+      "loss": 0.8868,
+      "step": 24880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.773957539927401,
+      "learning_rate": 7.715213563796752e-06,
+      "loss": 0.7077,
+      "step": 24881
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5281170992719235,
+      "learning_rate": 7.714470712644817e-06,
+      "loss": 0.7765,
+      "step": 24882
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7206899048115254,
+      "learning_rate": 7.713727874799954e-06,
+      "loss": 0.7788,
+      "step": 24883
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5764767347236028,
+      "learning_rate": 7.712985050266499e-06,
+      "loss": 0.9076,
+      "step": 24884
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6497310722648104,
+      "learning_rate": 7.712242239048761e-06,
+      "loss": 0.7324,
+      "step": 24885
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078450895038759,
+      "learning_rate": 7.711499441151077e-06,
+      "loss": 0.865,
+      "step": 24886
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4957863438986911,
+      "learning_rate": 7.710756656577764e-06,
+      "loss": 0.8461,
+      "step": 24887
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8128175151246675,
+      "learning_rate": 7.710013885333158e-06,
+      "loss": 0.7241,
+      "step": 24888
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5986845300711883,
+      "learning_rate": 7.70927112742157e-06,
+      "loss": 0.8228,
+      "step": 24889
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5819443506640105,
+      "learning_rate": 7.708528382847333e-06,
+      "loss": 0.8351,
+      "step": 24890
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9801540661720451,
+      "learning_rate": 7.70778565161477e-06,
+      "loss": 1.0024,
+      "step": 24891
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8353979086611132,
+      "learning_rate": 7.707042933728202e-06,
+      "loss": 0.9802,
+      "step": 24892
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4622359461561674,
+      "learning_rate": 7.706300229191956e-06,
+      "loss": 0.8714,
+      "step": 24893
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5094741307834332,
+      "learning_rate": 7.705557538010358e-06,
+      "loss": 0.8714,
+      "step": 24894
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7182835960052383,
+      "learning_rate": 7.704814860187731e-06,
+      "loss": 0.8711,
+      "step": 24895
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5657224825113545,
+      "learning_rate": 7.704072195728396e-06,
+      "loss": 0.7578,
+      "step": 24896
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6080313379901647,
+      "learning_rate": 7.70332954463668e-06,
+      "loss": 0.9078,
+      "step": 24897
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5183295218505282,
+      "learning_rate": 7.702586906916911e-06,
+      "loss": 0.7717,
+      "step": 24898
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5466493116586244,
+      "learning_rate": 7.701844282573405e-06,
+      "loss": 0.8954,
+      "step": 24899
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7649552696535014,
+      "learning_rate": 7.701101671610485e-06,
+      "loss": 0.8453,
+      "step": 24900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4917819123517877,
+      "learning_rate": 7.700359074032483e-06,
+      "loss": 0.7872,
+      "step": 24901
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4934276760508365,
+      "learning_rate": 7.699616489843723e-06,
+      "loss": 0.8291,
+      "step": 24902
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0788493768396945,
+      "learning_rate": 7.698873919048518e-06,
+      "loss": 0.8088,
+      "step": 24903
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5218735294841637,
+      "learning_rate": 7.698131361651196e-06,
+      "loss": 0.7541,
+      "step": 24904
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5386345964840669,
+      "learning_rate": 7.697388817656087e-06,
+      "loss": 0.8408,
+      "step": 24905
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6669916680281556,
+      "learning_rate": 7.696646287067509e-06,
+      "loss": 0.8367,
+      "step": 24906
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7563475250601222,
+      "learning_rate": 7.695903769889783e-06,
+      "loss": 0.8862,
+      "step": 24907
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4943849088280492,
+      "learning_rate": 7.695161266127234e-06,
+      "loss": 0.8146,
+      "step": 24908
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7074423609102978,
+      "learning_rate": 7.69441877578419e-06,
+      "loss": 0.6964,
+      "step": 24909
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6209043505809633,
+      "learning_rate": 7.693676298864967e-06,
+      "loss": 0.882,
+      "step": 24910
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8009119334196242,
+      "learning_rate": 7.69293383537389e-06,
+      "loss": 0.9581,
+      "step": 24911
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6731106047242932,
+      "learning_rate": 7.692191385315282e-06,
+      "loss": 0.8275,
+      "step": 24912
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6251937228194933,
+      "learning_rate": 7.691448948693469e-06,
+      "loss": 0.829,
+      "step": 24913
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5161678757221118,
+      "learning_rate": 7.690706525512771e-06,
+      "loss": 0.7299,
+      "step": 24914
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.4986826140229,
+      "learning_rate": 7.689964115777507e-06,
+      "loss": 0.8593,
+      "step": 24915
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7285259529050376,
+      "learning_rate": 7.68922171949201e-06,
+      "loss": 0.8407,
+      "step": 24916
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6068524015697325,
+      "learning_rate": 7.68847933666059e-06,
+      "loss": 0.9796,
+      "step": 24917
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4553656500003775,
+      "learning_rate": 7.687736967287578e-06,
+      "loss": 0.9249,
+      "step": 24918
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3311669312358851,
+      "learning_rate": 7.68699461137729e-06,
+      "loss": 0.8484,
+      "step": 24919
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5515159204738358,
+      "learning_rate": 7.686252268934058e-06,
+      "loss": 0.711,
+      "step": 24920
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9131347964981398,
+      "learning_rate": 7.685509939962191e-06,
+      "loss": 0.8708,
+      "step": 24921
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.491199156890132,
+      "learning_rate": 7.68476762446602e-06,
+      "loss": 0.8253,
+      "step": 24922
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.422478523722931,
+      "learning_rate": 7.684025322449869e-06,
+      "loss": 0.9816,
+      "step": 24923
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5782506988661298,
+      "learning_rate": 7.683283033918052e-06,
+      "loss": 0.8144,
+      "step": 24924
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6843862643631295,
+      "learning_rate": 7.682540758874894e-06,
+      "loss": 0.9359,
+      "step": 24925
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6438692796063754,
+      "learning_rate": 7.681798497324717e-06,
+      "loss": 0.7871,
+      "step": 24926
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5490411334633183,
+      "learning_rate": 7.681056249271848e-06,
+      "loss": 0.8036,
+      "step": 24927
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8002412483115324,
+      "learning_rate": 7.680314014720597e-06,
+      "loss": 0.7946,
+      "step": 24928
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4456986144134834,
+      "learning_rate": 7.679571793675295e-06,
+      "loss": 0.9528,
+      "step": 24929
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6526457605001608,
+      "learning_rate": 7.678829586140263e-06,
+      "loss": 0.8572,
+      "step": 24930
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.672293616013316,
+      "learning_rate": 7.678087392119818e-06,
+      "loss": 0.8135,
+      "step": 24931
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6980702160506955,
+      "learning_rate": 7.677345211618281e-06,
+      "loss": 0.7902,
+      "step": 24932
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4620079693291663,
+      "learning_rate": 7.676603044639978e-06,
+      "loss": 0.8613,
+      "step": 24933
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.531214557482027,
+      "learning_rate": 7.67586089118923e-06,
+      "loss": 0.9285,
+      "step": 24934
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.79063491599577,
+      "learning_rate": 7.675118751270355e-06,
+      "loss": 0.6738,
+      "step": 24935
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8518306201807865,
+      "learning_rate": 7.674376624887669e-06,
+      "loss": 0.7711,
+      "step": 24936
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6847271174725384,
+      "learning_rate": 7.673634512045506e-06,
+      "loss": 0.9065,
+      "step": 24937
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4988639672358335,
+      "learning_rate": 7.672892412748175e-06,
+      "loss": 0.911,
+      "step": 24938
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0275616631915865,
+      "learning_rate": 7.672150327000002e-06,
+      "loss": 1.002,
+      "step": 24939
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.914274833586084,
+      "learning_rate": 7.671408254805304e-06,
+      "loss": 0.7889,
+      "step": 24940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9880415339627382,
+      "learning_rate": 7.67066619616841e-06,
+      "loss": 0.8485,
+      "step": 24941
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.578596880300598,
+      "learning_rate": 7.669924151093632e-06,
+      "loss": 0.777,
+      "step": 24942
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5209314026018053,
+      "learning_rate": 7.669182119585292e-06,
+      "loss": 0.7327,
+      "step": 24943
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9208307608850557,
+      "learning_rate": 7.668440101647717e-06,
+      "loss": 0.8731,
+      "step": 24944
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4748060973895354,
+      "learning_rate": 7.667698097285214e-06,
+      "loss": 0.887,
+      "step": 24945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7364694316884595,
+      "learning_rate": 7.666956106502115e-06,
+      "loss": 0.9513,
+      "step": 24946
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.613517045484178,
+      "learning_rate": 7.666214129302737e-06,
+      "loss": 0.7978,
+      "step": 24947
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6793063263539534,
+      "learning_rate": 7.6654721656914e-06,
+      "loss": 0.946,
+      "step": 24948
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5711805129504914,
+      "learning_rate": 7.664730215672418e-06,
+      "loss": 0.927,
+      "step": 24949
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.873614623665094,
+      "learning_rate": 7.66398827925012e-06,
+      "loss": 0.8317,
+      "step": 24950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.108289283261523,
+      "learning_rate": 7.663246356428824e-06,
+      "loss": 0.8699,
+      "step": 24951
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6438543041433962,
+      "learning_rate": 7.662504447212845e-06,
+      "loss": 0.8608,
+      "step": 24952
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0014789731418516,
+      "learning_rate": 7.661762551606501e-06,
+      "loss": 0.8847,
+      "step": 24953
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1423954813844155,
+      "learning_rate": 7.66102066961412e-06,
+      "loss": 0.9397,
+      "step": 24954
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7594573054708071,
+      "learning_rate": 7.66027880124002e-06,
+      "loss": 0.6831,
+      "step": 24955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.147262048488461,
+      "learning_rate": 7.659536946488514e-06,
+      "loss": 0.8764,
+      "step": 24956
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.740206319834211,
+      "learning_rate": 7.658795105363922e-06,
+      "loss": 0.8885,
+      "step": 24957
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6776208924531224,
+      "learning_rate": 7.658053277870571e-06,
+      "loss": 0.8783,
+      "step": 24958
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.145752267363323,
+      "learning_rate": 7.657311464012774e-06,
+      "loss": 0.8128,
+      "step": 24959
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5603782080898676,
+      "learning_rate": 7.656569663794851e-06,
+      "loss": 0.8818,
+      "step": 24960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5509357593308524,
+      "learning_rate": 7.65582787722112e-06,
+      "loss": 0.9257,
+      "step": 24961
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6063427518800117,
+      "learning_rate": 7.655086104295904e-06,
+      "loss": 0.8435,
+      "step": 24962
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.54041459026855,
+      "learning_rate": 7.654344345023517e-06,
+      "loss": 0.8449,
+      "step": 24963
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5998451108113583,
+      "learning_rate": 7.653602599408279e-06,
+      "loss": 0.9161,
+      "step": 24964
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5446104664450129,
+      "learning_rate": 7.652860867454515e-06,
+      "loss": 0.7919,
+      "step": 24965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5480633326611168,
+      "learning_rate": 7.652119149166533e-06,
+      "loss": 0.8787,
+      "step": 24966
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.610484276857595,
+      "learning_rate": 7.651377444548657e-06,
+      "loss": 0.839,
+      "step": 24967
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8199187832412265,
+      "learning_rate": 7.650635753605203e-06,
+      "loss": 0.7896,
+      "step": 24968
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.023130249594112,
+      "learning_rate": 7.649894076340499e-06,
+      "loss": 0.8734,
+      "step": 24969
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.542041650106225,
+      "learning_rate": 7.649152412758848e-06,
+      "loss": 1.0078,
+      "step": 24970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4939850997194626,
+      "learning_rate": 7.64841076286458e-06,
+      "loss": 0.8718,
+      "step": 24971
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4700710332756965,
+      "learning_rate": 7.64766912666201e-06,
+      "loss": 0.9478,
+      "step": 24972
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.545626172396737,
+      "learning_rate": 7.646927504155452e-06,
+      "loss": 0.8729,
+      "step": 24973
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.469658383106088,
+      "learning_rate": 7.646185895349227e-06,
+      "loss": 0.8927,
+      "step": 24974
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5703761547097466,
+      "learning_rate": 7.645444300247654e-06,
+      "loss": 0.7688,
+      "step": 24975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5771918218878394,
+      "learning_rate": 7.64470271885505e-06,
+      "loss": 0.8099,
+      "step": 24976
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8111201245539307,
+      "learning_rate": 7.64396115117573e-06,
+      "loss": 0.7056,
+      "step": 24977
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5676812724837892,
+      "learning_rate": 7.643219597214017e-06,
+      "loss": 0.918,
+      "step": 24978
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5773408552280666,
+      "learning_rate": 7.642478056974228e-06,
+      "loss": 0.9712,
+      "step": 24979
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.40137841916793,
+      "learning_rate": 7.641736530460673e-06,
+      "loss": 0.8851,
+      "step": 24980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5978498993656158,
+      "learning_rate": 7.640995017677674e-06,
+      "loss": 0.7755,
+      "step": 24981
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6145508454808264,
+      "learning_rate": 7.640253518629553e-06,
+      "loss": 0.8366,
+      "step": 24982
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5891933869413886,
+      "learning_rate": 7.639512033320623e-06,
+      "loss": 0.8668,
+      "step": 24983
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7518162564669953,
+      "learning_rate": 7.6387705617552e-06,
+      "loss": 0.6834,
+      "step": 24984
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8323641123852326,
+      "learning_rate": 7.6380291039376e-06,
+      "loss": 0.9315,
+      "step": 24985
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6783855797278269,
+      "learning_rate": 7.637287659872147e-06,
+      "loss": 0.8512,
+      "step": 24986
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6090955620776324,
+      "learning_rate": 7.63654622956315e-06,
+      "loss": 0.8822,
+      "step": 24987
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7319624724889613,
+      "learning_rate": 7.63580481301493e-06,
+      "loss": 0.8687,
+      "step": 24988
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6056405519395085,
+      "learning_rate": 7.6350634102318e-06,
+      "loss": 0.95,
+      "step": 24989
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.700214339284849,
+      "learning_rate": 7.634322021218085e-06,
+      "loss": 0.7532,
+      "step": 24990
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7226791052223924,
+      "learning_rate": 7.633580645978093e-06,
+      "loss": 0.9625,
+      "step": 24991
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7571122309285647,
+      "learning_rate": 7.632839284516144e-06,
+      "loss": 0.7343,
+      "step": 24992
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8584626129290736,
+      "learning_rate": 7.632097936836555e-06,
+      "loss": 0.8126,
+      "step": 24993
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4705197922058018,
+      "learning_rate": 7.63135660294364e-06,
+      "loss": 0.7804,
+      "step": 24994
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6481560734447356,
+      "learning_rate": 7.630615282841717e-06,
+      "loss": 0.8707,
+      "step": 24995
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.824042977453474,
+      "learning_rate": 7.629873976535102e-06,
+      "loss": 0.7543,
+      "step": 24996
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7340961609013464,
+      "learning_rate": 7.629132684028113e-06,
+      "loss": 0.6795,
+      "step": 24997
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.541123253367524,
+      "learning_rate": 7.6283914053250605e-06,
+      "loss": 0.8629,
+      "step": 24998
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.719609114560205,
+      "learning_rate": 7.627650140430267e-06,
+      "loss": 0.8025,
+      "step": 24999
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9136992656044411,
+      "learning_rate": 7.626908889348046e-06,
+      "loss": 0.8301,
+      "step": 25000
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.545788105128429,
+      "learning_rate": 7.626167652082712e-06,
+      "loss": 0.9151,
+      "step": 25001
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6584702284603394,
+      "learning_rate": 7.625426428638579e-06,
+      "loss": 0.8431,
+      "step": 25002
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.581570471205462,
+      "learning_rate": 7.624685219019966e-06,
+      "loss": 0.8406,
+      "step": 25003
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7390956094029428,
+      "learning_rate": 7.623944023231191e-06,
+      "loss": 0.8616,
+      "step": 25004
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7443682466095856,
+      "learning_rate": 7.623202841276564e-06,
+      "loss": 0.7269,
+      "step": 25005
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9936659065561169,
+      "learning_rate": 7.6224616731604e-06,
+      "loss": 0.8169,
+      "step": 25006
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6046867137684029,
+      "learning_rate": 7.621720518887022e-06,
+      "loss": 0.8591,
+      "step": 25007
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5934933853260236,
+      "learning_rate": 7.620979378460736e-06,
+      "loss": 0.8084,
+      "step": 25008
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8684631875657027,
+      "learning_rate": 7.620238251885863e-06,
+      "loss": 0.8818,
+      "step": 25009
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7901821404955415,
+      "learning_rate": 7.6194971391667126e-06,
+      "loss": 0.7762,
+      "step": 25010
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5451816281583035,
+      "learning_rate": 7.618756040307609e-06,
+      "loss": 0.8444,
+      "step": 25011
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7701950600946178,
+      "learning_rate": 7.61801495531286e-06,
+      "loss": 0.6419,
+      "step": 25012
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3911374514138832,
+      "learning_rate": 7.617273884186779e-06,
+      "loss": 0.8488,
+      "step": 25013
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6637417948830158,
+      "learning_rate": 7.6165328269336905e-06,
+      "loss": 0.8636,
+      "step": 25014
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.854215592878207,
+      "learning_rate": 7.615791783557895e-06,
+      "loss": 0.7628,
+      "step": 25015
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8517641141518402,
+      "learning_rate": 7.615050754063717e-06,
+      "loss": 0.8359,
+      "step": 25016
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7379363094794957,
+      "learning_rate": 7.614309738455468e-06,
+      "loss": 0.7151,
+      "step": 25017
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4099769772845379,
+      "learning_rate": 7.613568736737469e-06,
+      "loss": 0.8001,
+      "step": 25018
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8148881846314726,
+      "learning_rate": 7.612827748914021e-06,
+      "loss": 0.8381,
+      "step": 25019
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4370907396984143,
+      "learning_rate": 7.612086774989449e-06,
+      "loss": 0.8487,
+      "step": 25020
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.778250539555748,
+      "learning_rate": 7.611345814968065e-06,
+      "loss": 0.6727,
+      "step": 25021
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6230167735145362,
+      "learning_rate": 7.610604868854179e-06,
+      "loss": 0.8876,
+      "step": 25022
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.251470270997788,
+      "learning_rate": 7.609863936652106e-06,
+      "loss": 0.8955,
+      "step": 25023
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.796539707216517,
+      "learning_rate": 7.6091230183661655e-06,
+      "loss": 0.9388,
+      "step": 25024
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5885088053680194,
+      "learning_rate": 7.60838211400067e-06,
+      "loss": 0.8595,
+      "step": 25025
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8303182180382043,
+      "learning_rate": 7.607641223559925e-06,
+      "loss": 0.8864,
+      "step": 25026
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.608844754289712,
+      "learning_rate": 7.606900347048254e-06,
+      "loss": 0.8157,
+      "step": 25027
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7830410689118668,
+      "learning_rate": 7.6061594844699645e-06,
+      "loss": 0.9061,
+      "step": 25028
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5917042642402595,
+      "learning_rate": 7.6054186358293765e-06,
+      "loss": 0.8784,
+      "step": 25029
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1517065498113768,
+      "learning_rate": 7.604677801130794e-06,
+      "loss": 0.9203,
+      "step": 25030
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.94673195632456,
+      "learning_rate": 7.603936980378539e-06,
+      "loss": 0.8504,
+      "step": 25031
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7447742785410565,
+      "learning_rate": 7.603196173576923e-06,
+      "loss": 0.8728,
+      "step": 25032
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5155302397663548,
+      "learning_rate": 7.602455380730256e-06,
+      "loss": 0.692,
+      "step": 25033
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.672018971536118,
+      "learning_rate": 7.601714601842849e-06,
+      "loss": 0.8263,
+      "step": 25034
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7346425993233251,
+      "learning_rate": 7.600973836919022e-06,
+      "loss": 0.7154,
+      "step": 25035
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5424736361255666,
+      "learning_rate": 7.600233085963088e-06,
+      "loss": 0.9102,
+      "step": 25036
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.132478903120827,
+      "learning_rate": 7.5994923489793535e-06,
+      "loss": 0.8232,
+      "step": 25037
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.914784143458396,
+      "learning_rate": 7.598751625972133e-06,
+      "loss": 0.7559,
+      "step": 25038
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4878870230411065,
+      "learning_rate": 7.598010916945744e-06,
+      "loss": 0.9407,
+      "step": 25039
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4936559833310488,
+      "learning_rate": 7.597270221904494e-06,
+      "loss": 0.7477,
+      "step": 25040
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7865964362798455,
+      "learning_rate": 7.596529540852699e-06,
+      "loss": 0.729,
+      "step": 25041
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9213787632023769,
+      "learning_rate": 7.595788873794666e-06,
+      "loss": 0.9305,
+      "step": 25042
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7492331870632176,
+      "learning_rate": 7.5950482207347165e-06,
+      "loss": 0.747,
+      "step": 25043
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.221233615083214,
+      "learning_rate": 7.594307581677155e-06,
+      "loss": 0.8387,
+      "step": 25044
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1243728621030677,
+      "learning_rate": 7.5935669566262974e-06,
+      "loss": 0.832,
+      "step": 25045
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7924377348764362,
+      "learning_rate": 7.592826345586456e-06,
+      "loss": 0.9468,
+      "step": 25046
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.540430928812516,
+      "learning_rate": 7.592085748561937e-06,
+      "loss": 0.9954,
+      "step": 25047
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.547060694707214,
+      "learning_rate": 7.5913451655570604e-06,
+      "loss": 0.8445,
+      "step": 25048
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2683077917414836,
+      "learning_rate": 7.590604596576134e-06,
+      "loss": 0.8479,
+      "step": 25049
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5688108402383778,
+      "learning_rate": 7.589864041623473e-06,
+      "loss": 0.7879,
+      "step": 25050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3739570694083139,
+      "learning_rate": 7.589123500703383e-06,
+      "loss": 0.8066,
+      "step": 25051
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.693595596823973,
+      "learning_rate": 7.588382973820181e-06,
+      "loss": 0.8972,
+      "step": 25052
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6776623771677908,
+      "learning_rate": 7.587642460978178e-06,
+      "loss": 0.8501,
+      "step": 25053
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8428561437692867,
+      "learning_rate": 7.586901962181683e-06,
+      "loss": 0.8303,
+      "step": 25054
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8563648243865833,
+      "learning_rate": 7.586161477435006e-06,
+      "loss": 0.9605,
+      "step": 25055
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5803728587809227,
+      "learning_rate": 7.585421006742464e-06,
+      "loss": 0.7705,
+      "step": 25056
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4916351705550608,
+      "learning_rate": 7.584680550108367e-06,
+      "loss": 0.7421,
+      "step": 25057
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5628197069652727,
+      "learning_rate": 7.5839401075370225e-06,
+      "loss": 0.8318,
+      "step": 25058
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0734536162933086,
+      "learning_rate": 7.583199679032741e-06,
+      "loss": 0.834,
+      "step": 25059
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8543748268234375,
+      "learning_rate": 7.582459264599842e-06,
+      "loss": 0.835,
+      "step": 25060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3759605869876188,
+      "learning_rate": 7.581718864242626e-06,
+      "loss": 0.761,
+      "step": 25061
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.099223622562588,
+      "learning_rate": 7.580978477965408e-06,
+      "loss": 0.8524,
+      "step": 25062
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.75985552899711,
+      "learning_rate": 7.5802381057725e-06,
+      "loss": 0.7388,
+      "step": 25063
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8031257037272146,
+      "learning_rate": 7.579497747668216e-06,
+      "loss": 0.9212,
+      "step": 25064
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6075363386211594,
+      "learning_rate": 7.578757403656859e-06,
+      "loss": 0.7717,
+      "step": 25065
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5313770785958716,
+      "learning_rate": 7.578017073742741e-06,
+      "loss": 0.8414,
+      "step": 25066
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6501102727141408,
+      "learning_rate": 7.577276757930181e-06,
+      "loss": 0.7937,
+      "step": 25067
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5130417245444305,
+      "learning_rate": 7.5765364562234756e-06,
+      "loss": 0.9444,
+      "step": 25068
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.717116593736033,
+      "learning_rate": 7.575796168626944e-06,
+      "loss": 0.7489,
+      "step": 25069
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.838605700494616,
+      "learning_rate": 7.575055895144894e-06,
+      "loss": 0.7356,
+      "step": 25070
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.588767111007216,
+      "learning_rate": 7.574315635781642e-06,
+      "loss": 0.803,
+      "step": 25071
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0802921265114587,
+      "learning_rate": 7.573575390541486e-06,
+      "loss": 0.9169,
+      "step": 25072
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9795184639465022,
+      "learning_rate": 7.572835159428744e-06,
+      "loss": 0.8231,
+      "step": 25073
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.561457811351153,
+      "learning_rate": 7.572094942447727e-06,
+      "loss": 0.9091,
+      "step": 25074
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6938946529927348,
+      "learning_rate": 7.571354739602736e-06,
+      "loss": 0.8352,
+      "step": 25075
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8380573446780695,
+      "learning_rate": 7.57061455089809e-06,
+      "loss": 0.9968,
+      "step": 25076
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.214262606498966,
+      "learning_rate": 7.569874376338095e-06,
+      "loss": 0.7885,
+      "step": 25077
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5564206743791447,
+      "learning_rate": 7.569134215927063e-06,
+      "loss": 0.985,
+      "step": 25078
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7788080369041509,
+      "learning_rate": 7.568394069669297e-06,
+      "loss": 0.7091,
+      "step": 25079
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5579492559549786,
+      "learning_rate": 7.567653937569113e-06,
+      "loss": 0.9259,
+      "step": 25080
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4766049183646515,
+      "learning_rate": 7.566913819630819e-06,
+      "loss": 0.8538,
+      "step": 25081
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7796377633787674,
+      "learning_rate": 7.566173715858722e-06,
+      "loss": 0.8275,
+      "step": 25082
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.160478123679188,
+      "learning_rate": 7.565433626257129e-06,
+      "loss": 0.8291,
+      "step": 25083
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.943614671968813,
+      "learning_rate": 7.564693550830355e-06,
+      "loss": 0.7008,
+      "step": 25084
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.819604235879912,
+      "learning_rate": 7.563953489582708e-06,
+      "loss": 0.8972,
+      "step": 25085
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4741364107413382,
+      "learning_rate": 7.5632134425184936e-06,
+      "loss": 0.7577,
+      "step": 25086
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.656064278172394,
+      "learning_rate": 7.562473409642019e-06,
+      "loss": 0.9486,
+      "step": 25087
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4600966292416366,
+      "learning_rate": 7.5617333909576006e-06,
+      "loss": 0.816,
+      "step": 25088
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5246569038290896,
+      "learning_rate": 7.56099338646954e-06,
+      "loss": 0.7075,
+      "step": 25089
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6648696986323328,
+      "learning_rate": 7.560253396182148e-06,
+      "loss": 0.828,
+      "step": 25090
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8399938558650613,
+      "learning_rate": 7.559513420099731e-06,
+      "loss": 0.8438,
+      "step": 25091
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7069129241151244,
+      "learning_rate": 7.558773458226605e-06,
+      "loss": 0.7613,
+      "step": 25092
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5804139136034212,
+      "learning_rate": 7.558033510567069e-06,
+      "loss": 0.7834,
+      "step": 25093
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8359522006341638,
+      "learning_rate": 7.557293577125435e-06,
+      "loss": 0.9686,
+      "step": 25094
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5421883416104099,
+      "learning_rate": 7.556553657906014e-06,
+      "loss": 0.8759,
+      "step": 25095
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6116171725708017,
+      "learning_rate": 7.555813752913107e-06,
+      "loss": 0.851,
+      "step": 25096
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9181560404396827,
+      "learning_rate": 7.555073862151028e-06,
+      "loss": 0.7502,
+      "step": 25097
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6298307870286484,
+      "learning_rate": 7.554333985624082e-06,
+      "loss": 0.9643,
+      "step": 25098
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.43139639544076,
+      "learning_rate": 7.553594123336583e-06,
+      "loss": 0.8253,
+      "step": 25099
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4798363035588962,
+      "learning_rate": 7.5528542752928265e-06,
+      "loss": 0.9182,
+      "step": 25100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6330773087901804,
+      "learning_rate": 7.552114441497131e-06,
+      "loss": 0.9262,
+      "step": 25101
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7652627967964409,
+      "learning_rate": 7.551374621953801e-06,
+      "loss": 0.7226,
+      "step": 25102
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2644240211902757,
+      "learning_rate": 7.550634816667142e-06,
+      "loss": 0.8243,
+      "step": 25103
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7121197122514291,
+      "learning_rate": 7.549895025641459e-06,
+      "loss": 0.8952,
+      "step": 25104
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7390613392655765,
+      "learning_rate": 7.549155248881066e-06,
+      "loss": 0.7501,
+      "step": 25105
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6663477286501398,
+      "learning_rate": 7.5484154863902705e-06,
+      "loss": 0.8181,
+      "step": 25106
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.037668921419396,
+      "learning_rate": 7.547675738173372e-06,
+      "loss": 0.8929,
+      "step": 25107
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7338975936221291,
+      "learning_rate": 7.546936004234679e-06,
+      "loss": 0.7063,
+      "step": 25108
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.592378217437298,
+      "learning_rate": 7.546196284578509e-06,
+      "loss": 0.8256,
+      "step": 25109
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.257819386929804,
+      "learning_rate": 7.545456579209156e-06,
+      "loss": 0.8059,
+      "step": 25110
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6626486259299451,
+      "learning_rate": 7.5447168881309295e-06,
+      "loss": 0.8268,
+      "step": 25111
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6477725799342657,
+      "learning_rate": 7.543977211348142e-06,
+      "loss": 0.9883,
+      "step": 25112
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6941007375481238,
+      "learning_rate": 7.543237548865099e-06,
+      "loss": 0.747,
+      "step": 25113
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.655398071711276,
+      "learning_rate": 7.542497900686102e-06,
+      "loss": 0.781,
+      "step": 25114
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6200250486020906,
+      "learning_rate": 7.5417582668154585e-06,
+      "loss": 0.7746,
+      "step": 25115
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8107070374273744,
+      "learning_rate": 7.5410186472574835e-06,
+      "loss": 0.8152,
+      "step": 25116
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.540595350997447,
+      "learning_rate": 7.540279042016469e-06,
+      "loss": 0.9003,
+      "step": 25117
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7050182760899208,
+      "learning_rate": 7.539539451096732e-06,
+      "loss": 0.7213,
+      "step": 25118
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6760271272942033,
+      "learning_rate": 7.538799874502572e-06,
+      "loss": 0.8804,
+      "step": 25119
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.981937896968665,
+      "learning_rate": 7.538060312238304e-06,
+      "loss": 0.825,
+      "step": 25120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7857715785372916,
+      "learning_rate": 7.5373207643082245e-06,
+      "loss": 0.9248,
+      "step": 25121
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.516807411996226,
+      "learning_rate": 7.536581230716643e-06,
+      "loss": 0.897,
+      "step": 25122
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6614972758919815,
+      "learning_rate": 7.535841711467868e-06,
+      "loss": 0.8587,
+      "step": 25123
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.553741520226107,
+      "learning_rate": 7.535102206566199e-06,
+      "loss": 0.8177,
+      "step": 25124
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9366685733088853,
+      "learning_rate": 7.534362716015948e-06,
+      "loss": 0.8989,
+      "step": 25125
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6469508941199669,
+      "learning_rate": 7.533623239821417e-06,
+      "loss": 0.8382,
+      "step": 25126
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.709280377247431,
+      "learning_rate": 7.532883777986915e-06,
+      "loss": 0.823,
+      "step": 25127
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5806672635188015,
+      "learning_rate": 7.53214433051674e-06,
+      "loss": 0.8387,
+      "step": 25128
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.608643317370781,
+      "learning_rate": 7.531404897415204e-06,
+      "loss": 0.874,
+      "step": 25129
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6144424477191714,
+      "learning_rate": 7.530665478686613e-06,
+      "loss": 1.0035,
+      "step": 25130
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7831230530288048,
+      "learning_rate": 7.5299260743352675e-06,
+      "loss": 0.7322,
+      "step": 25131
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5770972340645428,
+      "learning_rate": 7.529186684365471e-06,
+      "loss": 0.7812,
+      "step": 25132
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.553882584300526,
+      "learning_rate": 7.528447308781534e-06,
+      "loss": 0.7494,
+      "step": 25133
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3684297655185964,
+      "learning_rate": 7.527707947587764e-06,
+      "loss": 0.8308,
+      "step": 25134
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6576891836201504,
+      "learning_rate": 7.526968600788456e-06,
+      "loss": 0.786,
+      "step": 25135
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5502218807622485,
+      "learning_rate": 7.526229268387919e-06,
+      "loss": 0.8241,
+      "step": 25136
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6333675941064896,
+      "learning_rate": 7.525489950390461e-06,
+      "loss": 0.8598,
+      "step": 25137
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7087192906784294,
+      "learning_rate": 7.524750646800384e-06,
+      "loss": 0.888,
+      "step": 25138
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8689600430867677,
+      "learning_rate": 7.524011357621991e-06,
+      "loss": 0.8306,
+      "step": 25139
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7522390327103794,
+      "learning_rate": 7.523272082859587e-06,
+      "loss": 0.8929,
+      "step": 25140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5226156415625103,
+      "learning_rate": 7.52253282251748e-06,
+      "loss": 0.8263,
+      "step": 25141
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7909061277729312,
+      "learning_rate": 7.52179357659997e-06,
+      "loss": 0.7368,
+      "step": 25142
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.724205151451175,
+      "learning_rate": 7.521054345111361e-06,
+      "loss": 0.8296,
+      "step": 25143
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8627201175599288,
+      "learning_rate": 7.520315128055961e-06,
+      "loss": 0.712,
+      "step": 25144
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7309296326695003,
+      "learning_rate": 7.5195759254380674e-06,
+      "loss": 0.8959,
+      "step": 25145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5941402157047397,
+      "learning_rate": 7.518836737261991e-06,
+      "loss": 0.7434,
+      "step": 25146
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7494622125576118,
+      "learning_rate": 7.51809756353203e-06,
+      "loss": 0.9593,
+      "step": 25147
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5345988030084188,
+      "learning_rate": 7.517358404252497e-06,
+      "loss": 0.7858,
+      "step": 25148
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.62983834082854,
+      "learning_rate": 7.516619259427682e-06,
+      "loss": 0.8687,
+      "step": 25149
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.69654756581936,
+      "learning_rate": 7.5158801290619e-06,
+      "loss": 0.8787,
+      "step": 25150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6108348865655349,
+      "learning_rate": 7.515141013159447e-06,
+      "loss": 0.9222,
+      "step": 25151
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7463031519458883,
+      "learning_rate": 7.514401911724636e-06,
+      "loss": 0.9367,
+      "step": 25152
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6452894372602382,
+      "learning_rate": 7.513662824761758e-06,
+      "loss": 0.848,
+      "step": 25153
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.716887267759938,
+      "learning_rate": 7.512923752275124e-06,
+      "loss": 0.7397,
+      "step": 25154
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0927346830116353,
+      "learning_rate": 7.5121846942690376e-06,
+      "loss": 0.7573,
+      "step": 25155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1491457275885435,
+      "learning_rate": 7.511445650747797e-06,
+      "loss": 0.8665,
+      "step": 25156
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3811010101372982,
+      "learning_rate": 7.5107066217157065e-06,
+      "loss": 0.74,
+      "step": 25157
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8798460790061076,
+      "learning_rate": 7.509967607177072e-06,
+      "loss": 0.8753,
+      "step": 25158
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7873144905266197,
+      "learning_rate": 7.509228607136196e-06,
+      "loss": 0.7105,
+      "step": 25159
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6624651010539664,
+      "learning_rate": 7.508489621597376e-06,
+      "loss": 0.9336,
+      "step": 25160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5116098901138595,
+      "learning_rate": 7.50775065056492e-06,
+      "loss": 0.8278,
+      "step": 25161
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7519416081420505,
+      "learning_rate": 7.5070116940431315e-06,
+      "loss": 0.9741,
+      "step": 25162
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.593784533249547,
+      "learning_rate": 7.506272752036308e-06,
+      "loss": 0.8379,
+      "step": 25163
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5230548006574542,
+      "learning_rate": 7.5055338245487506e-06,
+      "loss": 0.7508,
+      "step": 25164
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4743329739432498,
+      "learning_rate": 7.504794911584767e-06,
+      "loss": 0.9777,
+      "step": 25165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5874412717610533,
+      "learning_rate": 7.504056013148661e-06,
+      "loss": 0.8538,
+      "step": 25166
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5658224127121525,
+      "learning_rate": 7.503317129244729e-06,
+      "loss": 0.8024,
+      "step": 25167
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5602070336508231,
+      "learning_rate": 7.502578259877272e-06,
+      "loss": 0.7216,
+      "step": 25168
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5083408221116383,
+      "learning_rate": 7.501839405050601e-06,
+      "loss": 0.6888,
+      "step": 25169
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6954506230045825,
+      "learning_rate": 7.501100564769007e-06,
+      "loss": 0.83,
+      "step": 25170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5379956924363583,
+      "learning_rate": 7.500361739036799e-06,
+      "loss": 0.8026,
+      "step": 25171
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.754852498341524,
+      "learning_rate": 7.499622927858272e-06,
+      "loss": 0.7606,
+      "step": 25172
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7525098113339237,
+      "learning_rate": 7.498884131237738e-06,
+      "loss": 0.8013,
+      "step": 25173
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6796054491008976,
+      "learning_rate": 7.49814534917949e-06,
+      "loss": 0.8321,
+      "step": 25174
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8973825250421803,
+      "learning_rate": 7.497406581687832e-06,
+      "loss": 0.8996,
+      "step": 25175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4428438986274434,
+      "learning_rate": 7.496667828767066e-06,
+      "loss": 0.9085,
+      "step": 25176
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5114990566900193,
+      "learning_rate": 7.4959290904214895e-06,
+      "loss": 0.792,
+      "step": 25177
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1491936237686198,
+      "learning_rate": 7.495190366655409e-06,
+      "loss": 0.778,
+      "step": 25178
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.827923409526676,
+      "learning_rate": 7.4944516574731215e-06,
+      "loss": 0.8657,
+      "step": 25179
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4967303683080286,
+      "learning_rate": 7.493712962878934e-06,
+      "loss": 0.8598,
+      "step": 25180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7459911766962598,
+      "learning_rate": 7.492974282877137e-06,
+      "loss": 0.6974,
+      "step": 25181
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5830193332527052,
+      "learning_rate": 7.492235617472041e-06,
+      "loss": 0.8337,
+      "step": 25182
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.636240110426562,
+      "learning_rate": 7.4914969666679435e-06,
+      "loss": 0.8255,
+      "step": 25183
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6305763522997,
+      "learning_rate": 7.490758330469144e-06,
+      "loss": 0.8874,
+      "step": 25184
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7395495107105392,
+      "learning_rate": 7.490019708879942e-06,
+      "loss": 0.6879,
+      "step": 25185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6072169577365414,
+      "learning_rate": 7.489281101904642e-06,
+      "loss": 0.7118,
+      "step": 25186
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.559889874397758,
+      "learning_rate": 7.488542509547544e-06,
+      "loss": 0.8509,
+      "step": 25187
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.768450982070273,
+      "learning_rate": 7.487803931812945e-06,
+      "loss": 0.8271,
+      "step": 25188
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6495563644271412,
+      "learning_rate": 7.487065368705144e-06,
+      "loss": 0.7716,
+      "step": 25189
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7212700154906215,
+      "learning_rate": 7.486326820228449e-06,
+      "loss": 0.8352,
+      "step": 25190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.964043944470164,
+      "learning_rate": 7.485588286387152e-06,
+      "loss": 0.9149,
+      "step": 25191
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6286984349942977,
+      "learning_rate": 7.484849767185556e-06,
+      "loss": 0.9315,
+      "step": 25192
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2161804423173925,
+      "learning_rate": 7.48411126262796e-06,
+      "loss": 0.9009,
+      "step": 25193
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.903352885976145,
+      "learning_rate": 7.483372772718668e-06,
+      "loss": 0.8947,
+      "step": 25194
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3331057186504665,
+      "learning_rate": 7.482634297461975e-06,
+      "loss": 0.8125,
+      "step": 25195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4722437695145936,
+      "learning_rate": 7.48189583686218e-06,
+      "loss": 0.8104,
+      "step": 25196
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5547993447854855,
+      "learning_rate": 7.48115739092359e-06,
+      "loss": 0.7797,
+      "step": 25197
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5010915635114577,
+      "learning_rate": 7.480418959650493e-06,
+      "loss": 0.823,
+      "step": 25198
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.669546409484222,
+      "learning_rate": 7.4796805430471975e-06,
+      "loss": 0.928,
+      "step": 25199
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6706039637876149,
+      "learning_rate": 7.478942141117998e-06,
+      "loss": 0.9429,
+      "step": 25200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6213410137940851,
+      "learning_rate": 7.478203753867202e-06,
+      "loss": 0.9062,
+      "step": 25201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7475271664917786,
+      "learning_rate": 7.477465381299094e-06,
+      "loss": 0.6903,
+      "step": 25202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.999862669967051,
+      "learning_rate": 7.476727023417985e-06,
+      "loss": 0.7923,
+      "step": 25203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5139920192225382,
+      "learning_rate": 7.47598868022817e-06,
+      "loss": 0.8161,
+      "step": 25204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.830712990890143,
+      "learning_rate": 7.475250351733947e-06,
+      "loss": 0.8886,
+      "step": 25205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4725508079603473,
+      "learning_rate": 7.474512037939614e-06,
+      "loss": 0.9094,
+      "step": 25206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7533923295855909,
+      "learning_rate": 7.473773738849475e-06,
+      "loss": 0.8754,
+      "step": 25207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5674612012316738,
+      "learning_rate": 7.473035454467825e-06,
+      "loss": 0.7649,
+      "step": 25208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5941649444545347,
+      "learning_rate": 7.472297184798959e-06,
+      "loss": 0.7734,
+      "step": 25209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.606113674366942,
+      "learning_rate": 7.471558929847181e-06,
+      "loss": 1.0295,
+      "step": 25210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6640628042669925,
+      "learning_rate": 7.4708206896167904e-06,
+      "loss": 0.8802,
+      "step": 25211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5635806351122554,
+      "learning_rate": 7.470082464112079e-06,
+      "loss": 0.9757,
+      "step": 25212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5890676951311742,
+      "learning_rate": 7.469344253337347e-06,
+      "loss": 1.0391,
+      "step": 25213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7062743295105345,
+      "learning_rate": 7.468606057296896e-06,
+      "loss": 0.841,
+      "step": 25214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6242366591732709,
+      "learning_rate": 7.467867875995025e-06,
+      "loss": 0.7671,
+      "step": 25215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9971272014959487,
+      "learning_rate": 7.467129709436025e-06,
+      "loss": 0.9724,
+      "step": 25216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.60487078665315,
+      "learning_rate": 7.466391557624197e-06,
+      "loss": 0.9412,
+      "step": 25217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.615680929811965,
+      "learning_rate": 7.465653420563846e-06,
+      "loss": 0.8016,
+      "step": 25218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5783162553266215,
+      "learning_rate": 7.464915298259256e-06,
+      "loss": 0.8483,
+      "step": 25219
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8125154976873332,
+      "learning_rate": 7.464177190714734e-06,
+      "loss": 0.7804,
+      "step": 25220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8328446023022495,
+      "learning_rate": 7.463439097934572e-06,
+      "loss": 0.8327,
+      "step": 25221
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.768431151432612,
+      "learning_rate": 7.462701019923078e-06,
+      "loss": 0.7264,
+      "step": 25222
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5426136435513322,
+      "learning_rate": 7.461962956684538e-06,
+      "loss": 0.8732,
+      "step": 25223
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4120468423661074,
+      "learning_rate": 7.4612249082232524e-06,
+      "loss": 0.8873,
+      "step": 25224
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5448232020863633,
+      "learning_rate": 7.460486874543522e-06,
+      "loss": 0.8427,
+      "step": 25225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.729657310735081,
+      "learning_rate": 7.459748855649637e-06,
+      "loss": 0.7422,
+      "step": 25226
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5684969619080253,
+      "learning_rate": 7.459010851545901e-06,
+      "loss": 0.8943,
+      "step": 25227
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.572658881991454,
+      "learning_rate": 7.458272862236606e-06,
+      "loss": 0.7589,
+      "step": 25228
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5176474453446602,
+      "learning_rate": 7.457534887726056e-06,
+      "loss": 0.9952,
+      "step": 25229
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.713502039721858,
+      "learning_rate": 7.456796928018537e-06,
+      "loss": 0.9154,
+      "step": 25230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6453922530589595,
+      "learning_rate": 7.456058983118354e-06,
+      "loss": 0.7523,
+      "step": 25231
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6348639484857836,
+      "learning_rate": 7.4553210530298025e-06,
+      "loss": 0.734,
+      "step": 25232
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6191181944628157,
+      "learning_rate": 7.454583137757176e-06,
+      "loss": 0.8188,
+      "step": 25233
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5175927467658983,
+      "learning_rate": 7.4538452373047694e-06,
+      "loss": 0.8507,
+      "step": 25234
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.649483377028303,
+      "learning_rate": 7.453107351676885e-06,
+      "loss": 0.8094,
+      "step": 25235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0922624044700284,
+      "learning_rate": 7.452369480877818e-06,
+      "loss": 0.8304,
+      "step": 25236
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8430062362950033,
+      "learning_rate": 7.45163162491186e-06,
+      "loss": 0.8359,
+      "step": 25237
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5292062973315947,
+      "learning_rate": 7.450893783783308e-06,
+      "loss": 0.9059,
+      "step": 25238
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5725293398869091,
+      "learning_rate": 7.4501559574964634e-06,
+      "loss": 0.7897,
+      "step": 25239
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5974626741017703,
+      "learning_rate": 7.449418146055616e-06,
+      "loss": 0.8687,
+      "step": 25240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7028552133071455,
+      "learning_rate": 7.448680349465063e-06,
+      "loss": 0.9343,
+      "step": 25241
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7783710133490119,
+      "learning_rate": 7.447942567729098e-06,
+      "loss": 0.7306,
+      "step": 25242
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6605482989622051,
+      "learning_rate": 7.447204800852025e-06,
+      "loss": 0.9389,
+      "step": 25243
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7853466558387636,
+      "learning_rate": 7.446467048838131e-06,
+      "loss": 0.7502,
+      "step": 25244
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9708020221308309,
+      "learning_rate": 7.445729311691713e-06,
+      "loss": 0.9603,
+      "step": 25245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6206708832311214,
+      "learning_rate": 7.444991589417073e-06,
+      "loss": 0.8532,
+      "step": 25246
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.213415912956377,
+      "learning_rate": 7.444253882018495e-06,
+      "loss": 0.8637,
+      "step": 25247
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7286169725208644,
+      "learning_rate": 7.4435161895002815e-06,
+      "loss": 0.9553,
+      "step": 25248
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5701950180753583,
+      "learning_rate": 7.442778511866725e-06,
+      "loss": 0.982,
+      "step": 25249
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6975182876704924,
+      "learning_rate": 7.442040849122127e-06,
+      "loss": 0.8,
+      "step": 25250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5044695610832592,
+      "learning_rate": 7.441303201270772e-06,
+      "loss": 0.8898,
+      "step": 25251
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6611387992694377,
+      "learning_rate": 7.4405655683169606e-06,
+      "loss": 0.8076,
+      "step": 25252
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4191014481014073,
+      "learning_rate": 7.43982795026499e-06,
+      "loss": 0.8352,
+      "step": 25253
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5587559626952974,
+      "learning_rate": 7.439090347119149e-06,
+      "loss": 0.9696,
+      "step": 25254
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9496915163067234,
+      "learning_rate": 7.438352758883732e-06,
+      "loss": 0.8013,
+      "step": 25255
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.544993148793596,
+      "learning_rate": 7.437615185563038e-06,
+      "loss": 0.7878,
+      "step": 25256
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5680213544255932,
+      "learning_rate": 7.436877627161364e-06,
+      "loss": 0.7718,
+      "step": 25257
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3920292784591672,
+      "learning_rate": 7.4361400836829945e-06,
+      "loss": 0.9227,
+      "step": 25258
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5663318817414913,
+      "learning_rate": 7.43540255513223e-06,
+      "loss": 0.82,
+      "step": 25259
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6632627580363046,
+      "learning_rate": 7.434665041513367e-06,
+      "loss": 0.8239,
+      "step": 25260
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.697530092170214,
+      "learning_rate": 7.433927542830695e-06,
+      "loss": 0.9489,
+      "step": 25261
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4012713459766002,
+      "learning_rate": 7.433190059088506e-06,
+      "loss": 0.8104,
+      "step": 25262
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5227806552448102,
+      "learning_rate": 7.4324525902910996e-06,
+      "loss": 0.829,
+      "step": 25263
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.778589370653793,
+      "learning_rate": 7.43171513644277e-06,
+      "loss": 0.9646,
+      "step": 25264
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.590803291203102,
+      "learning_rate": 7.4309776975478046e-06,
+      "loss": 0.8615,
+      "step": 25265
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3905348955772892,
+      "learning_rate": 7.4302402736104985e-06,
+      "loss": 0.8196,
+      "step": 25266
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5462674116022772,
+      "learning_rate": 7.429502864635154e-06,
+      "loss": 0.8054,
+      "step": 25267
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.616135267361928,
+      "learning_rate": 7.428765470626051e-06,
+      "loss": 0.8735,
+      "step": 25268
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.66696063849088,
+      "learning_rate": 7.428028091587491e-06,
+      "loss": 0.8082,
+      "step": 25269
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.569803814566293,
+      "learning_rate": 7.427290727523764e-06,
+      "loss": 0.9337,
+      "step": 25270
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4049747345005217,
+      "learning_rate": 7.42655337843917e-06,
+      "loss": 0.864,
+      "step": 25271
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8304709533697172,
+      "learning_rate": 7.4258160443379935e-06,
+      "loss": 0.8783,
+      "step": 25272
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5939854475071482,
+      "learning_rate": 7.4250787252245316e-06,
+      "loss": 0.8796,
+      "step": 25273
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9134060871516821,
+      "learning_rate": 7.424341421103074e-06,
+      "loss": 0.7048,
+      "step": 25274
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7761548785506638,
+      "learning_rate": 7.423604131977921e-06,
+      "loss": 0.7989,
+      "step": 25275
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.211787278852836,
+      "learning_rate": 7.422866857853358e-06,
+      "loss": 0.8677,
+      "step": 25276
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7457415698086252,
+      "learning_rate": 7.422129598733681e-06,
+      "loss": 0.8002,
+      "step": 25277
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5027040591598695,
+      "learning_rate": 7.421392354623181e-06,
+      "loss": 0.9493,
+      "step": 25278
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5269180545744878,
+      "learning_rate": 7.4206551255261495e-06,
+      "loss": 0.8334,
+      "step": 25279
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3943946251622648,
+      "learning_rate": 7.419917911446881e-06,
+      "loss": 0.7793,
+      "step": 25280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6638659270498457,
+      "learning_rate": 7.4191807123896686e-06,
+      "loss": 0.8383,
+      "step": 25281
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6420216259779594,
+      "learning_rate": 7.418443528358805e-06,
+      "loss": 0.8938,
+      "step": 25282
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.389531496195934,
+      "learning_rate": 7.417706359358576e-06,
+      "loss": 0.8949,
+      "step": 25283
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5409614603290824,
+      "learning_rate": 7.416969205393279e-06,
+      "loss": 0.7736,
+      "step": 25284
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4861034920726401,
+      "learning_rate": 7.4162320664672084e-06,
+      "loss": 0.9083,
+      "step": 25285
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6299283359384438,
+      "learning_rate": 7.415494942584651e-06,
+      "loss": 0.8735,
+      "step": 25286
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6532607550858596,
+      "learning_rate": 7.414757833749897e-06,
+      "loss": 0.9561,
+      "step": 25287
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5706423119645307,
+      "learning_rate": 7.414020739967245e-06,
+      "loss": 0.8989,
+      "step": 25288
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4207308506125254,
+      "learning_rate": 7.413283661240984e-06,
+      "loss": 0.8555,
+      "step": 25289
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7461284666769123,
+      "learning_rate": 7.412546597575403e-06,
+      "loss": 0.9259,
+      "step": 25290
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3115399658315674,
+      "learning_rate": 7.411809548974792e-06,
+      "loss": 0.7824,
+      "step": 25291
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.796141320893518,
+      "learning_rate": 7.411072515443449e-06,
+      "loss": 0.8582,
+      "step": 25292
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8638788709131415,
+      "learning_rate": 7.410335496985661e-06,
+      "loss": 0.8278,
+      "step": 25293
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.443918627290233,
+      "learning_rate": 7.409598493605716e-06,
+      "loss": 0.7913,
+      "step": 25294
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8504781607657066,
+      "learning_rate": 7.408861505307912e-06,
+      "loss": 0.8849,
+      "step": 25295
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9033080335671648,
+      "learning_rate": 7.408124532096539e-06,
+      "loss": 0.862,
+      "step": 25296
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.564805888766839,
+      "learning_rate": 7.407387573975881e-06,
+      "loss": 0.7976,
+      "step": 25297
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.695712825991363,
+      "learning_rate": 7.406650630950234e-06,
+      "loss": 0.8508,
+      "step": 25298
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1960644119706134,
+      "learning_rate": 7.405913703023894e-06,
+      "loss": 0.8018,
+      "step": 25299
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.77139809922408,
+      "learning_rate": 7.405176790201139e-06,
+      "loss": 0.6805,
+      "step": 25300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8152585995512565,
+      "learning_rate": 7.404439892486267e-06,
+      "loss": 0.7856,
+      "step": 25301
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5555913449859524,
+      "learning_rate": 7.403703009883567e-06,
+      "loss": 1.0072,
+      "step": 25302
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5717781107777113,
+      "learning_rate": 7.402966142397335e-06,
+      "loss": 0.8046,
+      "step": 25303
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6759652271780419,
+      "learning_rate": 7.402229290031851e-06,
+      "loss": 0.8975,
+      "step": 25304
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.70091327705169,
+      "learning_rate": 7.401492452791414e-06,
+      "loss": 0.7495,
+      "step": 25305
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.92195196959963,
+      "learning_rate": 7.4007556306803095e-06,
+      "loss": 0.8152,
+      "step": 25306
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.747396581331862,
+      "learning_rate": 7.400018823702827e-06,
+      "loss": 0.864,
+      "step": 25307
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.584987084993635,
+      "learning_rate": 7.39928203186326e-06,
+      "loss": 0.8514,
+      "step": 25308
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5996236787771902,
+      "learning_rate": 7.3985452551658956e-06,
+      "loss": 0.7534,
+      "step": 25309
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8737779518986202,
+      "learning_rate": 7.397808493615026e-06,
+      "loss": 0.8943,
+      "step": 25310
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4565136359359407,
+      "learning_rate": 7.3970717472149354e-06,
+      "loss": 0.9304,
+      "step": 25311
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5281513203327504,
+      "learning_rate": 7.39633501596992e-06,
+      "loss": 0.9075,
+      "step": 25312
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4862274605464738,
+      "learning_rate": 7.3955982998842675e-06,
+      "loss": 0.8819,
+      "step": 25313
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7565814731800773,
+      "learning_rate": 7.394861598962265e-06,
+      "loss": 0.7018,
+      "step": 25314
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6438703952353204,
+      "learning_rate": 7.394124913208201e-06,
+      "loss": 0.9371,
+      "step": 25315
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5826625087385422,
+      "learning_rate": 7.393388242626368e-06,
+      "loss": 0.8934,
+      "step": 25316
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8179107608086873,
+      "learning_rate": 7.3926515872210555e-06,
+      "loss": 0.8505,
+      "step": 25317
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4964478132680776,
+      "learning_rate": 7.3919149469965505e-06,
+      "loss": 0.8619,
+      "step": 25318
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6476378387218484,
+      "learning_rate": 7.39117832195714e-06,
+      "loss": 0.8309,
+      "step": 25319
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5708231190666417,
+      "learning_rate": 7.390441712107119e-06,
+      "loss": 0.825,
+      "step": 25320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7948716681510473,
+      "learning_rate": 7.389705117450769e-06,
+      "loss": 0.9269,
+      "step": 25321
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5850311346456547,
+      "learning_rate": 7.388968537992385e-06,
+      "loss": 0.9033,
+      "step": 25322
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6342898165842599,
+      "learning_rate": 7.388231973736248e-06,
+      "loss": 0.9204,
+      "step": 25323
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5227346008784985,
+      "learning_rate": 7.387495424686657e-06,
+      "loss": 0.8845,
+      "step": 25324
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8047950986664463,
+      "learning_rate": 7.3867588908478934e-06,
+      "loss": 0.7157,
+      "step": 25325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5672161966482867,
+      "learning_rate": 7.386022372224246e-06,
+      "loss": 0.9228,
+      "step": 25326
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7913829347106223,
+      "learning_rate": 7.385285868820006e-06,
+      "loss": 0.7459,
+      "step": 25327
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.607482308990439,
+      "learning_rate": 7.384549380639454e-06,
+      "loss": 0.8216,
+      "step": 25328
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.94488546569586,
+      "learning_rate": 7.383812907686888e-06,
+      "loss": 0.9015,
+      "step": 25329
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7987980421515648,
+      "learning_rate": 7.383076449966591e-06,
+      "loss": 0.9213,
+      "step": 25330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8539177010793813,
+      "learning_rate": 7.382340007482852e-06,
+      "loss": 0.9866,
+      "step": 25331
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.748956986116981,
+      "learning_rate": 7.381603580239956e-06,
+      "loss": 0.6962,
+      "step": 25332
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7405235263570922,
+      "learning_rate": 7.380867168242193e-06,
+      "loss": 0.9527,
+      "step": 25333
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8567032247329796,
+      "learning_rate": 7.380130771493854e-06,
+      "loss": 0.8051,
+      "step": 25334
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.578332345239343,
+      "learning_rate": 7.37939438999922e-06,
+      "loss": 0.8686,
+      "step": 25335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7639351570516375,
+      "learning_rate": 7.3786580237625804e-06,
+      "loss": 0.8725,
+      "step": 25336
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.493179609173798,
+      "learning_rate": 7.377921672788225e-06,
+      "loss": 0.9246,
+      "step": 25337
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.817792074588887,
+      "learning_rate": 7.377185337080443e-06,
+      "loss": 0.8252,
+      "step": 25338
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4561472425164559,
+      "learning_rate": 7.376449016643516e-06,
+      "loss": 0.7786,
+      "step": 25339
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.504334780227838,
+      "learning_rate": 7.375712711481731e-06,
+      "loss": 0.8604,
+      "step": 25340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8960262289619254,
+      "learning_rate": 7.3749764215993804e-06,
+      "loss": 0.7456,
+      "step": 25341
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9146824769151158,
+      "learning_rate": 7.374240147000748e-06,
+      "loss": 0.7647,
+      "step": 25342
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5270651304383858,
+      "learning_rate": 7.373503887690117e-06,
+      "loss": 0.8718,
+      "step": 25343
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5223385002014045,
+      "learning_rate": 7.372767643671781e-06,
+      "loss": 0.8643,
+      "step": 25344
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5738706956060344,
+      "learning_rate": 7.372031414950025e-06,
+      "loss": 0.7949,
+      "step": 25345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4319257972607538,
+      "learning_rate": 7.371295201529133e-06,
+      "loss": 0.825,
+      "step": 25346
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9672822799872254,
+      "learning_rate": 7.370559003413389e-06,
+      "loss": 0.8113,
+      "step": 25347
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4012131783296358,
+      "learning_rate": 7.369822820607091e-06,
+      "loss": 0.7231,
+      "step": 25348
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.525923833452668,
+      "learning_rate": 7.36908665311451e-06,
+      "loss": 0.9191,
+      "step": 25349
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9673227467310297,
+      "learning_rate": 7.3683505009399415e-06,
+      "loss": 0.8647,
+      "step": 25350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5544609407453989,
+      "learning_rate": 7.3676143640876675e-06,
+      "loss": 0.7774,
+      "step": 25351
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7393041349832143,
+      "learning_rate": 7.366878242561984e-06,
+      "loss": 0.9095,
+      "step": 25352
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6089426544484597,
+      "learning_rate": 7.36614213636716e-06,
+      "loss": 0.8679,
+      "step": 25353
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.023594834553786,
+      "learning_rate": 7.365406045507493e-06,
+      "loss": 0.7404,
+      "step": 25354
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.687035899810172,
+      "learning_rate": 7.364669969987269e-06,
+      "loss": 0.849,
+      "step": 25355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9613537015287892,
+      "learning_rate": 7.363933909810766e-06,
+      "loss": 0.7803,
+      "step": 25356
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7676572074145342,
+      "learning_rate": 7.363197864982277e-06,
+      "loss": 0.8443,
+      "step": 25357
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6907706604092236,
+      "learning_rate": 7.362461835506085e-06,
+      "loss": 0.7526,
+      "step": 25358
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6978244057840763,
+      "learning_rate": 7.361725821386476e-06,
+      "loss": 0.9241,
+      "step": 25359
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.490935667816902,
+      "learning_rate": 7.360989822627732e-06,
+      "loss": 0.8624,
+      "step": 25360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6363704358494913,
+      "learning_rate": 7.3602538392341415e-06,
+      "loss": 0.8631,
+      "step": 25361
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4256614964633987,
+      "learning_rate": 7.359517871209993e-06,
+      "loss": 0.7601,
+      "step": 25362
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7656988602769452,
+      "learning_rate": 7.358781918559563e-06,
+      "loss": 0.7212,
+      "step": 25363
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4710242613945115,
+      "learning_rate": 7.358045981287141e-06,
+      "loss": 0.8392,
+      "step": 25364
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1436629414920976,
+      "learning_rate": 7.357310059397011e-06,
+      "loss": 0.8834,
+      "step": 25365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4960151712732763,
+      "learning_rate": 7.356574152893463e-06,
+      "loss": 0.8611,
+      "step": 25366
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4585348678619205,
+      "learning_rate": 7.355838261780774e-06,
+      "loss": 0.857,
+      "step": 25367
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6256936887955389,
+      "learning_rate": 7.355102386063229e-06,
+      "loss": 0.7635,
+      "step": 25368
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4495478448834267,
+      "learning_rate": 7.354366525745121e-06,
+      "loss": 0.7644,
+      "step": 25369
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4865662568694975,
+      "learning_rate": 7.3536306808307256e-06,
+      "loss": 0.8208,
+      "step": 25370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7756751289827908,
+      "learning_rate": 7.35289485132433e-06,
+      "loss": 0.7532,
+      "step": 25371
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0750930865577795,
+      "learning_rate": 7.352159037230218e-06,
+      "loss": 0.7448,
+      "step": 25372
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7659314790774363,
+      "learning_rate": 7.351423238552677e-06,
+      "loss": 0.8367,
+      "step": 25373
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7760963048907222,
+      "learning_rate": 7.3506874552959886e-06,
+      "loss": 0.7127,
+      "step": 25374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5026354992509463,
+      "learning_rate": 7.349951687464434e-06,
+      "loss": 0.8654,
+      "step": 25375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7040686680527781,
+      "learning_rate": 7.349215935062302e-06,
+      "loss": 0.8511,
+      "step": 25376
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6933326685398835,
+      "learning_rate": 7.348480198093871e-06,
+      "loss": 0.6449,
+      "step": 25377
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5860424421213954,
+      "learning_rate": 7.34774447656343e-06,
+      "loss": 0.9079,
+      "step": 25378
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8350605946480643,
+      "learning_rate": 7.347008770475259e-06,
+      "loss": 0.8398,
+      "step": 25379
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7485904630985356,
+      "learning_rate": 7.346273079833645e-06,
+      "loss": 0.7587,
+      "step": 25380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8304889292244242,
+      "learning_rate": 7.3455374046428675e-06,
+      "loss": 0.933,
+      "step": 25381
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5389227184533125,
+      "learning_rate": 7.344801744907212e-06,
+      "loss": 0.8858,
+      "step": 25382
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5130352892992294,
+      "learning_rate": 7.344066100630963e-06,
+      "loss": 0.8265,
+      "step": 25383
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5469718038322302,
+      "learning_rate": 7.3433304718184004e-06,
+      "loss": 0.8496,
+      "step": 25384
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5111875121408305,
+      "learning_rate": 7.342594858473807e-06,
+      "loss": 0.7881,
+      "step": 25385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7719381513830038,
+      "learning_rate": 7.34185926060147e-06,
+      "loss": 0.7094,
+      "step": 25386
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.866359598569302,
+      "learning_rate": 7.34112367820567e-06,
+      "loss": 0.7894,
+      "step": 25387
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5869877918245243,
+      "learning_rate": 7.3403881112906905e-06,
+      "loss": 0.9161,
+      "step": 25388
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4765661079885224,
+      "learning_rate": 7.3396525598608105e-06,
+      "loss": 0.8212,
+      "step": 25389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5284737050299113,
+      "learning_rate": 7.338917023920319e-06,
+      "loss": 0.8428,
+      "step": 25390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.650527665995247,
+      "learning_rate": 7.338181503473493e-06,
+      "loss": 0.823,
+      "step": 25391
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5114415723091963,
+      "learning_rate": 7.3374459985246146e-06,
+      "loss": 0.8068,
+      "step": 25392
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4140207746923346,
+      "learning_rate": 7.336710509077971e-06,
+      "loss": 0.9213,
+      "step": 25393
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7565665083019872,
+      "learning_rate": 7.335975035137845e-06,
+      "loss": 0.8699,
+      "step": 25394
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5500872530142293,
+      "learning_rate": 7.335239576708513e-06,
+      "loss": 0.8578,
+      "step": 25395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.780428739374251,
+      "learning_rate": 7.334504133794258e-06,
+      "loss": 0.8957,
+      "step": 25396
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.739073887272121,
+      "learning_rate": 7.333768706399365e-06,
+      "loss": 0.7884,
+      "step": 25397
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5694929581739971,
+      "learning_rate": 7.333033294528118e-06,
+      "loss": 1.0054,
+      "step": 25398
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4507625213334014,
+      "learning_rate": 7.332297898184793e-06,
+      "loss": 0.9369,
+      "step": 25399
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6200583014779186,
+      "learning_rate": 7.331562517373672e-06,
+      "loss": 0.8257,
+      "step": 25400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5562412066697762,
+      "learning_rate": 7.330827152099045e-06,
+      "loss": 0.8539,
+      "step": 25401
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4227617697079846,
+      "learning_rate": 7.3300918023651805e-06,
+      "loss": 0.9495,
+      "step": 25402
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6731895312513083,
+      "learning_rate": 7.32935646817637e-06,
+      "loss": 0.8683,
+      "step": 25403
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.44249253797547,
+      "learning_rate": 7.32862114953689e-06,
+      "loss": 0.74,
+      "step": 25404
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8089591332177088,
+      "learning_rate": 7.327885846451026e-06,
+      "loss": 0.7636,
+      "step": 25405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5997919373207354,
+      "learning_rate": 7.327150558923055e-06,
+      "loss": 0.7792,
+      "step": 25406
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.494487116255993,
+      "learning_rate": 7.32641528695726e-06,
+      "loss": 0.8835,
+      "step": 25407
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6178172987955892,
+      "learning_rate": 7.3256800305579234e-06,
+      "loss": 0.9065,
+      "step": 25408
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7411716060881649,
+      "learning_rate": 7.3249447897293205e-06,
+      "loss": 0.634,
+      "step": 25409
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4823102904868124,
+      "learning_rate": 7.324209564475737e-06,
+      "loss": 0.8654,
+      "step": 25410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5124200665705967,
+      "learning_rate": 7.323474354801454e-06,
+      "loss": 0.8571,
+      "step": 25411
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3774029076191718,
+      "learning_rate": 7.322739160710752e-06,
+      "loss": 0.8023,
+      "step": 25412
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8563874051357585,
+      "learning_rate": 7.3220039822079066e-06,
+      "loss": 0.7528,
+      "step": 25413
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6325451162860576,
+      "learning_rate": 7.321268819297204e-06,
+      "loss": 0.8444,
+      "step": 25414
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4543582853839525,
+      "learning_rate": 7.320533671982924e-06,
+      "loss": 0.7596,
+      "step": 25415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.802052251862085,
+      "learning_rate": 7.319798540269343e-06,
+      "loss": 0.6766,
+      "step": 25416
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4489913806892727,
+      "learning_rate": 7.319063424160742e-06,
+      "loss": 0.8171,
+      "step": 25417
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4504407321411898,
+      "learning_rate": 7.318328323661403e-06,
+      "loss": 0.7881,
+      "step": 25418
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9627151188687122,
+      "learning_rate": 7.31759323877561e-06,
+      "loss": 0.8371,
+      "step": 25419
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7467507276013989,
+      "learning_rate": 7.316858169507635e-06,
+      "loss": 0.7117,
+      "step": 25420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1828180487373814,
+      "learning_rate": 7.31612311586176e-06,
+      "loss": 0.7597,
+      "step": 25421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.231807469735984,
+      "learning_rate": 7.315388077842269e-06,
+      "loss": 0.8356,
+      "step": 25422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6329607213792179,
+      "learning_rate": 7.3146530554534355e-06,
+      "loss": 0.8245,
+      "step": 25423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6574086967694397,
+      "learning_rate": 7.313918048699544e-06,
+      "loss": 0.8645,
+      "step": 25424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.935213249650836,
+      "learning_rate": 7.31318305758487e-06,
+      "loss": 0.8057,
+      "step": 25425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4921518339565065,
+      "learning_rate": 7.3124480821136974e-06,
+      "loss": 0.832,
+      "step": 25426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9333455450249135,
+      "learning_rate": 7.3117131222903025e-06,
+      "loss": 0.7727,
+      "step": 25427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8994241306018482,
+      "learning_rate": 7.3109781781189635e-06,
+      "loss": 0.8069,
+      "step": 25428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.529039073230126,
+      "learning_rate": 7.310243249603963e-06,
+      "loss": 0.8858,
+      "step": 25429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7763654205769344,
+      "learning_rate": 7.309508336749576e-06,
+      "loss": 0.9063,
+      "step": 25430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7512739536777113,
+      "learning_rate": 7.308773439560083e-06,
+      "loss": 0.7125,
+      "step": 25431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5571930968677608,
+      "learning_rate": 7.308038558039762e-06,
+      "loss": 0.9071,
+      "step": 25432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5787386305565623,
+      "learning_rate": 7.307303692192899e-06,
+      "loss": 0.9362,
+      "step": 25433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7626563917726445,
+      "learning_rate": 7.306568842023759e-06,
+      "loss": 0.7174,
+      "step": 25434
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6172580959136658,
+      "learning_rate": 7.305834007536632e-06,
+      "loss": 0.79,
+      "step": 25435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7332176783656026,
+      "learning_rate": 7.305099188735793e-06,
+      "loss": 0.6731,
+      "step": 25436
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5724655056522712,
+      "learning_rate": 7.3043643856255175e-06,
+      "loss": 0.858,
+      "step": 25437
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8261644494855069,
+      "learning_rate": 7.303629598210084e-06,
+      "loss": 0.7983,
+      "step": 25438
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.247698530701456,
+      "learning_rate": 7.302894826493775e-06,
+      "loss": 0.8592,
+      "step": 25439
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.894248474631495,
+      "learning_rate": 7.302160070480869e-06,
+      "loss": 0.8054,
+      "step": 25440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7263319598521187,
+      "learning_rate": 7.301425330175635e-06,
+      "loss": 0.8538,
+      "step": 25441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.507944045920637,
+      "learning_rate": 7.300690605582361e-06,
+      "loss": 0.7294,
+      "step": 25442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9845302543756556,
+      "learning_rate": 7.299955896705323e-06,
+      "loss": 0.7709,
+      "step": 25443
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5959404936768318,
+      "learning_rate": 7.299221203548794e-06,
+      "loss": 0.9499,
+      "step": 25444
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6376713447886952,
+      "learning_rate": 7.2984865261170525e-06,
+      "loss": 0.8792,
+      "step": 25445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.325154495739654,
+      "learning_rate": 7.29775186441438e-06,
+      "loss": 0.7437,
+      "step": 25446
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7282780448981974,
+      "learning_rate": 7.2970172184450525e-06,
+      "loss": 0.9273,
+      "step": 25447
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3485862137955955,
+      "learning_rate": 7.296282588213345e-06,
+      "loss": 1.0084,
+      "step": 25448
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5836818801413781,
+      "learning_rate": 7.295547973723533e-06,
+      "loss": 0.8953,
+      "step": 25449
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5949854100858873,
+      "learning_rate": 7.294813374979906e-06,
+      "loss": 0.9151,
+      "step": 25450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7673892670854755,
+      "learning_rate": 7.294078791986723e-06,
+      "loss": 0.6791,
+      "step": 25451
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7322286456545362,
+      "learning_rate": 7.293344224748274e-06,
+      "loss": 1.0435,
+      "step": 25452
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7232121500500144,
+      "learning_rate": 7.292609673268829e-06,
+      "loss": 0.8891,
+      "step": 25453
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4553016946878137,
+      "learning_rate": 7.291875137552671e-06,
+      "loss": 0.7233,
+      "step": 25454
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5816225515171702,
+      "learning_rate": 7.291140617604072e-06,
+      "loss": 0.8372,
+      "step": 25455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8126223672719761,
+      "learning_rate": 7.2904061134273105e-06,
+      "loss": 0.9179,
+      "step": 25456
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6290594934801732,
+      "learning_rate": 7.289671625026664e-06,
+      "loss": 0.9137,
+      "step": 25457
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6028348703765407,
+      "learning_rate": 7.288937152406404e-06,
+      "loss": 0.8458,
+      "step": 25458
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5127056997184525,
+      "learning_rate": 7.28820269557081e-06,
+      "loss": 0.7669,
+      "step": 25459
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5881324553271297,
+      "learning_rate": 7.2874682545241604e-06,
+      "loss": 0.7907,
+      "step": 25460
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.767434139423436,
+      "learning_rate": 7.286733829270731e-06,
+      "loss": 0.9392,
+      "step": 25461
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3462460089641974,
+      "learning_rate": 7.285999419814792e-06,
+      "loss": 0.9348,
+      "step": 25462
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6166451218322537,
+      "learning_rate": 7.285265026160626e-06,
+      "loss": 0.8526,
+      "step": 25463
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3728257718784667,
+      "learning_rate": 7.284530648312509e-06,
+      "loss": 0.8038,
+      "step": 25464
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7076631666416089,
+      "learning_rate": 7.283796286274711e-06,
+      "loss": 0.8144,
+      "step": 25465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5295565175300683,
+      "learning_rate": 7.283061940051509e-06,
+      "loss": 0.7896,
+      "step": 25466
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.933148487528754,
+      "learning_rate": 7.2823276096471825e-06,
+      "loss": 0.983,
+      "step": 25467
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5372925937760231,
+      "learning_rate": 7.281593295066008e-06,
+      "loss": 0.7178,
+      "step": 25468
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5713351174338943,
+      "learning_rate": 7.280858996312255e-06,
+      "loss": 0.8621,
+      "step": 25469
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.320721512365127,
+      "learning_rate": 7.280124713390201e-06,
+      "loss": 0.9212,
+      "step": 25470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5276326566867244,
+      "learning_rate": 7.279390446304125e-06,
+      "loss": 0.8285,
+      "step": 25471
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6096007031655228,
+      "learning_rate": 7.278656195058298e-06,
+      "loss": 0.9496,
+      "step": 25472
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.485988662220139,
+      "learning_rate": 7.277921959656995e-06,
+      "loss": 0.87,
+      "step": 25473
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.724238252599017,
+      "learning_rate": 7.277187740104492e-06,
+      "loss": 0.8719,
+      "step": 25474
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7141027156778912,
+      "learning_rate": 7.276453536405066e-06,
+      "loss": 0.8088,
+      "step": 25475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.091288173207551,
+      "learning_rate": 7.275719348562989e-06,
+      "loss": 0.8129,
+      "step": 25476
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4466918496865158,
+      "learning_rate": 7.274985176582536e-06,
+      "loss": 0.8744,
+      "step": 25477
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6633022159528412,
+      "learning_rate": 7.274251020467984e-06,
+      "loss": 0.9008,
+      "step": 25478
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8152469522341554,
+      "learning_rate": 7.273516880223601e-06,
+      "loss": 0.752,
+      "step": 25479
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7661174154230852,
+      "learning_rate": 7.272782755853669e-06,
+      "loss": 0.6992,
+      "step": 25480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4779855695584385,
+      "learning_rate": 7.272048647362457e-06,
+      "loss": 0.8363,
+      "step": 25481
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7391651183816426,
+      "learning_rate": 7.2713145547542475e-06,
+      "loss": 0.916,
+      "step": 25482
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7698583032151396,
+      "learning_rate": 7.270580478033302e-06,
+      "loss": 0.7063,
+      "step": 25483
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4661300556360568,
+      "learning_rate": 7.269846417203903e-06,
+      "loss": 0.8726,
+      "step": 25484
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6277708329777936,
+      "learning_rate": 7.269112372270325e-06,
+      "loss": 0.8665,
+      "step": 25485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4606636066011158,
+      "learning_rate": 7.268378343236837e-06,
+      "loss": 0.806,
+      "step": 25486
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4727225764009315,
+      "learning_rate": 7.267644330107714e-06,
+      "loss": 0.7921,
+      "step": 25487
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5678742249305202,
+      "learning_rate": 7.266910332887231e-06,
+      "loss": 0.7921,
+      "step": 25488
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5898934598444165,
+      "learning_rate": 7.266176351579666e-06,
+      "loss": 0.904,
+      "step": 25489
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6617909155896278,
+      "learning_rate": 7.265442386189281e-06,
+      "loss": 0.907,
+      "step": 25490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8010354495210397,
+      "learning_rate": 7.26470843672036e-06,
+      "loss": 0.8336,
+      "step": 25491
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4936916285787611,
+      "learning_rate": 7.263974503177174e-06,
+      "loss": 0.9624,
+      "step": 25492
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4606994993079836,
+      "learning_rate": 7.263240585563991e-06,
+      "loss": 0.8425,
+      "step": 25493
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7849030736296342,
+      "learning_rate": 7.262506683885088e-06,
+      "loss": 0.6865,
+      "step": 25494
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4864147355249304,
+      "learning_rate": 7.26177279814474e-06,
+      "loss": 0.8371,
+      "step": 25495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8091335632922996,
+      "learning_rate": 7.261038928347217e-06,
+      "loss": 0.8707,
+      "step": 25496
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.446872596239553,
+      "learning_rate": 7.260305074496793e-06,
+      "loss": 0.9107,
+      "step": 25497
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.406602369349781,
+      "learning_rate": 7.259571236597737e-06,
+      "loss": 0.8834,
+      "step": 25498
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4858323509395315,
+      "learning_rate": 7.2588374146543315e-06,
+      "loss": 0.957,
+      "step": 25499
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5520971720417112,
+      "learning_rate": 7.258103608670835e-06,
+      "loss": 0.7607,
+      "step": 25500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0454879106035175,
+      "learning_rate": 7.2573698186515316e-06,
+      "loss": 0.8592,
+      "step": 25501
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.024303565886224,
+      "learning_rate": 7.256636044600687e-06,
+      "loss": 0.8517,
+      "step": 25502
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6905013277594203,
+      "learning_rate": 7.255902286522578e-06,
+      "loss": 0.7491,
+      "step": 25503
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9577959770434823,
+      "learning_rate": 7.255168544421473e-06,
+      "loss": 0.9458,
+      "step": 25504
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5423386571773334,
+      "learning_rate": 7.254434818301648e-06,
+      "loss": 0.9526,
+      "step": 25505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4770742511263903,
+      "learning_rate": 7.253701108167373e-06,
+      "loss": 0.8404,
+      "step": 25506
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5574590411740596,
+      "learning_rate": 7.252967414022915e-06,
+      "loss": 0.8949,
+      "step": 25507
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7679094071245058,
+      "learning_rate": 7.252233735872554e-06,
+      "loss": 0.74,
+      "step": 25508
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5027180133243814,
+      "learning_rate": 7.251500073720558e-06,
+      "loss": 0.8842,
+      "step": 25509
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7765980461258943,
+      "learning_rate": 7.2507664275712e-06,
+      "loss": 0.8685,
+      "step": 25510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.477808045123796,
+      "learning_rate": 7.250032797428746e-06,
+      "loss": 1.0132,
+      "step": 25511
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7193813009038679,
+      "learning_rate": 7.249299183297475e-06,
+      "loss": 0.8898,
+      "step": 25512
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5867824966104878,
+      "learning_rate": 7.248565585181655e-06,
+      "loss": 0.8725,
+      "step": 25513
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8179132228114062,
+      "learning_rate": 7.2478320030855576e-06,
+      "loss": 0.7205,
+      "step": 25514
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8258201622997832,
+      "learning_rate": 7.247098437013449e-06,
+      "loss": 0.7764,
+      "step": 25515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6165081199691784,
+      "learning_rate": 7.246364886969609e-06,
+      "loss": 0.9076,
+      "step": 25516
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4622479938646535,
+      "learning_rate": 7.245631352958306e-06,
+      "loss": 0.8082,
+      "step": 25517
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5889895422956404,
+      "learning_rate": 7.244897834983806e-06,
+      "loss": 0.8809,
+      "step": 25518
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6681819530239699,
+      "learning_rate": 7.244164333050383e-06,
+      "loss": 0.9163,
+      "step": 25519
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6734545551574842,
+      "learning_rate": 7.243430847162308e-06,
+      "loss": 0.9089,
+      "step": 25520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5336332901837944,
+      "learning_rate": 7.242697377323854e-06,
+      "loss": 0.9986,
+      "step": 25521
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.932676944102785,
+      "learning_rate": 7.241963923539287e-06,
+      "loss": 0.7697,
+      "step": 25522
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5291241223802423,
+      "learning_rate": 7.241230485812877e-06,
+      "loss": 0.7548,
+      "step": 25523
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8823664634196922,
+      "learning_rate": 7.2404970641489015e-06,
+      "loss": 0.8894,
+      "step": 25524
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8583811716780637,
+      "learning_rate": 7.2397636585516215e-06,
+      "loss": 0.7066,
+      "step": 25525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.123526108125039,
+      "learning_rate": 7.239030269025311e-06,
+      "loss": 0.9163,
+      "step": 25526
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.494852046267675,
+      "learning_rate": 7.238296895574242e-06,
+      "loss": 0.7896,
+      "step": 25527
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6365956944478108,
+      "learning_rate": 7.237563538202685e-06,
+      "loss": 0.8733,
+      "step": 25528
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6443529813566562,
+      "learning_rate": 7.2368301969149054e-06,
+      "loss": 0.8644,
+      "step": 25529
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8544319040805242,
+      "learning_rate": 7.236096871715174e-06,
+      "loss": 0.8482,
+      "step": 25530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7707436119723599,
+      "learning_rate": 7.2353635626077664e-06,
+      "loss": 0.9197,
+      "step": 25531
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8408709958767744,
+      "learning_rate": 7.234630269596943e-06,
+      "loss": 0.826,
+      "step": 25532
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6692925231382287,
+      "learning_rate": 7.233896992686977e-06,
+      "loss": 0.7959,
+      "step": 25533
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5574755203426294,
+      "learning_rate": 7.233163731882139e-06,
+      "loss": 0.8099,
+      "step": 25534
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.634606454868502,
+      "learning_rate": 7.2324304871867036e-06,
+      "loss": 0.9623,
+      "step": 25535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5005115890647813,
+      "learning_rate": 7.231697258604926e-06,
+      "loss": 0.7867,
+      "step": 25536
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.782834160370677,
+      "learning_rate": 7.230964046141087e-06,
+      "loss": 0.8862,
+      "step": 25537
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6977690360768165,
+      "learning_rate": 7.230230849799454e-06,
+      "loss": 0.8841,
+      "step": 25538
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6623576624538876,
+      "learning_rate": 7.229497669584289e-06,
+      "loss": 0.8144,
+      "step": 25539
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.632812270226217,
+      "learning_rate": 7.228764505499869e-06,
+      "loss": 0.8528,
+      "step": 25540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6816813529801593,
+      "learning_rate": 7.228031357550458e-06,
+      "loss": 0.9844,
+      "step": 25541
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6846433256227997,
+      "learning_rate": 7.227298225740329e-06,
+      "loss": 0.8879,
+      "step": 25542
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7626711739395762,
+      "learning_rate": 7.226565110073743e-06,
+      "loss": 0.96,
+      "step": 25543
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8464580278242992,
+      "learning_rate": 7.225832010554976e-06,
+      "loss": 0.7943,
+      "step": 25544
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5646846405379455,
+      "learning_rate": 7.225098927188294e-06,
+      "loss": 0.7869,
+      "step": 25545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8811405138019937,
+      "learning_rate": 7.2243658599779645e-06,
+      "loss": 0.8125,
+      "step": 25546
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.081025092408679,
+      "learning_rate": 7.223632808928252e-06,
+      "loss": 0.8204,
+      "step": 25547
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9835368277751508,
+      "learning_rate": 7.22289977404343e-06,
+      "loss": 0.7403,
+      "step": 25548
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5000551782178528,
+      "learning_rate": 7.222166755327768e-06,
+      "loss": 0.756,
+      "step": 25549
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.552199386105149,
+      "learning_rate": 7.221433752785527e-06,
+      "loss": 0.7699,
+      "step": 25550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7190465108164477,
+      "learning_rate": 7.2207007664209785e-06,
+      "loss": 0.7307,
+      "step": 25551
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.492827024848564,
+      "learning_rate": 7.2199677962383915e-06,
+      "loss": 0.9065,
+      "step": 25552
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9045725513173075,
+      "learning_rate": 7.219234842242032e-06,
+      "loss": 0.7906,
+      "step": 25553
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6119062795315942,
+      "learning_rate": 7.2185019044361675e-06,
+      "loss": 0.9177,
+      "step": 25554
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.517879096501686,
+      "learning_rate": 7.217768982825062e-06,
+      "loss": 0.8242,
+      "step": 25555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7517205303090753,
+      "learning_rate": 7.217036077412992e-06,
+      "loss": 0.8405,
+      "step": 25556
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.439423588614053,
+      "learning_rate": 7.216303188204217e-06,
+      "loss": 0.7885,
+      "step": 25557
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.464013774066656,
+      "learning_rate": 7.215570315203006e-06,
+      "loss": 0.8383,
+      "step": 25558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.596700798614734,
+      "learning_rate": 7.2148374584136274e-06,
+      "loss": 0.8301,
+      "step": 25559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6146613818465714,
+      "learning_rate": 7.214104617840343e-06,
+      "loss": 0.7506,
+      "step": 25560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.572145837776528,
+      "learning_rate": 7.213371793487426e-06,
+      "loss": 0.893,
+      "step": 25561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5118481554439416,
+      "learning_rate": 7.212638985359139e-06,
+      "loss": 0.7542,
+      "step": 25562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6442978715756802,
+      "learning_rate": 7.2119061934597545e-06,
+      "loss": 0.8461,
+      "step": 25563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7384095151887016,
+      "learning_rate": 7.211173417793531e-06,
+      "loss": 0.6914,
+      "step": 25564
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6861706550001014,
+      "learning_rate": 7.210440658364739e-06,
+      "loss": 0.8811,
+      "step": 25565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.604411400131276,
+      "learning_rate": 7.209707915177648e-06,
+      "loss": 0.9305,
+      "step": 25566
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5221674677879256,
+      "learning_rate": 7.208975188236518e-06,
+      "loss": 0.8008,
+      "step": 25567
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7779299984661839,
+      "learning_rate": 7.208242477545618e-06,
+      "loss": 0.7003,
+      "step": 25568
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0037429306555588,
+      "learning_rate": 7.2075097831092135e-06,
+      "loss": 0.8426,
+      "step": 25569
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8370826975619043,
+      "learning_rate": 7.206777104931575e-06,
+      "loss": 0.8336,
+      "step": 25570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7824604707724312,
+      "learning_rate": 7.2060444430169615e-06,
+      "loss": 0.7263,
+      "step": 25571
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6670957177800128,
+      "learning_rate": 7.20531179736964e-06,
+      "loss": 0.9048,
+      "step": 25572
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5243897331102871,
+      "learning_rate": 7.204579167993881e-06,
+      "loss": 0.7881,
+      "step": 25573
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5247676583596927,
+      "learning_rate": 7.203846554893947e-06,
+      "loss": 0.7099,
+      "step": 25574
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6524531395727073,
+      "learning_rate": 7.203113958074101e-06,
+      "loss": 0.8708,
+      "step": 25575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3911825587421318,
+      "learning_rate": 7.202381377538612e-06,
+      "loss": 0.9194,
+      "step": 25576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5973960967706504,
+      "learning_rate": 7.201648813291748e-06,
+      "loss": 0.8307,
+      "step": 25577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6671104419062923,
+      "learning_rate": 7.200916265337767e-06,
+      "loss": 0.714,
+      "step": 25578
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4967463137729493,
+      "learning_rate": 7.200183733680937e-06,
+      "loss": 0.9114,
+      "step": 25579
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7498138064386783,
+      "learning_rate": 7.1994512183255284e-06,
+      "loss": 0.9255,
+      "step": 25580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5951880444410012,
+      "learning_rate": 7.1987187192757965e-06,
+      "loss": 0.773,
+      "step": 25581
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5545291585438896,
+      "learning_rate": 7.1979862365360135e-06,
+      "loss": 0.9132,
+      "step": 25582
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8025565766615914,
+      "learning_rate": 7.197253770110438e-06,
+      "loss": 0.7432,
+      "step": 25583
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9634839590050808,
+      "learning_rate": 7.196521320003346e-06,
+      "loss": 0.7338,
+      "step": 25584
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.371090779910919,
+      "learning_rate": 7.195788886218987e-06,
+      "loss": 0.7596,
+      "step": 25585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7319416592126236,
+      "learning_rate": 7.195056468761637e-06,
+      "loss": 0.81,
+      "step": 25586
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7829795854592247,
+      "learning_rate": 7.194324067635556e-06,
+      "loss": 0.8301,
+      "step": 25587
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.114801378530479,
+      "learning_rate": 7.193591682845006e-06,
+      "loss": 0.8012,
+      "step": 25588
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7999809007012326,
+      "learning_rate": 7.192859314394255e-06,
+      "loss": 0.7059,
+      "step": 25589
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1511307088351583,
+      "learning_rate": 7.1921269622875655e-06,
+      "loss": 0.9191,
+      "step": 25590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5614000950371871,
+      "learning_rate": 7.191394626529204e-06,
+      "loss": 0.8162,
+      "step": 25591
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7303234156637513,
+      "learning_rate": 7.190662307123428e-06,
+      "loss": 0.8539,
+      "step": 25592
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6199406791185598,
+      "learning_rate": 7.189930004074509e-06,
+      "loss": 0.784,
+      "step": 25593
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5822483481316894,
+      "learning_rate": 7.189197717386709e-06,
+      "loss": 0.7119,
+      "step": 25594
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5081147205034047,
+      "learning_rate": 7.188465447064286e-06,
+      "loss": 0.8951,
+      "step": 25595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.432814268926538,
+      "learning_rate": 7.187733193111505e-06,
+      "loss": 0.8654,
+      "step": 25596
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4612162292341537,
+      "learning_rate": 7.187000955532636e-06,
+      "loss": 0.8186,
+      "step": 25597
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6487284159708284,
+      "learning_rate": 7.186268734331938e-06,
+      "loss": 0.8718,
+      "step": 25598
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7332861791954117,
+      "learning_rate": 7.1855365295136746e-06,
+      "loss": 0.7726,
+      "step": 25599
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6824411775573256,
+      "learning_rate": 7.184804341082104e-06,
+      "loss": 0.8044,
+      "step": 25600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5486602189347214,
+      "learning_rate": 7.184072169041499e-06,
+      "loss": 0.769,
+      "step": 25601
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5583778251347826,
+      "learning_rate": 7.183340013396116e-06,
+      "loss": 0.9531,
+      "step": 25602
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.746509447814473,
+      "learning_rate": 7.182607874150218e-06,
+      "loss": 0.7799,
+      "step": 25603
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5951536585820685,
+      "learning_rate": 7.181875751308067e-06,
+      "loss": 0.9626,
+      "step": 25604
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4521654195336084,
+      "learning_rate": 7.181143644873932e-06,
+      "loss": 0.9144,
+      "step": 25605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5737271086016027,
+      "learning_rate": 7.1804115548520695e-06,
+      "loss": 0.9563,
+      "step": 25606
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6874030245612284,
+      "learning_rate": 7.179679481246743e-06,
+      "loss": 0.748,
+      "step": 25607
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4196191918946455,
+      "learning_rate": 7.178947424062217e-06,
+      "loss": 0.8484,
+      "step": 25608
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.683613420662507,
+      "learning_rate": 7.178215383302749e-06,
+      "loss": 0.846,
+      "step": 25609
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8087520688048474,
+      "learning_rate": 7.177483358972607e-06,
+      "loss": 0.7007,
+      "step": 25610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5028700332594978,
+      "learning_rate": 7.17675135107605e-06,
+      "loss": 0.8975,
+      "step": 25611
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6772609767468665,
+      "learning_rate": 7.176019359617341e-06,
+      "loss": 0.9075,
+      "step": 25612
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4174803848558954,
+      "learning_rate": 7.17528738460074e-06,
+      "loss": 0.7341,
+      "step": 25613
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8219259184757695,
+      "learning_rate": 7.17455542603051e-06,
+      "loss": 0.8083,
+      "step": 25614
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5139418004843253,
+      "learning_rate": 7.173823483910916e-06,
+      "loss": 0.8158,
+      "step": 25615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.742688746172299,
+      "learning_rate": 7.173091558246214e-06,
+      "loss": 1.0036,
+      "step": 25616
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.586215050947139,
+      "learning_rate": 7.172359649040666e-06,
+      "loss": 0.7403,
+      "step": 25617
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5560931065278365,
+      "learning_rate": 7.171627756298536e-06,
+      "loss": 0.9068,
+      "step": 25618
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6744930793495445,
+      "learning_rate": 7.170895880024088e-06,
+      "loss": 0.8503,
+      "step": 25619
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7080518598326855,
+      "learning_rate": 7.170164020221577e-06,
+      "loss": 0.8746,
+      "step": 25620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5640812779361275,
+      "learning_rate": 7.169432176895266e-06,
+      "loss": 0.899,
+      "step": 25621
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6782392263461545,
+      "learning_rate": 7.168700350049421e-06,
+      "loss": 0.897,
+      "step": 25622
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.751525494799163,
+      "learning_rate": 7.1679685396882966e-06,
+      "loss": 0.7645,
+      "step": 25623
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.485242998782556,
+      "learning_rate": 7.167236745816154e-06,
+      "loss": 0.897,
+      "step": 25624
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.427420607377171,
+      "learning_rate": 7.166504968437258e-06,
+      "loss": 0.8781,
+      "step": 25625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4839077923432524,
+      "learning_rate": 7.165773207555869e-06,
+      "loss": 0.8115,
+      "step": 25626
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7557708424400584,
+      "learning_rate": 7.165041463176243e-06,
+      "loss": 0.7569,
+      "step": 25627
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8855481796497713,
+      "learning_rate": 7.164309735302643e-06,
+      "loss": 0.7531,
+      "step": 25628
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4158779500345928,
+      "learning_rate": 7.163578023939335e-06,
+      "loss": 0.7113,
+      "step": 25629
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5198376350218783,
+      "learning_rate": 7.162846329090566e-06,
+      "loss": 0.8581,
+      "step": 25630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6016194011084937,
+      "learning_rate": 7.162114650760608e-06,
+      "loss": 0.8573,
+      "step": 25631
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5782235534610067,
+      "learning_rate": 7.161382988953714e-06,
+      "loss": 0.9299,
+      "step": 25632
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6629919070621475,
+      "learning_rate": 7.160651343674152e-06,
+      "loss": 0.9838,
+      "step": 25633
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7101276992157919,
+      "learning_rate": 7.159919714926171e-06,
+      "loss": 0.9161,
+      "step": 25634
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.986416181291756,
+      "learning_rate": 7.15918810271404e-06,
+      "loss": 0.7981,
+      "step": 25635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8830157906529013,
+      "learning_rate": 7.158456507042016e-06,
+      "loss": 0.9377,
+      "step": 25636
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.172454864689468,
+      "learning_rate": 7.157724927914355e-06,
+      "loss": 0.8239,
+      "step": 25637
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.779920488348297,
+      "learning_rate": 7.156993365335321e-06,
+      "loss": 0.8158,
+      "step": 25638
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5604143311142267,
+      "learning_rate": 7.156261819309171e-06,
+      "loss": 0.8434,
+      "step": 25639
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9900922265186787,
+      "learning_rate": 7.1555302898401665e-06,
+      "loss": 0.8562,
+      "step": 25640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.418048664481757,
+      "learning_rate": 7.154798776932563e-06,
+      "loss": 0.8244,
+      "step": 25641
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.774442638596583,
+      "learning_rate": 7.154067280590622e-06,
+      "loss": 0.8004,
+      "step": 25642
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.661967128767087,
+      "learning_rate": 7.153335800818604e-06,
+      "loss": 0.761,
+      "step": 25643
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4915603335792604,
+      "learning_rate": 7.1526043376207675e-06,
+      "loss": 0.7696,
+      "step": 25644
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5345137671626081,
+      "learning_rate": 7.1518728910013655e-06,
+      "loss": 0.8047,
+      "step": 25645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8094151720397427,
+      "learning_rate": 7.151141460964663e-06,
+      "loss": 0.8052,
+      "step": 25646
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.751295261622001,
+      "learning_rate": 7.15041004751492e-06,
+      "loss": 0.8772,
+      "step": 25647
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7469354835869432,
+      "learning_rate": 7.1496786506563885e-06,
+      "loss": 0.7327,
+      "step": 25648
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.514038393384599,
+      "learning_rate": 7.148947270393329e-06,
+      "loss": 0.8502,
+      "step": 25649
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4474784359828432,
+      "learning_rate": 7.148215906730004e-06,
+      "loss": 0.8563,
+      "step": 25650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.624589893833809,
+      "learning_rate": 7.14748455967067e-06,
+      "loss": 0.8308,
+      "step": 25651
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.62373741062469,
+      "learning_rate": 7.146753229219583e-06,
+      "loss": 0.884,
+      "step": 25652
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7587462972444113,
+      "learning_rate": 7.146021915380999e-06,
+      "loss": 0.709,
+      "step": 25653
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.891403365072674,
+      "learning_rate": 7.145290618159182e-06,
+      "loss": 0.8783,
+      "step": 25654
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6586712485279733,
+      "learning_rate": 7.1445593375583865e-06,
+      "loss": 0.8247,
+      "step": 25655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8480922738829355,
+      "learning_rate": 7.14382807358287e-06,
+      "loss": 0.7531,
+      "step": 25656
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5989705748553305,
+      "learning_rate": 7.143096826236888e-06,
+      "loss": 0.7832,
+      "step": 25657
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3923683421214457,
+      "learning_rate": 7.142365595524705e-06,
+      "loss": 0.846,
+      "step": 25658
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4814338267936626,
+      "learning_rate": 7.141634381450573e-06,
+      "loss": 1.0251,
+      "step": 25659
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.831886535358924,
+      "learning_rate": 7.14090318401875e-06,
+      "loss": 0.869,
+      "step": 25660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9242876450168616,
+      "learning_rate": 7.1401720032334965e-06,
+      "loss": 0.7296,
+      "step": 25661
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.618951969191145,
+      "learning_rate": 7.139440839099063e-06,
+      "loss": 0.8293,
+      "step": 25662
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9737878460844391,
+      "learning_rate": 7.138709691619712e-06,
+      "loss": 0.7988,
+      "step": 25663
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4637836660075136,
+      "learning_rate": 7.137978560799697e-06,
+      "loss": 0.9805,
+      "step": 25664
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9744333023988003,
+      "learning_rate": 7.137247446643283e-06,
+      "loss": 0.8105,
+      "step": 25665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7851285409500884,
+      "learning_rate": 7.136516349154714e-06,
+      "loss": 0.7026,
+      "step": 25666
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.599881728647707,
+      "learning_rate": 7.1357852683382565e-06,
+      "loss": 0.7493,
+      "step": 25667
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4957011487747074,
+      "learning_rate": 7.135054204198166e-06,
+      "loss": 0.8346,
+      "step": 25668
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6107984394900747,
+      "learning_rate": 7.134323156738694e-06,
+      "loss": 0.8267,
+      "step": 25669
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.176662995088297,
+      "learning_rate": 7.1335921259640996e-06,
+      "loss": 0.9686,
+      "step": 25670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.82687520342745,
+      "learning_rate": 7.132861111878639e-06,
+      "loss": 0.8753,
+      "step": 25671
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0300844501996105,
+      "learning_rate": 7.132130114486573e-06,
+      "loss": 0.7912,
+      "step": 25672
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6996667689913851,
+      "learning_rate": 7.131399133792147e-06,
+      "loss": 0.8994,
+      "step": 25673
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6835055589880044,
+      "learning_rate": 7.130668169799627e-06,
+      "loss": 0.9024,
+      "step": 25674
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.387265873368753,
+      "learning_rate": 7.129937222513268e-06,
+      "loss": 0.8328,
+      "step": 25675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.59618626580672,
+      "learning_rate": 7.129206291937319e-06,
+      "loss": 0.8864,
+      "step": 25676
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.727057536025423,
+      "learning_rate": 7.128475378076039e-06,
+      "loss": 0.821,
+      "step": 25677
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.65770877691656,
+      "learning_rate": 7.127744480933687e-06,
+      "loss": 0.8506,
+      "step": 25678
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.049732839188194,
+      "learning_rate": 7.127013600514517e-06,
+      "loss": 0.8584,
+      "step": 25679
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0293582633274982,
+      "learning_rate": 7.126282736822782e-06,
+      "loss": 0.809,
+      "step": 25680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.576681936142549,
+      "learning_rate": 7.125551889862736e-06,
+      "loss": 0.8803,
+      "step": 25681
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.783097044966493,
+      "learning_rate": 7.1248210596386444e-06,
+      "loss": 0.8746,
+      "step": 25682
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4770553083192792,
+      "learning_rate": 7.1240902461547475e-06,
+      "loss": 0.7875,
+      "step": 25683
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8320126298083226,
+      "learning_rate": 7.1233594494153104e-06,
+      "loss": 0.8788,
+      "step": 25684
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6699872571048533,
+      "learning_rate": 7.122628669424583e-06,
+      "loss": 0.8173,
+      "step": 25685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7208480713867491,
+      "learning_rate": 7.121897906186827e-06,
+      "loss": 0.7807,
+      "step": 25686
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4318051243688694,
+      "learning_rate": 7.12116715970629e-06,
+      "loss": 0.8355,
+      "step": 25687
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5882113076163238,
+      "learning_rate": 7.120436429987228e-06,
+      "loss": 0.9609,
+      "step": 25688
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5621192270726851,
+      "learning_rate": 7.1197057170339e-06,
+      "loss": 0.8561,
+      "step": 25689
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7669707309289668,
+      "learning_rate": 7.118975020850554e-06,
+      "loss": 0.9487,
+      "step": 25690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.55310311163952,
+      "learning_rate": 7.1182443414414495e-06,
+      "loss": 0.832,
+      "step": 25691
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7417392567031553,
+      "learning_rate": 7.117513678810837e-06,
+      "loss": 0.9258,
+      "step": 25692
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6800166606316993,
+      "learning_rate": 7.116783032962976e-06,
+      "loss": 0.8648,
+      "step": 25693
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6078577352372043,
+      "learning_rate": 7.116052403902112e-06,
+      "loss": 0.8154,
+      "step": 25694
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8122972297777096,
+      "learning_rate": 7.115321791632507e-06,
+      "loss": 0.87,
+      "step": 25695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6951408678746873,
+      "learning_rate": 7.114591196158413e-06,
+      "loss": 0.9779,
+      "step": 25696
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5087887377607434,
+      "learning_rate": 7.1138606174840806e-06,
+      "loss": 0.7956,
+      "step": 25697
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7086768185060202,
+      "learning_rate": 7.113130055613764e-06,
+      "loss": 0.8412,
+      "step": 25698
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5005101619691377,
+      "learning_rate": 7.11239951055172e-06,
+      "loss": 0.8447,
+      "step": 25699
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6523860391480563,
+      "learning_rate": 7.111668982302202e-06,
+      "loss": 0.6452,
+      "step": 25700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.588170742538371,
+      "learning_rate": 7.11093847086946e-06,
+      "loss": 0.809,
+      "step": 25701
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6340472454309494,
+      "learning_rate": 7.110207976257746e-06,
+      "loss": 0.9288,
+      "step": 25702
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6228138046782337,
+      "learning_rate": 7.10947749847132e-06,
+      "loss": 0.8982,
+      "step": 25703
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7342981325192437,
+      "learning_rate": 7.108747037514431e-06,
+      "loss": 0.8492,
+      "step": 25704
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5294849369678603,
+      "learning_rate": 7.10801659339133e-06,
+      "loss": 0.9057,
+      "step": 25705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.510302880008347,
+      "learning_rate": 7.107286166106271e-06,
+      "loss": 0.8585,
+      "step": 25706
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.516147964728888,
+      "learning_rate": 7.106555755663511e-06,
+      "loss": 0.9262,
+      "step": 25707
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7178418544929872,
+      "learning_rate": 7.105825362067299e-06,
+      "loss": 0.8958,
+      "step": 25708
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7061987492217425,
+      "learning_rate": 7.105094985321888e-06,
+      "loss": 0.8214,
+      "step": 25709
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.532787177785542,
+      "learning_rate": 7.10436462543153e-06,
+      "loss": 0.8725,
+      "step": 25710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.669158112365062,
+      "learning_rate": 7.103634282400475e-06,
+      "loss": 0.7829,
+      "step": 25711
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6697688867805014,
+      "learning_rate": 7.102903956232981e-06,
+      "loss": 0.7697,
+      "step": 25712
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6572813615318809,
+      "learning_rate": 7.102173646933295e-06,
+      "loss": 0.8604,
+      "step": 25713
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7200521361172085,
+      "learning_rate": 7.1014433545056785e-06,
+      "loss": 0.686,
+      "step": 25714
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5431915691767921,
+      "learning_rate": 7.100713078954369e-06,
+      "loss": 0.9817,
+      "step": 25715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5056429990392275,
+      "learning_rate": 7.099982820283629e-06,
+      "loss": 0.8588,
+      "step": 25716
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7974310707743368,
+      "learning_rate": 7.099252578497709e-06,
+      "loss": 0.7868,
+      "step": 25717
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.716804601209265,
+      "learning_rate": 7.0985223536008555e-06,
+      "loss": 0.7855,
+      "step": 25718
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.076789030302613,
+      "learning_rate": 7.097792145597322e-06,
+      "loss": 0.8635,
+      "step": 25719
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3688721465627303,
+      "learning_rate": 7.097061954491363e-06,
+      "loss": 0.9177,
+      "step": 25720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7356402997171255,
+      "learning_rate": 7.096331780287231e-06,
+      "loss": 0.6833,
+      "step": 25721
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5285336449910132,
+      "learning_rate": 7.0956016229891715e-06,
+      "loss": 1.1076,
+      "step": 25722
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6685830579910743,
+      "learning_rate": 7.09487148260144e-06,
+      "loss": 0.9296,
+      "step": 25723
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.499295272122565,
+      "learning_rate": 7.094141359128288e-06,
+      "loss": 0.8441,
+      "step": 25724
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.647908747583549,
+      "learning_rate": 7.093411252573964e-06,
+      "loss": 0.9768,
+      "step": 25725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7500101974868196,
+      "learning_rate": 7.092681162942717e-06,
+      "loss": 0.8577,
+      "step": 25726
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.505855566559885,
+      "learning_rate": 7.0919510902388034e-06,
+      "loss": 0.8848,
+      "step": 25727
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4647081206236416,
+      "learning_rate": 7.0912210344664736e-06,
+      "loss": 0.854,
+      "step": 25728
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6119537979663139,
+      "learning_rate": 7.090490995629973e-06,
+      "loss": 0.8203,
+      "step": 25729
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7243161635685762,
+      "learning_rate": 7.089760973733553e-06,
+      "loss": 0.8512,
+      "step": 25730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6674120336036264,
+      "learning_rate": 7.089030968781472e-06,
+      "loss": 1.0536,
+      "step": 25731
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5159070381840571,
+      "learning_rate": 7.088300980777969e-06,
+      "loss": 0.8642,
+      "step": 25732
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9211762728828512,
+      "learning_rate": 7.087571009727301e-06,
+      "loss": 0.8952,
+      "step": 25733
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.50580191572197,
+      "learning_rate": 7.086841055633713e-06,
+      "loss": 0.7861,
+      "step": 25734
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.645096641770948,
+      "learning_rate": 7.086111118501465e-06,
+      "loss": 0.7598,
+      "step": 25735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7935451774729034,
+      "learning_rate": 7.085381198334797e-06,
+      "loss": 0.7896,
+      "step": 25736
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.695146967983208,
+      "learning_rate": 7.084651295137964e-06,
+      "loss": 0.9939,
+      "step": 25737
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5350570934094663,
+      "learning_rate": 7.0839214089152155e-06,
+      "loss": 0.8212,
+      "step": 25738
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7344380760053335,
+      "learning_rate": 7.083191539670795e-06,
+      "loss": 0.8186,
+      "step": 25739
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6852301135954468,
+      "learning_rate": 7.08246168740896e-06,
+      "loss": 0.8472,
+      "step": 25740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.606743129666279,
+      "learning_rate": 7.081731852133958e-06,
+      "loss": 0.8301,
+      "step": 25741
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5725467575666345,
+      "learning_rate": 7.081002033850036e-06,
+      "loss": 0.8896,
+      "step": 25742
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.844517154827323,
+      "learning_rate": 7.080272232561441e-06,
+      "loss": 0.8935,
+      "step": 25743
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.737571387260418,
+      "learning_rate": 7.079542448272428e-06,
+      "loss": 0.8618,
+      "step": 25744
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.615229241857133,
+      "learning_rate": 7.078812680987246e-06,
+      "loss": 0.924,
+      "step": 25745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7865838103646269,
+      "learning_rate": 7.0780829307101395e-06,
+      "loss": 0.7279,
+      "step": 25746
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.845730041094457,
+      "learning_rate": 7.077353197445356e-06,
+      "loss": 0.8741,
+      "step": 25747
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9348677498940758,
+      "learning_rate": 7.07662348119715e-06,
+      "loss": 0.9073,
+      "step": 25748
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.626804912307557,
+      "learning_rate": 7.075893781969769e-06,
+      "loss": 0.8866,
+      "step": 25749
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.547814876496947,
+      "learning_rate": 7.07516409976746e-06,
+      "loss": 0.8469,
+      "step": 25750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0643821284725394,
+      "learning_rate": 7.074434434594469e-06,
+      "loss": 0.8776,
+      "step": 25751
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7398811036840447,
+      "learning_rate": 7.073704786455049e-06,
+      "loss": 0.688,
+      "step": 25752
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6016465012760999,
+      "learning_rate": 7.072975155353446e-06,
+      "loss": 0.8919,
+      "step": 25753
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6236306002175735,
+      "learning_rate": 7.072245541293908e-06,
+      "loss": 0.9058,
+      "step": 25754
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6711369771076354,
+      "learning_rate": 7.0715159442806815e-06,
+      "loss": 0.7931,
+      "step": 25755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.14850167024212,
+      "learning_rate": 7.0707863643180204e-06,
+      "loss": 0.8505,
+      "step": 25756
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4655866011712848,
+      "learning_rate": 7.070056801410165e-06,
+      "loss": 0.7106,
+      "step": 25757
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1457371086484307,
+      "learning_rate": 7.069327255561369e-06,
+      "loss": 0.8782,
+      "step": 25758
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9096260695938385,
+      "learning_rate": 7.068597726775879e-06,
+      "loss": 0.9256,
+      "step": 25759
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7768189263209315,
+      "learning_rate": 7.0678682150579355e-06,
+      "loss": 0.8465,
+      "step": 25760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6234970191369904,
+      "learning_rate": 7.067138720411795e-06,
+      "loss": 0.8922,
+      "step": 25761
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8310447057451928,
+      "learning_rate": 7.0664092428416976e-06,
+      "loss": 0.7917,
+      "step": 25762
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.771196154206886,
+      "learning_rate": 7.065679782351902e-06,
+      "loss": 0.7733,
+      "step": 25763
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7535303109267952,
+      "learning_rate": 7.064950338946641e-06,
+      "loss": 0.8581,
+      "step": 25764
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7720003907360762,
+      "learning_rate": 7.064220912630171e-06,
+      "loss": 0.9093,
+      "step": 25765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.677863646544194,
+      "learning_rate": 7.063491503406737e-06,
+      "loss": 0.8682,
+      "step": 25766
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5078706786277842,
+      "learning_rate": 7.062762111280585e-06,
+      "loss": 0.76,
+      "step": 25767
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.658429416232347,
+      "learning_rate": 7.062032736255959e-06,
+      "loss": 0.9593,
+      "step": 25768
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6675858924724027,
+      "learning_rate": 7.06130337833711e-06,
+      "loss": 0.7952,
+      "step": 25769
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5000636980266182,
+      "learning_rate": 7.060574037528285e-06,
+      "loss": 0.9393,
+      "step": 25770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5438057387297732,
+      "learning_rate": 7.059844713833726e-06,
+      "loss": 0.7649,
+      "step": 25771
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.185591674182868,
+      "learning_rate": 7.059115407257683e-06,
+      "loss": 0.9859,
+      "step": 25772
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6800392967289834,
+      "learning_rate": 7.058386117804401e-06,
+      "loss": 0.8245,
+      "step": 25773
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8123959042791108,
+      "learning_rate": 7.057656845478129e-06,
+      "loss": 0.7224,
+      "step": 25774
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4761082885531618,
+      "learning_rate": 7.056927590283104e-06,
+      "loss": 0.8707,
+      "step": 25775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4756645017661127,
+      "learning_rate": 7.056198352223585e-06,
+      "loss": 0.8714,
+      "step": 25776
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.591233390243006,
+      "learning_rate": 7.055469131303809e-06,
+      "loss": 0.8306,
+      "step": 25777
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.540220167915704,
+      "learning_rate": 7.0547399275280235e-06,
+      "loss": 0.9394,
+      "step": 25778
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5220530869309437,
+      "learning_rate": 7.054010740900472e-06,
+      "loss": 0.8501,
+      "step": 25779
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7432553200816403,
+      "learning_rate": 7.053281571425405e-06,
+      "loss": 0.8776,
+      "step": 25780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.533311275123345,
+      "learning_rate": 7.052552419107068e-06,
+      "loss": 0.9692,
+      "step": 25781
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8957098188894712,
+      "learning_rate": 7.051823283949702e-06,
+      "loss": 0.9497,
+      "step": 25782
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8030591615889006,
+      "learning_rate": 7.051094165957553e-06,
+      "loss": 0.7151,
+      "step": 25783
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8000466915724829,
+      "learning_rate": 7.05036506513487e-06,
+      "loss": 0.898,
+      "step": 25784
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5586841408477352,
+      "learning_rate": 7.049635981485894e-06,
+      "loss": 0.7382,
+      "step": 25785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4943417570171025,
+      "learning_rate": 7.048906915014871e-06,
+      "loss": 0.8566,
+      "step": 25786
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6639052160224184,
+      "learning_rate": 7.048177865726044e-06,
+      "loss": 0.8176,
+      "step": 25787
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7641720764385205,
+      "learning_rate": 7.0474488336236645e-06,
+      "loss": 0.6892,
+      "step": 25788
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.557410561599667,
+      "learning_rate": 7.0467198187119705e-06,
+      "loss": 0.8244,
+      "step": 25789
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6687646689992754,
+      "learning_rate": 7.045990820995208e-06,
+      "loss": 0.8629,
+      "step": 25790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7775492370223844,
+      "learning_rate": 7.0452618404776255e-06,
+      "loss": 0.8915,
+      "step": 25791
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3915832096905747,
+      "learning_rate": 7.044532877163459e-06,
+      "loss": 0.8089,
+      "step": 25792
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6173269042886336,
+      "learning_rate": 7.043803931056961e-06,
+      "loss": 0.6911,
+      "step": 25793
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.687817476938399,
+      "learning_rate": 7.04307500216237e-06,
+      "loss": 0.9287,
+      "step": 25794
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.579865726186852,
+      "learning_rate": 7.042346090483937e-06,
+      "loss": 0.7076,
+      "step": 25795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4137334350784636,
+      "learning_rate": 7.041617196025896e-06,
+      "loss": 0.9194,
+      "step": 25796
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6218394180312783,
+      "learning_rate": 7.040888318792499e-06,
+      "loss": 0.8687,
+      "step": 25797
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6258951272167632,
+      "learning_rate": 7.04015945878799e-06,
+      "loss": 0.9367,
+      "step": 25798
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7881391519042866,
+      "learning_rate": 7.039430616016605e-06,
+      "loss": 0.8512,
+      "step": 25799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.778107258283721,
+      "learning_rate": 7.038701790482592e-06,
+      "loss": 0.7264,
+      "step": 25800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7809937374401618,
+      "learning_rate": 7.037972982190197e-06,
+      "loss": 0.7084,
+      "step": 25801
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7337719338278823,
+      "learning_rate": 7.037244191143662e-06,
+      "loss": 0.8933,
+      "step": 25802
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6121860706697886,
+      "learning_rate": 7.036515417347228e-06,
+      "loss": 0.7176,
+      "step": 25803
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.329690459389721,
+      "learning_rate": 7.035786660805137e-06,
+      "loss": 0.9383,
+      "step": 25804
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.845023389313632,
+      "learning_rate": 7.03505792152164e-06,
+      "loss": 0.8574,
+      "step": 25805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8802220680604411,
+      "learning_rate": 7.034329199500971e-06,
+      "loss": 0.9245,
+      "step": 25806
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4086133602866773,
+      "learning_rate": 7.033600494747377e-06,
+      "loss": 0.9508,
+      "step": 25807
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7322557166989552,
+      "learning_rate": 7.032871807265097e-06,
+      "loss": 0.9406,
+      "step": 25808
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6844101343675888,
+      "learning_rate": 7.032143137058382e-06,
+      "loss": 0.8902,
+      "step": 25809
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5876724389747114,
+      "learning_rate": 7.031414484131467e-06,
+      "loss": 0.893,
+      "step": 25810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9253807018253206,
+      "learning_rate": 7.030685848488595e-06,
+      "loss": 0.8541,
+      "step": 25811
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5702092825100242,
+      "learning_rate": 7.029957230134015e-06,
+      "loss": 0.7225,
+      "step": 25812
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96696034158074,
+      "learning_rate": 7.029228629071958e-06,
+      "loss": 0.8812,
+      "step": 25813
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4463463934421132,
+      "learning_rate": 7.028500045306676e-06,
+      "loss": 0.8968,
+      "step": 25814
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.482420224778918,
+      "learning_rate": 7.027771478842405e-06,
+      "loss": 0.8216,
+      "step": 25815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6051831298806587,
+      "learning_rate": 7.027042929683394e-06,
+      "loss": 0.775,
+      "step": 25816
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7568112644495468,
+      "learning_rate": 7.0263143978338756e-06,
+      "loss": 0.8445,
+      "step": 25817
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5658594526784733,
+      "learning_rate": 7.025585883298099e-06,
+      "loss": 0.7934,
+      "step": 25818
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5487488401305516,
+      "learning_rate": 7.0248573860803035e-06,
+      "loss": 0.874,
+      "step": 25819
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0048240799141137,
+      "learning_rate": 7.024128906184726e-06,
+      "loss": 0.8619,
+      "step": 25820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.883709832943376,
+      "learning_rate": 7.0234004436156155e-06,
+      "loss": 0.7892,
+      "step": 25821
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7933257266852961,
+      "learning_rate": 7.02267199837721e-06,
+      "loss": 0.665,
+      "step": 25822
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8501479341375886,
+      "learning_rate": 7.021943570473751e-06,
+      "loss": 0.7956,
+      "step": 25823
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.549576575267109,
+      "learning_rate": 7.021215159909478e-06,
+      "loss": 0.9331,
+      "step": 25824
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8907115554774316,
+      "learning_rate": 7.020486766688633e-06,
+      "loss": 0.7859,
+      "step": 25825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.480912343364081,
+      "learning_rate": 7.019758390815461e-06,
+      "loss": 0.8569,
+      "step": 25826
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6922609693756425,
+      "learning_rate": 7.019030032294196e-06,
+      "loss": 0.9697,
+      "step": 25827
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5000565860535013,
+      "learning_rate": 7.0183016911290816e-06,
+      "loss": 0.8517,
+      "step": 25828
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4248104925679592,
+      "learning_rate": 7.0175733673243586e-06,
+      "loss": 0.8906,
+      "step": 25829
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.530742345273625,
+      "learning_rate": 7.016845060884271e-06,
+      "loss": 0.7217,
+      "step": 25830
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.738825728332619,
+      "learning_rate": 7.016116771813053e-06,
+      "loss": 0.8834,
+      "step": 25831
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6607608524140285,
+      "learning_rate": 7.015388500114947e-06,
+      "loss": 0.9622,
+      "step": 25832
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6834171695685811,
+      "learning_rate": 7.014660245794197e-06,
+      "loss": 0.977,
+      "step": 25833
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7808667169095553,
+      "learning_rate": 7.013932008855039e-06,
+      "loss": 0.9696,
+      "step": 25834
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5586861427517726,
+      "learning_rate": 7.013203789301713e-06,
+      "loss": 0.8306,
+      "step": 25835
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.863617280792671,
+      "learning_rate": 7.01247558713846e-06,
+      "loss": 0.8207,
+      "step": 25836
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.093427156937527,
+      "learning_rate": 7.011747402369522e-06,
+      "loss": 0.8617,
+      "step": 25837
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.61805207561616,
+      "learning_rate": 7.011019234999137e-06,
+      "loss": 0.77,
+      "step": 25838
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5899145124104839,
+      "learning_rate": 7.010291085031542e-06,
+      "loss": 0.9488,
+      "step": 25839
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.682722343637237,
+      "learning_rate": 7.0095629524709826e-06,
+      "loss": 0.9264,
+      "step": 25840
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8241107509244918,
+      "learning_rate": 7.008834837321689e-06,
+      "loss": 0.7439,
+      "step": 25841
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4543031028203206,
+      "learning_rate": 7.0081067395879085e-06,
+      "loss": 1.0044,
+      "step": 25842
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7038203980122006,
+      "learning_rate": 7.007378659273878e-06,
+      "loss": 0.8323,
+      "step": 25843
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.623199908036595,
+      "learning_rate": 7.006650596383839e-06,
+      "loss": 0.7132,
+      "step": 25844
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.00711104235773,
+      "learning_rate": 7.005922550922023e-06,
+      "loss": 0.8216,
+      "step": 25845
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4914147681756849,
+      "learning_rate": 7.005194522892676e-06,
+      "loss": 0.9212,
+      "step": 25846
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.516713151665285,
+      "learning_rate": 7.004466512300037e-06,
+      "loss": 0.8379,
+      "step": 25847
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7461487736099002,
+      "learning_rate": 7.00373851914834e-06,
+      "loss": 0.7002,
+      "step": 25848
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5919266600115116,
+      "learning_rate": 7.003010543441824e-06,
+      "loss": 0.8628,
+      "step": 25849
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9195953164606727,
+      "learning_rate": 7.002282585184731e-06,
+      "loss": 0.8752,
+      "step": 25850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.611991431248621,
+      "learning_rate": 7.0015546443813e-06,
+      "loss": 0.897,
+      "step": 25851
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4619831192802406,
+      "learning_rate": 7.000826721035765e-06,
+      "loss": 0.753,
+      "step": 25852
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7358912636555752,
+      "learning_rate": 7.000098815152365e-06,
+      "loss": 0.7624,
+      "step": 25853
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.841573678532922,
+      "learning_rate": 6.999370926735342e-06,
+      "loss": 0.6916,
+      "step": 25854
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.130895189810676,
+      "learning_rate": 6.998643055788932e-06,
+      "loss": 0.7396,
+      "step": 25855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8458141536816313,
+      "learning_rate": 6.997915202317372e-06,
+      "loss": 0.7946,
+      "step": 25856
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4872237885296518,
+      "learning_rate": 6.997187366324895e-06,
+      "loss": 0.8267,
+      "step": 25857
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.63419231030308,
+      "learning_rate": 6.99645954781575e-06,
+      "loss": 0.8512,
+      "step": 25858
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.574998867027941,
+      "learning_rate": 6.9957317467941675e-06,
+      "loss": 0.8053,
+      "step": 25859
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.39325634031009,
+      "learning_rate": 6.995003963264382e-06,
+      "loss": 0.8374,
+      "step": 25860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4956558834275615,
+      "learning_rate": 6.994276197230643e-06,
+      "loss": 0.7806,
+      "step": 25861
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6883639610898762,
+      "learning_rate": 6.993548448697171e-06,
+      "loss": 0.9185,
+      "step": 25862
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5121363857206405,
+      "learning_rate": 6.992820717668215e-06,
+      "loss": 0.8266,
+      "step": 25863
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9028640588794121,
+      "learning_rate": 6.992093004148006e-06,
+      "loss": 0.7536,
+      "step": 25864
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7920115991021808,
+      "learning_rate": 6.991365308140792e-06,
+      "loss": 0.7633,
+      "step": 25865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5553716303421534,
+      "learning_rate": 6.990637629650794e-06,
+      "loss": 0.7752,
+      "step": 25866
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.543555325890276,
+      "learning_rate": 6.989909968682258e-06,
+      "loss": 0.902,
+      "step": 25867
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5476112279357723,
+      "learning_rate": 6.989182325239423e-06,
+      "loss": 0.8311,
+      "step": 25868
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7864460840191336,
+      "learning_rate": 6.988454699326515e-06,
+      "loss": 0.9054,
+      "step": 25869
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0391159841049467,
+      "learning_rate": 6.987727090947781e-06,
+      "loss": 0.7901,
+      "step": 25870
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.664499406306401,
+      "learning_rate": 6.986999500107454e-06,
+      "loss": 0.902,
+      "step": 25871
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6901000234882377,
+      "learning_rate": 6.9862719268097715e-06,
+      "loss": 0.9397,
+      "step": 25872
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6821613540481934,
+      "learning_rate": 6.985544371058963e-06,
+      "loss": 0.7787,
+      "step": 25873
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5730739171602945,
+      "learning_rate": 6.984816832859271e-06,
+      "loss": 0.834,
+      "step": 25874
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.504043292522253,
+      "learning_rate": 6.9840893122149346e-06,
+      "loss": 0.8734,
+      "step": 25875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.156522119334764,
+      "learning_rate": 6.98336180913018e-06,
+      "loss": 0.8946,
+      "step": 25876
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5278233315758705,
+      "learning_rate": 6.982634323609249e-06,
+      "loss": 0.7868,
+      "step": 25877
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.097998801321762,
+      "learning_rate": 6.981906855656376e-06,
+      "loss": 0.9392,
+      "step": 25878
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4730079738207706,
+      "learning_rate": 6.9811794052758e-06,
+      "loss": 0.7858,
+      "step": 25879
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5889890771440667,
+      "learning_rate": 6.980451972471751e-06,
+      "loss": 0.8127,
+      "step": 25880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7348868320545772,
+      "learning_rate": 6.979724557248464e-06,
+      "loss": 0.8256,
+      "step": 25881
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5293846526142636,
+      "learning_rate": 6.978997159610181e-06,
+      "loss": 0.8149,
+      "step": 25882
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6150382466759052,
+      "learning_rate": 6.978269779561131e-06,
+      "loss": 0.8681,
+      "step": 25883
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3602073936889112,
+      "learning_rate": 6.977542417105552e-06,
+      "loss": 0.7697,
+      "step": 25884
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7010190420992888,
+      "learning_rate": 6.9768150722476744e-06,
+      "loss": 0.7033,
+      "step": 25885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.464498287651572,
+      "learning_rate": 6.976087744991742e-06,
+      "loss": 0.825,
+      "step": 25886
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5221626237492567,
+      "learning_rate": 6.975360435341981e-06,
+      "loss": 0.833,
+      "step": 25887
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9257838928082895,
+      "learning_rate": 6.974633143302629e-06,
+      "loss": 0.8387,
+      "step": 25888
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.099598810341227,
+      "learning_rate": 6.973905868877923e-06,
+      "loss": 0.8765,
+      "step": 25889
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4813043592370927,
+      "learning_rate": 6.973178612072092e-06,
+      "loss": 0.8355,
+      "step": 25890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.618141086562881,
+      "learning_rate": 6.972451372889375e-06,
+      "loss": 0.8608,
+      "step": 25891
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5507523273646366,
+      "learning_rate": 6.971724151334005e-06,
+      "loss": 0.8312,
+      "step": 25892
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6053870207427359,
+      "learning_rate": 6.970996947410217e-06,
+      "loss": 1.0281,
+      "step": 25893
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.609676530078336,
+      "learning_rate": 6.9702697611222394e-06,
+      "loss": 0.7608,
+      "step": 25894
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.695799794383521,
+      "learning_rate": 6.9695425924743125e-06,
+      "loss": 0.8965,
+      "step": 25895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5520275341363248,
+      "learning_rate": 6.968815441470666e-06,
+      "loss": 0.8262,
+      "step": 25896
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.563308344675428,
+      "learning_rate": 6.968088308115543e-06,
+      "loss": 0.8821,
+      "step": 25897
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6420119864873979,
+      "learning_rate": 6.967361192413163e-06,
+      "loss": 0.8697,
+      "step": 25898
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.627363465507511,
+      "learning_rate": 6.966634094367768e-06,
+      "loss": 0.9168,
+      "step": 25899
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6000726354723251,
+      "learning_rate": 6.96590701398359e-06,
+      "loss": 0.8641,
+      "step": 25900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7925269181734824,
+      "learning_rate": 6.965179951264862e-06,
+      "loss": 0.6969,
+      "step": 25901
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0620133684557316,
+      "learning_rate": 6.964452906215815e-06,
+      "loss": 0.8577,
+      "step": 25902
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5213415576723373,
+      "learning_rate": 6.963725878840686e-06,
+      "loss": 0.7573,
+      "step": 25903
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.497306314546873,
+      "learning_rate": 6.9629988691437086e-06,
+      "loss": 0.9344,
+      "step": 25904
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6295450751432918,
+      "learning_rate": 6.962271877129111e-06,
+      "loss": 0.8011,
+      "step": 25905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8057917546459166,
+      "learning_rate": 6.961544902801126e-06,
+      "loss": 0.7095,
+      "step": 25906
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.449308229092182,
+      "learning_rate": 6.960817946163993e-06,
+      "loss": 0.7956,
+      "step": 25907
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7019113118595273,
+      "learning_rate": 6.960091007221938e-06,
+      "loss": 0.9019,
+      "step": 25908
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7570855362025437,
+      "learning_rate": 6.959364085979192e-06,
+      "loss": 0.8257,
+      "step": 25909
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7852844367092251,
+      "learning_rate": 6.958637182439996e-06,
+      "loss": 0.7325,
+      "step": 25910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.656491443049231,
+      "learning_rate": 6.957910296608576e-06,
+      "loss": 0.9071,
+      "step": 25911
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6817874578510976,
+      "learning_rate": 6.957183428489165e-06,
+      "loss": 0.8385,
+      "step": 25912
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1787039544740057,
+      "learning_rate": 6.956456578085994e-06,
+      "loss": 0.7203,
+      "step": 25913
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5238422426209317,
+      "learning_rate": 6.955729745403301e-06,
+      "loss": 0.8662,
+      "step": 25914
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5629408769627278,
+      "learning_rate": 6.955002930445308e-06,
+      "loss": 0.9187,
+      "step": 25915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8362065152713964,
+      "learning_rate": 6.954276133216254e-06,
+      "loss": 0.8718,
+      "step": 25916
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8219461704050017,
+      "learning_rate": 6.953549353720366e-06,
+      "loss": 0.7724,
+      "step": 25917
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7913554075751326,
+      "learning_rate": 6.952822591961883e-06,
+      "loss": 0.8933,
+      "step": 25918
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7809172606965422,
+      "learning_rate": 6.9520958479450285e-06,
+      "loss": 0.8475,
+      "step": 25919
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8790253684765956,
+      "learning_rate": 6.9513691216740385e-06,
+      "loss": 0.9415,
+      "step": 25920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0240270183647833,
+      "learning_rate": 6.950642413153144e-06,
+      "loss": 0.8374,
+      "step": 25921
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.597260206306168,
+      "learning_rate": 6.949915722386571e-06,
+      "loss": 0.794,
+      "step": 25922
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.629413847380817,
+      "learning_rate": 6.949189049378555e-06,
+      "loss": 0.8181,
+      "step": 25923
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6055343550280046,
+      "learning_rate": 6.9484623941333274e-06,
+      "loss": 0.9167,
+      "step": 25924
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.336034331295258,
+      "learning_rate": 6.94773575665512e-06,
+      "loss": 0.9364,
+      "step": 25925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.502138264721998,
+      "learning_rate": 6.947009136948157e-06,
+      "loss": 0.8421,
+      "step": 25926
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5920240099752871,
+      "learning_rate": 6.9462825350166744e-06,
+      "loss": 0.909,
+      "step": 25927
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.633948734400826,
+      "learning_rate": 6.945555950864904e-06,
+      "loss": 0.9364,
+      "step": 25928
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6571979455606984,
+      "learning_rate": 6.944829384497073e-06,
+      "loss": 0.9109,
+      "step": 25929
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5437830939524049,
+      "learning_rate": 6.94410283591741e-06,
+      "loss": 0.9106,
+      "step": 25930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.833057917981894,
+      "learning_rate": 6.943376305130149e-06,
+      "loss": 0.7985,
+      "step": 25931
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4310235407559537,
+      "learning_rate": 6.942649792139522e-06,
+      "loss": 0.7156,
+      "step": 25932
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5439568688774779,
+      "learning_rate": 6.941923296949753e-06,
+      "loss": 0.7194,
+      "step": 25933
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.192279266671268,
+      "learning_rate": 6.941196819565074e-06,
+      "loss": 0.8917,
+      "step": 25934
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5900496198825624,
+      "learning_rate": 6.940470359989719e-06,
+      "loss": 0.8049,
+      "step": 25935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5308078995040186,
+      "learning_rate": 6.939743918227911e-06,
+      "loss": 0.9041,
+      "step": 25936
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8254951900226857,
+      "learning_rate": 6.939017494283884e-06,
+      "loss": 0.8371,
+      "step": 25937
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6508432681935727,
+      "learning_rate": 6.938291088161864e-06,
+      "loss": 0.8241,
+      "step": 25938
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7255192194191293,
+      "learning_rate": 6.937564699866086e-06,
+      "loss": 0.8517,
+      "step": 25939
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7366174551619953,
+      "learning_rate": 6.936838329400774e-06,
+      "loss": 0.9235,
+      "step": 25940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3211211154104427,
+      "learning_rate": 6.93611197677016e-06,
+      "loss": 0.9086,
+      "step": 25941
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9882626074913976,
+      "learning_rate": 6.9353856419784734e-06,
+      "loss": 0.9183,
+      "step": 25942
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7452310785869152,
+      "learning_rate": 6.934659325029938e-06,
+      "loss": 0.7889,
+      "step": 25943
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6772877399092538,
+      "learning_rate": 6.933933025928788e-06,
+      "loss": 0.8657,
+      "step": 25944
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5206320653065115,
+      "learning_rate": 6.933206744679249e-06,
+      "loss": 0.9391,
+      "step": 25945
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3885070888320765,
+      "learning_rate": 6.932480481285558e-06,
+      "loss": 0.8383,
+      "step": 25946
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7272921825023266,
+      "learning_rate": 6.931754235751929e-06,
+      "loss": 0.8806,
+      "step": 25947
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5441267867056896,
+      "learning_rate": 6.931028008082602e-06,
+      "loss": 0.9548,
+      "step": 25948
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.49919641058654,
+      "learning_rate": 6.930301798281803e-06,
+      "loss": 0.9374,
+      "step": 25949
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.636748588876532,
+      "learning_rate": 6.929575606353757e-06,
+      "loss": 0.8566,
+      "step": 25950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.585086764199919,
+      "learning_rate": 6.928849432302691e-06,
+      "loss": 0.9076,
+      "step": 25951
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7097538602295992,
+      "learning_rate": 6.9281232761328384e-06,
+      "loss": 0.9029,
+      "step": 25952
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4251718479251354,
+      "learning_rate": 6.927397137848427e-06,
+      "loss": 0.8668,
+      "step": 25953
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7859843252917875,
+      "learning_rate": 6.926671017453678e-06,
+      "loss": 0.8367,
+      "step": 25954
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.023679976095039,
+      "learning_rate": 6.925944914952825e-06,
+      "loss": 0.9154,
+      "step": 25955
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6630520272388778,
+      "learning_rate": 6.925218830350097e-06,
+      "loss": 0.9247,
+      "step": 25956
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5007584700713823,
+      "learning_rate": 6.924492763649715e-06,
+      "loss": 0.8483,
+      "step": 25957
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5730136675253694,
+      "learning_rate": 6.923766714855909e-06,
+      "loss": 0.7997,
+      "step": 25958
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.852593639205927,
+      "learning_rate": 6.923040683972908e-06,
+      "loss": 0.9998,
+      "step": 25959
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7151356342814,
+      "learning_rate": 6.922314671004942e-06,
+      "loss": 0.8861,
+      "step": 25960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.602013164900212,
+      "learning_rate": 6.92158867595623e-06,
+      "loss": 0.9602,
+      "step": 25961
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7188364565640855,
+      "learning_rate": 6.920862698831002e-06,
+      "loss": 0.8953,
+      "step": 25962
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7603334768324551,
+      "learning_rate": 6.920136739633494e-06,
+      "loss": 0.7506,
+      "step": 25963
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.622434506771323,
+      "learning_rate": 6.919410798367917e-06,
+      "loss": 0.8843,
+      "step": 25964
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.789808432654621,
+      "learning_rate": 6.9186848750385085e-06,
+      "loss": 0.8818,
+      "step": 25965
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6235514498911234,
+      "learning_rate": 6.91795896964949e-06,
+      "loss": 0.8053,
+      "step": 25966
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6707357364189386,
+      "learning_rate": 6.917233082205094e-06,
+      "loss": 0.9679,
+      "step": 25967
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6025573582065413,
+      "learning_rate": 6.916507212709541e-06,
+      "loss": 0.9195,
+      "step": 25968
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.428640588822179,
+      "learning_rate": 6.91578136116706e-06,
+      "loss": 0.8863,
+      "step": 25969
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.646349838850563,
+      "learning_rate": 6.915055527581878e-06,
+      "loss": 0.7639,
+      "step": 25970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.59967501275894,
+      "learning_rate": 6.914329711958214e-06,
+      "loss": 0.8708,
+      "step": 25971
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4559210662442759,
+      "learning_rate": 6.913603914300302e-06,
+      "loss": 0.9231,
+      "step": 25972
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.499671068531624,
+      "learning_rate": 6.912878134612367e-06,
+      "loss": 0.9168,
+      "step": 25973
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8625553525915912,
+      "learning_rate": 6.912152372898633e-06,
+      "loss": 0.927,
+      "step": 25974
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7902926792860284,
+      "learning_rate": 6.911426629163323e-06,
+      "loss": 0.6669,
+      "step": 25975
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8140881737290822,
+      "learning_rate": 6.9107009034106655e-06,
+      "loss": 0.8278,
+      "step": 25976
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.459391558539647,
+      "learning_rate": 6.909975195644887e-06,
+      "loss": 0.7626,
+      "step": 25977
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6645326803102476,
+      "learning_rate": 6.9092495058702105e-06,
+      "loss": 0.88,
+      "step": 25978
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.732715771040539,
+      "learning_rate": 6.9085238340908585e-06,
+      "loss": 0.7752,
+      "step": 25979
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.757579023167503,
+      "learning_rate": 6.907798180311063e-06,
+      "loss": 0.8873,
+      "step": 25980
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7993147198773458,
+      "learning_rate": 6.907072544535048e-06,
+      "loss": 0.6875,
+      "step": 25981
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.010214776295769,
+      "learning_rate": 6.906346926767031e-06,
+      "loss": 0.8867,
+      "step": 25982
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.497702179139444,
+      "learning_rate": 6.905621327011241e-06,
+      "loss": 0.894,
+      "step": 25983
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6468485642117427,
+      "learning_rate": 6.904895745271909e-06,
+      "loss": 0.841,
+      "step": 25984
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4302421200270183,
+      "learning_rate": 6.904170181553251e-06,
+      "loss": 0.85,
+      "step": 25985
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5807323089960308,
+      "learning_rate": 6.903444635859493e-06,
+      "loss": 0.7545,
+      "step": 25986
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6524133805846917,
+      "learning_rate": 6.902719108194859e-06,
+      "loss": 0.8996,
+      "step": 25987
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9162014252385053,
+      "learning_rate": 6.90199359856358e-06,
+      "loss": 0.9141,
+      "step": 25988
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5897213725791493,
+      "learning_rate": 6.901268106969872e-06,
+      "loss": 0.9035,
+      "step": 25989
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4900046236656634,
+      "learning_rate": 6.9005426334179635e-06,
+      "loss": 0.8398,
+      "step": 25990
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.774306990874981,
+      "learning_rate": 6.899817177912078e-06,
+      "loss": 0.7327,
+      "step": 25991
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9314409150392167,
+      "learning_rate": 6.899091740456434e-06,
+      "loss": 0.8412,
+      "step": 25992
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6064042595209314,
+      "learning_rate": 6.898366321055263e-06,
+      "loss": 0.9008,
+      "step": 25993
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.638587766975748,
+      "learning_rate": 6.897640919712782e-06,
+      "loss": 0.9206,
+      "step": 25994
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.221719048002056,
+      "learning_rate": 6.896915536433224e-06,
+      "loss": 0.8505,
+      "step": 25995
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6074758487431458,
+      "learning_rate": 6.8961901712208005e-06,
+      "loss": 0.9314,
+      "step": 25996
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7068440044918285,
+      "learning_rate": 6.895464824079743e-06,
+      "loss": 0.9146,
+      "step": 25997
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6171827567086825,
+      "learning_rate": 6.894739495014273e-06,
+      "loss": 0.8612,
+      "step": 25998
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7924069014883933,
+      "learning_rate": 6.894014184028611e-06,
+      "loss": 0.6888,
+      "step": 25999
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6699255390085213,
+      "learning_rate": 6.893288891126981e-06,
+      "loss": 0.8187,
+      "step": 26000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4682810483277942,
+      "learning_rate": 6.892563616313608e-06,
+      "loss": 0.8209,
+      "step": 26001
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.000038390465195,
+      "learning_rate": 6.891838359592714e-06,
+      "loss": 0.8087,
+      "step": 26002
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5358803436689892,
+      "learning_rate": 6.891113120968518e-06,
+      "loss": 0.9045,
+      "step": 26003
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8319726799738152,
+      "learning_rate": 6.890387900445248e-06,
+      "loss": 0.7329,
+      "step": 26004
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7504783923709075,
+      "learning_rate": 6.8896626980271265e-06,
+      "loss": 0.7542,
+      "step": 26005
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7707229719968418,
+      "learning_rate": 6.8889375137183714e-06,
+      "loss": 0.8227,
+      "step": 26006
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5303699955187167,
+      "learning_rate": 6.888212347523204e-06,
+      "loss": 0.8125,
+      "step": 26007
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5394701467764536,
+      "learning_rate": 6.887487199445852e-06,
+      "loss": 0.8199,
+      "step": 26008
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.528147890873533,
+      "learning_rate": 6.886762069490538e-06,
+      "loss": 0.8105,
+      "step": 26009
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6309819189019148,
+      "learning_rate": 6.886036957661477e-06,
+      "loss": 0.8607,
+      "step": 26010
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6735593590500233,
+      "learning_rate": 6.885311863962894e-06,
+      "loss": 0.7021,
+      "step": 26011
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5806592033912348,
+      "learning_rate": 6.884586788399017e-06,
+      "loss": 0.8006,
+      "step": 26012
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6070254653356322,
+      "learning_rate": 6.883861730974057e-06,
+      "loss": 1.001,
+      "step": 26013
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6576159930918761,
+      "learning_rate": 6.883136691692241e-06,
+      "loss": 0.7708,
+      "step": 26014
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6641748897870015,
+      "learning_rate": 6.882411670557788e-06,
+      "loss": 0.8668,
+      "step": 26015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4441421498398164,
+      "learning_rate": 6.881686667574926e-06,
+      "loss": 0.8119,
+      "step": 26016
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5588159416971585,
+      "learning_rate": 6.8809616827478695e-06,
+      "loss": 0.875,
+      "step": 26017
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7906605848897279,
+      "learning_rate": 6.880236716080841e-06,
+      "loss": 0.7151,
+      "step": 26018
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7182272426304264,
+      "learning_rate": 6.879511767578061e-06,
+      "loss": 0.9768,
+      "step": 26019
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5779849899107652,
+      "learning_rate": 6.878786837243755e-06,
+      "loss": 0.9311,
+      "step": 26020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5370007652897293,
+      "learning_rate": 6.878061925082138e-06,
+      "loss": 0.8611,
+      "step": 26021
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7441490170534837,
+      "learning_rate": 6.877337031097433e-06,
+      "loss": 0.8477,
+      "step": 26022
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.767985622431878,
+      "learning_rate": 6.8766121552938625e-06,
+      "loss": 0.9171,
+      "step": 26023
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.669248072657922,
+      "learning_rate": 6.8758872976756406e-06,
+      "loss": 0.8585,
+      "step": 26024
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.817111280054935,
+      "learning_rate": 6.875162458246994e-06,
+      "loss": 0.9314,
+      "step": 26025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.611582312294721,
+      "learning_rate": 6.874437637012141e-06,
+      "loss": 0.9063,
+      "step": 26026
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6048146484762054,
+      "learning_rate": 6.873712833975304e-06,
+      "loss": 0.8547,
+      "step": 26027
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7252899546969909,
+      "learning_rate": 6.872988049140695e-06,
+      "loss": 0.9459,
+      "step": 26028
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9180340591753284,
+      "learning_rate": 6.872263282512544e-06,
+      "loss": 0.7563,
+      "step": 26029
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7495519096853538,
+      "learning_rate": 6.8715385340950655e-06,
+      "loss": 0.9536,
+      "step": 26030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7755828373450466,
+      "learning_rate": 6.87081380389248e-06,
+      "loss": 0.8773,
+      "step": 26031
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.672589692187606,
+      "learning_rate": 6.8700890919090044e-06,
+      "loss": 0.8281,
+      "step": 26032
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.971794395722507,
+      "learning_rate": 6.869364398148865e-06,
+      "loss": 0.7347,
+      "step": 26033
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5286894357853513,
+      "learning_rate": 6.868639722616277e-06,
+      "loss": 0.8895,
+      "step": 26034
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7281592080416122,
+      "learning_rate": 6.8679150653154595e-06,
+      "loss": 0.9186,
+      "step": 26035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6423502710264108,
+      "learning_rate": 6.867190426250629e-06,
+      "loss": 0.9157,
+      "step": 26036
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.797667197628964,
+      "learning_rate": 6.8664658054260124e-06,
+      "loss": 0.8658,
+      "step": 26037
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6106649102111348,
+      "learning_rate": 6.865741202845821e-06,
+      "loss": 0.877,
+      "step": 26038
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6208858377128639,
+      "learning_rate": 6.865016618514277e-06,
+      "loss": 0.8045,
+      "step": 26039
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8339278184924486,
+      "learning_rate": 6.864292052435597e-06,
+      "loss": 0.8741,
+      "step": 26040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6081734952834674,
+      "learning_rate": 6.863567504614006e-06,
+      "loss": 0.8055,
+      "step": 26041
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5204187707556192,
+      "learning_rate": 6.862842975053714e-06,
+      "loss": 0.9407,
+      "step": 26042
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5884285258195348,
+      "learning_rate": 6.862118463758944e-06,
+      "loss": 0.8615,
+      "step": 26043
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.581421510780213,
+      "learning_rate": 6.8613939707339185e-06,
+      "loss": 0.7892,
+      "step": 26044
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8452302208325595,
+      "learning_rate": 6.860669495982844e-06,
+      "loss": 0.8693,
+      "step": 26045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7637466973412954,
+      "learning_rate": 6.859945039509948e-06,
+      "loss": 0.8074,
+      "step": 26046
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6819841136923035,
+      "learning_rate": 6.859220601319445e-06,
+      "loss": 0.9524,
+      "step": 26047
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8176748230383946,
+      "learning_rate": 6.8584961814155595e-06,
+      "loss": 0.8468,
+      "step": 26048
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5740853950433236,
+      "learning_rate": 6.857771779802498e-06,
+      "loss": 0.7745,
+      "step": 26049
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6442238960136986,
+      "learning_rate": 6.857047396484486e-06,
+      "loss": 0.727,
+      "step": 26050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7303529425018893,
+      "learning_rate": 6.856323031465741e-06,
+      "loss": 0.6527,
+      "step": 26051
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5263252985503142,
+      "learning_rate": 6.855598684750475e-06,
+      "loss": 0.8416,
+      "step": 26052
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6740543675679376,
+      "learning_rate": 6.854874356342911e-06,
+      "loss": 0.9686,
+      "step": 26053
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7059632205094757,
+      "learning_rate": 6.854150046247263e-06,
+      "loss": 0.7969,
+      "step": 26054
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3991370184540237,
+      "learning_rate": 6.853425754467753e-06,
+      "loss": 0.9257,
+      "step": 26055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8119410035230852,
+      "learning_rate": 6.852701481008589e-06,
+      "loss": 0.7256,
+      "step": 26056
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4168834486815902,
+      "learning_rate": 6.851977225873996e-06,
+      "loss": 0.7952,
+      "step": 26057
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6708485241524151,
+      "learning_rate": 6.851252989068191e-06,
+      "loss": 0.8405,
+      "step": 26058
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6646181328737266,
+      "learning_rate": 6.8505287705953835e-06,
+      "loss": 0.888,
+      "step": 26059
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6786801567932508,
+      "learning_rate": 6.849804570459795e-06,
+      "loss": 0.8109,
+      "step": 26060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8168488433709462,
+      "learning_rate": 6.849080388665643e-06,
+      "loss": 0.8595,
+      "step": 26061
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3304572350075743,
+      "learning_rate": 6.848356225217146e-06,
+      "loss": 0.8315,
+      "step": 26062
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7567064657524708,
+      "learning_rate": 6.8476320801185116e-06,
+      "loss": 0.7398,
+      "step": 26063
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6402475250279245,
+      "learning_rate": 6.846907953373962e-06,
+      "loss": 0.7786,
+      "step": 26064
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8643141587074,
+      "learning_rate": 6.846183844987715e-06,
+      "loss": 0.8382,
+      "step": 26065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7422084206217544,
+      "learning_rate": 6.845459754963983e-06,
+      "loss": 0.6947,
+      "step": 26066
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4532721808089366,
+      "learning_rate": 6.844735683306982e-06,
+      "loss": 0.8828,
+      "step": 26067
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2003743217225975,
+      "learning_rate": 6.844011630020928e-06,
+      "loss": 0.7486,
+      "step": 26068
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7576025955631388,
+      "learning_rate": 6.843287595110041e-06,
+      "loss": 0.7092,
+      "step": 26069
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3885386068500432,
+      "learning_rate": 6.842563578578531e-06,
+      "loss": 0.8827,
+      "step": 26070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7462401281049635,
+      "learning_rate": 6.841839580430616e-06,
+      "loss": 0.8905,
+      "step": 26071
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7272421347400513,
+      "learning_rate": 6.8411156006705116e-06,
+      "loss": 0.7677,
+      "step": 26072
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7131619972857838,
+      "learning_rate": 6.8403916393024285e-06,
+      "loss": 0.6781,
+      "step": 26073
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5468238893016,
+      "learning_rate": 6.839667696330589e-06,
+      "loss": 0.8222,
+      "step": 26074
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.551810608769754,
+      "learning_rate": 6.838943771759204e-06,
+      "loss": 0.8082,
+      "step": 26075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8156424690565093,
+      "learning_rate": 6.838219865592491e-06,
+      "loss": 0.6959,
+      "step": 26076
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7447879736660872,
+      "learning_rate": 6.837495977834658e-06,
+      "loss": 0.6772,
+      "step": 26077
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5311807247782097,
+      "learning_rate": 6.836772108489928e-06,
+      "loss": 0.8346,
+      "step": 26078
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4610394789520431,
+      "learning_rate": 6.836048257562513e-06,
+      "loss": 0.8852,
+      "step": 26079
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4981271254340116,
+      "learning_rate": 6.835324425056626e-06,
+      "loss": 0.8361,
+      "step": 26080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4924125837293145,
+      "learning_rate": 6.834600610976478e-06,
+      "loss": 0.9575,
+      "step": 26081
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.490727644055166,
+      "learning_rate": 6.8338768153262925e-06,
+      "loss": 0.7827,
+      "step": 26082
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.572147591377844,
+      "learning_rate": 6.833153038110279e-06,
+      "loss": 0.8868,
+      "step": 26083
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8188298180684683,
+      "learning_rate": 6.8324292793326485e-06,
+      "loss": 0.7541,
+      "step": 26084
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7435374407954355,
+      "learning_rate": 6.831705538997617e-06,
+      "loss": 0.7558,
+      "step": 26085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.55216426881568,
+      "learning_rate": 6.830981817109401e-06,
+      "loss": 1.0582,
+      "step": 26086
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6548569998322933,
+      "learning_rate": 6.830258113672213e-06,
+      "loss": 0.9042,
+      "step": 26087
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.510596467863747,
+      "learning_rate": 6.829534428690263e-06,
+      "loss": 0.8322,
+      "step": 26088
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5376253757752298,
+      "learning_rate": 6.8288107621677655e-06,
+      "loss": 0.9394,
+      "step": 26089
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6701970678140872,
+      "learning_rate": 6.8280871141089415e-06,
+      "loss": 0.8987,
+      "step": 26090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4902916556618546,
+      "learning_rate": 6.8273634845179955e-06,
+      "loss": 0.7915,
+      "step": 26091
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4473481724445432,
+      "learning_rate": 6.8266398733991415e-06,
+      "loss": 0.8291,
+      "step": 26092
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5010025686498911,
+      "learning_rate": 6.825916280756601e-06,
+      "loss": 0.7655,
+      "step": 26093
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.497984219960427,
+      "learning_rate": 6.8251927065945755e-06,
+      "loss": 0.7827,
+      "step": 26094
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8043450828775175,
+      "learning_rate": 6.824469150917286e-06,
+      "loss": 0.694,
+      "step": 26095
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.519404851657204,
+      "learning_rate": 6.8237456137289385e-06,
+      "loss": 0.8298,
+      "step": 26096
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7512447819114436,
+      "learning_rate": 6.823022095033756e-06,
+      "loss": 0.7393,
+      "step": 26097
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9013953045051617,
+      "learning_rate": 6.822298594835939e-06,
+      "loss": 0.8639,
+      "step": 26098
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5503398107225395,
+      "learning_rate": 6.821575113139707e-06,
+      "loss": 0.8508,
+      "step": 26099
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6475402480857735,
+      "learning_rate": 6.820851649949274e-06,
+      "loss": 0.8718,
+      "step": 26100
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.540074200133334,
+      "learning_rate": 6.820128205268845e-06,
+      "loss": 0.9078,
+      "step": 26101
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5770009071848492,
+      "learning_rate": 6.8194047791026375e-06,
+      "loss": 0.9124,
+      "step": 26102
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7269269969418273,
+      "learning_rate": 6.8186813714548625e-06,
+      "loss": 0.9247,
+      "step": 26103
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5717345673085419,
+      "learning_rate": 6.817957982329735e-06,
+      "loss": 0.9463,
+      "step": 26104
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.601753254177974,
+      "learning_rate": 6.817234611731458e-06,
+      "loss": 0.8694,
+      "step": 26105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7887466305517763,
+      "learning_rate": 6.816511259664251e-06,
+      "loss": 0.826,
+      "step": 26106
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7893467145040065,
+      "learning_rate": 6.8157879261323265e-06,
+      "loss": 0.7934,
+      "step": 26107
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5416387039540904,
+      "learning_rate": 6.81506461113989e-06,
+      "loss": 0.9703,
+      "step": 26108
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8659526657713372,
+      "learning_rate": 6.814341314691153e-06,
+      "loss": 0.9646,
+      "step": 26109
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.071576695879168,
+      "learning_rate": 6.813618036790333e-06,
+      "loss": 0.9514,
+      "step": 26110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6616907720630003,
+      "learning_rate": 6.812894777441638e-06,
+      "loss": 0.7573,
+      "step": 26111
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5353505892789034,
+      "learning_rate": 6.812171536649278e-06,
+      "loss": 0.8216,
+      "step": 26112
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.995722548934547,
+      "learning_rate": 6.811448314417461e-06,
+      "loss": 0.9105,
+      "step": 26113
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0815025777864014,
+      "learning_rate": 6.8107251107504065e-06,
+      "loss": 0.8358,
+      "step": 26114
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6362303434219356,
+      "learning_rate": 6.810001925652319e-06,
+      "loss": 0.8355,
+      "step": 26115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7555821475257318,
+      "learning_rate": 6.809278759127409e-06,
+      "loss": 0.7753,
+      "step": 26116
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5318103515453076,
+      "learning_rate": 6.808555611179886e-06,
+      "loss": 0.9187,
+      "step": 26117
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.560290038711213,
+      "learning_rate": 6.807832481813967e-06,
+      "loss": 0.9244,
+      "step": 26118
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6380740889314558,
+      "learning_rate": 6.807109371033856e-06,
+      "loss": 0.6797,
+      "step": 26119
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5500072023258398,
+      "learning_rate": 6.806386278843764e-06,
+      "loss": 0.7735,
+      "step": 26120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7527945966852656,
+      "learning_rate": 6.8056632052479055e-06,
+      "loss": 0.9252,
+      "step": 26121
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.463053354315804,
+      "learning_rate": 6.804940150250483e-06,
+      "loss": 0.8381,
+      "step": 26122
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9319132198126576,
+      "learning_rate": 6.804217113855712e-06,
+      "loss": 0.8522,
+      "step": 26123
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8147340118813229,
+      "learning_rate": 6.803494096067802e-06,
+      "loss": 0.8793,
+      "step": 26124
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.159389667513035,
+      "learning_rate": 6.802771096890962e-06,
+      "loss": 0.8286,
+      "step": 26125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.575607793141436,
+      "learning_rate": 6.8020481163293975e-06,
+      "loss": 0.9438,
+      "step": 26126
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.747095951837864,
+      "learning_rate": 6.801325154387323e-06,
+      "loss": 0.914,
+      "step": 26127
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6267624525929636,
+      "learning_rate": 6.800602211068947e-06,
+      "loss": 0.8784,
+      "step": 26128
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4723003389225806,
+      "learning_rate": 6.7998792863784775e-06,
+      "loss": 0.8489,
+      "step": 26129
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6008017399516445,
+      "learning_rate": 6.799156380320122e-06,
+      "loss": 0.9453,
+      "step": 26130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7879590943322687,
+      "learning_rate": 6.798433492898093e-06,
+      "loss": 0.7655,
+      "step": 26131
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8630184153868794,
+      "learning_rate": 6.797710624116601e-06,
+      "loss": 0.885,
+      "step": 26132
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4890609710400058,
+      "learning_rate": 6.796987773979846e-06,
+      "loss": 0.9837,
+      "step": 26133
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8095778732311999,
+      "learning_rate": 6.796264942492044e-06,
+      "loss": 0.7473,
+      "step": 26134
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7478658358876074,
+      "learning_rate": 6.7955421296574045e-06,
+      "loss": 0.7233,
+      "step": 26135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.360434388223447,
+      "learning_rate": 6.7948193354801306e-06,
+      "loss": 1.0047,
+      "step": 26136
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5403116094409888,
+      "learning_rate": 6.794096559964434e-06,
+      "loss": 0.8102,
+      "step": 26137
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4806317840244783,
+      "learning_rate": 6.7933738031145205e-06,
+      "loss": 0.9178,
+      "step": 26138
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4078705844079389,
+      "learning_rate": 6.792651064934602e-06,
+      "loss": 0.7859,
+      "step": 26139
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5145748293168586,
+      "learning_rate": 6.791928345428884e-06,
+      "loss": 0.8037,
+      "step": 26140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7024419440395153,
+      "learning_rate": 6.791205644601573e-06,
+      "loss": 0.7644,
+      "step": 26141
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.112577816558648,
+      "learning_rate": 6.790482962456879e-06,
+      "loss": 0.8059,
+      "step": 26142
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7421505202544527,
+      "learning_rate": 6.789760298999013e-06,
+      "loss": 0.753,
+      "step": 26143
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.025832105662269,
+      "learning_rate": 6.789037654232177e-06,
+      "loss": 0.8364,
+      "step": 26144
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5844091362630432,
+      "learning_rate": 6.788315028160577e-06,
+      "loss": 0.8087,
+      "step": 26145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8082283320802589,
+      "learning_rate": 6.787592420788431e-06,
+      "loss": 0.7207,
+      "step": 26146
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4431661879545714,
+      "learning_rate": 6.786869832119933e-06,
+      "loss": 0.7717,
+      "step": 26147
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6887520455708023,
+      "learning_rate": 6.786147262159297e-06,
+      "loss": 0.9089,
+      "step": 26148
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7449005174674154,
+      "learning_rate": 6.7854247109107285e-06,
+      "loss": 0.6594,
+      "step": 26149
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.550217485880568,
+      "learning_rate": 6.784702178378437e-06,
+      "loss": 0.8297,
+      "step": 26150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6103060105873528,
+      "learning_rate": 6.783979664566627e-06,
+      "loss": 0.845,
+      "step": 26151
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.62615954909503,
+      "learning_rate": 6.783257169479504e-06,
+      "loss": 0.7894,
+      "step": 26152
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8184094163480824,
+      "learning_rate": 6.782534693121281e-06,
+      "loss": 0.8786,
+      "step": 26153
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5768499900271222,
+      "learning_rate": 6.781812235496153e-06,
+      "loss": 0.9285,
+      "step": 26154
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5163249822638907,
+      "learning_rate": 6.7810897966083365e-06,
+      "loss": 0.981,
+      "step": 26155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5627872921159354,
+      "learning_rate": 6.780367376462033e-06,
+      "loss": 0.8591,
+      "step": 26156
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.857563490372103,
+      "learning_rate": 6.779644975061454e-06,
+      "loss": 0.9021,
+      "step": 26157
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1088393396186746,
+      "learning_rate": 6.778922592410796e-06,
+      "loss": 0.8628,
+      "step": 26158
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8231965092738267,
+      "learning_rate": 6.778200228514274e-06,
+      "loss": 0.8335,
+      "step": 26159
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5138874676242298,
+      "learning_rate": 6.777477883376089e-06,
+      "loss": 0.7348,
+      "step": 26160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7747165061532664,
+      "learning_rate": 6.776755557000449e-06,
+      "loss": 0.8927,
+      "step": 26161
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5945101760420943,
+      "learning_rate": 6.776033249391555e-06,
+      "loss": 0.8102,
+      "step": 26162
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5252686931019293,
+      "learning_rate": 6.775310960553618e-06,
+      "loss": 0.8678,
+      "step": 26163
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5436272136065379,
+      "learning_rate": 6.774588690490844e-06,
+      "loss": 0.805,
+      "step": 26164
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.514212570929323,
+      "learning_rate": 6.7738664392074345e-06,
+      "loss": 0.801,
+      "step": 26165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.683233425686288,
+      "learning_rate": 6.773144206707594e-06,
+      "loss": 0.8976,
+      "step": 26166
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7827123415593734,
+      "learning_rate": 6.772421992995533e-06,
+      "loss": 0.6479,
+      "step": 26167
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4526518400591313,
+      "learning_rate": 6.7716997980754505e-06,
+      "loss": 0.9099,
+      "step": 26168
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.523022516979075,
+      "learning_rate": 6.770977621951554e-06,
+      "loss": 0.8613,
+      "step": 26169
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8723021788026273,
+      "learning_rate": 6.770255464628047e-06,
+      "loss": 0.8675,
+      "step": 26170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7450201566164515,
+      "learning_rate": 6.769533326109139e-06,
+      "loss": 0.7862,
+      "step": 26171
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7818437017466652,
+      "learning_rate": 6.768811206399028e-06,
+      "loss": 0.7123,
+      "step": 26172
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9086164197246867,
+      "learning_rate": 6.7680891055019215e-06,
+      "loss": 0.8029,
+      "step": 26173
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.23262295925846,
+      "learning_rate": 6.767367023422025e-06,
+      "loss": 0.8318,
+      "step": 26174
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6412929310243953,
+      "learning_rate": 6.766644960163538e-06,
+      "loss": 0.8219,
+      "step": 26175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.728628482912583,
+      "learning_rate": 6.765922915730669e-06,
+      "loss": 0.8235,
+      "step": 26176
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5348454101481654,
+      "learning_rate": 6.765200890127619e-06,
+      "loss": 0.8008,
+      "step": 26177
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5523869560460435,
+      "learning_rate": 6.7644788833586e-06,
+      "loss": 0.8179,
+      "step": 26178
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.405183666931076,
+      "learning_rate": 6.763756895427802e-06,
+      "loss": 0.79,
+      "step": 26179
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6931634021843665,
+      "learning_rate": 6.763034926339438e-06,
+      "loss": 0.9365,
+      "step": 26180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5871730594599036,
+      "learning_rate": 6.762312976097714e-06,
+      "loss": 0.9067,
+      "step": 26181
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4425167277507458,
+      "learning_rate": 6.7615910447068235e-06,
+      "loss": 0.8544,
+      "step": 26182
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.62223372742696,
+      "learning_rate": 6.7608691321709755e-06,
+      "loss": 0.8437,
+      "step": 26183
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3106211194448805,
+      "learning_rate": 6.760147238494374e-06,
+      "loss": 0.8418,
+      "step": 26184
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7662749730595666,
+      "learning_rate": 6.759425363681223e-06,
+      "loss": 0.8224,
+      "step": 26185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7271591312593593,
+      "learning_rate": 6.758703507735721e-06,
+      "loss": 0.9365,
+      "step": 26186
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3444338193478718,
+      "learning_rate": 6.7579816706620724e-06,
+      "loss": 0.7396,
+      "step": 26187
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7830466726356337,
+      "learning_rate": 6.757259852464484e-06,
+      "loss": 0.7059,
+      "step": 26188
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.57787290584691,
+      "learning_rate": 6.7565380531471545e-06,
+      "loss": 0.818,
+      "step": 26189
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8400972261580473,
+      "learning_rate": 6.755816272714284e-06,
+      "loss": 0.8831,
+      "step": 26190
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4944170164047474,
+      "learning_rate": 6.755094511170082e-06,
+      "loss": 0.9515,
+      "step": 26191
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4917032475893146,
+      "learning_rate": 6.754372768518746e-06,
+      "loss": 0.9287,
+      "step": 26192
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6533092019878906,
+      "learning_rate": 6.7536510447644795e-06,
+      "loss": 0.8149,
+      "step": 26193
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7486748001453858,
+      "learning_rate": 6.752929339911483e-06,
+      "loss": 0.9361,
+      "step": 26194
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.710532871168271,
+      "learning_rate": 6.752207653963965e-06,
+      "loss": 0.7201,
+      "step": 26195
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9046752912702023,
+      "learning_rate": 6.7514859869261164e-06,
+      "loss": 0.8961,
+      "step": 26196
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.197861333892022,
+      "learning_rate": 6.750764338802147e-06,
+      "loss": 0.7856,
+      "step": 26197
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6714524553729777,
+      "learning_rate": 6.750042709596254e-06,
+      "loss": 0.8733,
+      "step": 26198
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9484326458038714,
+      "learning_rate": 6.749321099312644e-06,
+      "loss": 0.8427,
+      "step": 26199
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4326751945857539,
+      "learning_rate": 6.748599507955516e-06,
+      "loss": 0.8401,
+      "step": 26200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.174531352194189,
+      "learning_rate": 6.74787793552907e-06,
+      "loss": 0.8725,
+      "step": 26201
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.670567611163484,
+      "learning_rate": 6.747156382037511e-06,
+      "loss": 0.8987,
+      "step": 26202
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6918859716917236,
+      "learning_rate": 6.7464348474850324e-06,
+      "loss": 0.813,
+      "step": 26203
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.670321476036023,
+      "learning_rate": 6.745713331875843e-06,
+      "loss": 0.9082,
+      "step": 26204
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5829648981368083,
+      "learning_rate": 6.74499183521414e-06,
+      "loss": 0.8722,
+      "step": 26205
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.995281014263687,
+      "learning_rate": 6.744270357504128e-06,
+      "loss": 0.9204,
+      "step": 26206
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8703173966751248,
+      "learning_rate": 6.743548898749999e-06,
+      "loss": 0.833,
+      "step": 26207
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.477459734743052,
+      "learning_rate": 6.7428274589559626e-06,
+      "loss": 0.921,
+      "step": 26208
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.362308960870731,
+      "learning_rate": 6.742106038126218e-06,
+      "loss": 0.8017,
+      "step": 26209
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5583433510895068,
+      "learning_rate": 6.741384636264961e-06,
+      "loss": 0.9331,
+      "step": 26210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.403149038614186,
+      "learning_rate": 6.740663253376393e-06,
+      "loss": 0.8577,
+      "step": 26211
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1942013617757405,
+      "learning_rate": 6.7399418894647164e-06,
+      "loss": 0.7513,
+      "step": 26212
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6878459747224508,
+      "learning_rate": 6.739220544534133e-06,
+      "loss": 0.9025,
+      "step": 26213
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6645073935386723,
+      "learning_rate": 6.738499218588839e-06,
+      "loss": 0.9048,
+      "step": 26214
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.665379402155701,
+      "learning_rate": 6.737777911633031e-06,
+      "loss": 0.93,
+      "step": 26215
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.348954879754029,
+      "learning_rate": 6.737056623670918e-06,
+      "loss": 0.7787,
+      "step": 26216
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0060687342251904,
+      "learning_rate": 6.736335354706691e-06,
+      "loss": 0.8182,
+      "step": 26217
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7832414667281251,
+      "learning_rate": 6.7356141047445545e-06,
+      "loss": 0.9445,
+      "step": 26218
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.041556852137011,
+      "learning_rate": 6.734892873788703e-06,
+      "loss": 0.7578,
+      "step": 26219
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6918702284469769,
+      "learning_rate": 6.7341716618433435e-06,
+      "loss": 0.8235,
+      "step": 26220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.570216858450944,
+      "learning_rate": 6.733450468912669e-06,
+      "loss": 0.9079,
+      "step": 26221
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7023430634000701,
+      "learning_rate": 6.732729295000879e-06,
+      "loss": 0.6959,
+      "step": 26222
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5315047613890322,
+      "learning_rate": 6.732008140112176e-06,
+      "loss": 0.7741,
+      "step": 26223
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5835044653408323,
+      "learning_rate": 6.731287004250752e-06,
+      "loss": 0.8857,
+      "step": 26224
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.649380632601533,
+      "learning_rate": 6.730565887420813e-06,
+      "loss": 0.8633,
+      "step": 26225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.664283273306785,
+      "learning_rate": 6.729844789626552e-06,
+      "loss": 0.7801,
+      "step": 26226
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6518776878308183,
+      "learning_rate": 6.729123710872175e-06,
+      "loss": 0.9263,
+      "step": 26227
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7748384773775787,
+      "learning_rate": 6.728402651161869e-06,
+      "loss": 1.026,
+      "step": 26228
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7137067987717742,
+      "learning_rate": 6.727681610499842e-06,
+      "loss": 0.8094,
+      "step": 26229
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8607697018998515,
+      "learning_rate": 6.72696058889029e-06,
+      "loss": 0.8208,
+      "step": 26230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4853615500260444,
+      "learning_rate": 6.726239586337408e-06,
+      "loss": 0.8185,
+      "step": 26231
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.151488843620206,
+      "learning_rate": 6.725518602845394e-06,
+      "loss": 0.8262,
+      "step": 26232
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.879279222225355,
+      "learning_rate": 6.7247976384184475e-06,
+      "loss": 0.8556,
+      "step": 26233
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.589281315784002,
+      "learning_rate": 6.72407669306077e-06,
+      "loss": 0.9118,
+      "step": 26234
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.907034284215859,
+      "learning_rate": 6.723355766776553e-06,
+      "loss": 0.8683,
+      "step": 26235
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6155474474259701,
+      "learning_rate": 6.722634859569992e-06,
+      "loss": 0.8451,
+      "step": 26236
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5730237142036831,
+      "learning_rate": 6.721913971445295e-06,
+      "loss": 0.7699,
+      "step": 26237
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6248791925836494,
+      "learning_rate": 6.721193102406649e-06,
+      "loss": 0.8877,
+      "step": 26238
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6814818004046006,
+      "learning_rate": 6.720472252458254e-06,
+      "loss": 0.9309,
+      "step": 26239
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6431323316677104,
+      "learning_rate": 6.719751421604309e-06,
+      "loss": 0.7856,
+      "step": 26240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5266677886649729,
+      "learning_rate": 6.719030609849012e-06,
+      "loss": 0.7914,
+      "step": 26241
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7577750553341716,
+      "learning_rate": 6.718309817196556e-06,
+      "loss": 0.8858,
+      "step": 26242
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6961594278586254,
+      "learning_rate": 6.717589043651136e-06,
+      "loss": 0.8571,
+      "step": 26243
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.812819905214564,
+      "learning_rate": 6.716868289216959e-06,
+      "loss": 0.6774,
+      "step": 26244
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8906904712128312,
+      "learning_rate": 6.716147553898206e-06,
+      "loss": 0.9369,
+      "step": 26245
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6982413489388748,
+      "learning_rate": 6.715426837699086e-06,
+      "loss": 0.9745,
+      "step": 26246
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.756622792463329,
+      "learning_rate": 6.714706140623789e-06,
+      "loss": 0.7258,
+      "step": 26247
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5878611550317854,
+      "learning_rate": 6.7139854626765155e-06,
+      "loss": 0.9549,
+      "step": 26248
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4622703464017286,
+      "learning_rate": 6.713264803861459e-06,
+      "loss": 0.9042,
+      "step": 26249
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.80189755668007,
+      "learning_rate": 6.712544164182813e-06,
+      "loss": 0.7518,
+      "step": 26250
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7479226303133684,
+      "learning_rate": 6.711823543644779e-06,
+      "loss": 0.8403,
+      "step": 26251
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6639749519758207,
+      "learning_rate": 6.711102942251544e-06,
+      "loss": 0.7706,
+      "step": 26252
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7748491586195132,
+      "learning_rate": 6.710382360007313e-06,
+      "loss": 0.7756,
+      "step": 26253
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8886390197803298,
+      "learning_rate": 6.7096617969162756e-06,
+      "loss": 0.8637,
+      "step": 26254
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.489586588573539,
+      "learning_rate": 6.708941252982632e-06,
+      "loss": 0.857,
+      "step": 26255
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.60737742396806,
+      "learning_rate": 6.70822072821057e-06,
+      "loss": 0.8565,
+      "step": 26256
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7164869909307041,
+      "learning_rate": 6.707500222604291e-06,
+      "loss": 0.867,
+      "step": 26257
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6222451987636544,
+      "learning_rate": 6.70677973616799e-06,
+      "loss": 0.8616,
+      "step": 26258
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6414328098325648,
+      "learning_rate": 6.706059268905858e-06,
+      "loss": 0.7686,
+      "step": 26259
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5248902406763392,
+      "learning_rate": 6.705338820822089e-06,
+      "loss": 0.7778,
+      "step": 26260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5917268972980512,
+      "learning_rate": 6.704618391920884e-06,
+      "loss": 0.8432,
+      "step": 26261
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6345341682654175,
+      "learning_rate": 6.7038979822064355e-06,
+      "loss": 0.8524,
+      "step": 26262
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7113676284059042,
+      "learning_rate": 6.703177591682935e-06,
+      "loss": 0.6752,
+      "step": 26263
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7371651926853919,
+      "learning_rate": 6.702457220354575e-06,
+      "loss": 0.7138,
+      "step": 26264
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.473143091630185,
+      "learning_rate": 6.701736868225556e-06,
+      "loss": 0.8514,
+      "step": 26265
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5889830823618452,
+      "learning_rate": 6.701016535300071e-06,
+      "loss": 0.8524,
+      "step": 26266
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8454336560982356,
+      "learning_rate": 6.7002962215823095e-06,
+      "loss": 0.8879,
+      "step": 26267
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7527660772387544,
+      "learning_rate": 6.699575927076467e-06,
+      "loss": 0.7611,
+      "step": 26268
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.687061093859617,
+      "learning_rate": 6.698855651786742e-06,
+      "loss": 0.8941,
+      "step": 26269
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7480296928884184,
+      "learning_rate": 6.698135395717323e-06,
+      "loss": 0.6876,
+      "step": 26270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7170012387550357,
+      "learning_rate": 6.697415158872405e-06,
+      "loss": 0.8787,
+      "step": 26271
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7076080272641407,
+      "learning_rate": 6.696694941256179e-06,
+      "loss": 0.8475,
+      "step": 26272
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.719767485086475,
+      "learning_rate": 6.695974742872844e-06,
+      "loss": 0.7435,
+      "step": 26273
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9327169599676177,
+      "learning_rate": 6.69525456372659e-06,
+      "loss": 0.9685,
+      "step": 26274
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7478502318126443,
+      "learning_rate": 6.694534403821606e-06,
+      "loss": 0.8119,
+      "step": 26275
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6162000226461868,
+      "learning_rate": 6.6938142631620975e-06,
+      "loss": 0.8541,
+      "step": 26276
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7694459593063323,
+      "learning_rate": 6.693094141752241e-06,
+      "loss": 0.8695,
+      "step": 26277
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6442409844148862,
+      "learning_rate": 6.692374039596241e-06,
+      "loss": 0.7976,
+      "step": 26278
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4839882230082764,
+      "learning_rate": 6.691653956698283e-06,
+      "loss": 0.9587,
+      "step": 26279
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.811544799131423,
+      "learning_rate": 6.69093389306257e-06,
+      "loss": 0.898,
+      "step": 26280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6572825481585272,
+      "learning_rate": 6.690213848693282e-06,
+      "loss": 0.9159,
+      "step": 26281
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7575151209154054,
+      "learning_rate": 6.689493823594618e-06,
+      "loss": 0.9021,
+      "step": 26282
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.518084146033868,
+      "learning_rate": 6.688773817770772e-06,
+      "loss": 0.794,
+      "step": 26283
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.65044011473519,
+      "learning_rate": 6.68805383122593e-06,
+      "loss": 0.8184,
+      "step": 26284
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5759668194306777,
+      "learning_rate": 6.687333863964285e-06,
+      "loss": 0.8864,
+      "step": 26285
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.378034068646356,
+      "learning_rate": 6.686613915990034e-06,
+      "loss": 0.9039,
+      "step": 26286
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7221812532170513,
+      "learning_rate": 6.685893987307368e-06,
+      "loss": 0.8661,
+      "step": 26287
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.667437244803851,
+      "learning_rate": 6.685174077920471e-06,
+      "loss": 0.9865,
+      "step": 26288
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.542375508229737,
+      "learning_rate": 6.684454187833543e-06,
+      "loss": 0.9914,
+      "step": 26289
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5396616520486048,
+      "learning_rate": 6.683734317050774e-06,
+      "loss": 0.7919,
+      "step": 26290
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5938038590449017,
+      "learning_rate": 6.683014465576353e-06,
+      "loss": 0.7803,
+      "step": 26291
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5603508006630729,
+      "learning_rate": 6.6822946334144675e-06,
+      "loss": 0.9195,
+      "step": 26292
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.676797811946708,
+      "learning_rate": 6.681574820569317e-06,
+      "loss": 0.9169,
+      "step": 26293
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5585180385728454,
+      "learning_rate": 6.68085502704509e-06,
+      "loss": 0.7782,
+      "step": 26294
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5694734580280938,
+      "learning_rate": 6.680135252845974e-06,
+      "loss": 0.9108,
+      "step": 26295
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.411988745483609,
+      "learning_rate": 6.6794154979761595e-06,
+      "loss": 0.8129,
+      "step": 26296
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7735996330958685,
+      "learning_rate": 6.678695762439843e-06,
+      "loss": 0.8597,
+      "step": 26297
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7236282346039244,
+      "learning_rate": 6.677976046241209e-06,
+      "loss": 0.8397,
+      "step": 26298
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4703321624883356,
+      "learning_rate": 6.677256349384451e-06,
+      "loss": 0.8406,
+      "step": 26299
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6374910892953527,
+      "learning_rate": 6.676536671873755e-06,
+      "loss": 0.8001,
+      "step": 26300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8791128277849833,
+      "learning_rate": 6.675817013713321e-06,
+      "loss": 0.7698,
+      "step": 26301
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9673276650718612,
+      "learning_rate": 6.675097374907329e-06,
+      "loss": 0.8135,
+      "step": 26302
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.020083479221207,
+      "learning_rate": 6.6743777554599735e-06,
+      "loss": 0.8472,
+      "step": 26303
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5604865183298926,
+      "learning_rate": 6.673658155375446e-06,
+      "loss": 0.9435,
+      "step": 26304
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8147389477283017,
+      "learning_rate": 6.672938574657928e-06,
+      "loss": 0.7279,
+      "step": 26305
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.748330719292155,
+      "learning_rate": 6.672219013311618e-06,
+      "loss": 0.6935,
+      "step": 26306
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7937210735177787,
+      "learning_rate": 6.671499471340704e-06,
+      "loss": 0.9089,
+      "step": 26307
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7075975919150785,
+      "learning_rate": 6.670779948749373e-06,
+      "loss": 0.878,
+      "step": 26308
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5991277870373024,
+      "learning_rate": 6.6700604455418125e-06,
+      "loss": 0.8685,
+      "step": 26309
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.76303859936227,
+      "learning_rate": 6.669340961722217e-06,
+      "loss": 0.7929,
+      "step": 26310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.556258878683456,
+      "learning_rate": 6.668621497294774e-06,
+      "loss": 0.8628,
+      "step": 26311
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6922884849877942,
+      "learning_rate": 6.667902052263671e-06,
+      "loss": 0.7394,
+      "step": 26312
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6503784887972248,
+      "learning_rate": 6.667182626633093e-06,
+      "loss": 0.7121,
+      "step": 26313
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.237682341459534,
+      "learning_rate": 6.666463220407235e-06,
+      "loss": 0.8864,
+      "step": 26314
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7543558090921517,
+      "learning_rate": 6.665743833590287e-06,
+      "loss": 0.7314,
+      "step": 26315
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5937717335606383,
+      "learning_rate": 6.665024466186431e-06,
+      "loss": 0.9469,
+      "step": 26316
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5330203466555163,
+      "learning_rate": 6.664305118199857e-06,
+      "loss": 0.8704,
+      "step": 26317
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6358918912727118,
+      "learning_rate": 6.663585789634758e-06,
+      "loss": 0.8488,
+      "step": 26318
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.714533636347372,
+      "learning_rate": 6.662866480495317e-06,
+      "loss": 0.8514,
+      "step": 26319
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6822037221011161,
+      "learning_rate": 6.6621471907857234e-06,
+      "loss": 0.8381,
+      "step": 26320
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.74448325642925,
+      "learning_rate": 6.661427920510164e-06,
+      "loss": 0.7972,
+      "step": 26321
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4592589788749806,
+      "learning_rate": 6.660708669672833e-06,
+      "loss": 0.7863,
+      "step": 26322
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5821411230056979,
+      "learning_rate": 6.6599894382779104e-06,
+      "loss": 0.8492,
+      "step": 26323
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8579936595409243,
+      "learning_rate": 6.659270226329586e-06,
+      "loss": 0.9715,
+      "step": 26324
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6173039966455252,
+      "learning_rate": 6.658551033832052e-06,
+      "loss": 0.7793,
+      "step": 26325
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0485589291454582,
+      "learning_rate": 6.657831860789488e-06,
+      "loss": 0.7874,
+      "step": 26326
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9306448460719043,
+      "learning_rate": 6.657112707206086e-06,
+      "loss": 0.9202,
+      "step": 26327
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6643308376440762,
+      "learning_rate": 6.656393573086029e-06,
+      "loss": 0.8795,
+      "step": 26328
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.561228729488703,
+      "learning_rate": 6.655674458433514e-06,
+      "loss": 0.8503,
+      "step": 26329
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8581462363924215,
+      "learning_rate": 6.6549553632527154e-06,
+      "loss": 0.9185,
+      "step": 26330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7780131970244701,
+      "learning_rate": 6.6542362875478285e-06,
+      "loss": 0.8536,
+      "step": 26331
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6439829471224572,
+      "learning_rate": 6.6535172313230375e-06,
+      "loss": 0.7717,
+      "step": 26332
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5893509032657054,
+      "learning_rate": 6.652798194582528e-06,
+      "loss": 0.8001,
+      "step": 26333
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5426881079227472,
+      "learning_rate": 6.652079177330484e-06,
+      "loss": 0.8039,
+      "step": 26334
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8151984539804469,
+      "learning_rate": 6.6513601795710984e-06,
+      "loss": 0.9606,
+      "step": 26335
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7710696953292413,
+      "learning_rate": 6.650641201308554e-06,
+      "loss": 0.7426,
+      "step": 26336
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6917059780236414,
+      "learning_rate": 6.649922242547033e-06,
+      "loss": 0.8074,
+      "step": 26337
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.680254828175585,
+      "learning_rate": 6.64920330329073e-06,
+      "loss": 0.7864,
+      "step": 26338
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.55726042756815,
+      "learning_rate": 6.648484383543826e-06,
+      "loss": 0.7669,
+      "step": 26339
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7570225465348488,
+      "learning_rate": 6.647765483310506e-06,
+      "loss": 0.6524,
+      "step": 26340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1359454892944103,
+      "learning_rate": 6.647046602594953e-06,
+      "loss": 0.8067,
+      "step": 26341
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4741905603889083,
+      "learning_rate": 6.64632774140136e-06,
+      "loss": 0.7841,
+      "step": 26342
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5314571347677004,
+      "learning_rate": 6.645608899733911e-06,
+      "loss": 0.9299,
+      "step": 26343
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6208572727778652,
+      "learning_rate": 6.644890077596785e-06,
+      "loss": 0.8893,
+      "step": 26344
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7647885020649345,
+      "learning_rate": 6.64417127499417e-06,
+      "loss": 0.7495,
+      "step": 26345
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7191032712084144,
+      "learning_rate": 6.643452491930257e-06,
+      "loss": 0.7823,
+      "step": 26346
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.436419138896108,
+      "learning_rate": 6.642733728409223e-06,
+      "loss": 0.8225,
+      "step": 26347
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8470372295674127,
+      "learning_rate": 6.642014984435257e-06,
+      "loss": 0.8961,
+      "step": 26348
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7381908902246006,
+      "learning_rate": 6.641296260012541e-06,
+      "loss": 0.8916,
+      "step": 26349
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9006114954770355,
+      "learning_rate": 6.640577555145266e-06,
+      "loss": 0.7994,
+      "step": 26350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5924085544136157,
+      "learning_rate": 6.639858869837608e-06,
+      "loss": 0.7922,
+      "step": 26351
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.668122314193506,
+      "learning_rate": 6.639140204093757e-06,
+      "loss": 0.7974,
+      "step": 26352
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8283214536926813,
+      "learning_rate": 6.638421557917898e-06,
+      "loss": 0.8203,
+      "step": 26353
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6922392477295198,
+      "learning_rate": 6.637702931314208e-06,
+      "loss": 0.9031,
+      "step": 26354
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4107702212416386,
+      "learning_rate": 6.6369843242868795e-06,
+      "loss": 0.7754,
+      "step": 26355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6125646040120927,
+      "learning_rate": 6.63626573684009e-06,
+      "loss": 0.8695,
+      "step": 26356
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.608475664302148,
+      "learning_rate": 6.635547168978032e-06,
+      "loss": 0.9292,
+      "step": 26357
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.726460239893776,
+      "learning_rate": 6.6348286207048776e-06,
+      "loss": 0.7905,
+      "step": 26358
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.454638014289824,
+      "learning_rate": 6.634110092024819e-06,
+      "loss": 0.916,
+      "step": 26359
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8236549426694835,
+      "learning_rate": 6.633391582942038e-06,
+      "loss": 0.7478,
+      "step": 26360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7766391904446015,
+      "learning_rate": 6.632673093460715e-06,
+      "loss": 0.8573,
+      "step": 26361
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5244354544416412,
+      "learning_rate": 6.631954623585035e-06,
+      "loss": 0.9534,
+      "step": 26362
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.72177177600524,
+      "learning_rate": 6.631236173319182e-06,
+      "loss": 0.7781,
+      "step": 26363
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7875832966828464,
+      "learning_rate": 6.630517742667341e-06,
+      "loss": 0.8422,
+      "step": 26364
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5658156137101809,
+      "learning_rate": 6.629799331633691e-06,
+      "loss": 0.9182,
+      "step": 26365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7008400659087477,
+      "learning_rate": 6.629080940222414e-06,
+      "loss": 0.9533,
+      "step": 26366
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6165967708173912,
+      "learning_rate": 6.628362568437698e-06,
+      "loss": 0.7384,
+      "step": 26367
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6299945602177266,
+      "learning_rate": 6.6276442162837215e-06,
+      "loss": 0.8333,
+      "step": 26368
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6215695333767381,
+      "learning_rate": 6.6269258837646675e-06,
+      "loss": 0.8671,
+      "step": 26369
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6767422257782316,
+      "learning_rate": 6.626207570884717e-06,
+      "loss": 0.7741,
+      "step": 26370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.402781978513231,
+      "learning_rate": 6.625489277648059e-06,
+      "loss": 0.8041,
+      "step": 26371
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.65620268604632,
+      "learning_rate": 6.624771004058869e-06,
+      "loss": 0.8856,
+      "step": 26372
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7059962142390652,
+      "learning_rate": 6.624052750121328e-06,
+      "loss": 0.8687,
+      "step": 26373
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6498016211384228,
+      "learning_rate": 6.623334515839627e-06,
+      "loss": 0.7533,
+      "step": 26374
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7261031090060062,
+      "learning_rate": 6.6226163012179345e-06,
+      "loss": 0.9008,
+      "step": 26375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8754680589190502,
+      "learning_rate": 6.621898106260443e-06,
+      "loss": 0.9186,
+      "step": 26376
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.026459986891397,
+      "learning_rate": 6.621179930971327e-06,
+      "loss": 0.97,
+      "step": 26377
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.808381353707772,
+      "learning_rate": 6.620461775354776e-06,
+      "loss": 0.8715,
+      "step": 26378
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5037324238694576,
+      "learning_rate": 6.619743639414961e-06,
+      "loss": 0.906,
+      "step": 26379
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.656813330476001,
+      "learning_rate": 6.619025523156073e-06,
+      "loss": 0.8829,
+      "step": 26380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4504341721699445,
+      "learning_rate": 6.618307426582289e-06,
+      "loss": 0.9307,
+      "step": 26381
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5924686140237165,
+      "learning_rate": 6.617589349697786e-06,
+      "loss": 0.7383,
+      "step": 26382
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7050521878893452,
+      "learning_rate": 6.61687129250675e-06,
+      "loss": 0.9756,
+      "step": 26383
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6400614168002932,
+      "learning_rate": 6.616153255013363e-06,
+      "loss": 0.8287,
+      "step": 26384
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6058029246097887,
+      "learning_rate": 6.6154352372218015e-06,
+      "loss": 0.9509,
+      "step": 26385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.936952350507676,
+      "learning_rate": 6.614717239136246e-06,
+      "loss": 0.7312,
+      "step": 26386
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8622446396813108,
+      "learning_rate": 6.6139992607608795e-06,
+      "loss": 0.8179,
+      "step": 26387
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9920252388113342,
+      "learning_rate": 6.613281302099881e-06,
+      "loss": 0.8472,
+      "step": 26388
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7958368110577976,
+      "learning_rate": 6.612563363157434e-06,
+      "loss": 0.8592,
+      "step": 26389
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5319003943600031,
+      "learning_rate": 6.61184544393771e-06,
+      "loss": 1.0003,
+      "step": 26390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6041095088680692,
+      "learning_rate": 6.611127544444897e-06,
+      "loss": 0.8957,
+      "step": 26391
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5095420488881395,
+      "learning_rate": 6.610409664683176e-06,
+      "loss": 0.9352,
+      "step": 26392
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5971995629081157,
+      "learning_rate": 6.609691804656719e-06,
+      "loss": 0.8004,
+      "step": 26393
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7883954146788288,
+      "learning_rate": 6.608973964369709e-06,
+      "loss": 0.7434,
+      "step": 26394
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5845344645467918,
+      "learning_rate": 6.608256143826328e-06,
+      "loss": 0.8088,
+      "step": 26395
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6675881995744906,
+      "learning_rate": 6.607538343030755e-06,
+      "loss": 0.8925,
+      "step": 26396
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7597009028227827,
+      "learning_rate": 6.6068205619871664e-06,
+      "loss": 0.7484,
+      "step": 26397
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6705459775348275,
+      "learning_rate": 6.606102800699741e-06,
+      "loss": 0.8381,
+      "step": 26398
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.527451104444052,
+      "learning_rate": 6.605385059172662e-06,
+      "loss": 0.8307,
+      "step": 26399
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.592680020797128,
+      "learning_rate": 6.6046673374101054e-06,
+      "loss": 0.7952,
+      "step": 26400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7547705708648392,
+      "learning_rate": 6.60394963541625e-06,
+      "loss": 0.7156,
+      "step": 26401
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.434317106603829,
+      "learning_rate": 6.603231953195272e-06,
+      "loss": 0.8682,
+      "step": 26402
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6378260762625858,
+      "learning_rate": 6.6025142907513596e-06,
+      "loss": 0.7801,
+      "step": 26403
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4842880949819037,
+      "learning_rate": 6.601796648088681e-06,
+      "loss": 0.8041,
+      "step": 26404
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6231452956918508,
+      "learning_rate": 6.601079025211418e-06,
+      "loss": 0.902,
+      "step": 26405
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5331132433924677,
+      "learning_rate": 6.600361422123751e-06,
+      "loss": 0.7723,
+      "step": 26406
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6423984736687254,
+      "learning_rate": 6.599643838829852e-06,
+      "loss": 0.7944,
+      "step": 26407
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9648432127878848,
+      "learning_rate": 6.598926275333906e-06,
+      "loss": 0.7845,
+      "step": 26408
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7258111047642681,
+      "learning_rate": 6.598208731640085e-06,
+      "loss": 0.7271,
+      "step": 26409
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5256521642255687,
+      "learning_rate": 6.597491207752577e-06,
+      "loss": 0.9483,
+      "step": 26410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.763143746447505,
+      "learning_rate": 6.596773703675545e-06,
+      "loss": 0.7956,
+      "step": 26411
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5281306168483093,
+      "learning_rate": 6.596056219413176e-06,
+      "loss": 0.8206,
+      "step": 26412
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.584620661740388,
+      "learning_rate": 6.595338754969649e-06,
+      "loss": 0.8873,
+      "step": 26413
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6105571837341655,
+      "learning_rate": 6.594621310349134e-06,
+      "loss": 0.7648,
+      "step": 26414
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6742135310698028,
+      "learning_rate": 6.59390388555581e-06,
+      "loss": 0.7827,
+      "step": 26415
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0739755874522587,
+      "learning_rate": 6.593186480593858e-06,
+      "loss": 0.9714,
+      "step": 26416
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6248893905755892,
+      "learning_rate": 6.592469095467456e-06,
+      "loss": 0.7856,
+      "step": 26417
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7482748978421463,
+      "learning_rate": 6.591751730180774e-06,
+      "loss": 0.8991,
+      "step": 26418
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7994642383619663,
+      "learning_rate": 6.591034384737992e-06,
+      "loss": 0.8782,
+      "step": 26419
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9753465149700173,
+      "learning_rate": 6.5903170591432896e-06,
+      "loss": 0.9363,
+      "step": 26420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4481559270826734,
+      "learning_rate": 6.58959975340084e-06,
+      "loss": 0.9412,
+      "step": 26421
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5512842704499292,
+      "learning_rate": 6.588882467514819e-06,
+      "loss": 0.8403,
+      "step": 26422
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8390307299993874,
+      "learning_rate": 6.588165201489405e-06,
+      "loss": 0.7943,
+      "step": 26423
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0564075438622,
+      "learning_rate": 6.587447955328775e-06,
+      "loss": 0.9218,
+      "step": 26424
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7300848625551575,
+      "learning_rate": 6.586730729037102e-06,
+      "loss": 0.921,
+      "step": 26425
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7694164504741536,
+      "learning_rate": 6.586013522618562e-06,
+      "loss": 0.734,
+      "step": 26426
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5675410407332582,
+      "learning_rate": 6.585296336077337e-06,
+      "loss": 1.0469,
+      "step": 26427
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.502358889925027,
+      "learning_rate": 6.584579169417592e-06,
+      "loss": 0.883,
+      "step": 26428
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7118075979665728,
+      "learning_rate": 6.58386202264351e-06,
+      "loss": 0.8929,
+      "step": 26429
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6077039818238086,
+      "learning_rate": 6.583144895759264e-06,
+      "loss": 0.8662,
+      "step": 26430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4966212924214846,
+      "learning_rate": 6.582427788769033e-06,
+      "loss": 0.7625,
+      "step": 26431
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7508201631993017,
+      "learning_rate": 6.581710701676989e-06,
+      "loss": 0.8222,
+      "step": 26432
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6069969594282865,
+      "learning_rate": 6.580993634487306e-06,
+      "loss": 0.9624,
+      "step": 26433
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5753187668405195,
+      "learning_rate": 6.580276587204163e-06,
+      "loss": 0.8389,
+      "step": 26434
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.368830483575791,
+      "learning_rate": 6.579559559831729e-06,
+      "loss": 0.8709,
+      "step": 26435
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6192736225032673,
+      "learning_rate": 6.5788425523741825e-06,
+      "loss": 0.9317,
+      "step": 26436
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.749438119272973,
+      "learning_rate": 6.578125564835699e-06,
+      "loss": 0.8884,
+      "step": 26437
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6662699212852883,
+      "learning_rate": 6.5774085972204535e-06,
+      "loss": 1.0148,
+      "step": 26438
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.632212872281307,
+      "learning_rate": 6.576691649532615e-06,
+      "loss": 0.8239,
+      "step": 26439
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6528217525951432,
+      "learning_rate": 6.575974721776364e-06,
+      "loss": 0.8412,
+      "step": 26440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.01139357711281,
+      "learning_rate": 6.575257813955873e-06,
+      "loss": 0.8428,
+      "step": 26441
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6482436319079976,
+      "learning_rate": 6.574540926075314e-06,
+      "loss": 0.8432,
+      "step": 26442
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5163478975199585,
+      "learning_rate": 6.57382405813886e-06,
+      "loss": 0.8459,
+      "step": 26443
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5229438594249578,
+      "learning_rate": 6.573107210150688e-06,
+      "loss": 0.8785,
+      "step": 26444
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6535358887625282,
+      "learning_rate": 6.5723903821149746e-06,
+      "loss": 0.9157,
+      "step": 26445
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6224992726809897,
+      "learning_rate": 6.5716735740358864e-06,
+      "loss": 0.8685,
+      "step": 26446
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4932978242813897,
+      "learning_rate": 6.570956785917599e-06,
+      "loss": 0.8375,
+      "step": 26447
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6881311041266476,
+      "learning_rate": 6.57024001776429e-06,
+      "loss": 0.8729,
+      "step": 26448
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4768090537982168,
+      "learning_rate": 6.569523269580127e-06,
+      "loss": 0.941,
+      "step": 26449
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4750800511308095,
+      "learning_rate": 6.568806541369287e-06,
+      "loss": 0.836,
+      "step": 26450
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6633395077053017,
+      "learning_rate": 6.568089833135939e-06,
+      "loss": 0.9341,
+      "step": 26451
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6439575346989155,
+      "learning_rate": 6.567373144884262e-06,
+      "loss": 0.7214,
+      "step": 26452
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.182780673866917,
+      "learning_rate": 6.566656476618424e-06,
+      "loss": 0.9212,
+      "step": 26453
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5465129019045494,
+      "learning_rate": 6.565939828342598e-06,
+      "loss": 0.839,
+      "step": 26454
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8195063142085576,
+      "learning_rate": 6.565223200060961e-06,
+      "loss": 0.6797,
+      "step": 26455
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7883046348007958,
+      "learning_rate": 6.564506591777677e-06,
+      "loss": 0.722,
+      "step": 26456
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.597797921226838,
+      "learning_rate": 6.563790003496926e-06,
+      "loss": 0.6938,
+      "step": 26457
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7438701987121709,
+      "learning_rate": 6.563073435222874e-06,
+      "loss": 0.7142,
+      "step": 26458
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7669489643597989,
+      "learning_rate": 6.562356886959704e-06,
+      "loss": 0.7763,
+      "step": 26459
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7231652461125855,
+      "learning_rate": 6.561640358711574e-06,
+      "loss": 0.8366,
+      "step": 26460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6697731915429004,
+      "learning_rate": 6.560923850482665e-06,
+      "loss": 0.7996,
+      "step": 26461
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.499987031975584,
+      "learning_rate": 6.560207362277146e-06,
+      "loss": 0.7816,
+      "step": 26462
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7389273820069957,
+      "learning_rate": 6.55949089409919e-06,
+      "loss": 0.7974,
+      "step": 26463
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.582273977660443,
+      "learning_rate": 6.558774445952962e-06,
+      "loss": 0.7709,
+      "step": 26464
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.676289893743228,
+      "learning_rate": 6.558058017842643e-06,
+      "loss": 0.8673,
+      "step": 26465
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8601824342200226,
+      "learning_rate": 6.5573416097724e-06,
+      "loss": 0.8897,
+      "step": 26466
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.558962762292614,
+      "learning_rate": 6.5566252217464045e-06,
+      "loss": 0.8248,
+      "step": 26467
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8491728950009556,
+      "learning_rate": 6.555908853768824e-06,
+      "loss": 0.7966,
+      "step": 26468
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.577438304750828,
+      "learning_rate": 6.555192505843836e-06,
+      "loss": 0.954,
+      "step": 26469
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7321422296874762,
+      "learning_rate": 6.554476177975606e-06,
+      "loss": 0.9031,
+      "step": 26470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8076345882371687,
+      "learning_rate": 6.553759870168306e-06,
+      "loss": 0.8977,
+      "step": 26471
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.474816665937353,
+      "learning_rate": 6.553043582426106e-06,
+      "loss": 0.8034,
+      "step": 26472
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5871677696370714,
+      "learning_rate": 6.552327314753182e-06,
+      "loss": 0.7774,
+      "step": 26473
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.576835708422886,
+      "learning_rate": 6.551611067153698e-06,
+      "loss": 0.8962,
+      "step": 26474
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7366419567961908,
+      "learning_rate": 6.550894839631823e-06,
+      "loss": 0.7955,
+      "step": 26475
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7798418041897622,
+      "learning_rate": 6.550178632191737e-06,
+      "loss": 0.7146,
+      "step": 26476
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5385355238586993,
+      "learning_rate": 6.5494624448375964e-06,
+      "loss": 1.0007,
+      "step": 26477
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.557825661253788,
+      "learning_rate": 6.548746277573581e-06,
+      "loss": 0.973,
+      "step": 26478
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5347572735943698,
+      "learning_rate": 6.548030130403854e-06,
+      "loss": 0.8363,
+      "step": 26479
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7828622482521426,
+      "learning_rate": 6.5473140033325935e-06,
+      "loss": 0.7677,
+      "step": 26480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.814850944927068,
+      "learning_rate": 6.546597896363962e-06,
+      "loss": 0.8395,
+      "step": 26481
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6766530660589007,
+      "learning_rate": 6.54588180950213e-06,
+      "loss": 0.9052,
+      "step": 26482
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3569366178657156,
+      "learning_rate": 6.545165742751269e-06,
+      "loss": 0.8289,
+      "step": 26483
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.119882338973692,
+      "learning_rate": 6.544449696115545e-06,
+      "loss": 0.9677,
+      "step": 26484
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5229987139943608,
+      "learning_rate": 6.543733669599131e-06,
+      "loss": 0.8624,
+      "step": 26485
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8041669346050707,
+      "learning_rate": 6.543017663206192e-06,
+      "loss": 0.63,
+      "step": 26486
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3421030709747102,
+      "learning_rate": 6.542301676940901e-06,
+      "loss": 0.8238,
+      "step": 26487
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.570338607358738,
+      "learning_rate": 6.541585710807421e-06,
+      "loss": 0.8875,
+      "step": 26488
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8975185250203082,
+      "learning_rate": 6.540869764809925e-06,
+      "loss": 1.0295,
+      "step": 26489
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5405306137272428,
+      "learning_rate": 6.5401538389525835e-06,
+      "loss": 0.8021,
+      "step": 26490
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7928419393283388,
+      "learning_rate": 6.539437933239559e-06,
+      "loss": 0.9308,
+      "step": 26491
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5482329910775336,
+      "learning_rate": 6.53872204767502e-06,
+      "loss": 0.7837,
+      "step": 26492
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1380377007658562,
+      "learning_rate": 6.538006182263139e-06,
+      "loss": 0.7723,
+      "step": 26493
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4898748739529812,
+      "learning_rate": 6.537290337008085e-06,
+      "loss": 0.846,
+      "step": 26494
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8044875593350067,
+      "learning_rate": 6.536574511914021e-06,
+      "loss": 0.9156,
+      "step": 26495
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.717461683153009,
+      "learning_rate": 6.535858706985114e-06,
+      "loss": 0.9261,
+      "step": 26496
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7354972481003085,
+      "learning_rate": 6.5351429222255394e-06,
+      "loss": 0.7493,
+      "step": 26497
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7570682121138556,
+      "learning_rate": 6.5344271576394555e-06,
+      "loss": 0.7357,
+      "step": 26498
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6085116295049484,
+      "learning_rate": 6.533711413231036e-06,
+      "loss": 0.7392,
+      "step": 26499
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.671204346569947,
+      "learning_rate": 6.532995689004444e-06,
+      "loss": 0.8768,
+      "step": 26500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.363181828047402,
+      "learning_rate": 6.532279984963852e-06,
+      "loss": 0.8904,
+      "step": 26501
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5709988376294746,
+      "learning_rate": 6.531564301113422e-06,
+      "loss": 0.9376,
+      "step": 26502
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6033114731017306,
+      "learning_rate": 6.530848637457322e-06,
+      "loss": 0.8658,
+      "step": 26503
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6108227163482833,
+      "learning_rate": 6.530132993999723e-06,
+      "loss": 0.787,
+      "step": 26504
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5596221906316812,
+      "learning_rate": 6.529417370744783e-06,
+      "loss": 0.858,
+      "step": 26505
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8114356123772486,
+      "learning_rate": 6.528701767696676e-06,
+      "loss": 0.7939,
+      "step": 26506
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0134358688350678,
+      "learning_rate": 6.527986184859565e-06,
+      "loss": 0.777,
+      "step": 26507
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9507398278255574,
+      "learning_rate": 6.527270622237623e-06,
+      "loss": 0.9498,
+      "step": 26508
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.556273643358216,
+      "learning_rate": 6.526555079835004e-06,
+      "loss": 0.8882,
+      "step": 26509
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.69434473000397,
+      "learning_rate": 6.525839557655884e-06,
+      "loss": 0.9076,
+      "step": 26510
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4872332643922703,
+      "learning_rate": 6.525124055704425e-06,
+      "loss": 0.9136,
+      "step": 26511
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5465370646942123,
+      "learning_rate": 6.5244085739848e-06,
+      "loss": 0.798,
+      "step": 26512
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1940693016338018,
+      "learning_rate": 6.52369311250116e-06,
+      "loss": 0.9121,
+      "step": 26513
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.751773968930021,
+      "learning_rate": 6.522977671257683e-06,
+      "loss": 0.9741,
+      "step": 26514
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7488025114651131,
+      "learning_rate": 6.522262250258533e-06,
+      "loss": 0.7279,
+      "step": 26515
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7554853049687835,
+      "learning_rate": 6.521546849507872e-06,
+      "loss": 0.9259,
+      "step": 26516
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9926726448543366,
+      "learning_rate": 6.520831469009863e-06,
+      "loss": 0.8592,
+      "step": 26517
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0031425699559167,
+      "learning_rate": 6.520116108768677e-06,
+      "loss": 0.8261,
+      "step": 26518
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7664803999770042,
+      "learning_rate": 6.519400768788478e-06,
+      "loss": 0.8126,
+      "step": 26519
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8156489144788666,
+      "learning_rate": 6.518685449073427e-06,
+      "loss": 0.84,
+      "step": 26520
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6499471269924124,
+      "learning_rate": 6.517970149627692e-06,
+      "loss": 0.7745,
+      "step": 26521
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.699075137769631,
+      "learning_rate": 6.5172548704554405e-06,
+      "loss": 0.8134,
+      "step": 26522
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4113907142575566,
+      "learning_rate": 6.5165396115608315e-06,
+      "loss": 0.6942,
+      "step": 26523
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8559341869118786,
+      "learning_rate": 6.51582437294803e-06,
+      "loss": 0.689,
+      "step": 26524
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6056486553982874,
+      "learning_rate": 6.5151091546212034e-06,
+      "loss": 0.9199,
+      "step": 26525
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.907856655793657,
+      "learning_rate": 6.5143939565845175e-06,
+      "loss": 0.8427,
+      "step": 26526
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.786897329480874,
+      "learning_rate": 6.513678778842129e-06,
+      "loss": 0.7286,
+      "step": 26527
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.497499802729416,
+      "learning_rate": 6.512963621398207e-06,
+      "loss": 0.7487,
+      "step": 26528
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7391195817987288,
+      "learning_rate": 6.512248484256918e-06,
+      "loss": 0.6964,
+      "step": 26529
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.572220570536254,
+      "learning_rate": 6.51153336742242e-06,
+      "loss": 0.8108,
+      "step": 26530
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6193205951773766,
+      "learning_rate": 6.51081827089888e-06,
+      "loss": 0.8321,
+      "step": 26531
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5581800994681871,
+      "learning_rate": 6.510103194690459e-06,
+      "loss": 0.9548,
+      "step": 26532
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7948334636332868,
+      "learning_rate": 6.509388138801326e-06,
+      "loss": 0.8842,
+      "step": 26533
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4924670080986195,
+      "learning_rate": 6.508673103235637e-06,
+      "loss": 0.9295,
+      "step": 26534
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.644499539598518,
+      "learning_rate": 6.507958087997559e-06,
+      "loss": 0.9612,
+      "step": 26535
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6572684073475963,
+      "learning_rate": 6.507243093091258e-06,
+      "loss": 0.757,
+      "step": 26536
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5756357224768074,
+      "learning_rate": 6.506528118520888e-06,
+      "loss": 0.9722,
+      "step": 26537
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9799140710924834,
+      "learning_rate": 6.50581316429062e-06,
+      "loss": 0.891,
+      "step": 26538
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8008953818538439,
+      "learning_rate": 6.505098230404614e-06,
+      "loss": 0.7865,
+      "step": 26539
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9261878371894878,
+      "learning_rate": 6.504383316867035e-06,
+      "loss": 0.8942,
+      "step": 26540
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5852296301097653,
+      "learning_rate": 6.503668423682037e-06,
+      "loss": 0.7073,
+      "step": 26541
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7822045799264863,
+      "learning_rate": 6.502953550853792e-06,
+      "loss": 0.8315,
+      "step": 26542
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5104048198288107,
+      "learning_rate": 6.50223869838646e-06,
+      "loss": 0.8411,
+      "step": 26543
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9982070105610963,
+      "learning_rate": 6.501523866284201e-06,
+      "loss": 0.8796,
+      "step": 26544
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7427700023566235,
+      "learning_rate": 6.500809054551173e-06,
+      "loss": 0.8021,
+      "step": 26545
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5219952888086417,
+      "learning_rate": 6.500094263191548e-06,
+      "loss": 0.8079,
+      "step": 26546
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4698616703957854,
+      "learning_rate": 6.499379492209481e-06,
+      "loss": 1.0378,
+      "step": 26547
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7789371799877377,
+      "learning_rate": 6.498664741609135e-06,
+      "loss": 0.7777,
+      "step": 26548
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8490801774709191,
+      "learning_rate": 6.497950011394669e-06,
+      "loss": 0.9037,
+      "step": 26549
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6635709641967527,
+      "learning_rate": 6.497235301570251e-06,
+      "loss": 0.8243,
+      "step": 26550
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4584206427244286,
+      "learning_rate": 6.496520612140033e-06,
+      "loss": 0.7499,
+      "step": 26551
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7033066101111289,
+      "learning_rate": 6.4958059431081845e-06,
+      "loss": 0.7198,
+      "step": 26552
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.769071217901533,
+      "learning_rate": 6.4950912944788614e-06,
+      "loss": 0.866,
+      "step": 26553
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.634849737754667,
+      "learning_rate": 6.494376666256229e-06,
+      "loss": 0.8755,
+      "step": 26554
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6736643694681421,
+      "learning_rate": 6.493662058444443e-06,
+      "loss": 0.7833,
+      "step": 26555
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.428907773148795,
+      "learning_rate": 6.492947471047666e-06,
+      "loss": 0.8336,
+      "step": 26556
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.45611359989288,
+      "learning_rate": 6.492232904070065e-06,
+      "loss": 0.8074,
+      "step": 26557
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6200857015703818,
+      "learning_rate": 6.491518357515788e-06,
+      "loss": 0.9381,
+      "step": 26558
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7498084462317922,
+      "learning_rate": 6.490803831389003e-06,
+      "loss": 0.8302,
+      "step": 26559
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6525126849348013,
+      "learning_rate": 6.490089325693869e-06,
+      "loss": 0.872,
+      "step": 26560
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6743885798179927,
+      "learning_rate": 6.489374840434551e-06,
+      "loss": 0.8601,
+      "step": 26561
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4487381781995539,
+      "learning_rate": 6.4886603756151985e-06,
+      "loss": 0.8068,
+      "step": 26562
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1029565843870217,
+      "learning_rate": 6.487945931239978e-06,
+      "loss": 0.8413,
+      "step": 26563
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6134296947441131,
+      "learning_rate": 6.487231507313052e-06,
+      "loss": 0.7648,
+      "step": 26564
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4799550725484663,
+      "learning_rate": 6.486517103838574e-06,
+      "loss": 0.8791,
+      "step": 26565
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6339987592566436,
+      "learning_rate": 6.485802720820702e-06,
+      "loss": 1.0307,
+      "step": 26566
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6791788736151074,
+      "learning_rate": 6.485088358263603e-06,
+      "loss": 0.8681,
+      "step": 26567
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.643790323592242,
+      "learning_rate": 6.4843740161714334e-06,
+      "loss": 0.8724,
+      "step": 26568
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2250972384288565,
+      "learning_rate": 6.483659694548348e-06,
+      "loss": 0.7781,
+      "step": 26569
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5491464267200328,
+      "learning_rate": 6.4829453933985096e-06,
+      "loss": 0.8692,
+      "step": 26570
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6606142349018687,
+      "learning_rate": 6.482231112726081e-06,
+      "loss": 0.798,
+      "step": 26571
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4956867166448156,
+      "learning_rate": 6.4815168525352124e-06,
+      "loss": 0.8177,
+      "step": 26572
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.843638564454651,
+      "learning_rate": 6.4808026128300665e-06,
+      "loss": 0.8856,
+      "step": 26573
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6576247014494228,
+      "learning_rate": 6.480088393614802e-06,
+      "loss": 0.7732,
+      "step": 26574
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8052740512375025,
+      "learning_rate": 6.4793741948935795e-06,
+      "loss": 0.7605,
+      "step": 26575
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6016909427500063,
+      "learning_rate": 6.478660016670555e-06,
+      "loss": 0.839,
+      "step": 26576
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6110790669075261,
+      "learning_rate": 6.4779458589498835e-06,
+      "loss": 0.8504,
+      "step": 26577
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8651113033416826,
+      "learning_rate": 6.477231721735729e-06,
+      "loss": 0.8526,
+      "step": 26578
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7086260110596285,
+      "learning_rate": 6.4765176050322465e-06,
+      "loss": 0.8638,
+      "step": 26579
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6169785893591075,
+      "learning_rate": 6.4758035088435935e-06,
+      "loss": 0.8105,
+      "step": 26580
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6228250927509005,
+      "learning_rate": 6.475089433173926e-06,
+      "loss": 0.7119,
+      "step": 26581
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7691880314872612,
+      "learning_rate": 6.474375378027408e-06,
+      "loss": 0.6832,
+      "step": 26582
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.694480399705279,
+      "learning_rate": 6.473661343408191e-06,
+      "loss": 0.7576,
+      "step": 26583
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6027076050390079,
+      "learning_rate": 6.472947329320434e-06,
+      "loss": 0.8988,
+      "step": 26584
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4010969276040957,
+      "learning_rate": 6.472233335768295e-06,
+      "loss": 0.8896,
+      "step": 26585
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5502620182927738,
+      "learning_rate": 6.471519362755927e-06,
+      "loss": 0.8428,
+      "step": 26586
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6979881495984599,
+      "learning_rate": 6.470805410287493e-06,
+      "loss": 0.5995,
+      "step": 26587
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0475453756338258,
+      "learning_rate": 6.470091478367147e-06,
+      "loss": 0.7754,
+      "step": 26588
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5679610849739234,
+      "learning_rate": 6.469377566999049e-06,
+      "loss": 0.8641,
+      "step": 26589
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4439855706186298,
+      "learning_rate": 6.468663676187346e-06,
+      "loss": 0.8673,
+      "step": 26590
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.560100624028711,
+      "learning_rate": 6.467949805936206e-06,
+      "loss": 0.8949,
+      "step": 26591
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0070500779209617,
+      "learning_rate": 6.46723595624978e-06,
+      "loss": 0.8069,
+      "step": 26592
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8140648552545394,
+      "learning_rate": 6.4665221271322245e-06,
+      "loss": 0.7231,
+      "step": 26593
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7286523100520237,
+      "learning_rate": 6.465808318587693e-06,
+      "loss": 0.8968,
+      "step": 26594
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7360011946088247,
+      "learning_rate": 6.465094530620346e-06,
+      "loss": 0.8541,
+      "step": 26595
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5557193302899355,
+      "learning_rate": 6.464380763234341e-06,
+      "loss": 0.9252,
+      "step": 26596
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5192245176916603,
+      "learning_rate": 6.463667016433828e-06,
+      "loss": 0.8637,
+      "step": 26597
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0135448071416913,
+      "learning_rate": 6.462953290222963e-06,
+      "loss": 0.84,
+      "step": 26598
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6755141298768481,
+      "learning_rate": 6.462239584605907e-06,
+      "loss": 0.865,
+      "step": 26599
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3333270501765266,
+      "learning_rate": 6.461525899586812e-06,
+      "loss": 0.875,
+      "step": 26600
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8565429138727167,
+      "learning_rate": 6.460812235169832e-06,
+      "loss": 0.9446,
+      "step": 26601
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7460006839083768,
+      "learning_rate": 6.460098591359123e-06,
+      "loss": 0.6934,
+      "step": 26602
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.001910084285013,
+      "learning_rate": 6.459384968158844e-06,
+      "loss": 0.8402,
+      "step": 26603
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5443365896562395,
+      "learning_rate": 6.458671365573144e-06,
+      "loss": 0.8902,
+      "step": 26604
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5937180878394424,
+      "learning_rate": 6.457957783606179e-06,
+      "loss": 0.8793,
+      "step": 26605
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.584506207505449,
+      "learning_rate": 6.457244222262111e-06,
+      "loss": 0.8324,
+      "step": 26606
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5324695836739661,
+      "learning_rate": 6.456530681545083e-06,
+      "loss": 0.8901,
+      "step": 26607
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8804696145822601,
+      "learning_rate": 6.455817161459257e-06,
+      "loss": 0.8176,
+      "step": 26608
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7499330808319675,
+      "learning_rate": 6.455103662008784e-06,
+      "loss": 0.806,
+      "step": 26609
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6024813699769156,
+      "learning_rate": 6.4543901831978254e-06,
+      "loss": 0.9333,
+      "step": 26610
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8190194327513781,
+      "learning_rate": 6.453676725030524e-06,
+      "loss": 0.8047,
+      "step": 26611
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4703407420167542,
+      "learning_rate": 6.452963287511041e-06,
+      "loss": 0.8568,
+      "step": 26612
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8953253256464764,
+      "learning_rate": 6.452249870643531e-06,
+      "loss": 0.8478,
+      "step": 26613
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.440942545488605,
+      "learning_rate": 6.451536474432143e-06,
+      "loss": 0.7566,
+      "step": 26614
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6183287957055168,
+      "learning_rate": 6.450823098881032e-06,
+      "loss": 0.8281,
+      "step": 26615
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5263349284046095,
+      "learning_rate": 6.450109743994353e-06,
+      "loss": 0.7116,
+      "step": 26616
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7598902028973121,
+      "learning_rate": 6.449396409776263e-06,
+      "loss": 0.7667,
+      "step": 26617
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6835085146870257,
+      "learning_rate": 6.448683096230906e-06,
+      "loss": 0.8782,
+      "step": 26618
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7828592076355698,
+      "learning_rate": 6.447969803362442e-06,
+      "loss": 0.8185,
+      "step": 26619
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1747499823407868,
+      "learning_rate": 6.447256531175026e-06,
+      "loss": 0.7806,
+      "step": 26620
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.77751317320396,
+      "learning_rate": 6.446543279672803e-06,
+      "loss": 0.9346,
+      "step": 26621
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5395617691087113,
+      "learning_rate": 6.445830048859929e-06,
+      "loss": 0.9728,
+      "step": 26622
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5665864375206127,
+      "learning_rate": 6.445116838740559e-06,
+      "loss": 0.9559,
+      "step": 26623
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.608702486832898,
+      "learning_rate": 6.444403649318849e-06,
+      "loss": 0.7296,
+      "step": 26624
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.710041072024793,
+      "learning_rate": 6.443690480598942e-06,
+      "loss": 0.7857,
+      "step": 26625
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4037495663920982,
+      "learning_rate": 6.4429773325849935e-06,
+      "loss": 0.7718,
+      "step": 26626
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.58475434129591,
+      "learning_rate": 6.442264205281161e-06,
+      "loss": 0.801,
+      "step": 26627
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5418987236784993,
+      "learning_rate": 6.441551098691591e-06,
+      "loss": 0.7447,
+      "step": 26628
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8586764482399507,
+      "learning_rate": 6.440838012820437e-06,
+      "loss": 0.7617,
+      "step": 26629
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5081198050953144,
+      "learning_rate": 6.440124947671848e-06,
+      "loss": 0.8008,
+      "step": 26630
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8585245732167446,
+      "learning_rate": 6.439411903249983e-06,
+      "loss": 0.8282,
+      "step": 26631
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.622098987606935,
+      "learning_rate": 6.438698879558989e-06,
+      "loss": 0.8834,
+      "step": 26632
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7358478104589479,
+      "learning_rate": 6.437985876603017e-06,
+      "loss": 0.6855,
+      "step": 26633
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.604078829988128,
+      "learning_rate": 6.437272894386219e-06,
+      "loss": 0.8933,
+      "step": 26634
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4769966377036987,
+      "learning_rate": 6.4365599329127435e-06,
+      "loss": 0.8071,
+      "step": 26635
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7126309274155864,
+      "learning_rate": 6.435846992186746e-06,
+      "loss": 0.8955,
+      "step": 26636
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7023073847918222,
+      "learning_rate": 6.435134072212377e-06,
+      "loss": 0.7143,
+      "step": 26637
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7788854910129737,
+      "learning_rate": 6.4344211729937855e-06,
+      "loss": 0.8133,
+      "step": 26638
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.377552098149914,
+      "learning_rate": 6.43370829453512e-06,
+      "loss": 0.8554,
+      "step": 26639
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8070062446448456,
+      "learning_rate": 6.432995436840537e-06,
+      "loss": 0.908,
+      "step": 26640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6867445250925044,
+      "learning_rate": 6.43228259991418e-06,
+      "loss": 0.8769,
+      "step": 26641
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6479194966390969,
+      "learning_rate": 6.43156978376021e-06,
+      "loss": 0.877,
+      "step": 26642
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5850225690835926,
+      "learning_rate": 6.430856988382766e-06,
+      "loss": 0.8923,
+      "step": 26643
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.161432571737168,
+      "learning_rate": 6.430144213786002e-06,
+      "loss": 0.8492,
+      "step": 26644
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7598582187269607,
+      "learning_rate": 6.4294314599740715e-06,
+      "loss": 0.8762,
+      "step": 26645
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7502260737799412,
+      "learning_rate": 6.42871872695112e-06,
+      "loss": 0.7351,
+      "step": 26646
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6337411156669965,
+      "learning_rate": 6.428006014721295e-06,
+      "loss": 0.9236,
+      "step": 26647
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7843087507688826,
+      "learning_rate": 6.427293323288754e-06,
+      "loss": 0.711,
+      "step": 26648
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6765527008406924,
+      "learning_rate": 6.426580652657643e-06,
+      "loss": 0.9135,
+      "step": 26649
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.768605014069436,
+      "learning_rate": 6.4258680028321094e-06,
+      "loss": 0.8253,
+      "step": 26650
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8933352349194748,
+      "learning_rate": 6.425155373816303e-06,
+      "loss": 0.8498,
+      "step": 26651
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5642330334316406,
+      "learning_rate": 6.424442765614376e-06,
+      "loss": 0.8395,
+      "step": 26652
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7589813777340261,
+      "learning_rate": 6.423730178230474e-06,
+      "loss": 0.7182,
+      "step": 26653
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9092164565043301,
+      "learning_rate": 6.423017611668745e-06,
+      "loss": 0.8897,
+      "step": 26654
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7498057777982812,
+      "learning_rate": 6.422305065933342e-06,
+      "loss": 0.9457,
+      "step": 26655
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.479609456132025,
+      "learning_rate": 6.421592541028414e-06,
+      "loss": 0.8807,
+      "step": 26656
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.482480460574859,
+      "learning_rate": 6.420880036958103e-06,
+      "loss": 0.9048,
+      "step": 26657
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.945557913540827,
+      "learning_rate": 6.4201675537265615e-06,
+      "loss": 0.9796,
+      "step": 26658
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.57786293752238,
+      "learning_rate": 6.419455091337944e-06,
+      "loss": 0.8489,
+      "step": 26659
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7805859196478708,
+      "learning_rate": 6.4187426497963855e-06,
+      "loss": 1.0315,
+      "step": 26660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5236560570575863,
+      "learning_rate": 6.418030229106045e-06,
+      "loss": 0.866,
+      "step": 26661
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5061769805724887,
+      "learning_rate": 6.417317829271062e-06,
+      "loss": 0.7693,
+      "step": 26662
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6202347547433633,
+      "learning_rate": 6.4166054502955965e-06,
+      "loss": 0.9297,
+      "step": 26663
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6417546578338484,
+      "learning_rate": 6.415893092183782e-06,
+      "loss": 0.7934,
+      "step": 26664
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7633056082039951,
+      "learning_rate": 6.415180754939775e-06,
+      "loss": 0.6605,
+      "step": 26665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8121834359116289,
+      "learning_rate": 6.414468438567722e-06,
+      "loss": 0.7106,
+      "step": 26666
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.720147702637674,
+      "learning_rate": 6.413756143071765e-06,
+      "loss": 0.7983,
+      "step": 26667
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5711718469090843,
+      "learning_rate": 6.413043868456057e-06,
+      "loss": 0.8366,
+      "step": 26668
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8168982166265977,
+      "learning_rate": 6.4123316147247435e-06,
+      "loss": 0.8236,
+      "step": 26669
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6467279289101757,
+      "learning_rate": 6.4116193818819735e-06,
+      "loss": 0.8722,
+      "step": 26670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9427128263811249,
+      "learning_rate": 6.410907169931888e-06,
+      "loss": 0.8402,
+      "step": 26671
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7581467510805914,
+      "learning_rate": 6.410194978878638e-06,
+      "loss": 0.7249,
+      "step": 26672
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6853901544983356,
+      "learning_rate": 6.409482808726372e-06,
+      "loss": 0.7978,
+      "step": 26673
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1608023271618455,
+      "learning_rate": 6.408770659479231e-06,
+      "loss": 0.772,
+      "step": 26674
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4403077075517556,
+      "learning_rate": 6.408058531141363e-06,
+      "loss": 0.8365,
+      "step": 26675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7036884320464618,
+      "learning_rate": 6.407346423716918e-06,
+      "loss": 0.8199,
+      "step": 26676
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7373693333242077,
+      "learning_rate": 6.406634337210041e-06,
+      "loss": 0.7617,
+      "step": 26677
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5118671977437554,
+      "learning_rate": 6.405922271624874e-06,
+      "loss": 0.9681,
+      "step": 26678
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.747824009680792,
+      "learning_rate": 6.405210226965564e-06,
+      "loss": 0.7685,
+      "step": 26679
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9696528199197816,
+      "learning_rate": 6.404498203236261e-06,
+      "loss": 0.9212,
+      "step": 26680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4323889613994532,
+      "learning_rate": 6.403786200441106e-06,
+      "loss": 0.806,
+      "step": 26681
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7986783461273366,
+      "learning_rate": 6.403074218584248e-06,
+      "loss": 0.7473,
+      "step": 26682
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9071231673685773,
+      "learning_rate": 6.402362257669827e-06,
+      "loss": 0.7647,
+      "step": 26683
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.639255762667785,
+      "learning_rate": 6.401650317701997e-06,
+      "loss": 0.8758,
+      "step": 26684
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8543701904504784,
+      "learning_rate": 6.400938398684895e-06,
+      "loss": 0.8804,
+      "step": 26685
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8259348016917845,
+      "learning_rate": 6.40022650062267e-06,
+      "loss": 0.86,
+      "step": 26686
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6953162319470785,
+      "learning_rate": 6.399514623519467e-06,
+      "loss": 0.9234,
+      "step": 26687
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7749751997257914,
+      "learning_rate": 6.398802767379426e-06,
+      "loss": 0.9295,
+      "step": 26688
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7198347241822591,
+      "learning_rate": 6.398090932206698e-06,
+      "loss": 0.9046,
+      "step": 26689
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4235140977008653,
+      "learning_rate": 6.397379118005423e-06,
+      "loss": 0.9278,
+      "step": 26690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4061638686919664,
+      "learning_rate": 6.396667324779754e-06,
+      "loss": 0.7765,
+      "step": 26691
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0562499767560003,
+      "learning_rate": 6.395955552533821e-06,
+      "loss": 0.9,
+      "step": 26692
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.161010704073927,
+      "learning_rate": 6.395243801271779e-06,
+      "loss": 0.7962,
+      "step": 26693
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5972682399304456,
+      "learning_rate": 6.39453207099777e-06,
+      "loss": 0.8365,
+      "step": 26694
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7226723086632847,
+      "learning_rate": 6.393820361715935e-06,
+      "loss": 0.9261,
+      "step": 26695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4664908110951338,
+      "learning_rate": 6.393108673430418e-06,
+      "loss": 1.0196,
+      "step": 26696
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.533107149235577,
+      "learning_rate": 6.392397006145366e-06,
+      "loss": 0.7245,
+      "step": 26697
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6599725945650605,
+      "learning_rate": 6.391685359864923e-06,
+      "loss": 0.8325,
+      "step": 26698
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.636390436986459,
+      "learning_rate": 6.390973734593227e-06,
+      "loss": 0.8477,
+      "step": 26699
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6178873018947095,
+      "learning_rate": 6.390262130334424e-06,
+      "loss": 0.8664,
+      "step": 26700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5563627135595979,
+      "learning_rate": 6.389550547092661e-06,
+      "loss": 0.9435,
+      "step": 26701
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5063659438873305,
+      "learning_rate": 6.388838984872077e-06,
+      "loss": 0.736,
+      "step": 26702
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5008691775172454,
+      "learning_rate": 6.388127443676811e-06,
+      "loss": 0.8278,
+      "step": 26703
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4612167998316845,
+      "learning_rate": 6.3874159235110144e-06,
+      "loss": 0.7097,
+      "step": 26704
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5477689068330167,
+      "learning_rate": 6.386704424378828e-06,
+      "loss": 0.878,
+      "step": 26705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5489799973857679,
+      "learning_rate": 6.38599294628439e-06,
+      "loss": 0.7661,
+      "step": 26706
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9919588334667393,
+      "learning_rate": 6.385281489231843e-06,
+      "loss": 0.8592,
+      "step": 26707
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5447960170019672,
+      "learning_rate": 6.384570053225338e-06,
+      "loss": 0.6839,
+      "step": 26708
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5745639995000684,
+      "learning_rate": 6.383858638269005e-06,
+      "loss": 0.8768,
+      "step": 26709
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4214356097912626,
+      "learning_rate": 6.383147244366994e-06,
+      "loss": 0.8605,
+      "step": 26710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7921474892151974,
+      "learning_rate": 6.382435871523442e-06,
+      "loss": 0.8264,
+      "step": 26711
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5643523236010979,
+      "learning_rate": 6.3817245197425005e-06,
+      "loss": 0.83,
+      "step": 26712
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.220003881078588,
+      "learning_rate": 6.3810131890282975e-06,
+      "loss": 1.0772,
+      "step": 26713
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5042754127938127,
+      "learning_rate": 6.3803018793849835e-06,
+      "loss": 0.8182,
+      "step": 26714
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6142499248354627,
+      "learning_rate": 6.3795905908167e-06,
+      "loss": 0.9163,
+      "step": 26715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6785767191699144,
+      "learning_rate": 6.378879323327583e-06,
+      "loss": 0.9084,
+      "step": 26716
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2384392289543755,
+      "learning_rate": 6.378168076921778e-06,
+      "loss": 0.9427,
+      "step": 26717
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8475747234286484,
+      "learning_rate": 6.377456851603426e-06,
+      "loss": 0.8798,
+      "step": 26718
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7774090044748259,
+      "learning_rate": 6.37674564737667e-06,
+      "loss": 0.8536,
+      "step": 26719
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.621566613298339,
+      "learning_rate": 6.376034464245641e-06,
+      "loss": 0.8969,
+      "step": 26720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7917187243388883,
+      "learning_rate": 6.375323302214491e-06,
+      "loss": 0.7244,
+      "step": 26721
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7685116890389885,
+      "learning_rate": 6.374612161287358e-06,
+      "loss": 0.7716,
+      "step": 26722
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7924119284322227,
+      "learning_rate": 6.373901041468378e-06,
+      "loss": 0.6893,
+      "step": 26723
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5636044424420963,
+      "learning_rate": 6.373189942761693e-06,
+      "loss": 0.8139,
+      "step": 26724
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6025295085116584,
+      "learning_rate": 6.372478865171446e-06,
+      "loss": 0.8144,
+      "step": 26725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5100638780403968,
+      "learning_rate": 6.371767808701778e-06,
+      "loss": 0.9352,
+      "step": 26726
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4331468464484634,
+      "learning_rate": 6.371056773356823e-06,
+      "loss": 0.7716,
+      "step": 26727
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.520972188789286,
+      "learning_rate": 6.370345759140724e-06,
+      "loss": 0.7926,
+      "step": 26728
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5175817667090172,
+      "learning_rate": 6.369634766057624e-06,
+      "loss": 0.9474,
+      "step": 26729
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9544550893671113,
+      "learning_rate": 6.368923794111659e-06,
+      "loss": 0.8345,
+      "step": 26730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5843725984012489,
+      "learning_rate": 6.368212843306968e-06,
+      "loss": 0.839,
+      "step": 26731
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5677786348641862,
+      "learning_rate": 6.36750191364769e-06,
+      "loss": 0.8949,
+      "step": 26732
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5759125795049376,
+      "learning_rate": 6.36679100513797e-06,
+      "loss": 0.9207,
+      "step": 26733
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6090494047221557,
+      "learning_rate": 6.366080117781941e-06,
+      "loss": 0.9667,
+      "step": 26734
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7850289656401143,
+      "learning_rate": 6.365369251583744e-06,
+      "loss": 0.7118,
+      "step": 26735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.862415810840571,
+      "learning_rate": 6.364658406547519e-06,
+      "loss": 0.736,
+      "step": 26736
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8947866761675496,
+      "learning_rate": 6.363947582677401e-06,
+      "loss": 0.8097,
+      "step": 26737
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.070626795152611,
+      "learning_rate": 6.363236779977533e-06,
+      "loss": 0.8959,
+      "step": 26738
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6138838607755948,
+      "learning_rate": 6.362525998452051e-06,
+      "loss": 0.6774,
+      "step": 26739
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5525418003191065,
+      "learning_rate": 6.3618152381050975e-06,
+      "loss": 0.7708,
+      "step": 26740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6405420468105685,
+      "learning_rate": 6.3611044989408046e-06,
+      "loss": 0.8682,
+      "step": 26741
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6531864314737261,
+      "learning_rate": 6.360393780963313e-06,
+      "loss": 0.9499,
+      "step": 26742
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6889567954322988,
+      "learning_rate": 6.359683084176764e-06,
+      "loss": 0.9321,
+      "step": 26743
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.684381942617773,
+      "learning_rate": 6.358972408585291e-06,
+      "loss": 0.9025,
+      "step": 26744
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6036814046807637,
+      "learning_rate": 6.358261754193032e-06,
+      "loss": 0.8638,
+      "step": 26745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9540728610740268,
+      "learning_rate": 6.357551121004127e-06,
+      "loss": 0.8427,
+      "step": 26746
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6661419879768038,
+      "learning_rate": 6.3568405090227125e-06,
+      "loss": 0.8176,
+      "step": 26747
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6073594857958524,
+      "learning_rate": 6.356129918252927e-06,
+      "loss": 0.7754,
+      "step": 26748
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.931371880481767,
+      "learning_rate": 6.3554193486989035e-06,
+      "loss": 0.7767,
+      "step": 26749
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7722640239804985,
+      "learning_rate": 6.354708800364786e-06,
+      "loss": 0.7129,
+      "step": 26750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6991991184459834,
+      "learning_rate": 6.353998273254707e-06,
+      "loss": 0.7006,
+      "step": 26751
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7009369167165171,
+      "learning_rate": 6.353287767372802e-06,
+      "loss": 1.0766,
+      "step": 26752
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5253492516415121,
+      "learning_rate": 6.3525772827232115e-06,
+      "loss": 0.808,
+      "step": 26753
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6632395238427666,
+      "learning_rate": 6.351866819310073e-06,
+      "loss": 0.7695,
+      "step": 26754
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.302413709764508,
+      "learning_rate": 6.35115637713752e-06,
+      "loss": 0.9548,
+      "step": 26755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5292451145505208,
+      "learning_rate": 6.350445956209687e-06,
+      "loss": 0.8793,
+      "step": 26756
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4360914837639558,
+      "learning_rate": 6.349735556530718e-06,
+      "loss": 0.8916,
+      "step": 26757
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5136203731150657,
+      "learning_rate": 6.34902517810474e-06,
+      "loss": 0.8046,
+      "step": 26758
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8011042370775652,
+      "learning_rate": 6.348314820935895e-06,
+      "loss": 0.819,
+      "step": 26759
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3630861313974227,
+      "learning_rate": 6.347604485028314e-06,
+      "loss": 0.8343,
+      "step": 26760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7454676109010565,
+      "learning_rate": 6.346894170386143e-06,
+      "loss": 0.9136,
+      "step": 26761
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.797964021437683,
+      "learning_rate": 6.346183877013505e-06,
+      "loss": 0.856,
+      "step": 26762
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5166832355899074,
+      "learning_rate": 6.345473604914542e-06,
+      "loss": 0.8429,
+      "step": 26763
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7655974920573062,
+      "learning_rate": 6.344763354093388e-06,
+      "loss": 0.6798,
+      "step": 26764
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5679390435773735,
+      "learning_rate": 6.344053124554183e-06,
+      "loss": 0.9324,
+      "step": 26765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.345175599760703,
+      "learning_rate": 6.343342916301055e-06,
+      "loss": 0.6899,
+      "step": 26766
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5911827587058274,
+      "learning_rate": 6.342632729338143e-06,
+      "loss": 0.8932,
+      "step": 26767
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5497035139514685,
+      "learning_rate": 6.341922563669582e-06,
+      "loss": 0.7869,
+      "step": 26768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.028157423396737,
+      "learning_rate": 6.341212419299503e-06,
+      "loss": 0.9372,
+      "step": 26769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.514404997351929,
+      "learning_rate": 6.340502296232046e-06,
+      "loss": 0.9357,
+      "step": 26770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7623472873955826,
+      "learning_rate": 6.3397921944713435e-06,
+      "loss": 0.7197,
+      "step": 26771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7816387253252792,
+      "learning_rate": 6.339082114021531e-06,
+      "loss": 0.7339,
+      "step": 26772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6840494105165178,
+      "learning_rate": 6.3383720548867366e-06,
+      "loss": 0.912,
+      "step": 26773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.443405534228132,
+      "learning_rate": 6.3376620170711026e-06,
+      "loss": 0.8007,
+      "step": 26774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4803749286282164,
+      "learning_rate": 6.336952000578761e-06,
+      "loss": 0.8813,
+      "step": 26775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6322458307691594,
+      "learning_rate": 6.336242005413843e-06,
+      "loss": 0.9191,
+      "step": 26776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5741999664381066,
+      "learning_rate": 6.335532031580482e-06,
+      "loss": 0.8438,
+      "step": 26777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.800192179970868,
+      "learning_rate": 6.334822079082815e-06,
+      "loss": 0.9464,
+      "step": 26778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.6403144524202355,
+      "learning_rate": 6.334112147924978e-06,
+      "loss": 0.9869,
+      "step": 26779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4467809866091068,
+      "learning_rate": 6.333402238111097e-06,
+      "loss": 0.7984,
+      "step": 26780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5083212076176624,
+      "learning_rate": 6.332692349645307e-06,
+      "loss": 0.9321,
+      "step": 26781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4878792177703755,
+      "learning_rate": 6.331982482531748e-06,
+      "loss": 0.8239,
+      "step": 26782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9763876949350527,
+      "learning_rate": 6.331272636774546e-06,
+      "loss": 0.8491,
+      "step": 26783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.605587507771773,
+      "learning_rate": 6.330562812377836e-06,
+      "loss": 0.8199,
+      "step": 26784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.918177147638858,
+      "learning_rate": 6.329853009345748e-06,
+      "loss": 0.9154,
+      "step": 26785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.836467392031511,
+      "learning_rate": 6.329143227682423e-06,
+      "loss": 0.9943,
+      "step": 26786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6051834815163166,
+      "learning_rate": 6.328433467391986e-06,
+      "loss": 0.7986,
+      "step": 26787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.798748760311861,
+      "learning_rate": 6.3277237284785695e-06,
+      "loss": 1.0483,
+      "step": 26788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6989427042910556,
+      "learning_rate": 6.327014010946315e-06,
+      "loss": 0.834,
+      "step": 26789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6366154834378124,
+      "learning_rate": 6.326304314799342e-06,
+      "loss": 0.8083,
+      "step": 26790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5378064968183671,
+      "learning_rate": 6.32559464004179e-06,
+      "loss": 0.852,
+      "step": 26791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5569163777681974,
+      "learning_rate": 6.324884986677786e-06,
+      "loss": 0.9253,
+      "step": 26792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.103142399737182,
+      "learning_rate": 6.324175354711473e-06,
+      "loss": 0.8909,
+      "step": 26793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5242768431862572,
+      "learning_rate": 6.323465744146967e-06,
+      "loss": 0.8792,
+      "step": 26794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7502743934330642,
+      "learning_rate": 6.32275615498841e-06,
+      "loss": 0.9121,
+      "step": 26795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.567098263546286,
+      "learning_rate": 6.322046587239934e-06,
+      "loss": 0.8121,
+      "step": 26796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4939495865008623,
+      "learning_rate": 6.321337040905664e-06,
+      "loss": 0.7067,
+      "step": 26797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4142980958871485,
+      "learning_rate": 6.320627515989733e-06,
+      "loss": 0.9054,
+      "step": 26798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4959301762249562,
+      "learning_rate": 6.319918012496276e-06,
+      "loss": 0.8673,
+      "step": 26799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5121385247732102,
+      "learning_rate": 6.319208530429422e-06,
+      "loss": 0.8412,
+      "step": 26800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4651664467256396,
+      "learning_rate": 6.318499069793298e-06,
+      "loss": 0.8755,
+      "step": 26801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.799977425405085,
+      "learning_rate": 6.317789630592039e-06,
+      "loss": 0.8315,
+      "step": 26802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6305902619293975,
+      "learning_rate": 6.317080212829778e-06,
+      "loss": 0.8971,
+      "step": 26803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9950813638386817,
+      "learning_rate": 6.31637081651064e-06,
+      "loss": 0.8514,
+      "step": 26804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.14484573023931,
+      "learning_rate": 6.315661441638754e-06,
+      "loss": 0.7798,
+      "step": 26805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6241649689596132,
+      "learning_rate": 6.314952088218256e-06,
+      "loss": 0.8248,
+      "step": 26806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7674824431409821,
+      "learning_rate": 6.314242756253275e-06,
+      "loss": 0.7744,
+      "step": 26807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7540576105262669,
+      "learning_rate": 6.313533445747939e-06,
+      "loss": 0.9869,
+      "step": 26808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8148936445306487,
+      "learning_rate": 6.312824156706376e-06,
+      "loss": 0.7421,
+      "step": 26809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4629565157744686,
+      "learning_rate": 6.312114889132721e-06,
+      "loss": 0.8685,
+      "step": 26810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7601535509756703,
+      "learning_rate": 6.311405643031101e-06,
+      "loss": 0.998,
+      "step": 26811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.605564559486699,
+      "learning_rate": 6.310696418405643e-06,
+      "loss": 0.8279,
+      "step": 26812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7515445581383866,
+      "learning_rate": 6.309987215260477e-06,
+      "loss": 1.0042,
+      "step": 26813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8187348868466808,
+      "learning_rate": 6.309278033599738e-06,
+      "loss": 0.8586,
+      "step": 26814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7787350567193058,
+      "learning_rate": 6.3085688734275485e-06,
+      "loss": 0.7036,
+      "step": 26815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6703043131328255,
+      "learning_rate": 6.307859734748039e-06,
+      "loss": 0.8402,
+      "step": 26816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5364431331434576,
+      "learning_rate": 6.307150617565343e-06,
+      "loss": 0.8988,
+      "step": 26817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.544500825854828,
+      "learning_rate": 6.3064415218835795e-06,
+      "loss": 0.7775,
+      "step": 26818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6579295985419717,
+      "learning_rate": 6.305732447706885e-06,
+      "loss": 0.9034,
+      "step": 26819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5242394858489696,
+      "learning_rate": 6.305023395039387e-06,
+      "loss": 0.8934,
+      "step": 26820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6727461019095138,
+      "learning_rate": 6.3043143638852134e-06,
+      "loss": 0.8609,
+      "step": 26821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.533953208968131,
+      "learning_rate": 6.303605354248488e-06,
+      "loss": 0.817,
+      "step": 26822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.584903445298454,
+      "learning_rate": 6.302896366133344e-06,
+      "loss": 0.8618,
+      "step": 26823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7924890581792607,
+      "learning_rate": 6.302187399543911e-06,
+      "loss": 1.0161,
+      "step": 26824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2244557864386594,
+      "learning_rate": 6.301478454484311e-06,
+      "loss": 0.8029,
+      "step": 26825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5673427341401096,
+      "learning_rate": 6.300769530958673e-06,
+      "loss": 0.8005,
+      "step": 26826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0754786380166776,
+      "learning_rate": 6.300060628971127e-06,
+      "loss": 0.6857,
+      "step": 26827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5762814169230535,
+      "learning_rate": 6.299351748525802e-06,
+      "loss": 0.9745,
+      "step": 26828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5266782961015504,
+      "learning_rate": 6.298642889626821e-06,
+      "loss": 0.7696,
+      "step": 26829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.173959893755783,
+      "learning_rate": 6.297934052278309e-06,
+      "loss": 0.8179,
+      "step": 26830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9696718955054502,
+      "learning_rate": 6.297225236484402e-06,
+      "loss": 0.9209,
+      "step": 26831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9860235872680612,
+      "learning_rate": 6.296516442249221e-06,
+      "loss": 0.8304,
+      "step": 26832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6288654630100545,
+      "learning_rate": 6.295807669576892e-06,
+      "loss": 0.8081,
+      "step": 26833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8754836073617067,
+      "learning_rate": 6.295098918471543e-06,
+      "loss": 0.8939,
+      "step": 26834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6589889430757223,
+      "learning_rate": 6.294390188937305e-06,
+      "loss": 0.7747,
+      "step": 26835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6886512673790492,
+      "learning_rate": 6.293681480978298e-06,
+      "loss": 0.8195,
+      "step": 26836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.098162581600511,
+      "learning_rate": 6.292972794598649e-06,
+      "loss": 0.9095,
+      "step": 26837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.595669038950146,
+      "learning_rate": 6.292264129802492e-06,
+      "loss": 0.7813,
+      "step": 26838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7060170651076323,
+      "learning_rate": 6.291555486593941e-06,
+      "loss": 0.7933,
+      "step": 26839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7739626394897844,
+      "learning_rate": 6.290846864977129e-06,
+      "loss": 0.6797,
+      "step": 26840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6024066799468704,
+      "learning_rate": 6.290138264956181e-06,
+      "loss": 0.8852,
+      "step": 26841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6428844231560136,
+      "learning_rate": 6.289429686535226e-06,
+      "loss": 0.8472,
+      "step": 26842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7374502052915188,
+      "learning_rate": 6.2887211297183825e-06,
+      "loss": 0.7695,
+      "step": 26843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.693831527884859,
+      "learning_rate": 6.28801259450978e-06,
+      "loss": 0.8995,
+      "step": 26844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4949672351137768,
+      "learning_rate": 6.2873040809135465e-06,
+      "loss": 0.9158,
+      "step": 26845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4317014756377198,
+      "learning_rate": 6.286595588933802e-06,
+      "loss": 0.9056,
+      "step": 26846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4884986786473524,
+      "learning_rate": 6.285887118574671e-06,
+      "loss": 0.8507,
+      "step": 26847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.618598627746178,
+      "learning_rate": 6.285178669840284e-06,
+      "loss": 0.8689,
+      "step": 26848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5400245372860388,
+      "learning_rate": 6.284470242734764e-06,
+      "loss": 0.8897,
+      "step": 26849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5528691361535985,
+      "learning_rate": 6.28376183726223e-06,
+      "loss": 0.8772,
+      "step": 26850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.836560070026356,
+      "learning_rate": 6.283053453426813e-06,
+      "loss": 0.8427,
+      "step": 26851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7869306993701584,
+      "learning_rate": 6.282345091232639e-06,
+      "loss": 0.8208,
+      "step": 26852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6821041635484664,
+      "learning_rate": 6.2816367506838265e-06,
+      "loss": 0.8426,
+      "step": 26853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5949824081825223,
+      "learning_rate": 6.280928431784499e-06,
+      "loss": 0.756,
+      "step": 26854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4886616706169962,
+      "learning_rate": 6.280220134538787e-06,
+      "loss": 0.9124,
+      "step": 26855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4428686391862817,
+      "learning_rate": 6.279511858950813e-06,
+      "loss": 0.7589,
+      "step": 26856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.577238338597949,
+      "learning_rate": 6.278803605024695e-06,
+      "loss": 0.8838,
+      "step": 26857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.870668339438252,
+      "learning_rate": 6.278095372764561e-06,
+      "loss": 0.7438,
+      "step": 26858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.455570191209205,
+      "learning_rate": 6.277387162174536e-06,
+      "loss": 0.8795,
+      "step": 26859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7440992320993631,
+      "learning_rate": 6.276678973258741e-06,
+      "loss": 0.7077,
+      "step": 26860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1240957330171204,
+      "learning_rate": 6.2759708060212985e-06,
+      "loss": 0.892,
+      "step": 26861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5533590704324962,
+      "learning_rate": 6.275262660466333e-06,
+      "loss": 0.8197,
+      "step": 26862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9784530182590507,
+      "learning_rate": 6.274554536597969e-06,
+      "loss": 0.8544,
+      "step": 26863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5406101596558022,
+      "learning_rate": 6.273846434420328e-06,
+      "loss": 0.8017,
+      "step": 26864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4982145045783415,
+      "learning_rate": 6.2731383539375315e-06,
+      "loss": 0.877,
+      "step": 26865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7454927991248108,
+      "learning_rate": 6.272430295153705e-06,
+      "loss": 0.7422,
+      "step": 26866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.574757527920494,
+      "learning_rate": 6.2717222580729665e-06,
+      "loss": 0.8607,
+      "step": 26867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7994311930954974,
+      "learning_rate": 6.271014242699442e-06,
+      "loss": 0.8383,
+      "step": 26868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5603057265047808,
+      "learning_rate": 6.270306249037254e-06,
+      "loss": 0.9535,
+      "step": 26869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8280782328715564,
+      "learning_rate": 6.269598277090527e-06,
+      "loss": 0.9133,
+      "step": 26870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4709287430954348,
+      "learning_rate": 6.268890326863375e-06,
+      "loss": 0.8669,
+      "step": 26871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4791722823453044,
+      "learning_rate": 6.268182398359927e-06,
+      "loss": 0.8859,
+      "step": 26872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6432837376542162,
+      "learning_rate": 6.267474491584304e-06,
+      "loss": 0.911,
+      "step": 26873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.693272096924664,
+      "learning_rate": 6.266766606540624e-06,
+      "loss": 0.7975,
+      "step": 26874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6046853750204597,
+      "learning_rate": 6.26605874323301e-06,
+      "loss": 0.9014,
+      "step": 26875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7708383014757691,
+      "learning_rate": 6.265350901665586e-06,
+      "loss": 0.7649,
+      "step": 26876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.81714308080734,
+      "learning_rate": 6.264643081842472e-06,
+      "loss": 0.8404,
+      "step": 26877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.726345068893949,
+      "learning_rate": 6.263935283767787e-06,
+      "loss": 0.8705,
+      "step": 26878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7596707884934726,
+      "learning_rate": 6.2632275074456515e-06,
+      "loss": 0.6897,
+      "step": 26879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.671373076874974,
+      "learning_rate": 6.262519752880193e-06,
+      "loss": 0.8281,
+      "step": 26880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6548093243939233,
+      "learning_rate": 6.2618120200755264e-06,
+      "loss": 0.8241,
+      "step": 26881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5525281494210637,
+      "learning_rate": 6.261104309035772e-06,
+      "loss": 0.9495,
+      "step": 26882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5741873984950947,
+      "learning_rate": 6.260396619765051e-06,
+      "loss": 0.7858,
+      "step": 26883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5636233674881337,
+      "learning_rate": 6.259688952267489e-06,
+      "loss": 0.9687,
+      "step": 26884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.665985834919165,
+      "learning_rate": 6.2589813065472005e-06,
+      "loss": 0.741,
+      "step": 26885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.618589339946817,
+      "learning_rate": 6.258273682608304e-06,
+      "loss": 0.5664,
+      "step": 26886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.637826069296827,
+      "learning_rate": 6.257566080454925e-06,
+      "loss": 0.7629,
+      "step": 26887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3524536024906213,
+      "learning_rate": 6.256858500091184e-06,
+      "loss": 0.7844,
+      "step": 26888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6146919741687038,
+      "learning_rate": 6.256150941521195e-06,
+      "loss": 0.9367,
+      "step": 26889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6894846231943477,
+      "learning_rate": 6.25544340474908e-06,
+      "loss": 0.8409,
+      "step": 26890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.533760434883234,
+      "learning_rate": 6.2547358897789625e-06,
+      "loss": 0.94,
+      "step": 26891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.802917349552013,
+      "learning_rate": 6.254028396614954e-06,
+      "loss": 0.8952,
+      "step": 26892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4947192512313707,
+      "learning_rate": 6.2533209252611795e-06,
+      "loss": 0.935,
+      "step": 26893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3983833418463867,
+      "learning_rate": 6.252613475721756e-06,
+      "loss": 0.8075,
+      "step": 26894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6082337562193065,
+      "learning_rate": 6.251906048000807e-06,
+      "loss": 0.7148,
+      "step": 26895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.472022632601972,
+      "learning_rate": 6.251198642102444e-06,
+      "loss": 0.7418,
+      "step": 26896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.602796459324762,
+      "learning_rate": 6.2504912580307905e-06,
+      "loss": 0.6794,
+      "step": 26897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6823941844123818,
+      "learning_rate": 6.2497838957899645e-06,
+      "loss": 0.8261,
+      "step": 26898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7326047779217113,
+      "learning_rate": 6.249076555384082e-06,
+      "loss": 0.967,
+      "step": 26899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7571314894060095,
+      "learning_rate": 6.248369236817265e-06,
+      "loss": 0.8262,
+      "step": 26900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3646919585437025,
+      "learning_rate": 6.247661940093629e-06,
+      "loss": 0.7786,
+      "step": 26901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6367145249996757,
+      "learning_rate": 6.246954665217295e-06,
+      "loss": 0.8887,
+      "step": 26902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6298507051008189,
+      "learning_rate": 6.2462474121923765e-06,
+      "loss": 0.8272,
+      "step": 26903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5632178727258879,
+      "learning_rate": 6.245540181022996e-06,
+      "loss": 0.8257,
+      "step": 26904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8321854958475192,
+      "learning_rate": 6.2448329717132705e-06,
+      "loss": 0.8487,
+      "step": 26905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6950916559517282,
+      "learning_rate": 6.244125784267314e-06,
+      "loss": 0.7769,
+      "step": 26906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6252978114606236,
+      "learning_rate": 6.243418618689246e-06,
+      "loss": 0.7411,
+      "step": 26907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.687605801757431,
+      "learning_rate": 6.242711474983183e-06,
+      "loss": 0.7669,
+      "step": 26908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8045647759917955,
+      "learning_rate": 6.242004353153248e-06,
+      "loss": 0.909,
+      "step": 26909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7330933462211102,
+      "learning_rate": 6.241297253203551e-06,
+      "loss": 0.9514,
+      "step": 26910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5982338094847173,
+      "learning_rate": 6.240590175138208e-06,
+      "loss": 0.807,
+      "step": 26911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6077443522866324,
+      "learning_rate": 6.239883118961344e-06,
+      "loss": 0.898,
+      "step": 26912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4783551938590067,
+      "learning_rate": 6.23917608467707e-06,
+      "loss": 0.8264,
+      "step": 26913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4173618832746504,
+      "learning_rate": 6.2384690722895016e-06,
+      "loss": 0.9447,
+      "step": 26914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.563533175891396,
+      "learning_rate": 6.237762081802757e-06,
+      "loss": 0.8182,
+      "step": 26915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0953169979951296,
+      "learning_rate": 6.237055113220955e-06,
+      "loss": 0.8522,
+      "step": 26916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0183727510454594,
+      "learning_rate": 6.236348166548208e-06,
+      "loss": 0.8623,
+      "step": 26917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7725134529454112,
+      "learning_rate": 6.2356412417886335e-06,
+      "loss": 0.752,
+      "step": 26918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.746371704288437,
+      "learning_rate": 6.23493433894635e-06,
+      "loss": 0.7414,
+      "step": 26919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5998810658861693,
+      "learning_rate": 6.234227458025466e-06,
+      "loss": 0.9164,
+      "step": 26920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.76271961709224,
+      "learning_rate": 6.233520599030105e-06,
+      "loss": 0.8642,
+      "step": 26921
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7273671552631333,
+      "learning_rate": 6.232813761964377e-06,
+      "loss": 0.854,
+      "step": 26922
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6632028281454923,
+      "learning_rate": 6.232106946832406e-06,
+      "loss": 0.9409,
+      "step": 26923
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.744884750996277,
+      "learning_rate": 6.231400153638294e-06,
+      "loss": 0.7946,
+      "step": 26924
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.987189858470433,
+      "learning_rate": 6.230693382386167e-06,
+      "loss": 0.725,
+      "step": 26925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5653883104536106,
+      "learning_rate": 6.229986633080138e-06,
+      "loss": 0.8062,
+      "step": 26926
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0160249386194207,
+      "learning_rate": 6.229279905724319e-06,
+      "loss": 0.9338,
+      "step": 26927
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7688068825411332,
+      "learning_rate": 6.228573200322823e-06,
+      "loss": 0.8076,
+      "step": 26928
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6295012936389957,
+      "learning_rate": 6.22786651687977e-06,
+      "loss": 0.8498,
+      "step": 26929
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.440464999849056,
+      "learning_rate": 6.227159855399276e-06,
+      "loss": 0.891,
+      "step": 26930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6424602421310612,
+      "learning_rate": 6.226453215885448e-06,
+      "loss": 0.8895,
+      "step": 26931
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5974205377210047,
+      "learning_rate": 6.225746598342402e-06,
+      "loss": 0.8979,
+      "step": 26932
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.586707575660579,
+      "learning_rate": 6.22504000277426e-06,
+      "loss": 0.7708,
+      "step": 26933
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6105789313911119,
+      "learning_rate": 6.224333429185127e-06,
+      "loss": 0.7528,
+      "step": 26934
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.566279550725694,
+      "learning_rate": 6.223626877579119e-06,
+      "loss": 0.8163,
+      "step": 26935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.678991512453388,
+      "learning_rate": 6.22292034796035e-06,
+      "loss": 0.8791,
+      "step": 26936
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2070721030937785,
+      "learning_rate": 6.22221384033294e-06,
+      "loss": 0.8588,
+      "step": 26937
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.920386404932779,
+      "learning_rate": 6.221507354700993e-06,
+      "loss": 0.8585,
+      "step": 26938
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8239092627970532,
+      "learning_rate": 6.2208008910686255e-06,
+      "loss": 0.9809,
+      "step": 26939
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6922344845139261,
+      "learning_rate": 6.220094449439957e-06,
+      "loss": 0.7326,
+      "step": 26940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7604443501649931,
+      "learning_rate": 6.21938802981909e-06,
+      "loss": 0.8433,
+      "step": 26941
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7486349130616139,
+      "learning_rate": 6.218681632210144e-06,
+      "loss": 0.6678,
+      "step": 26942
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7984304944656404,
+      "learning_rate": 6.21797525661723e-06,
+      "loss": 0.7974,
+      "step": 26943
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5296137082361332,
+      "learning_rate": 6.217268903044467e-06,
+      "loss": 0.931,
+      "step": 26944
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5273078472690327,
+      "learning_rate": 6.2165625714959545e-06,
+      "loss": 0.8301,
+      "step": 26945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7945572123473168,
+      "learning_rate": 6.215856261975816e-06,
+      "loss": 0.7654,
+      "step": 26946
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5615771088828172,
+      "learning_rate": 6.215149974488161e-06,
+      "loss": 0.7742,
+      "step": 26947
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8579643213419457,
+      "learning_rate": 6.214443709037099e-06,
+      "loss": 0.855,
+      "step": 26948
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8005022394068888,
+      "learning_rate": 6.213737465626746e-06,
+      "loss": 0.875,
+      "step": 26949
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6262097399496265,
+      "learning_rate": 6.213031244261212e-06,
+      "loss": 0.8756,
+      "step": 26950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5497919589828792,
+      "learning_rate": 6.212325044944609e-06,
+      "loss": 0.7882,
+      "step": 26951
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7227642844574473,
+      "learning_rate": 6.2116188676810466e-06,
+      "loss": 0.9274,
+      "step": 26952
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7255533276998556,
+      "learning_rate": 6.2109127124746405e-06,
+      "loss": 0.7797,
+      "step": 26953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7499396660311634,
+      "learning_rate": 6.210206579329502e-06,
+      "loss": 0.7139,
+      "step": 26954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7819644973265499,
+      "learning_rate": 6.209500468249738e-06,
+      "loss": 0.7398,
+      "step": 26955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0694881579166586,
+      "learning_rate": 6.208794379239462e-06,
+      "loss": 0.8249,
+      "step": 26956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5922963792112355,
+      "learning_rate": 6.208088312302785e-06,
+      "loss": 0.7463,
+      "step": 26957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.591270598874332,
+      "learning_rate": 6.207382267443823e-06,
+      "loss": 0.8663,
+      "step": 26958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.446166308108094,
+      "learning_rate": 6.2066762446666785e-06,
+      "loss": 0.8771,
+      "step": 26959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4630495159479442,
+      "learning_rate": 6.205970243975464e-06,
+      "loss": 0.7703,
+      "step": 26960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6131181601650486,
+      "learning_rate": 6.205264265374295e-06,
+      "loss": 0.8227,
+      "step": 26961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6915145483273253,
+      "learning_rate": 6.2045583088672785e-06,
+      "loss": 0.9197,
+      "step": 26962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3385547560293407,
+      "learning_rate": 6.2038523744585254e-06,
+      "loss": 0.7688,
+      "step": 26963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2165695329341135,
+      "learning_rate": 6.203146462152142e-06,
+      "loss": 0.8525,
+      "step": 26964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6566143238264044,
+      "learning_rate": 6.202440571952246e-06,
+      "loss": 0.893,
+      "step": 26965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6451369170031418,
+      "learning_rate": 6.201734703862942e-06,
+      "loss": 0.8636,
+      "step": 26966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2327700697833293,
+      "learning_rate": 6.201028857888342e-06,
+      "loss": 0.8807,
+      "step": 26967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.629626074056473,
+      "learning_rate": 6.200323034032556e-06,
+      "loss": 0.8832,
+      "step": 26968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7603104970157761,
+      "learning_rate": 6.1996172322996885e-06,
+      "loss": 0.7762,
+      "step": 26969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6437985167785638,
+      "learning_rate": 6.1989114526938535e-06,
+      "loss": 0.8949,
+      "step": 26970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8534234249986996,
+      "learning_rate": 6.198205695219157e-06,
+      "loss": 0.8142,
+      "step": 26971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5349426931265209,
+      "learning_rate": 6.197499959879718e-06,
+      "loss": 0.8664,
+      "step": 26972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.040130648444855,
+      "learning_rate": 6.1967942466796315e-06,
+      "loss": 0.8301,
+      "step": 26973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.490455688384551,
+      "learning_rate": 6.196088555623016e-06,
+      "loss": 1.0034,
+      "step": 26974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6118088235679628,
+      "learning_rate": 6.1953828867139784e-06,
+      "loss": 0.9015,
+      "step": 26975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7866685120283887,
+      "learning_rate": 6.194677239956625e-06,
+      "loss": 0.7594,
+      "step": 26976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.091976243592423,
+      "learning_rate": 6.193971615355062e-06,
+      "loss": 0.8356,
+      "step": 26977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.324951375785033,
+      "learning_rate": 6.1932660129134035e-06,
+      "loss": 0.7464,
+      "step": 26978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.105649382697799,
+      "learning_rate": 6.1925604326357595e-06,
+      "loss": 0.8932,
+      "step": 26979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1193035947660883,
+      "learning_rate": 6.191854874526229e-06,
+      "loss": 0.7374,
+      "step": 26980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.428055535531366,
+      "learning_rate": 6.191149338588924e-06,
+      "loss": 0.8158,
+      "step": 26981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.100385599243297,
+      "learning_rate": 6.190443824827958e-06,
+      "loss": 0.8872,
+      "step": 26982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.131781519101684,
+      "learning_rate": 6.189738333247433e-06,
+      "loss": 0.8361,
+      "step": 26983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7173314831954567,
+      "learning_rate": 6.189032863851454e-06,
+      "loss": 0.8189,
+      "step": 26984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4994616808348074,
+      "learning_rate": 6.1883274166441356e-06,
+      "loss": 0.8346,
+      "step": 26985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6592776136908582,
+      "learning_rate": 6.187621991629582e-06,
+      "loss": 0.8697,
+      "step": 26986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5121355558933798,
+      "learning_rate": 6.1869165888119e-06,
+      "loss": 0.8021,
+      "step": 26987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7025818453545805,
+      "learning_rate": 6.186211208195194e-06,
+      "loss": 0.7147,
+      "step": 26988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4719371381408732,
+      "learning_rate": 6.18550584978358e-06,
+      "loss": 0.8257,
+      "step": 26989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6514579817886392,
+      "learning_rate": 6.184800513581153e-06,
+      "loss": 0.8685,
+      "step": 26990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6273784343071436,
+      "learning_rate": 6.184095199592026e-06,
+      "loss": 0.8896,
+      "step": 26991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6188101216737414,
+      "learning_rate": 6.183389907820303e-06,
+      "loss": 0.8057,
+      "step": 26992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.531299999859005,
+      "learning_rate": 6.1826846382701e-06,
+      "loss": 0.8189,
+      "step": 26993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8511913692850455,
+      "learning_rate": 6.1819793909455076e-06,
+      "loss": 0.8299,
+      "step": 26994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7027690260066182,
+      "learning_rate": 6.181274165850644e-06,
+      "loss": 0.7774,
+      "step": 26995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3454425203487232,
+      "learning_rate": 6.18056896298961e-06,
+      "loss": 0.8458,
+      "step": 26996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5187753053305137,
+      "learning_rate": 6.179863782366512e-06,
+      "loss": 0.8496,
+      "step": 26997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.193661517109868,
+      "learning_rate": 6.179158623985457e-06,
+      "loss": 0.8277,
+      "step": 26998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1803850561559357,
+      "learning_rate": 6.178453487850551e-06,
+      "loss": 0.8627,
+      "step": 26999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7817210977233564,
+      "learning_rate": 6.1777483739659e-06,
+      "loss": 0.7093,
+      "step": 27000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8275918535936995,
+      "learning_rate": 6.177043282335603e-06,
+      "loss": 0.8614,
+      "step": 27001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6604527735436518,
+      "learning_rate": 6.1763382129637725e-06,
+      "loss": 0.8184,
+      "step": 27002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6683794803684602,
+      "learning_rate": 6.175633165854515e-06,
+      "loss": 0.8974,
+      "step": 27003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5540897344566942,
+      "learning_rate": 6.17492814101193e-06,
+      "loss": 0.702,
+      "step": 27004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4198476705049872,
+      "learning_rate": 6.174223138440121e-06,
+      "loss": 0.9196,
+      "step": 27005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6167827892147824,
+      "learning_rate": 6.173518158143199e-06,
+      "loss": 0.8179,
+      "step": 27006
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9308945921465075,
+      "learning_rate": 6.172813200125267e-06,
+      "loss": 0.7916,
+      "step": 27007
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5587128568912847,
+      "learning_rate": 6.172108264390428e-06,
+      "loss": 0.8356,
+      "step": 27008
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.76244796922263,
+      "learning_rate": 6.171403350942783e-06,
+      "loss": 0.8453,
+      "step": 27009
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7431861428089857,
+      "learning_rate": 6.170698459786442e-06,
+      "loss": 0.9023,
+      "step": 27010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.746243538448183,
+      "learning_rate": 6.169993590925509e-06,
+      "loss": 0.9052,
+      "step": 27011
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7700123868666533,
+      "learning_rate": 6.169288744364084e-06,
+      "loss": 0.8099,
+      "step": 27012
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4804824587296095,
+      "learning_rate": 6.16858392010627e-06,
+      "loss": 0.8635,
+      "step": 27013
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6672325941067183,
+      "learning_rate": 6.167879118156179e-06,
+      "loss": 0.7488,
+      "step": 27014
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6405500536607396,
+      "learning_rate": 6.167174338517905e-06,
+      "loss": 0.8777,
+      "step": 27015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6356268334636963,
+      "learning_rate": 6.166469581195558e-06,
+      "loss": 0.8173,
+      "step": 27016
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6128164089047379,
+      "learning_rate": 6.165764846193234e-06,
+      "loss": 0.8781,
+      "step": 27017
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.229162097161136,
+      "learning_rate": 6.165060133515046e-06,
+      "loss": 0.8549,
+      "step": 27018
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.693644405412758,
+      "learning_rate": 6.164355443165091e-06,
+      "loss": 0.9066,
+      "step": 27019
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.550809320718554,
+      "learning_rate": 6.163650775147471e-06,
+      "loss": 0.9249,
+      "step": 27020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5668859448699908,
+      "learning_rate": 6.162946129466295e-06,
+      "loss": 0.8354,
+      "step": 27021
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4506655734145721,
+      "learning_rate": 6.162241506125657e-06,
+      "loss": 0.7823,
+      "step": 27022
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5445930873555258,
+      "learning_rate": 6.161536905129666e-06,
+      "loss": 0.7651,
+      "step": 27023
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4659946088130364,
+      "learning_rate": 6.160832326482418e-06,
+      "loss": 0.9366,
+      "step": 27024
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6411775355194163,
+      "learning_rate": 6.160127770188029e-06,
+      "loss": 0.8189,
+      "step": 27025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7220903117119425,
+      "learning_rate": 6.1594232362505835e-06,
+      "loss": 0.6802,
+      "step": 27026
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6057752925120692,
+      "learning_rate": 6.158718724674194e-06,
+      "loss": 0.9428,
+      "step": 27027
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.505112729794294,
+      "learning_rate": 6.158014235462964e-06,
+      "loss": 0.6594,
+      "step": 27028
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7122298922516823,
+      "learning_rate": 6.157309768620987e-06,
+      "loss": 0.9509,
+      "step": 27029
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.710153278836577,
+      "learning_rate": 6.156605324152369e-06,
+      "loss": 0.8691,
+      "step": 27030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6418949112732604,
+      "learning_rate": 6.155900902061212e-06,
+      "loss": 0.8853,
+      "step": 27031
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5905319207894664,
+      "learning_rate": 6.155196502351621e-06,
+      "loss": 0.9153,
+      "step": 27032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.619532196273058,
+      "learning_rate": 6.154492125027688e-06,
+      "loss": 0.9331,
+      "step": 27033
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6871528369106246,
+      "learning_rate": 6.1537877700935215e-06,
+      "loss": 0.8805,
+      "step": 27034
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.493773548199695,
+      "learning_rate": 6.153083437553222e-06,
+      "loss": 0.8176,
+      "step": 27035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6370787984874118,
+      "learning_rate": 6.152379127410887e-06,
+      "loss": 0.9764,
+      "step": 27036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5158559297244631,
+      "learning_rate": 6.151674839670616e-06,
+      "loss": 0.7884,
+      "step": 27037
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6883301605265189,
+      "learning_rate": 6.1509705743365154e-06,
+      "loss": 0.9331,
+      "step": 27038
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.304061015824182,
+      "learning_rate": 6.150266331412684e-06,
+      "loss": 0.7282,
+      "step": 27039
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7505406233378602,
+      "learning_rate": 6.14956211090322e-06,
+      "loss": 0.8231,
+      "step": 27040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8412247932446275,
+      "learning_rate": 6.148857912812222e-06,
+      "loss": 0.8671,
+      "step": 27041
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6431181384412334,
+      "learning_rate": 6.148153737143798e-06,
+      "loss": 0.859,
+      "step": 27042
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9574798933436135,
+      "learning_rate": 6.1474495839020365e-06,
+      "loss": 0.7834,
+      "step": 27043
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6120896488080576,
+      "learning_rate": 6.1467454530910454e-06,
+      "loss": 0.9062,
+      "step": 27044
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6545489093297543,
+      "learning_rate": 6.146041344714919e-06,
+      "loss": 0.8101,
+      "step": 27045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4243984731676065,
+      "learning_rate": 6.145337258777766e-06,
+      "loss": 0.8615,
+      "step": 27046
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6826984241444516,
+      "learning_rate": 6.144633195283677e-06,
+      "loss": 0.9801,
+      "step": 27047
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5604495529664384,
+      "learning_rate": 6.1439291542367544e-06,
+      "loss": 0.7744,
+      "step": 27048
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5503158738103762,
+      "learning_rate": 6.143225135641097e-06,
+      "loss": 0.7386,
+      "step": 27049
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4296028413299964,
+      "learning_rate": 6.142521139500803e-06,
+      "loss": 0.756,
+      "step": 27050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9386195983261065,
+      "learning_rate": 6.1418171658199725e-06,
+      "loss": 0.8238,
+      "step": 27051
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5442075882099338,
+      "learning_rate": 6.141113214602705e-06,
+      "loss": 0.827,
+      "step": 27052
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.50637860287621,
+      "learning_rate": 6.1404092858531e-06,
+      "loss": 0.7863,
+      "step": 27053
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5185519111859374,
+      "learning_rate": 6.139705379575248e-06,
+      "loss": 0.8219,
+      "step": 27054
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6191059649235353,
+      "learning_rate": 6.139001495773258e-06,
+      "loss": 1.0114,
+      "step": 27055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4489523113819998,
+      "learning_rate": 6.138297634451224e-06,
+      "loss": 0.7042,
+      "step": 27056
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.53936924549941,
+      "learning_rate": 6.137593795613242e-06,
+      "loss": 0.8442,
+      "step": 27057
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7093740740839536,
+      "learning_rate": 6.136889979263409e-06,
+      "loss": 0.7968,
+      "step": 27058
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3751371378820232,
+      "learning_rate": 6.13618618540583e-06,
+      "loss": 0.7632,
+      "step": 27059
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6151588806557742,
+      "learning_rate": 6.135482414044598e-06,
+      "loss": 0.905,
+      "step": 27060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5581246509046658,
+      "learning_rate": 6.1347786651838105e-06,
+      "loss": 0.938,
+      "step": 27061
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7844936954592021,
+      "learning_rate": 6.134074938827562e-06,
+      "loss": 0.7576,
+      "step": 27062
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.724988343199283,
+      "learning_rate": 6.133371234979958e-06,
+      "loss": 0.8826,
+      "step": 27063
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9171668168283422,
+      "learning_rate": 6.1326675536450885e-06,
+      "loss": 0.9145,
+      "step": 27064
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5997083332068869,
+      "learning_rate": 6.1319638948270536e-06,
+      "loss": 0.8772,
+      "step": 27065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.512258024425231,
+      "learning_rate": 6.131260258529947e-06,
+      "loss": 0.9298,
+      "step": 27066
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6569641350655615,
+      "learning_rate": 6.130556644757871e-06,
+      "loss": 0.8033,
+      "step": 27067
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8951393757682666,
+      "learning_rate": 6.129853053514918e-06,
+      "loss": 1.0071,
+      "step": 27068
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7026790869205408,
+      "learning_rate": 6.129149484805183e-06,
+      "loss": 0.8958,
+      "step": 27069
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7753505178627185,
+      "learning_rate": 6.1284459386327735e-06,
+      "loss": 0.7081,
+      "step": 27070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6019799807943391,
+      "learning_rate": 6.1277424150017695e-06,
+      "loss": 0.8502,
+      "step": 27071
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8738319243113897,
+      "learning_rate": 6.127038913916278e-06,
+      "loss": 0.9126,
+      "step": 27072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5773113294310819,
+      "learning_rate": 6.12633543538039e-06,
+      "loss": 1.0635,
+      "step": 27073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7297722992796751,
+      "learning_rate": 6.125631979398209e-06,
+      "loss": 0.6952,
+      "step": 27074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5307612114164737,
+      "learning_rate": 6.124928545973819e-06,
+      "loss": 0.816,
+      "step": 27075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8183018358725886,
+      "learning_rate": 6.124225135111325e-06,
+      "loss": 0.8475,
+      "step": 27076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3716318192795554,
+      "learning_rate": 6.12352174681482e-06,
+      "loss": 0.7191,
+      "step": 27077
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8513399110566116,
+      "learning_rate": 6.122818381088398e-06,
+      "loss": 0.9204,
+      "step": 27078
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0892367909716234,
+      "learning_rate": 6.1221150379361514e-06,
+      "loss": 0.8265,
+      "step": 27079
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.449537922326887,
+      "learning_rate": 6.1214117173621815e-06,
+      "loss": 0.7173,
+      "step": 27080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4010923843897205,
+      "learning_rate": 6.120708419370583e-06,
+      "loss": 0.9277,
+      "step": 27081
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.481351713119913,
+      "learning_rate": 6.120005143965443e-06,
+      "loss": 0.8762,
+      "step": 27082
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.457925484443869,
+      "learning_rate": 6.1193018911508635e-06,
+      "loss": 0.718,
+      "step": 27083
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7531461179797354,
+      "learning_rate": 6.118598660930941e-06,
+      "loss": 0.8326,
+      "step": 27084
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.532703775437012,
+      "learning_rate": 6.117895453309762e-06,
+      "loss": 0.8988,
+      "step": 27085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.50485232787722,
+      "learning_rate": 6.117192268291423e-06,
+      "loss": 0.7811,
+      "step": 27086
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3419331157255847,
+      "learning_rate": 6.116489105880023e-06,
+      "loss": 0.7265,
+      "step": 27087
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.760969723364445,
+      "learning_rate": 6.115785966079653e-06,
+      "loss": 1.0059,
+      "step": 27088
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8257681583118861,
+      "learning_rate": 6.115082848894407e-06,
+      "loss": 0.7088,
+      "step": 27089
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5582932686791977,
+      "learning_rate": 6.114379754328375e-06,
+      "loss": 0.7891,
+      "step": 27090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9995549190748563,
+      "learning_rate": 6.113676682385661e-06,
+      "loss": 0.7515,
+      "step": 27091
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7671007059840993,
+      "learning_rate": 6.112973633070346e-06,
+      "loss": 0.714,
+      "step": 27092
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5311144616681935,
+      "learning_rate": 6.11227060638653e-06,
+      "loss": 0.8428,
+      "step": 27093
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8156155102723143,
+      "learning_rate": 6.111567602338304e-06,
+      "loss": 0.7155,
+      "step": 27094
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9553834889441575,
+      "learning_rate": 6.110864620929766e-06,
+      "loss": 0.8344,
+      "step": 27095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6984398550233946,
+      "learning_rate": 6.110161662165005e-06,
+      "loss": 0.8899,
+      "step": 27096
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7324876470737902,
+      "learning_rate": 6.109458726048113e-06,
+      "loss": 0.8269,
+      "step": 27097
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5587688435420635,
+      "learning_rate": 6.108755812583187e-06,
+      "loss": 0.8128,
+      "step": 27098
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4863598982732535,
+      "learning_rate": 6.108052921774312e-06,
+      "loss": 0.7081,
+      "step": 27099
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.524395328618306,
+      "learning_rate": 6.107350053625587e-06,
+      "loss": 0.8595,
+      "step": 27100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8099063724740838,
+      "learning_rate": 6.106647208141101e-06,
+      "loss": 0.9233,
+      "step": 27101
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5920971755731315,
+      "learning_rate": 6.105944385324951e-06,
+      "loss": 0.8733,
+      "step": 27102
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7697777113976236,
+      "learning_rate": 6.105241585181222e-06,
+      "loss": 0.8986,
+      "step": 27103
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6757174700101605,
+      "learning_rate": 6.104538807714011e-06,
+      "loss": 0.9222,
+      "step": 27104
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6107717639241024,
+      "learning_rate": 6.10383605292741e-06,
+      "loss": 0.8414,
+      "step": 27105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.719477210764102,
+      "learning_rate": 6.103133320825507e-06,
+      "loss": 0.9802,
+      "step": 27106
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.67749360658398,
+      "learning_rate": 6.102430611412395e-06,
+      "loss": 0.8896,
+      "step": 27107
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8604630032024152,
+      "learning_rate": 6.101727924692167e-06,
+      "loss": 0.7117,
+      "step": 27108
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4757604925924843,
+      "learning_rate": 6.101025260668916e-06,
+      "loss": 0.8246,
+      "step": 27109
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6414572426028524,
+      "learning_rate": 6.100322619346726e-06,
+      "loss": 0.871,
+      "step": 27110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.424468229006639,
+      "learning_rate": 6.099620000729692e-06,
+      "loss": 0.8435,
+      "step": 27111
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.752388106592846,
+      "learning_rate": 6.098917404821909e-06,
+      "loss": 0.7269,
+      "step": 27112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.559930850020242,
+      "learning_rate": 6.098214831627462e-06,
+      "loss": 0.8964,
+      "step": 27113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5733225253322096,
+      "learning_rate": 6.097512281150444e-06,
+      "loss": 0.9148,
+      "step": 27114
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4865016322960636,
+      "learning_rate": 6.096809753394941e-06,
+      "loss": 0.8059,
+      "step": 27115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.932589869585635,
+      "learning_rate": 6.0961072483650526e-06,
+      "loss": 0.8191,
+      "step": 27116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7554426901378727,
+      "learning_rate": 6.095404766064863e-06,
+      "loss": 0.7495,
+      "step": 27117
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5418373967169645,
+      "learning_rate": 6.094702306498459e-06,
+      "loss": 0.9104,
+      "step": 27118
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5667347711042916,
+      "learning_rate": 6.093999869669942e-06,
+      "loss": 0.8624,
+      "step": 27119
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.239074023796353,
+      "learning_rate": 6.0932974555833875e-06,
+      "loss": 0.9307,
+      "step": 27120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.453874742010804,
+      "learning_rate": 6.092595064242894e-06,
+      "loss": 0.8239,
+      "step": 27121
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.608041255204226,
+      "learning_rate": 6.091892695652548e-06,
+      "loss": 0.884,
+      "step": 27122
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.648627006171199,
+      "learning_rate": 6.091190349816446e-06,
+      "loss": 0.7998,
+      "step": 27123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5898549071412236,
+      "learning_rate": 6.090488026738666e-06,
+      "loss": 0.8999,
+      "step": 27124
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.680687340082559,
+      "learning_rate": 6.089785726423303e-06,
+      "loss": 0.8604,
+      "step": 27125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7035879559382512,
+      "learning_rate": 6.089083448874446e-06,
+      "loss": 0.831,
+      "step": 27126
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8743109439659356,
+      "learning_rate": 6.088381194096184e-06,
+      "loss": 0.8686,
+      "step": 27127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7909193782473551,
+      "learning_rate": 6.087678962092603e-06,
+      "loss": 0.6449,
+      "step": 27128
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.537639188660571,
+      "learning_rate": 6.0869767528677944e-06,
+      "loss": 0.7987,
+      "step": 27129
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5782190601314574,
+      "learning_rate": 6.086274566425848e-06,
+      "loss": 0.8197,
+      "step": 27130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.440081713376581,
+      "learning_rate": 6.085572402770848e-06,
+      "loss": 0.9167,
+      "step": 27131
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.705424082951629,
+      "learning_rate": 6.084870261906883e-06,
+      "loss": 0.8463,
+      "step": 27132
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5273056600228494,
+      "learning_rate": 6.084168143838046e-06,
+      "loss": 1.0199,
+      "step": 27133
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6609442088424746,
+      "learning_rate": 6.083466048568423e-06,
+      "loss": 0.8847,
+      "step": 27134
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5749760476322332,
+      "learning_rate": 6.082763976102097e-06,
+      "loss": 0.8456,
+      "step": 27135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6731053562837697,
+      "learning_rate": 6.082061926443162e-06,
+      "loss": 0.8776,
+      "step": 27136
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4258962936520518,
+      "learning_rate": 6.081359899595702e-06,
+      "loss": 0.8511,
+      "step": 27137
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5245744773646575,
+      "learning_rate": 6.080657895563807e-06,
+      "loss": 0.8755,
+      "step": 27138
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8711088076561246,
+      "learning_rate": 6.079955914351558e-06,
+      "loss": 0.9333,
+      "step": 27139
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.769648036908114,
+      "learning_rate": 6.079253955963049e-06,
+      "loss": 0.8475,
+      "step": 27140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.673762439340451,
+      "learning_rate": 6.078552020402367e-06,
+      "loss": 0.8965,
+      "step": 27141
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7221733892461764,
+      "learning_rate": 6.0778501076735956e-06,
+      "loss": 0.6889,
+      "step": 27142
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5424203215283996,
+      "learning_rate": 6.077148217780818e-06,
+      "loss": 0.7574,
+      "step": 27143
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7759997910896712,
+      "learning_rate": 6.076446350728131e-06,
+      "loss": 0.898,
+      "step": 27144
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5494352224732497,
+      "learning_rate": 6.0757445065196145e-06,
+      "loss": 0.9352,
+      "step": 27145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7446888739835619,
+      "learning_rate": 6.075042685159356e-06,
+      "loss": 0.6726,
+      "step": 27146
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5816466694455733,
+      "learning_rate": 6.074340886651438e-06,
+      "loss": 0.7339,
+      "step": 27147
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.560112671985653,
+      "learning_rate": 6.073639110999955e-06,
+      "loss": 0.8719,
+      "step": 27148
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6188673763544792,
+      "learning_rate": 6.072937358208986e-06,
+      "loss": 0.8439,
+      "step": 27149
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8227555549977194,
+      "learning_rate": 6.072235628282619e-06,
+      "loss": 0.8044,
+      "step": 27150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5232836763349622,
+      "learning_rate": 6.071533921224941e-06,
+      "loss": 0.8533,
+      "step": 27151
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4440530307835462,
+      "learning_rate": 6.0708322370400345e-06,
+      "loss": 0.7067,
+      "step": 27152
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7159449006818133,
+      "learning_rate": 6.0701305757319866e-06,
+      "loss": 0.8556,
+      "step": 27153
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6532753977304937,
+      "learning_rate": 6.069428937304883e-06,
+      "loss": 0.817,
+      "step": 27154
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5023531679607656,
+      "learning_rate": 6.06872732176281e-06,
+      "loss": 0.8468,
+      "step": 27155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6777022720752701,
+      "learning_rate": 6.068025729109849e-06,
+      "loss": 0.811,
+      "step": 27156
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3535830311006265,
+      "learning_rate": 6.067324159350088e-06,
+      "loss": 0.7976,
+      "step": 27157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7738514523233955,
+      "learning_rate": 6.066622612487611e-06,
+      "loss": 0.8161,
+      "step": 27158
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5435882974594315,
+      "learning_rate": 6.065921088526503e-06,
+      "loss": 0.7337,
+      "step": 27159
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8507391075433061,
+      "learning_rate": 6.065219587470845e-06,
+      "loss": 0.6833,
+      "step": 27160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.806886439901864,
+      "learning_rate": 6.064518109324727e-06,
+      "loss": 0.7146,
+      "step": 27161
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6409768800082039,
+      "learning_rate": 6.063816654092232e-06,
+      "loss": 0.8373,
+      "step": 27162
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7486917692449184,
+      "learning_rate": 6.06311522177744e-06,
+      "loss": 0.8888,
+      "step": 27163
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4482686011398542,
+      "learning_rate": 6.062413812384436e-06,
+      "loss": 0.9698,
+      "step": 27164
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5874972355784613,
+      "learning_rate": 6.06171242591731e-06,
+      "loss": 0.8939,
+      "step": 27165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4383479600742022,
+      "learning_rate": 6.061011062380139e-06,
+      "loss": 0.8124,
+      "step": 27166
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.665498509373428,
+      "learning_rate": 6.060309721777006e-06,
+      "loss": 0.8765,
+      "step": 27167
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.807941421269025,
+      "learning_rate": 6.059608404111998e-06,
+      "loss": 0.7154,
+      "step": 27168
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6434505724540582,
+      "learning_rate": 6.058907109389202e-06,
+      "loss": 0.9498,
+      "step": 27169
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.890716919980919,
+      "learning_rate": 6.058205837612694e-06,
+      "loss": 0.7817,
+      "step": 27170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5985314197812388,
+      "learning_rate": 6.057504588786556e-06,
+      "loss": 0.8349,
+      "step": 27171
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4312922736072389,
+      "learning_rate": 6.0568033629148814e-06,
+      "loss": 0.8603,
+      "step": 27172
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8710841957803523,
+      "learning_rate": 6.05610216000174e-06,
+      "loss": 0.8924,
+      "step": 27173
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.673406599150993,
+      "learning_rate": 6.055400980051223e-06,
+      "loss": 0.8766,
+      "step": 27174
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5821223569394434,
+      "learning_rate": 6.054699823067407e-06,
+      "loss": 0.8514,
+      "step": 27175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5794327457685315,
+      "learning_rate": 6.053998689054384e-06,
+      "loss": 0.8148,
+      "step": 27176
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9458299073464869,
+      "learning_rate": 6.053297578016224e-06,
+      "loss": 0.6981,
+      "step": 27177
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5098637714983294,
+      "learning_rate": 6.052596489957017e-06,
+      "loss": 0.8175,
+      "step": 27178
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6122941904416557,
+      "learning_rate": 6.051895424880847e-06,
+      "loss": 0.8827,
+      "step": 27179
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7420763029340967,
+      "learning_rate": 6.051194382791785e-06,
+      "loss": 0.6609,
+      "step": 27180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6428071706904337,
+      "learning_rate": 6.050493363693924e-06,
+      "loss": 0.669,
+      "step": 27181
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5860540661789353,
+      "learning_rate": 6.049792367591338e-06,
+      "loss": 0.8104,
+      "step": 27182
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4915414169103318,
+      "learning_rate": 6.049091394488116e-06,
+      "loss": 0.8401,
+      "step": 27183
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7842998312766909,
+      "learning_rate": 6.048390444388329e-06,
+      "loss": 0.7465,
+      "step": 27184
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6106260499246334,
+      "learning_rate": 6.047689517296067e-06,
+      "loss": 0.7902,
+      "step": 27185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8832870554417651,
+      "learning_rate": 6.04698861321541e-06,
+      "loss": 0.8598,
+      "step": 27186
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5128142884274445,
+      "learning_rate": 6.046287732150433e-06,
+      "loss": 0.823,
+      "step": 27187
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9908971986148216,
+      "learning_rate": 6.045586874105221e-06,
+      "loss": 0.7947,
+      "step": 27188
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6818176083738,
+      "learning_rate": 6.044886039083855e-06,
+      "loss": 0.9508,
+      "step": 27189
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7858490200651796,
+      "learning_rate": 6.044185227090415e-06,
+      "loss": 0.75,
+      "step": 27190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.017013969217315,
+      "learning_rate": 6.0434844381289814e-06,
+      "loss": 1.0726,
+      "step": 27191
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.87931566696833,
+      "learning_rate": 6.042783672203632e-06,
+      "loss": 0.7873,
+      "step": 27192
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.507983308281021,
+      "learning_rate": 6.042082929318452e-06,
+      "loss": 0.8336,
+      "step": 27193
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6170743155831018,
+      "learning_rate": 6.041382209477516e-06,
+      "loss": 0.9104,
+      "step": 27194
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6055125986784362,
+      "learning_rate": 6.040681512684907e-06,
+      "loss": 0.8386,
+      "step": 27195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5385223887536625,
+      "learning_rate": 6.039980838944701e-06,
+      "loss": 0.7747,
+      "step": 27196
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5753317811227394,
+      "learning_rate": 6.039280188260985e-06,
+      "loss": 0.8414,
+      "step": 27197
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6837754506592615,
+      "learning_rate": 6.0385795606378315e-06,
+      "loss": 0.8377,
+      "step": 27198
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5133714499671125,
+      "learning_rate": 6.037878956079322e-06,
+      "loss": 0.9578,
+      "step": 27199
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.442700889859105,
+      "learning_rate": 6.037178374589536e-06,
+      "loss": 0.8508,
+      "step": 27200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.528436113234549,
+      "learning_rate": 6.0364778161725505e-06,
+      "loss": 0.7703,
+      "step": 27201
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7459713903157206,
+      "learning_rate": 6.035777280832447e-06,
+      "loss": 0.7093,
+      "step": 27202
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5291406918555621,
+      "learning_rate": 6.035076768573303e-06,
+      "loss": 0.924,
+      "step": 27203
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8408681785497203,
+      "learning_rate": 6.034376279399199e-06,
+      "loss": 0.7925,
+      "step": 27204
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5880868141784494,
+      "learning_rate": 6.033675813314208e-06,
+      "loss": 0.8546,
+      "step": 27205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7791061278840663,
+      "learning_rate": 6.032975370322414e-06,
+      "loss": 0.8337,
+      "step": 27206
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9754665574129313,
+      "learning_rate": 6.032274950427896e-06,
+      "loss": 0.8984,
+      "step": 27207
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.620816725933709,
+      "learning_rate": 6.031574553634727e-06,
+      "loss": 0.7891,
+      "step": 27208
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.451087431633613,
+      "learning_rate": 6.030874179946986e-06,
+      "loss": 0.9607,
+      "step": 27209
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.226868540609599,
+      "learning_rate": 6.0301738293687525e-06,
+      "loss": 0.8756,
+      "step": 27210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7779939137947207,
+      "learning_rate": 6.029473501904108e-06,
+      "loss": 0.7929,
+      "step": 27211
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7666895048684107,
+      "learning_rate": 6.028773197557121e-06,
+      "loss": 0.7529,
+      "step": 27212
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6742764296727821,
+      "learning_rate": 6.028072916331874e-06,
+      "loss": 0.8513,
+      "step": 27213
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3737799253770764,
+      "learning_rate": 6.027372658232448e-06,
+      "loss": 0.8312,
+      "step": 27214
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.649283283666066,
+      "learning_rate": 6.026672423262914e-06,
+      "loss": 0.8111,
+      "step": 27215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7742434519448813,
+      "learning_rate": 6.025972211427346e-06,
+      "loss": 0.7847,
+      "step": 27216
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7282399028024336,
+      "learning_rate": 6.025272022729831e-06,
+      "loss": 0.8421,
+      "step": 27217
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7631885890116112,
+      "learning_rate": 6.024571857174443e-06,
+      "loss": 0.8102,
+      "step": 27218
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5350657361618727,
+      "learning_rate": 6.0238717147652525e-06,
+      "loss": 0.8708,
+      "step": 27219
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6200246632696944,
+      "learning_rate": 6.023171595506338e-06,
+      "loss": 0.8947,
+      "step": 27220
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.602696603616684,
+      "learning_rate": 6.022471499401785e-06,
+      "loss": 0.7795,
+      "step": 27221
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5203338139255995,
+      "learning_rate": 6.021771426455656e-06,
+      "loss": 0.8972,
+      "step": 27222
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6091410687293024,
+      "learning_rate": 6.0210713766720344e-06,
+      "loss": 0.8021,
+      "step": 27223
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1363819381011266,
+      "learning_rate": 6.020371350054992e-06,
+      "loss": 0.9291,
+      "step": 27224
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5235413449455568,
+      "learning_rate": 6.019671346608615e-06,
+      "loss": 0.8849,
+      "step": 27225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5612885709760034,
+      "learning_rate": 6.018971366336965e-06,
+      "loss": 0.7282,
+      "step": 27226
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5845520354152784,
+      "learning_rate": 6.018271409244127e-06,
+      "loss": 0.7649,
+      "step": 27227
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.679537710972642,
+      "learning_rate": 6.017571475334174e-06,
+      "loss": 0.8906,
+      "step": 27228
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.957905595357561,
+      "learning_rate": 6.016871564611177e-06,
+      "loss": 0.7698,
+      "step": 27229
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5897626622749523,
+      "learning_rate": 6.016171677079218e-06,
+      "loss": 0.9157,
+      "step": 27230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8553903537269163,
+      "learning_rate": 6.0154718127423685e-06,
+      "loss": 0.8185,
+      "step": 27231
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7043949085122179,
+      "learning_rate": 6.014771971604705e-06,
+      "loss": 0.9115,
+      "step": 27232
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.509592237295889,
+      "learning_rate": 6.014072153670296e-06,
+      "loss": 0.9457,
+      "step": 27233
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.549727007461883,
+      "learning_rate": 6.0133723589432235e-06,
+      "loss": 0.8464,
+      "step": 27234
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.455051267259223,
+      "learning_rate": 6.012672587427562e-06,
+      "loss": 0.8593,
+      "step": 27235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5549604584597005,
+      "learning_rate": 6.01197283912738e-06,
+      "loss": 0.7788,
+      "step": 27236
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8433577783293082,
+      "learning_rate": 6.011273114046754e-06,
+      "loss": 0.9163,
+      "step": 27237
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8168140998174305,
+      "learning_rate": 6.010573412189759e-06,
+      "loss": 0.9932,
+      "step": 27238
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4610345103508033,
+      "learning_rate": 6.009873733560472e-06,
+      "loss": 0.8786,
+      "step": 27239
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8710943103333513,
+      "learning_rate": 6.00917407816296e-06,
+      "loss": 0.8231,
+      "step": 27240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7284313947486571,
+      "learning_rate": 6.0084744460013e-06,
+      "loss": 0.9641,
+      "step": 27241
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7500971377078065,
+      "learning_rate": 6.007774837079567e-06,
+      "loss": 0.8764,
+      "step": 27242
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7011385163278556,
+      "learning_rate": 6.007075251401833e-06,
+      "loss": 0.8426,
+      "step": 27243
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7957729376907383,
+      "learning_rate": 6.00637568897217e-06,
+      "loss": 0.8331,
+      "step": 27244
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4664142239048459,
+      "learning_rate": 6.0056761497946505e-06,
+      "loss": 0.8489,
+      "step": 27245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.62597561226207,
+      "learning_rate": 6.004976633873354e-06,
+      "loss": 1.024,
+      "step": 27246
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.884357053738744,
+      "learning_rate": 6.004277141212345e-06,
+      "loss": 0.8347,
+      "step": 27247
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5489986113209337,
+      "learning_rate": 6.003577671815699e-06,
+      "loss": 0.7808,
+      "step": 27248
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7794238858450984,
+      "learning_rate": 6.002878225687492e-06,
+      "loss": 0.6487,
+      "step": 27249
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.076406415996193,
+      "learning_rate": 6.002178802831789e-06,
+      "loss": 0.8674,
+      "step": 27250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7334682417530508,
+      "learning_rate": 6.001479403252669e-06,
+      "loss": 0.6906,
+      "step": 27251
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.389076567575322,
+      "learning_rate": 6.000780026954199e-06,
+      "loss": 0.6915,
+      "step": 27252
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8650011622409233,
+      "learning_rate": 6.000080673940461e-06,
+      "loss": 0.8599,
+      "step": 27253
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8214192488690029,
+      "learning_rate": 5.999381344215514e-06,
+      "loss": 0.9794,
+      "step": 27254
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5146178489982187,
+      "learning_rate": 5.998682037783437e-06,
+      "loss": 0.842,
+      "step": 27255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4634074709552707,
+      "learning_rate": 5.997982754648297e-06,
+      "loss": 0.9102,
+      "step": 27256
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5458025555417017,
+      "learning_rate": 5.997283494814175e-06,
+      "loss": 0.9185,
+      "step": 27257
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7296513955034862,
+      "learning_rate": 5.99658425828513e-06,
+      "loss": 0.8726,
+      "step": 27258
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7320118913966406,
+      "learning_rate": 5.99588504506524e-06,
+      "loss": 0.6635,
+      "step": 27259
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5310832130341099,
+      "learning_rate": 5.995185855158577e-06,
+      "loss": 0.9442,
+      "step": 27260
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7026663484619422,
+      "learning_rate": 5.994486688569209e-06,
+      "loss": 0.8585,
+      "step": 27261
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7819465501470441,
+      "learning_rate": 5.993787545301204e-06,
+      "loss": 0.7124,
+      "step": 27262
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6626983299808282,
+      "learning_rate": 5.99308842535864e-06,
+      "loss": 0.9585,
+      "step": 27263
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6933215877085344,
+      "learning_rate": 5.9923893287455845e-06,
+      "loss": 0.8165,
+      "step": 27264
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7840121404107098,
+      "learning_rate": 5.991690255466105e-06,
+      "loss": 0.6787,
+      "step": 27265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.125788146204356,
+      "learning_rate": 5.990991205524274e-06,
+      "loss": 0.8723,
+      "step": 27266
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7142125347918453,
+      "learning_rate": 5.990292178924164e-06,
+      "loss": 0.9194,
+      "step": 27267
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.508334504848625,
+      "learning_rate": 5.98959317566984e-06,
+      "loss": 0.8875,
+      "step": 27268
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5501718686143817,
+      "learning_rate": 5.988894195765372e-06,
+      "loss": 0.8024,
+      "step": 27269
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.714422625622733,
+      "learning_rate": 5.988195239214835e-06,
+      "loss": 0.9022,
+      "step": 27270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5126860355030547,
+      "learning_rate": 5.987496306022297e-06,
+      "loss": 0.7395,
+      "step": 27271
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4335738304197139,
+      "learning_rate": 5.986797396191824e-06,
+      "loss": 0.7667,
+      "step": 27272
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.644738403390401,
+      "learning_rate": 5.986098509727485e-06,
+      "loss": 0.7476,
+      "step": 27273
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.785073769482686,
+      "learning_rate": 5.985399646633356e-06,
+      "loss": 0.7694,
+      "step": 27274
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6098056885741248,
+      "learning_rate": 5.984700806913497e-06,
+      "loss": 0.8392,
+      "step": 27275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6017846325965843,
+      "learning_rate": 5.9840019905719815e-06,
+      "loss": 0.8733,
+      "step": 27276
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1575646881530957,
+      "learning_rate": 5.983303197612878e-06,
+      "loss": 0.8826,
+      "step": 27277
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.886874792456181,
+      "learning_rate": 5.9826044280402574e-06,
+      "loss": 0.8222,
+      "step": 27278
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5751043273729342,
+      "learning_rate": 5.981905681858184e-06,
+      "loss": 0.8876,
+      "step": 27279
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6361478915902312,
+      "learning_rate": 5.981206959070728e-06,
+      "loss": 0.8939,
+      "step": 27280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5539672692076145,
+      "learning_rate": 5.980508259681958e-06,
+      "loss": 0.776,
+      "step": 27281
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4479747275131087,
+      "learning_rate": 5.9798095836959405e-06,
+      "loss": 0.8555,
+      "step": 27282
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.563296415064575,
+      "learning_rate": 5.979110931116744e-06,
+      "loss": 0.8399,
+      "step": 27283
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.545987766956818,
+      "learning_rate": 5.978412301948438e-06,
+      "loss": 0.8827,
+      "step": 27284
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6232793776441636,
+      "learning_rate": 5.977713696195089e-06,
+      "loss": 0.8526,
+      "step": 27285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9117473730797625,
+      "learning_rate": 5.977015113860762e-06,
+      "loss": 0.8786,
+      "step": 27286
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.779609071674085,
+      "learning_rate": 5.9763165549495286e-06,
+      "loss": 0.8276,
+      "step": 27287
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1507765952245728,
+      "learning_rate": 5.975618019465454e-06,
+      "loss": 0.8507,
+      "step": 27288
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6151221370678843,
+      "learning_rate": 5.974919507412607e-06,
+      "loss": 0.8941,
+      "step": 27289
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.645615112148415,
+      "learning_rate": 5.974221018795048e-06,
+      "loss": 0.8282,
+      "step": 27290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6718823746136706,
+      "learning_rate": 5.9735225536168505e-06,
+      "loss": 0.7762,
+      "step": 27291
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7346552669889288,
+      "learning_rate": 5.972824111882083e-06,
+      "loss": 0.8125,
+      "step": 27292
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5422303150548016,
+      "learning_rate": 5.972125693594806e-06,
+      "loss": 0.7873,
+      "step": 27293
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7689758969948137,
+      "learning_rate": 5.971427298759087e-06,
+      "loss": 0.6904,
+      "step": 27294
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7816579433707695,
+      "learning_rate": 5.970728927378997e-06,
+      "loss": 0.864,
+      "step": 27295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5765668663325019,
+      "learning_rate": 5.970030579458597e-06,
+      "loss": 0.8792,
+      "step": 27296
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.658877669728995,
+      "learning_rate": 5.969332255001955e-06,
+      "loss": 0.8728,
+      "step": 27297
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4906232813937326,
+      "learning_rate": 5.968633954013134e-06,
+      "loss": 0.7527,
+      "step": 27298
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7024456069181437,
+      "learning_rate": 5.967935676496208e-06,
+      "loss": 0.7938,
+      "step": 27299
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5646229551989845,
+      "learning_rate": 5.967237422455234e-06,
+      "loss": 0.9334,
+      "step": 27300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.246642006376361,
+      "learning_rate": 5.966539191894279e-06,
+      "loss": 0.7824,
+      "step": 27301
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.552547609370124,
+      "learning_rate": 5.965840984817415e-06,
+      "loss": 0.8763,
+      "step": 27302
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6404011601176787,
+      "learning_rate": 5.965142801228696e-06,
+      "loss": 0.8725,
+      "step": 27303
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.775211455331118,
+      "learning_rate": 5.964444641132194e-06,
+      "loss": 0.8435,
+      "step": 27304
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.680214140428375,
+      "learning_rate": 5.9637465045319734e-06,
+      "loss": 0.8624,
+      "step": 27305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8409336387852073,
+      "learning_rate": 5.963048391432102e-06,
+      "loss": 0.7532,
+      "step": 27306
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6798659766959871,
+      "learning_rate": 5.962350301836637e-06,
+      "loss": 0.9484,
+      "step": 27307
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8515013006740615,
+      "learning_rate": 5.961652235749648e-06,
+      "loss": 0.8984,
+      "step": 27308
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8078563304339013,
+      "learning_rate": 5.9609541931752015e-06,
+      "loss": 0.7192,
+      "step": 27309
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.57119858826838,
+      "learning_rate": 5.960256174117355e-06,
+      "loss": 0.7271,
+      "step": 27310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4840022500993826,
+      "learning_rate": 5.959558178580174e-06,
+      "loss": 0.7525,
+      "step": 27311
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7418163635074062,
+      "learning_rate": 5.958860206567727e-06,
+      "loss": 0.7221,
+      "step": 27312
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7560391994030797,
+      "learning_rate": 5.958162258084077e-06,
+      "loss": 0.6925,
+      "step": 27313
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5719425358137633,
+      "learning_rate": 5.957464333133282e-06,
+      "loss": 0.9577,
+      "step": 27314
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.077909433383316,
+      "learning_rate": 5.9567664317194115e-06,
+      "loss": 0.7715,
+      "step": 27315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7759975472954009,
+      "learning_rate": 5.956068553846528e-06,
+      "loss": 0.6941,
+      "step": 27316
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9059447548164723,
+      "learning_rate": 5.955370699518694e-06,
+      "loss": 0.8798,
+      "step": 27317
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8640507389597496,
+      "learning_rate": 5.954672868739968e-06,
+      "loss": 0.8175,
+      "step": 27318
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4462038046584682,
+      "learning_rate": 5.95397506151442e-06,
+      "loss": 0.8275,
+      "step": 27319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7018010689310898,
+      "learning_rate": 5.953277277846112e-06,
+      "loss": 0.7751,
+      "step": 27320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.618905684479937,
+      "learning_rate": 5.952579517739103e-06,
+      "loss": 0.7882,
+      "step": 27321
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6058405512498948,
+      "learning_rate": 5.951881781197456e-06,
+      "loss": 0.7951,
+      "step": 27322
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8305544431501035,
+      "learning_rate": 5.95118406822524e-06,
+      "loss": 0.8201,
+      "step": 27323
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6952056298467553,
+      "learning_rate": 5.950486378826508e-06,
+      "loss": 0.9775,
+      "step": 27324
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5965845095340305,
+      "learning_rate": 5.949788713005326e-06,
+      "loss": 0.7229,
+      "step": 27325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4793551236145221,
+      "learning_rate": 5.949091070765755e-06,
+      "loss": 0.9531,
+      "step": 27326
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7867615341937644,
+      "learning_rate": 5.948393452111863e-06,
+      "loss": 0.6018,
+      "step": 27327
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6476274354213936,
+      "learning_rate": 5.9476958570477036e-06,
+      "loss": 0.7131,
+      "step": 27328
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5053685040487237,
+      "learning_rate": 5.946998285577343e-06,
+      "loss": 0.7299,
+      "step": 27329
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7208626177433333,
+      "learning_rate": 5.946300737704845e-06,
+      "loss": 0.8424,
+      "step": 27330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5559536514471173,
+      "learning_rate": 5.945603213434261e-06,
+      "loss": 0.9303,
+      "step": 27331
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.513049281675913,
+      "learning_rate": 5.9449057127696616e-06,
+      "loss": 0.8968,
+      "step": 27332
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6619848855645782,
+      "learning_rate": 5.944208235715105e-06,
+      "loss": 0.749,
+      "step": 27333
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6407564894601177,
+      "learning_rate": 5.943510782274655e-06,
+      "loss": 0.7745,
+      "step": 27334
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5036743047320023,
+      "learning_rate": 5.942813352452365e-06,
+      "loss": 0.9003,
+      "step": 27335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0848384712889376,
+      "learning_rate": 5.9421159462523e-06,
+      "loss": 0.8586,
+      "step": 27336
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.914847213137379,
+      "learning_rate": 5.941418563678525e-06,
+      "loss": 0.9034,
+      "step": 27337
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5331628356031746,
+      "learning_rate": 5.940721204735094e-06,
+      "loss": 0.9005,
+      "step": 27338
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6260156487157351,
+      "learning_rate": 5.940023869426067e-06,
+      "loss": 0.806,
+      "step": 27339
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7394840050825455,
+      "learning_rate": 5.939326557755508e-06,
+      "loss": 0.8936,
+      "step": 27340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8953153244657042,
+      "learning_rate": 5.938629269727477e-06,
+      "loss": 0.806,
+      "step": 27341
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6585854318421307,
+      "learning_rate": 5.937932005346031e-06,
+      "loss": 0.9311,
+      "step": 27342
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5794220590952124,
+      "learning_rate": 5.93723476461523e-06,
+      "loss": 0.8952,
+      "step": 27343
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6479674577233516,
+      "learning_rate": 5.936537547539136e-06,
+      "loss": 0.794,
+      "step": 27344
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7517661200601332,
+      "learning_rate": 5.9358403541218055e-06,
+      "loss": 0.683,
+      "step": 27345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6369936546710515,
+      "learning_rate": 5.9351431843673e-06,
+      "loss": 0.7579,
+      "step": 27346
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4890362961173502,
+      "learning_rate": 5.934446038279676e-06,
+      "loss": 0.8688,
+      "step": 27347
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7028732346804976,
+      "learning_rate": 5.933748915862997e-06,
+      "loss": 0.8889,
+      "step": 27348
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5467183486908884,
+      "learning_rate": 5.933051817121318e-06,
+      "loss": 0.7992,
+      "step": 27349
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6225565774866617,
+      "learning_rate": 5.9323547420586965e-06,
+      "loss": 0.8244,
+      "step": 27350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7675812149547736,
+      "learning_rate": 5.9316576906792e-06,
+      "loss": 0.8207,
+      "step": 27351
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.618397262879812,
+      "learning_rate": 5.930960662986873e-06,
+      "loss": 0.8901,
+      "step": 27352
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4037217675392433,
+      "learning_rate": 5.930263658985785e-06,
+      "loss": 0.8674,
+      "step": 27353
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4667506773790566,
+      "learning_rate": 5.9295666786799874e-06,
+      "loss": 0.8833,
+      "step": 27354
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9759583583281024,
+      "learning_rate": 5.928869722073549e-06,
+      "loss": 0.8032,
+      "step": 27355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6356293399159878,
+      "learning_rate": 5.928172789170512e-06,
+      "loss": 0.8545,
+      "step": 27356
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7772192688738755,
+      "learning_rate": 5.927475879974946e-06,
+      "loss": 0.7479,
+      "step": 27357
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6622314967186054,
+      "learning_rate": 5.926778994490905e-06,
+      "loss": 0.937,
+      "step": 27358
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5960867884102963,
+      "learning_rate": 5.926082132722445e-06,
+      "loss": 0.8812,
+      "step": 27359
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1352490982365295,
+      "learning_rate": 5.925385294673623e-06,
+      "loss": 0.9081,
+      "step": 27360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.590934954811551,
+      "learning_rate": 5.924688480348498e-06,
+      "loss": 0.9942,
+      "step": 27361
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5605630624819884,
+      "learning_rate": 5.92399168975113e-06,
+      "loss": 0.8492,
+      "step": 27362
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5661191783902113,
+      "learning_rate": 5.923294922885569e-06,
+      "loss": 0.8004,
+      "step": 27363
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5365619571526323,
+      "learning_rate": 5.922598179755879e-06,
+      "loss": 0.8069,
+      "step": 27364
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7968449454184752,
+      "learning_rate": 5.921901460366113e-06,
+      "loss": 0.8579,
+      "step": 27365
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.536022810140892,
+      "learning_rate": 5.921204764720327e-06,
+      "loss": 0.8194,
+      "step": 27366
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6101358926879035,
+      "learning_rate": 5.920508092822575e-06,
+      "loss": 0.7839,
+      "step": 27367
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8090965417393556,
+      "learning_rate": 5.919811444676919e-06,
+      "loss": 0.7388,
+      "step": 27368
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.789473952690968,
+      "learning_rate": 5.919114820287416e-06,
+      "loss": 0.753,
+      "step": 27369
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9089833450852474,
+      "learning_rate": 5.918418219658115e-06,
+      "loss": 0.8901,
+      "step": 27370
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6224932828388443,
+      "learning_rate": 5.917721642793074e-06,
+      "loss": 0.8431,
+      "step": 27371
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.620322345086822,
+      "learning_rate": 5.917025089696355e-06,
+      "loss": 0.8366,
+      "step": 27372
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.807783099010661,
+      "learning_rate": 5.916328560372004e-06,
+      "loss": 0.9314,
+      "step": 27373
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7488946639631644,
+      "learning_rate": 5.915632054824083e-06,
+      "loss": 0.7684,
+      "step": 27374
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6797111322090612,
+      "learning_rate": 5.914935573056641e-06,
+      "loss": 0.7176,
+      "step": 27375
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7714771287387653,
+      "learning_rate": 5.914239115073743e-06,
+      "loss": 0.6832,
+      "step": 27376
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6837614971180461,
+      "learning_rate": 5.913542680879437e-06,
+      "loss": 0.9481,
+      "step": 27377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7941702987316189,
+      "learning_rate": 5.912846270477779e-06,
+      "loss": 0.731,
+      "step": 27378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.611734279339704,
+      "learning_rate": 5.912149883872822e-06,
+      "loss": 0.8103,
+      "step": 27379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5312973858717138,
+      "learning_rate": 5.911453521068626e-06,
+      "loss": 0.8314,
+      "step": 27380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7756329317997275,
+      "learning_rate": 5.91075718206924e-06,
+      "loss": 0.7668,
+      "step": 27381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4979743776394632,
+      "learning_rate": 5.910060866878719e-06,
+      "loss": 0.8692,
+      "step": 27382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6596616528727,
+      "learning_rate": 5.909364575501123e-06,
+      "loss": 0.8499,
+      "step": 27383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7840338969831597,
+      "learning_rate": 5.908668307940497e-06,
+      "loss": 0.8849,
+      "step": 27384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6428189826450093,
+      "learning_rate": 5.9079720642009004e-06,
+      "loss": 0.9166,
+      "step": 27385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7774228075805614,
+      "learning_rate": 5.907275844286386e-06,
+      "loss": 0.8518,
+      "step": 27386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.709228915497965,
+      "learning_rate": 5.9065796482010104e-06,
+      "loss": 0.8683,
+      "step": 27387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7417459808485576,
+      "learning_rate": 5.905883475948819e-06,
+      "loss": 0.7941,
+      "step": 27388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5626607838994873,
+      "learning_rate": 5.905187327533873e-06,
+      "loss": 0.8945,
+      "step": 27389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7276045689359025,
+      "learning_rate": 5.904491202960225e-06,
+      "loss": 0.8648,
+      "step": 27390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.55578193933364,
+      "learning_rate": 5.903795102231923e-06,
+      "loss": 0.9234,
+      "step": 27391
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6055393055204132,
+      "learning_rate": 5.903099025353021e-06,
+      "loss": 0.8233,
+      "step": 27392
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5717041574671875,
+      "learning_rate": 5.902402972327575e-06,
+      "loss": 0.7808,
+      "step": 27393
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5781033521261816,
+      "learning_rate": 5.9017069431596395e-06,
+      "loss": 0.8978,
+      "step": 27394
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5742886356157777,
+      "learning_rate": 5.901010937853261e-06,
+      "loss": 0.8116,
+      "step": 27395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7664917595397,
+      "learning_rate": 5.900314956412492e-06,
+      "loss": 0.7239,
+      "step": 27396
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.546086916963691,
+      "learning_rate": 5.8996189988413925e-06,
+      "loss": 0.7484,
+      "step": 27397
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.661294470809465,
+      "learning_rate": 5.898923065144006e-06,
+      "loss": 0.9214,
+      "step": 27398
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4575879445944915,
+      "learning_rate": 5.8982271553243866e-06,
+      "loss": 0.8434,
+      "step": 27399
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4844342004662634,
+      "learning_rate": 5.89753126938659e-06,
+      "loss": 0.9185,
+      "step": 27400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5357952478343404,
+      "learning_rate": 5.896835407334665e-06,
+      "loss": 0.9235,
+      "step": 27401
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1851819504116663,
+      "learning_rate": 5.8961395691726635e-06,
+      "loss": 0.7241,
+      "step": 27402
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.863487274122943,
+      "learning_rate": 5.8954437549046325e-06,
+      "loss": 0.766,
+      "step": 27403
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7479080204849384,
+      "learning_rate": 5.894747964534635e-06,
+      "loss": 0.7041,
+      "step": 27404
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8135017702589027,
+      "learning_rate": 5.8940521980667086e-06,
+      "loss": 0.8463,
+      "step": 27405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.220185112630899,
+      "learning_rate": 5.893356455504911e-06,
+      "loss": 0.7628,
+      "step": 27406
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.009800164854692,
+      "learning_rate": 5.892660736853291e-06,
+      "loss": 0.9597,
+      "step": 27407
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5953009338686739,
+      "learning_rate": 5.891965042115907e-06,
+      "loss": 1.0133,
+      "step": 27408
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.62403152312054,
+      "learning_rate": 5.891269371296798e-06,
+      "loss": 0.8933,
+      "step": 27409
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7526235497246356,
+      "learning_rate": 5.89057372440002e-06,
+      "loss": 0.9112,
+      "step": 27410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6859231568007313,
+      "learning_rate": 5.889878101429624e-06,
+      "loss": 0.8297,
+      "step": 27411
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7873943333575777,
+      "learning_rate": 5.889182502389657e-06,
+      "loss": 0.9329,
+      "step": 27412
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.585058623784777,
+      "learning_rate": 5.88848692728417e-06,
+      "loss": 0.8177,
+      "step": 27413
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5203326694003285,
+      "learning_rate": 5.887791376117216e-06,
+      "loss": 0.8177,
+      "step": 27414
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4442183034267846,
+      "learning_rate": 5.887095848892846e-06,
+      "loss": 0.844,
+      "step": 27415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7317937127537408,
+      "learning_rate": 5.886400345615098e-06,
+      "loss": 0.8958,
+      "step": 27416
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7797187511983301,
+      "learning_rate": 5.885704866288034e-06,
+      "loss": 0.7448,
+      "step": 27417
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6895095339113275,
+      "learning_rate": 5.885009410915701e-06,
+      "loss": 0.8501,
+      "step": 27418
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4134267769437452,
+      "learning_rate": 5.8843139795021446e-06,
+      "loss": 0.8911,
+      "step": 27419
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7735967968017738,
+      "learning_rate": 5.883618572051411e-06,
+      "loss": 0.7687,
+      "step": 27420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7017708083589858,
+      "learning_rate": 5.882923188567558e-06,
+      "loss": 0.6927,
+      "step": 27421
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7172660570496678,
+      "learning_rate": 5.88222782905463e-06,
+      "loss": 0.8149,
+      "step": 27422
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7910873300009252,
+      "learning_rate": 5.881532493516673e-06,
+      "loss": 0.7246,
+      "step": 27423
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6234933035159802,
+      "learning_rate": 5.880837181957737e-06,
+      "loss": 0.7791,
+      "step": 27424
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7554667461368547,
+      "learning_rate": 5.880141894381873e-06,
+      "loss": 0.8812,
+      "step": 27425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7514963477201733,
+      "learning_rate": 5.879446630793126e-06,
+      "loss": 0.8536,
+      "step": 27426
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7616770941167793,
+      "learning_rate": 5.878751391195546e-06,
+      "loss": 0.6756,
+      "step": 27427
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.767080487937093,
+      "learning_rate": 5.8780561755931785e-06,
+      "loss": 0.8533,
+      "step": 27428
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5186574402655186,
+      "learning_rate": 5.877360983990077e-06,
+      "loss": 0.8919,
+      "step": 27429
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6837349809296052,
+      "learning_rate": 5.876665816390281e-06,
+      "loss": 0.8393,
+      "step": 27430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4944867182632262,
+      "learning_rate": 5.875970672797844e-06,
+      "loss": 0.8225,
+      "step": 27431
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5567985578359276,
+      "learning_rate": 5.8752755532168125e-06,
+      "loss": 0.7115,
+      "step": 27432
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7624793979945013,
+      "learning_rate": 5.874580457651229e-06,
+      "loss": 0.9151,
+      "step": 27433
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.428803193263215,
+      "learning_rate": 5.8738853861051455e-06,
+      "loss": 0.84,
+      "step": 27434
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5218560023344134,
+      "learning_rate": 5.873190338582608e-06,
+      "loss": 0.7376,
+      "step": 27435
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8164537057971126,
+      "learning_rate": 5.8724953150876655e-06,
+      "loss": 0.8682,
+      "step": 27436
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.503873164142758,
+      "learning_rate": 5.871800315624357e-06,
+      "loss": 0.9493,
+      "step": 27437
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7783192645086345,
+      "learning_rate": 5.871105340196737e-06,
+      "loss": 0.7501,
+      "step": 27438
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8541541888363324,
+      "learning_rate": 5.87041038880885e-06,
+      "loss": 0.8039,
+      "step": 27439
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.618031579152851,
+      "learning_rate": 5.86971546146474e-06,
+      "loss": 0.7525,
+      "step": 27440
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5799000274463884,
+      "learning_rate": 5.869020558168452e-06,
+      "loss": 0.844,
+      "step": 27441
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1116811421933317,
+      "learning_rate": 5.868325678924036e-06,
+      "loss": 0.8003,
+      "step": 27442
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5098396879689595,
+      "learning_rate": 5.867630823735537e-06,
+      "loss": 0.9155,
+      "step": 27443
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.694743568432465,
+      "learning_rate": 5.866935992606999e-06,
+      "loss": 0.8407,
+      "step": 27444
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6094962454219783,
+      "learning_rate": 5.866241185542466e-06,
+      "loss": 0.8472,
+      "step": 27445
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4937512284209848,
+      "learning_rate": 5.865546402545989e-06,
+      "loss": 0.8645,
+      "step": 27446
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5868261321539074,
+      "learning_rate": 5.864851643621607e-06,
+      "loss": 0.8722,
+      "step": 27447
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8405023428747884,
+      "learning_rate": 5.864156908773369e-06,
+      "loss": 0.7932,
+      "step": 27448
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.216402225728071,
+      "learning_rate": 5.863462198005318e-06,
+      "loss": 0.7788,
+      "step": 27449
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8071862449272912,
+      "learning_rate": 5.862767511321502e-06,
+      "loss": 0.7156,
+      "step": 27450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2509101091502353,
+      "learning_rate": 5.862072848725961e-06,
+      "loss": 0.8013,
+      "step": 27451
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6288074335077782,
+      "learning_rate": 5.8613782102227416e-06,
+      "loss": 0.9116,
+      "step": 27452
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6165280825427644,
+      "learning_rate": 5.860683595815894e-06,
+      "loss": 0.8113,
+      "step": 27453
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5146563579220347,
+      "learning_rate": 5.85998900550945e-06,
+      "loss": 0.8619,
+      "step": 27454
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.670424405149115,
+      "learning_rate": 5.8592944393074634e-06,
+      "loss": 0.7849,
+      "step": 27455
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5232743886977524,
+      "learning_rate": 5.858599897213975e-06,
+      "loss": 0.8799,
+      "step": 27456
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.476396012397006,
+      "learning_rate": 5.857905379233033e-06,
+      "loss": 0.8296,
+      "step": 27457
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7311239882671823,
+      "learning_rate": 5.857210885368672e-06,
+      "loss": 0.7091,
+      "step": 27458
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6046076200405253,
+      "learning_rate": 5.856516415624942e-06,
+      "loss": 0.9621,
+      "step": 27459
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8427457865172946,
+      "learning_rate": 5.855821970005888e-06,
+      "loss": 0.7203,
+      "step": 27460
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7713517782401822,
+      "learning_rate": 5.855127548515547e-06,
+      "loss": 0.9567,
+      "step": 27461
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9604004681952127,
+      "learning_rate": 5.8544331511579664e-06,
+      "loss": 0.9312,
+      "step": 27462
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7940023912321226,
+      "learning_rate": 5.85373877793719e-06,
+      "loss": 0.894,
+      "step": 27463
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6517539986891945,
+      "learning_rate": 5.8530444288572595e-06,
+      "loss": 0.8662,
+      "step": 27464
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.487090210179574,
+      "learning_rate": 5.852350103922214e-06,
+      "loss": 0.7628,
+      "step": 27465
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.74140644736342,
+      "learning_rate": 5.851655803136101e-06,
+      "loss": 0.7243,
+      "step": 27466
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0710828566825783,
+      "learning_rate": 5.850961526502964e-06,
+      "loss": 0.7612,
+      "step": 27467
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6062389706502465,
+      "learning_rate": 5.850267274026839e-06,
+      "loss": 0.7123,
+      "step": 27468
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.829836219973277,
+      "learning_rate": 5.849573045711771e-06,
+      "loss": 1.033,
+      "step": 27469
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.92415257100561,
+      "learning_rate": 5.848878841561805e-06,
+      "loss": 0.7123,
+      "step": 27470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5475843914242262,
+      "learning_rate": 5.848184661580982e-06,
+      "loss": 0.8194,
+      "step": 27471
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7697001000442143,
+      "learning_rate": 5.847490505773341e-06,
+      "loss": 0.665,
+      "step": 27472
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6684165183689637,
+      "learning_rate": 5.846796374142923e-06,
+      "loss": 0.8007,
+      "step": 27473
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.664567385105894,
+      "learning_rate": 5.8461022666937735e-06,
+      "loss": 0.7516,
+      "step": 27474
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6960831238264915,
+      "learning_rate": 5.8454081834299324e-06,
+      "loss": 0.8241,
+      "step": 27475
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7469491819006409,
+      "learning_rate": 5.844714124355438e-06,
+      "loss": 0.9469,
+      "step": 27476
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7152016757742992,
+      "learning_rate": 5.844020089474334e-06,
+      "loss": 0.9285,
+      "step": 27477
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5397003296398002,
+      "learning_rate": 5.8433260787906624e-06,
+      "loss": 0.7545,
+      "step": 27478
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5481194703346102,
+      "learning_rate": 5.84263209230846e-06,
+      "loss": 0.9837,
+      "step": 27479
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4166988469726047,
+      "learning_rate": 5.841938130031772e-06,
+      "loss": 0.8759,
+      "step": 27480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5979635354395996,
+      "learning_rate": 5.841244191964637e-06,
+      "loss": 0.778,
+      "step": 27481
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8363036979878875,
+      "learning_rate": 5.8405502781110925e-06,
+      "loss": 0.8293,
+      "step": 27482
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4355028901935991,
+      "learning_rate": 5.839856388475182e-06,
+      "loss": 0.7926,
+      "step": 27483
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6299875282115641,
+      "learning_rate": 5.839162523060944e-06,
+      "loss": 0.749,
+      "step": 27484
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5019445004970517,
+      "learning_rate": 5.8384686818724205e-06,
+      "loss": 0.9077,
+      "step": 27485
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6696848927563483,
+      "learning_rate": 5.83777486491365e-06,
+      "loss": 0.9772,
+      "step": 27486
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7598807631027042,
+      "learning_rate": 5.837081072188669e-06,
+      "loss": 0.7001,
+      "step": 27487
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6969181460168603,
+      "learning_rate": 5.836387303701525e-06,
+      "loss": 1.0235,
+      "step": 27488
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9961033596219955,
+      "learning_rate": 5.835693559456248e-06,
+      "loss": 0.8397,
+      "step": 27489
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.0027465907227,
+      "learning_rate": 5.8349998394568845e-06,
+      "loss": 0.889,
+      "step": 27490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3146038994438434,
+      "learning_rate": 5.834306143707467e-06,
+      "loss": 0.9438,
+      "step": 27491
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6807266417961482,
+      "learning_rate": 5.833612472212043e-06,
+      "loss": 0.9414,
+      "step": 27492
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8102889549505206,
+      "learning_rate": 5.83291882497464e-06,
+      "loss": 0.7356,
+      "step": 27493
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8905553103103394,
+      "learning_rate": 5.832225201999308e-06,
+      "loss": 0.8776,
+      "step": 27494
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.922077524473447,
+      "learning_rate": 5.83153160329008e-06,
+      "loss": 0.8368,
+      "step": 27495
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5489743189173657,
+      "learning_rate": 5.830838028850992e-06,
+      "loss": 0.9063,
+      "step": 27496
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5125369093841876,
+      "learning_rate": 5.830144478686084e-06,
+      "loss": 0.8068,
+      "step": 27497
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4982778245355481,
+      "learning_rate": 5.829450952799398e-06,
+      "loss": 0.8024,
+      "step": 27498
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6633908882388715,
+      "learning_rate": 5.828757451194971e-06,
+      "loss": 0.8381,
+      "step": 27499
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.383153112002245,
+      "learning_rate": 5.828063973876834e-06,
+      "loss": 0.9092,
+      "step": 27500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5258056719943993,
+      "learning_rate": 5.827370520849029e-06,
+      "loss": 0.9473,
+      "step": 27501
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8165928158712534,
+      "learning_rate": 5.826677092115598e-06,
+      "loss": 0.7021,
+      "step": 27502
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7296385612432028,
+      "learning_rate": 5.8259836876805745e-06,
+      "loss": 1.0042,
+      "step": 27503
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7717391085633276,
+      "learning_rate": 5.82529030754799e-06,
+      "loss": 0.8608,
+      "step": 27504
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7625327882708735,
+      "learning_rate": 5.824596951721888e-06,
+      "loss": 1.0123,
+      "step": 27505
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6537892497386812,
+      "learning_rate": 5.823903620206308e-06,
+      "loss": 0.7949,
+      "step": 27506
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6454571226028185,
+      "learning_rate": 5.8232103130052855e-06,
+      "loss": 0.8507,
+      "step": 27507
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9572026274729981,
+      "learning_rate": 5.8225170301228485e-06,
+      "loss": 0.8098,
+      "step": 27508
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1189381600217563,
+      "learning_rate": 5.8218237715630396e-06,
+      "loss": 0.8024,
+      "step": 27509
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3992592913962998,
+      "learning_rate": 5.821130537329902e-06,
+      "loss": 0.844,
+      "step": 27510
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5301717298626485,
+      "learning_rate": 5.820437327427464e-06,
+      "loss": 0.8308,
+      "step": 27511
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.725129049138597,
+      "learning_rate": 5.819744141859759e-06,
+      "loss": 0.8352,
+      "step": 27512
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6740801795767126,
+      "learning_rate": 5.8190509806308316e-06,
+      "loss": 0.691,
+      "step": 27513
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6085897256762123,
+      "learning_rate": 5.8183578437447085e-06,
+      "loss": 0.7532,
+      "step": 27514
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4952947740730513,
+      "learning_rate": 5.817664731205429e-06,
+      "loss": 0.927,
+      "step": 27515
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.913589014332937,
+      "learning_rate": 5.816971643017036e-06,
+      "loss": 0.8007,
+      "step": 27516
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5479835158994386,
+      "learning_rate": 5.816278579183556e-06,
+      "loss": 0.8372,
+      "step": 27517
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5692320812782583,
+      "learning_rate": 5.815585539709023e-06,
+      "loss": 0.8923,
+      "step": 27518
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6190892387255345,
+      "learning_rate": 5.814892524597477e-06,
+      "loss": 0.7486,
+      "step": 27519
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7400206053761722,
+      "learning_rate": 5.814199533852959e-06,
+      "loss": 0.6909,
+      "step": 27520
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4312125883393465,
+      "learning_rate": 5.8135065674794885e-06,
+      "loss": 0.8356,
+      "step": 27521
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.991697980963235,
+      "learning_rate": 5.812813625481108e-06,
+      "loss": 0.803,
+      "step": 27522
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6409102915424096,
+      "learning_rate": 5.812120707861852e-06,
+      "loss": 0.8429,
+      "step": 27523
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7503621895185723,
+      "learning_rate": 5.811427814625765e-06,
+      "loss": 0.7691,
+      "step": 27524
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7938634372437876,
+      "learning_rate": 5.8107349457768615e-06,
+      "loss": 0.7218,
+      "step": 27525
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5528415716403698,
+      "learning_rate": 5.810042101319185e-06,
+      "loss": 0.8233,
+      "step": 27526
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6406362008275843,
+      "learning_rate": 5.8093492812567754e-06,
+      "loss": 0.8699,
+      "step": 27527
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9521918395753142,
+      "learning_rate": 5.808656485593657e-06,
+      "loss": 0.8319,
+      "step": 27528
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.151227465197668,
+      "learning_rate": 5.80796371433387e-06,
+      "loss": 0.9202,
+      "step": 27529
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6210793533317875,
+      "learning_rate": 5.807270967481442e-06,
+      "loss": 0.7678,
+      "step": 27530
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4986777911065159,
+      "learning_rate": 5.8065782450404155e-06,
+      "loss": 0.8752,
+      "step": 27531
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7287271291771085,
+      "learning_rate": 5.805885547014812e-06,
+      "loss": 0.7822,
+      "step": 27532
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.526672571398265,
+      "learning_rate": 5.805192873408676e-06,
+      "loss": 0.8167,
+      "step": 27533
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9264163338753728,
+      "learning_rate": 5.804500224226034e-06,
+      "loss": 0.7597,
+      "step": 27534
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6110436564460753,
+      "learning_rate": 5.803807599470917e-06,
+      "loss": 0.7928,
+      "step": 27535
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9403830823475448,
+      "learning_rate": 5.80311499914736e-06,
+      "loss": 0.8961,
+      "step": 27536
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7382896871627102,
+      "learning_rate": 5.8024224232594e-06,
+      "loss": 0.752,
+      "step": 27537
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5147235737281155,
+      "learning_rate": 5.801729871811066e-06,
+      "loss": 0.9678,
+      "step": 27538
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7031486534334603,
+      "learning_rate": 5.801037344806386e-06,
+      "loss": 0.9315,
+      "step": 27539
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5863801807481805,
+      "learning_rate": 5.800344842249396e-06,
+      "loss": 0.9516,
+      "step": 27540
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7711079907623024,
+      "learning_rate": 5.79965236414413e-06,
+      "loss": 0.872,
+      "step": 27541
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7891965625468392,
+      "learning_rate": 5.798959910494622e-06,
+      "loss": 0.6871,
+      "step": 27542
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5792402809097372,
+      "learning_rate": 5.798267481304891e-06,
+      "loss": 0.8325,
+      "step": 27543
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6492022637705004,
+      "learning_rate": 5.797575076578978e-06,
+      "loss": 0.7477,
+      "step": 27544
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8874027143652117,
+      "learning_rate": 5.79688269632092e-06,
+      "loss": 0.7943,
+      "step": 27545
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.695336823739414,
+      "learning_rate": 5.79619034053474e-06,
+      "loss": 0.8216,
+      "step": 27546
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5662692355597378,
+      "learning_rate": 5.795498009224465e-06,
+      "loss": 0.7925,
+      "step": 27547
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.544891783648658,
+      "learning_rate": 5.794805702394138e-06,
+      "loss": 0.8259,
+      "step": 27548
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7436220120770926,
+      "learning_rate": 5.794113420047779e-06,
+      "loss": 0.6856,
+      "step": 27549
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6309328548608588,
+      "learning_rate": 5.793421162189427e-06,
+      "loss": 0.8934,
+      "step": 27550
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9770148413392865,
+      "learning_rate": 5.792728928823105e-06,
+      "loss": 0.8561,
+      "step": 27551
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7477551707493652,
+      "learning_rate": 5.792036719952851e-06,
+      "loss": 0.742,
+      "step": 27552
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.450040497587963,
+      "learning_rate": 5.791344535582687e-06,
+      "loss": 0.6671,
+      "step": 27553
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4354314822372456,
+      "learning_rate": 5.790652375716653e-06,
+      "loss": 0.9295,
+      "step": 27554
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7768503353232648,
+      "learning_rate": 5.789960240358772e-06,
+      "loss": 0.8928,
+      "step": 27555
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6693294083789612,
+      "learning_rate": 5.78926812951307e-06,
+      "loss": 0.777,
+      "step": 27556
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4744027980089454,
+      "learning_rate": 5.788576043183585e-06,
+      "loss": 0.7903,
+      "step": 27557
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.803731824962969,
+      "learning_rate": 5.787883981374347e-06,
+      "loss": 0.7344,
+      "step": 27558
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3520205665882346,
+      "learning_rate": 5.787191944089381e-06,
+      "loss": 0.9013,
+      "step": 27559
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.75540088010195,
+      "learning_rate": 5.786499931332714e-06,
+      "loss": 0.8871,
+      "step": 27560
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5822696192183947,
+      "learning_rate": 5.785807943108377e-06,
+      "loss": 0.7946,
+      "step": 27561
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6043970367647868,
+      "learning_rate": 5.785115979420406e-06,
+      "loss": 0.9273,
+      "step": 27562
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7436173177660663,
+      "learning_rate": 5.784424040272823e-06,
+      "loss": 0.8665,
+      "step": 27563
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.1878573728395985,
+      "learning_rate": 5.7837321256696555e-06,
+      "loss": 0.8434,
+      "step": 27564
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.656112189821583,
+      "learning_rate": 5.783040235614933e-06,
+      "loss": 0.8364,
+      "step": 27565
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6696669767608843,
+      "learning_rate": 5.782348370112689e-06,
+      "loss": 0.8084,
+      "step": 27566
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7549935793550903,
+      "learning_rate": 5.7816565291669504e-06,
+      "loss": 0.7113,
+      "step": 27567
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.753104741840036,
+      "learning_rate": 5.780964712781736e-06,
+      "loss": 0.686,
+      "step": 27568
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5878472386042308,
+      "learning_rate": 5.780272920961087e-06,
+      "loss": 0.7982,
+      "step": 27569
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7762664659485552,
+      "learning_rate": 5.779581153709019e-06,
+      "loss": 0.6708,
+      "step": 27570
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4152242007203735,
+      "learning_rate": 5.778889411029571e-06,
+      "loss": 0.7893,
+      "step": 27571
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7206836907992056,
+      "learning_rate": 5.778197692926761e-06,
+      "loss": 0.7788,
+      "step": 27572
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.486645664784343,
+      "learning_rate": 5.777505999404625e-06,
+      "loss": 0.8452,
+      "step": 27573
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.821115902855222,
+      "learning_rate": 5.776814330467182e-06,
+      "loss": 0.8755,
+      "step": 27574
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6637128003199773,
+      "learning_rate": 5.776122686118466e-06,
+      "loss": 0.8589,
+      "step": 27575
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6210620247224112,
+      "learning_rate": 5.775431066362503e-06,
+      "loss": 0.7313,
+      "step": 27576
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7439878428851,
+      "learning_rate": 5.774739471203311e-06,
+      "loss": 0.673,
+      "step": 27577
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7568772417576997,
+      "learning_rate": 5.774047900644926e-06,
+      "loss": 0.7181,
+      "step": 27578
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5567249498829367,
+      "learning_rate": 5.773356354691374e-06,
+      "loss": 0.862,
+      "step": 27579
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8356208103925509,
+      "learning_rate": 5.772664833346681e-06,
+      "loss": 0.809,
+      "step": 27580
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5968342134488551,
+      "learning_rate": 5.771973336614867e-06,
+      "loss": 0.8812,
+      "step": 27581
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6097596498611988,
+      "learning_rate": 5.7712818644999624e-06,
+      "loss": 0.8595,
+      "step": 27582
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.449684216102821,
+      "learning_rate": 5.7705904170059976e-06,
+      "loss": 0.9456,
+      "step": 27583
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.776272957315386,
+      "learning_rate": 5.769898994136996e-06,
+      "loss": 0.9829,
+      "step": 27584
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8933155723814978,
+      "learning_rate": 5.769207595896975e-06,
+      "loss": 1.0006,
+      "step": 27585
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1202800623705147,
+      "learning_rate": 5.768516222289968e-06,
+      "loss": 0.881,
+      "step": 27586
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4435458628614168,
+      "learning_rate": 5.767824873320005e-06,
+      "loss": 0.8994,
+      "step": 27587
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6067465408073833,
+      "learning_rate": 5.767133548991104e-06,
+      "loss": 0.7657,
+      "step": 27588
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8158782125616606,
+      "learning_rate": 5.766442249307287e-06,
+      "loss": 0.9565,
+      "step": 27589
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5766984179066714,
+      "learning_rate": 5.765750974272588e-06,
+      "loss": 0.8115,
+      "step": 27590
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7791091128186606,
+      "learning_rate": 5.765059723891024e-06,
+      "loss": 0.7191,
+      "step": 27591
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7377522714616727,
+      "learning_rate": 5.764368498166626e-06,
+      "loss": 0.68,
+      "step": 27592
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8138297265078596,
+      "learning_rate": 5.763677297103411e-06,
+      "loss": 0.8679,
+      "step": 27593
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.797988644962055,
+      "learning_rate": 5.7629861207054135e-06,
+      "loss": 0.8739,
+      "step": 27594
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7188906415988974,
+      "learning_rate": 5.7622949689766496e-06,
+      "loss": 0.7364,
+      "step": 27595
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7846799527156965,
+      "learning_rate": 5.761603841921147e-06,
+      "loss": 0.7095,
+      "step": 27596
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6357083170145499,
+      "learning_rate": 5.76091273954293e-06,
+      "loss": 0.838,
+      "step": 27597
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7177764512180457,
+      "learning_rate": 5.760221661846017e-06,
+      "loss": 0.7312,
+      "step": 27598
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5435113484162029,
+      "learning_rate": 5.7595306088344364e-06,
+      "loss": 0.9456,
+      "step": 27599
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.650229351022605,
+      "learning_rate": 5.758839580512214e-06,
+      "loss": 0.7626,
+      "step": 27600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5716586042942935,
+      "learning_rate": 5.7581485768833715e-06,
+      "loss": 0.6941,
+      "step": 27601
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.753112601622951,
+      "learning_rate": 5.757457597951925e-06,
+      "loss": 0.8583,
+      "step": 27602
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5036990350781008,
+      "learning_rate": 5.756766643721905e-06,
+      "loss": 0.8886,
+      "step": 27603
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0209151355993766,
+      "learning_rate": 5.756075714197338e-06,
+      "loss": 0.8111,
+      "step": 27604
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8488333187044423,
+      "learning_rate": 5.755384809382239e-06,
+      "loss": 0.8774,
+      "step": 27605
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.44775929385952,
+      "learning_rate": 5.754693929280633e-06,
+      "loss": 0.8896,
+      "step": 27606
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6393287546132413,
+      "learning_rate": 5.754003073896539e-06,
+      "loss": 0.7948,
+      "step": 27607
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1463050330633617,
+      "learning_rate": 5.75331224323399e-06,
+      "loss": 0.9574,
+      "step": 27608
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6385038267521244,
+      "learning_rate": 5.7526214372970014e-06,
+      "loss": 0.8136,
+      "step": 27609
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6439416381913958,
+      "learning_rate": 5.751930656089591e-06,
+      "loss": 0.6942,
+      "step": 27610
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.760318257174734,
+      "learning_rate": 5.751239899615789e-06,
+      "loss": 0.7837,
+      "step": 27611
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7105169484656595,
+      "learning_rate": 5.75054916787961e-06,
+      "loss": 0.7756,
+      "step": 27612
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5280126779817242,
+      "learning_rate": 5.7498584608850786e-06,
+      "loss": 0.8954,
+      "step": 27613
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6695174265629735,
+      "learning_rate": 5.749167778636221e-06,
+      "loss": 0.9242,
+      "step": 27614
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5944997954528195,
+      "learning_rate": 5.748477121137055e-06,
+      "loss": 0.8997,
+      "step": 27615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.014915331573145,
+      "learning_rate": 5.7477864883915956e-06,
+      "loss": 0.9149,
+      "step": 27616
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6384194079642134,
+      "learning_rate": 5.74709588040387e-06,
+      "loss": 0.6863,
+      "step": 27617
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6938812355142354,
+      "learning_rate": 5.746405297177908e-06,
+      "loss": 0.8654,
+      "step": 27618
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.629204805675312,
+      "learning_rate": 5.745714738717711e-06,
+      "loss": 0.8212,
+      "step": 27619
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6304800848490915,
+      "learning_rate": 5.745024205027312e-06,
+      "loss": 0.8657,
+      "step": 27620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8415838046400617,
+      "learning_rate": 5.744333696110729e-06,
+      "loss": 0.9662,
+      "step": 27621
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.811306036547266,
+      "learning_rate": 5.743643211971989e-06,
+      "loss": 0.9481,
+      "step": 27622
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5057823905024386,
+      "learning_rate": 5.742952752615099e-06,
+      "loss": 0.7368,
+      "step": 27623
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.690304225672492,
+      "learning_rate": 5.742262318044084e-06,
+      "loss": 0.9317,
+      "step": 27624
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4090596399538111,
+      "learning_rate": 5.74157190826297e-06,
+      "loss": 0.8079,
+      "step": 27625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3978769988726116,
+      "learning_rate": 5.740881523275771e-06,
+      "loss": 0.8521,
+      "step": 27626
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5365693459329568,
+      "learning_rate": 5.74019116308651e-06,
+      "loss": 0.8242,
+      "step": 27627
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.485452349613995,
+      "learning_rate": 5.7395008276992015e-06,
+      "loss": 0.8805,
+      "step": 27628
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5248650732525872,
+      "learning_rate": 5.738810517117872e-06,
+      "loss": 0.8778,
+      "step": 27629
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5826469113642976,
+      "learning_rate": 5.7381202313465334e-06,
+      "loss": 0.8188,
+      "step": 27630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.435315583635166,
+      "learning_rate": 5.73742997038921e-06,
+      "loss": 0.7898,
+      "step": 27631
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.148687820229963,
+      "learning_rate": 5.736739734249916e-06,
+      "loss": 0.8581,
+      "step": 27632
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6242333516001388,
+      "learning_rate": 5.736049522932679e-06,
+      "loss": 0.9928,
+      "step": 27633
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6236729871711837,
+      "learning_rate": 5.735359336441504e-06,
+      "loss": 0.7074,
+      "step": 27634
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5648330962130108,
+      "learning_rate": 5.734669174780423e-06,
+      "loss": 0.8105,
+      "step": 27635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.597714276217501,
+      "learning_rate": 5.733979037953449e-06,
+      "loss": 0.9272,
+      "step": 27636
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6554950197448228,
+      "learning_rate": 5.733288925964595e-06,
+      "loss": 0.7698,
+      "step": 27637
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.596756015420159,
+      "learning_rate": 5.732598838817883e-06,
+      "loss": 0.7761,
+      "step": 27638
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.492095489260797,
+      "learning_rate": 5.731908776517336e-06,
+      "loss": 0.857,
+      "step": 27639
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2046264490281215,
+      "learning_rate": 5.731218739066968e-06,
+      "loss": 0.8335,
+      "step": 27640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8515237303879384,
+      "learning_rate": 5.730528726470792e-06,
+      "loss": 0.8493,
+      "step": 27641
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3340685741822003,
+      "learning_rate": 5.7298387387328295e-06,
+      "loss": 0.7358,
+      "step": 27642
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5664866273838878,
+      "learning_rate": 5.729148775857101e-06,
+      "loss": 0.9128,
+      "step": 27643
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5283675988944148,
+      "learning_rate": 5.72845883784762e-06,
+      "loss": 0.7516,
+      "step": 27644
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4215926640867886,
+      "learning_rate": 5.7277689247084016e-06,
+      "loss": 0.76,
+      "step": 27645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6228900932969847,
+      "learning_rate": 5.727079036443464e-06,
+      "loss": 0.7758,
+      "step": 27646
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5048784874051562,
+      "learning_rate": 5.726389173056829e-06,
+      "loss": 0.8134,
+      "step": 27647
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4488461008024467,
+      "learning_rate": 5.725699334552509e-06,
+      "loss": 0.8949,
+      "step": 27648
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6713195995250303,
+      "learning_rate": 5.725009520934517e-06,
+      "loss": 0.9673,
+      "step": 27649
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.94009027661788,
+      "learning_rate": 5.724319732206878e-06,
+      "loss": 0.8596,
+      "step": 27650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.698790640106601,
+      "learning_rate": 5.723629968373598e-06,
+      "loss": 0.838,
+      "step": 27651
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.64406541779132,
+      "learning_rate": 5.722940229438703e-06,
+      "loss": 0.9853,
+      "step": 27652
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7908767416751221,
+      "learning_rate": 5.722250515406199e-06,
+      "loss": 0.7234,
+      "step": 27653
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5114426347763197,
+      "learning_rate": 5.721560826280112e-06,
+      "loss": 0.9419,
+      "step": 27654
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7433447884187105,
+      "learning_rate": 5.720871162064447e-06,
+      "loss": 0.6901,
+      "step": 27655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5989739341217781,
+      "learning_rate": 5.7201815227632305e-06,
+      "loss": 0.8081,
+      "step": 27656
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5044591777822705,
+      "learning_rate": 5.719491908380471e-06,
+      "loss": 0.8323,
+      "step": 27657
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3815693644706062,
+      "learning_rate": 5.718802318920181e-06,
+      "loss": 0.8749,
+      "step": 27658
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5574165070154267,
+      "learning_rate": 5.71811275438638e-06,
+      "loss": 0.8756,
+      "step": 27659
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8479008520274192,
+      "learning_rate": 5.717423214783085e-06,
+      "loss": 0.7538,
+      "step": 27660
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4424028665188706,
+      "learning_rate": 5.7167337001143085e-06,
+      "loss": 0.7929,
+      "step": 27661
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.456300811614343,
+      "learning_rate": 5.716044210384061e-06,
+      "loss": 1.0446,
+      "step": 27662
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5324527714892469,
+      "learning_rate": 5.7153547455963595e-06,
+      "loss": 0.876,
+      "step": 27663
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4471780703276558,
+      "learning_rate": 5.714665305755223e-06,
+      "loss": 0.9621,
+      "step": 27664
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5389168036187406,
+      "learning_rate": 5.7139758908646625e-06,
+      "loss": 0.9445,
+      "step": 27665
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6808057161639705,
+      "learning_rate": 5.7132865009286875e-06,
+      "loss": 0.8763,
+      "step": 27666
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4092872882808962,
+      "learning_rate": 5.712597135951314e-06,
+      "loss": 0.7798,
+      "step": 27667
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6373368295910797,
+      "learning_rate": 5.711907795936564e-06,
+      "loss": 0.7873,
+      "step": 27668
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6956275915827637,
+      "learning_rate": 5.711218480888443e-06,
+      "loss": 0.9773,
+      "step": 27669
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6187886371847815,
+      "learning_rate": 5.710529190810962e-06,
+      "loss": 0.9093,
+      "step": 27670
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5727966654968561,
+      "learning_rate": 5.709839925708141e-06,
+      "loss": 0.8102,
+      "step": 27671
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6904447351438252,
+      "learning_rate": 5.709150685583987e-06,
+      "loss": 0.8142,
+      "step": 27672
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.565383385309936,
+      "learning_rate": 5.708461470442521e-06,
+      "loss": 0.9947,
+      "step": 27673
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.425067164813848,
+      "learning_rate": 5.7077722802877465e-06,
+      "loss": 0.7357,
+      "step": 27674
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6022976121192563,
+      "learning_rate": 5.7070831151236836e-06,
+      "loss": 0.9307,
+      "step": 27675
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7284703761646465,
+      "learning_rate": 5.706393974954338e-06,
+      "loss": 0.8071,
+      "step": 27676
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6954203815238647,
+      "learning_rate": 5.705704859783732e-06,
+      "loss": 0.9629,
+      "step": 27677
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.551518677437794,
+      "learning_rate": 5.70501576961587e-06,
+      "loss": 0.9173,
+      "step": 27678
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5741587121136569,
+      "learning_rate": 5.7043267044547636e-06,
+      "loss": 0.8452,
+      "step": 27679
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5539124178068768,
+      "learning_rate": 5.703637664304425e-06,
+      "loss": 0.7069,
+      "step": 27680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8741459137468097,
+      "learning_rate": 5.702948649168874e-06,
+      "loss": 0.7535,
+      "step": 27681
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.537290590801328,
+      "learning_rate": 5.702259659052115e-06,
+      "loss": 0.7808,
+      "step": 27682
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8969485494316638,
+      "learning_rate": 5.701570693958159e-06,
+      "loss": 0.9356,
+      "step": 27683
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.671566207185525,
+      "learning_rate": 5.700881753891017e-06,
+      "loss": 0.8398,
+      "step": 27684
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5238766907728674,
+      "learning_rate": 5.700192838854707e-06,
+      "loss": 0.9042,
+      "step": 27685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4594859759827534,
+      "learning_rate": 5.6995039488532355e-06,
+      "loss": 0.8475,
+      "step": 27686
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.561589577256314,
+      "learning_rate": 5.6988150838906096e-06,
+      "loss": 0.9051,
+      "step": 27687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6302173521806562,
+      "learning_rate": 5.698126243970845e-06,
+      "loss": 0.8894,
+      "step": 27688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8428658890761866,
+      "learning_rate": 5.697437429097954e-06,
+      "loss": 0.9094,
+      "step": 27689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7974540816373676,
+      "learning_rate": 5.6967486392759455e-06,
+      "loss": 0.8937,
+      "step": 27690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.012131712884775,
+      "learning_rate": 5.696059874508823e-06,
+      "loss": 0.8842,
+      "step": 27691
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7622470427863935,
+      "learning_rate": 5.695371134800609e-06,
+      "loss": 0.7005,
+      "step": 27692
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8585282512896635,
+      "learning_rate": 5.694682420155302e-06,
+      "loss": 0.8218,
+      "step": 27693
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7609989544524998,
+      "learning_rate": 5.693993730576921e-06,
+      "loss": 0.7404,
+      "step": 27694
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.005996878376622,
+      "learning_rate": 5.6933050660694674e-06,
+      "loss": 0.878,
+      "step": 27695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.67989610617075,
+      "learning_rate": 5.69261642663696e-06,
+      "loss": 0.908,
+      "step": 27696
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8015478649880254,
+      "learning_rate": 5.6919278122834e-06,
+      "loss": 0.9044,
+      "step": 27697
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.53460934963885,
+      "learning_rate": 5.691239223012802e-06,
+      "loss": 0.7796,
+      "step": 27698
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6577699292192272,
+      "learning_rate": 5.690550658829175e-06,
+      "loss": 0.7536,
+      "step": 27699
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5493880017025696,
+      "learning_rate": 5.689862119736522e-06,
+      "loss": 0.8878,
+      "step": 27700
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.088792482748866,
+      "learning_rate": 5.689173605738856e-06,
+      "loss": 0.799,
+      "step": 27701
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3941761296469688,
+      "learning_rate": 5.688485116840187e-06,
+      "loss": 0.9216,
+      "step": 27702
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8240092266838128,
+      "learning_rate": 5.68779665304453e-06,
+      "loss": 0.8765,
+      "step": 27703
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.653305796381641,
+      "learning_rate": 5.687108214355877e-06,
+      "loss": 0.8315,
+      "step": 27704
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.03868059922095,
+      "learning_rate": 5.686419800778246e-06,
+      "loss": 0.8005,
+      "step": 27705
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.90246235189651,
+      "learning_rate": 5.685731412315649e-06,
+      "loss": 0.9423,
+      "step": 27706
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5285906669868212,
+      "learning_rate": 5.685043048972091e-06,
+      "loss": 0.7782,
+      "step": 27707
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6162114107175924,
+      "learning_rate": 5.684354710751573e-06,
+      "loss": 0.8063,
+      "step": 27708
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3046715291477509,
+      "learning_rate": 5.683666397658108e-06,
+      "loss": 0.8337,
+      "step": 27709
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8282504194083404,
+      "learning_rate": 5.6829781096957086e-06,
+      "loss": 0.8237,
+      "step": 27710
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7223169355065264,
+      "learning_rate": 5.682289846868372e-06,
+      "loss": 0.9489,
+      "step": 27711
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8658771136320944,
+      "learning_rate": 5.681601609180117e-06,
+      "loss": 0.7438,
+      "step": 27712
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8463968615815236,
+      "learning_rate": 5.680913396634943e-06,
+      "loss": 0.8023,
+      "step": 27713
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.54552396489695,
+      "learning_rate": 5.680225209236855e-06,
+      "loss": 0.9154,
+      "step": 27714
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5306984076495758,
+      "learning_rate": 5.6795370469898625e-06,
+      "loss": 1.0367,
+      "step": 27715
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3210022563773658,
+      "learning_rate": 5.678848909897979e-06,
+      "loss": 0.8052,
+      "step": 27716
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6043729357958423,
+      "learning_rate": 5.678160797965204e-06,
+      "loss": 0.7844,
+      "step": 27717
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.673089973380613,
+      "learning_rate": 5.677472711195542e-06,
+      "loss": 0.8441,
+      "step": 27718
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7079334893435798,
+      "learning_rate": 5.676784649593001e-06,
+      "loss": 0.8494,
+      "step": 27719
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7342675149628775,
+      "learning_rate": 5.676096613161597e-06,
+      "loss": 0.9452,
+      "step": 27720
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5918367046656385,
+      "learning_rate": 5.67540860190532e-06,
+      "loss": 0.7725,
+      "step": 27721
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7641958185709619,
+      "learning_rate": 5.674720615828182e-06,
+      "loss": 0.8931,
+      "step": 27722
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8003413387485727,
+      "learning_rate": 5.674032654934192e-06,
+      "loss": 0.7273,
+      "step": 27723
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6855162649570257,
+      "learning_rate": 5.673344719227356e-06,
+      "loss": 0.8535,
+      "step": 27724
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6726985237928442,
+      "learning_rate": 5.672656808711677e-06,
+      "loss": 0.8599,
+      "step": 27725
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8329562106799104,
+      "learning_rate": 5.671968923391156e-06,
+      "loss": 0.9246,
+      "step": 27726
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6563313538436328,
+      "learning_rate": 5.671281063269807e-06,
+      "loss": 0.8177,
+      "step": 27727
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6176789366408943,
+      "learning_rate": 5.670593228351625e-06,
+      "loss": 0.8016,
+      "step": 27728
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5795459376030305,
+      "learning_rate": 5.669905418640624e-06,
+      "loss": 0.8116,
+      "step": 27729
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6280392374728676,
+      "learning_rate": 5.669217634140802e-06,
+      "loss": 0.7408,
+      "step": 27730
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4832660901554424,
+      "learning_rate": 5.668529874856168e-06,
+      "loss": 0.871,
+      "step": 27731
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.780022258327559,
+      "learning_rate": 5.667842140790722e-06,
+      "loss": 0.7334,
+      "step": 27732
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.459768385552938,
+      "learning_rate": 5.667154431948474e-06,
+      "loss": 0.8629,
+      "step": 27733
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6454159841539544,
+      "learning_rate": 5.666466748333424e-06,
+      "loss": 0.9473,
+      "step": 27734
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7790403585517908,
+      "learning_rate": 5.665779089949574e-06,
+      "loss": 0.7469,
+      "step": 27735
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7575375397362283,
+      "learning_rate": 5.665091456800929e-06,
+      "loss": 0.7268,
+      "step": 27736
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5922584917673486,
+      "learning_rate": 5.664403848891499e-06,
+      "loss": 0.9226,
+      "step": 27737
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4368705050865707,
+      "learning_rate": 5.663716266225281e-06,
+      "loss": 0.8866,
+      "step": 27738
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4611282361181617,
+      "learning_rate": 5.663028708806277e-06,
+      "loss": 0.8347,
+      "step": 27739
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8330374887995282,
+      "learning_rate": 5.662341176638493e-06,
+      "loss": 0.8631,
+      "step": 27740
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9135937543425685,
+      "learning_rate": 5.661653669725935e-06,
+      "loss": 0.938,
+      "step": 27741
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7667014112144483,
+      "learning_rate": 5.660966188072603e-06,
+      "loss": 0.8611,
+      "step": 27742
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6792371377777497,
+      "learning_rate": 5.660278731682496e-06,
+      "loss": 0.7202,
+      "step": 27743
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6234149851996937,
+      "learning_rate": 5.65959130055962e-06,
+      "loss": 0.9001,
+      "step": 27744
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.702484698803816,
+      "learning_rate": 5.658903894707981e-06,
+      "loss": 0.7901,
+      "step": 27745
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7071954563424907,
+      "learning_rate": 5.65821651413158e-06,
+      "loss": 0.7346,
+      "step": 27746
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.776851648676184,
+      "learning_rate": 5.65752915883441e-06,
+      "loss": 0.8433,
+      "step": 27747
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4844371369729792,
+      "learning_rate": 5.656841828820486e-06,
+      "loss": 0.784,
+      "step": 27748
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7107365158939516,
+      "learning_rate": 5.6561545240937995e-06,
+      "loss": 0.8155,
+      "step": 27749
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7752473924504311,
+      "learning_rate": 5.655467244658361e-06,
+      "loss": 0.6972,
+      "step": 27750
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5305514751934997,
+      "learning_rate": 5.654779990518163e-06,
+      "loss": 0.9587,
+      "step": 27751
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.678969666421939,
+      "learning_rate": 5.6540927616772165e-06,
+      "loss": 0.8136,
+      "step": 27752
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7653820562319047,
+      "learning_rate": 5.653405558139513e-06,
+      "loss": 0.9537,
+      "step": 27753
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8383837770323045,
+      "learning_rate": 5.652718379909062e-06,
+      "loss": 0.7638,
+      "step": 27754
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5549666255094439,
+      "learning_rate": 5.652031226989858e-06,
+      "loss": 0.7609,
+      "step": 27755
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5122194466521568,
+      "learning_rate": 5.651344099385908e-06,
+      "loss": 0.8781,
+      "step": 27756
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.696839896445273,
+      "learning_rate": 5.650656997101205e-06,
+      "loss": 0.8481,
+      "step": 27757
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4742604671839064,
+      "learning_rate": 5.64996992013976e-06,
+      "loss": 0.8744,
+      "step": 27758
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.594933460963659,
+      "learning_rate": 5.6492828685055656e-06,
+      "loss": 0.8744,
+      "step": 27759
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5267376362089151,
+      "learning_rate": 5.64859584220262e-06,
+      "loss": 0.8511,
+      "step": 27760
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7443336167877392,
+      "learning_rate": 5.647908841234927e-06,
+      "loss": 0.6861,
+      "step": 27761
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8944371762650596,
+      "learning_rate": 5.64722186560649e-06,
+      "loss": 0.7547,
+      "step": 27762
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.661059127926254,
+      "learning_rate": 5.6465349153213065e-06,
+      "loss": 0.8169,
+      "step": 27763
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.481130987044701,
+      "learning_rate": 5.6458479903833705e-06,
+      "loss": 1.087,
+      "step": 27764
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.422397434035429,
+      "learning_rate": 5.645161090796687e-06,
+      "loss": 0.7757,
+      "step": 27765
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5959347716146315,
+      "learning_rate": 5.644474216565257e-06,
+      "loss": 0.9789,
+      "step": 27766
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.555043025934199,
+      "learning_rate": 5.643787367693077e-06,
+      "loss": 0.7959,
+      "step": 27767
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5673252086771936,
+      "learning_rate": 5.643100544184142e-06,
+      "loss": 0.8684,
+      "step": 27768
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7706154025904969,
+      "learning_rate": 5.642413746042457e-06,
+      "loss": 0.6893,
+      "step": 27769
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5697541612429395,
+      "learning_rate": 5.6417269732720204e-06,
+      "loss": 0.7614,
+      "step": 27770
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6049027289917372,
+      "learning_rate": 5.64104022587683e-06,
+      "loss": 0.898,
+      "step": 27771
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4575231346807873,
+      "learning_rate": 5.64035350386088e-06,
+      "loss": 0.9148,
+      "step": 27772
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4560646449595431,
+      "learning_rate": 5.6396668072281745e-06,
+      "loss": 0.6807,
+      "step": 27773
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.205745562222619,
+      "learning_rate": 5.6389801359827075e-06,
+      "loss": 0.7846,
+      "step": 27774
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8399033686432642,
+      "learning_rate": 5.638293490128481e-06,
+      "loss": 0.9231,
+      "step": 27775
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7903199089009239,
+      "learning_rate": 5.6376068696694875e-06,
+      "loss": 0.7032,
+      "step": 27776
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.471572674642997,
+      "learning_rate": 5.636920274609731e-06,
+      "loss": 0.7871,
+      "step": 27777
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6099701068375372,
+      "learning_rate": 5.636233704953205e-06,
+      "loss": 0.8205,
+      "step": 27778
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4467557509563822,
+      "learning_rate": 5.635547160703909e-06,
+      "loss": 0.7884,
+      "step": 27779
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.788084245537859,
+      "learning_rate": 5.63486064186584e-06,
+      "loss": 0.8877,
+      "step": 27780
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5145060686889755,
+      "learning_rate": 5.63417414844299e-06,
+      "loss": 0.8707,
+      "step": 27781
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.040504890161693,
+      "learning_rate": 5.633487680439362e-06,
+      "loss": 0.9696,
+      "step": 27782
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.465104274113058,
+      "learning_rate": 5.632801237858954e-06,
+      "loss": 0.8529,
+      "step": 27783
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4651239741234217,
+      "learning_rate": 5.632114820705759e-06,
+      "loss": 1.0136,
+      "step": 27784
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7236385344857452,
+      "learning_rate": 5.6314284289837705e-06,
+      "loss": 0.9172,
+      "step": 27785
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5826552331986892,
+      "learning_rate": 5.630742062696989e-06,
+      "loss": 0.8099,
+      "step": 27786
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.582011488986812,
+      "learning_rate": 5.630055721849414e-06,
+      "loss": 0.7891,
+      "step": 27787
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6662013973590775,
+      "learning_rate": 5.62936940644504e-06,
+      "loss": 0.7199,
+      "step": 27788
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6630911135760973,
+      "learning_rate": 5.6286831164878556e-06,
+      "loss": 0.9147,
+      "step": 27789
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5963289891007253,
+      "learning_rate": 5.627996851981863e-06,
+      "loss": 0.8135,
+      "step": 27790
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.001276357062924,
+      "learning_rate": 5.627310612931058e-06,
+      "loss": 0.7947,
+      "step": 27791
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7594487081376737,
+      "learning_rate": 5.626624399339438e-06,
+      "loss": 0.7704,
+      "step": 27792
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5909857394992044,
+      "learning_rate": 5.62593821121099e-06,
+      "loss": 0.755,
+      "step": 27793
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.7485130625844665,
+      "learning_rate": 5.62525204854972e-06,
+      "loss": 0.8767,
+      "step": 27794
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.65066778672634,
+      "learning_rate": 5.624565911359612e-06,
+      "loss": 0.7707,
+      "step": 27795
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8070162154607978,
+      "learning_rate": 5.6238797996446714e-06,
+      "loss": 0.8758,
+      "step": 27796
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.75510910095807,
+      "learning_rate": 5.623193713408883e-06,
+      "loss": 0.856,
+      "step": 27797
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6980775684244263,
+      "learning_rate": 5.6225076526562525e-06,
+      "loss": 0.9282,
+      "step": 27798
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6280244981550556,
+      "learning_rate": 5.621821617390763e-06,
+      "loss": 0.7621,
+      "step": 27799
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5098500901297305,
+      "learning_rate": 5.621135607616413e-06,
+      "loss": 0.9001,
+      "step": 27800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8203093584104164,
+      "learning_rate": 5.620449623337208e-06,
+      "loss": 0.841,
+      "step": 27801
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4445161022826647,
+      "learning_rate": 5.619763664557123e-06,
+      "loss": 0.8715,
+      "step": 27802
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4887893394483045,
+      "learning_rate": 5.61907773128016e-06,
+      "loss": 0.9431,
+      "step": 27803
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.498994105157775,
+      "learning_rate": 5.6183918235103144e-06,
+      "loss": 0.8811,
+      "step": 27804
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.598056352605056,
+      "learning_rate": 5.617705941251586e-06,
+      "loss": 0.9035,
+      "step": 27805
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5350162376128795,
+      "learning_rate": 5.617020084507954e-06,
+      "loss": 0.83,
+      "step": 27806
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5278391023179945,
+      "learning_rate": 5.616334253283418e-06,
+      "loss": 0.8114,
+      "step": 27807
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.438208113416195,
+      "learning_rate": 5.615648447581976e-06,
+      "loss": 0.7651,
+      "step": 27808
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.530931138914033,
+      "learning_rate": 5.614962667407613e-06,
+      "loss": 0.9349,
+      "step": 27809
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7269364765354599,
+      "learning_rate": 5.614276912764329e-06,
+      "loss": 0.7385,
+      "step": 27810
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4836669533506661,
+      "learning_rate": 5.61359118365611e-06,
+      "loss": 0.7569,
+      "step": 27811
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9177387935737917,
+      "learning_rate": 5.612905480086954e-06,
+      "loss": 0.8341,
+      "step": 27812
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8082048235989518,
+      "learning_rate": 5.61221980206085e-06,
+      "loss": 0.9272,
+      "step": 27813
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8481380970551877,
+      "learning_rate": 5.611534149581794e-06,
+      "loss": 0.8681,
+      "step": 27814
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7767260360346363,
+      "learning_rate": 5.610848522653776e-06,
+      "loss": 0.7931,
+      "step": 27815
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7785525583282346,
+      "learning_rate": 5.610162921280783e-06,
+      "loss": 0.7127,
+      "step": 27816
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7123949744109759,
+      "learning_rate": 5.609477345466811e-06,
+      "loss": 0.8655,
+      "step": 27817
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4196133012450893,
+      "learning_rate": 5.608791795215857e-06,
+      "loss": 0.8371,
+      "step": 27818
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6351279784467134,
+      "learning_rate": 5.608106270531907e-06,
+      "loss": 0.8723,
+      "step": 27819
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5424897930279142,
+      "learning_rate": 5.6074207714189475e-06,
+      "loss": 0.8319,
+      "step": 27820
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.631787024628283,
+      "learning_rate": 5.606735297880977e-06,
+      "loss": 0.8117,
+      "step": 27821
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.772494485402316,
+      "learning_rate": 5.606049849921987e-06,
+      "loss": 0.7464,
+      "step": 27822
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.52348136557873,
+      "learning_rate": 5.6053644275459665e-06,
+      "loss": 0.8253,
+      "step": 27823
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8999595614632092,
+      "learning_rate": 5.604679030756901e-06,
+      "loss": 0.9272,
+      "step": 27824
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9129714728939289,
+      "learning_rate": 5.603993659558786e-06,
+      "loss": 0.9752,
+      "step": 27825
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9565165404342941,
+      "learning_rate": 5.603308313955617e-06,
+      "loss": 0.8586,
+      "step": 27826
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7396474621296407,
+      "learning_rate": 5.602622993951377e-06,
+      "loss": 0.9673,
+      "step": 27827
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.841949436361487,
+      "learning_rate": 5.601937699550055e-06,
+      "loss": 0.8069,
+      "step": 27828
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9320971459770977,
+      "learning_rate": 5.601252430755648e-06,
+      "loss": 0.8526,
+      "step": 27829
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4722137709351792,
+      "learning_rate": 5.600567187572139e-06,
+      "loss": 0.8824,
+      "step": 27830
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8467797145288187,
+      "learning_rate": 5.599881970003522e-06,
+      "loss": 0.7559,
+      "step": 27831
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5152052982484403,
+      "learning_rate": 5.599196778053784e-06,
+      "loss": 0.7908,
+      "step": 27832
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7989906894685659,
+      "learning_rate": 5.5985116117269175e-06,
+      "loss": 0.8042,
+      "step": 27833
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5869139299392645,
+      "learning_rate": 5.5978264710269085e-06,
+      "loss": 0.8759,
+      "step": 27834
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6886403023957897,
+      "learning_rate": 5.597141355957749e-06,
+      "loss": 0.9299,
+      "step": 27835
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6096581747661651,
+      "learning_rate": 5.5964562665234275e-06,
+      "loss": 0.8773,
+      "step": 27836
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7509722549205446,
+      "learning_rate": 5.595771202727928e-06,
+      "loss": 0.8207,
+      "step": 27837
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6229125066178627,
+      "learning_rate": 5.595086164575243e-06,
+      "loss": 0.932,
+      "step": 27838
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4787428216567373,
+      "learning_rate": 5.5944011520693645e-06,
+      "loss": 1.0009,
+      "step": 27839
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.640309906756561,
+      "learning_rate": 5.593716165214276e-06,
+      "loss": 0.92,
+      "step": 27840
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.571314595029402,
+      "learning_rate": 5.593031204013965e-06,
+      "loss": 0.7884,
+      "step": 27841
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6561515257009145,
+      "learning_rate": 5.59234626847242e-06,
+      "loss": 0.8569,
+      "step": 27842
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6710184558334977,
+      "learning_rate": 5.591661358593636e-06,
+      "loss": 0.8697,
+      "step": 27843
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7149293841224145,
+      "learning_rate": 5.590976474381595e-06,
+      "loss": 0.8069,
+      "step": 27844
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.093483430044397,
+      "learning_rate": 5.59029161584028e-06,
+      "loss": 0.737,
+      "step": 27845
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5668889318581896,
+      "learning_rate": 5.589606782973683e-06,
+      "loss": 0.7882,
+      "step": 27846
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.843860386865416,
+      "learning_rate": 5.588921975785797e-06,
+      "loss": 0.8473,
+      "step": 27847
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8245626817042198,
+      "learning_rate": 5.588237194280603e-06,
+      "loss": 0.8866,
+      "step": 27848
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5472464366124858,
+      "learning_rate": 5.5875524384620846e-06,
+      "loss": 0.7258,
+      "step": 27849
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7115026606957577,
+      "learning_rate": 5.586867708334236e-06,
+      "loss": 0.7088,
+      "step": 27850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.650382481310153,
+      "learning_rate": 5.5861830039010375e-06,
+      "loss": 0.8738,
+      "step": 27851
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5573691216145074,
+      "learning_rate": 5.585498325166484e-06,
+      "loss": 0.8748,
+      "step": 27852
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.823947005426198,
+      "learning_rate": 5.584813672134553e-06,
+      "loss": 0.7221,
+      "step": 27853
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5798288388397634,
+      "learning_rate": 5.584129044809237e-06,
+      "loss": 0.8181,
+      "step": 27854
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7140346431537312,
+      "learning_rate": 5.583444443194517e-06,
+      "loss": 0.8596,
+      "step": 27855
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.700051652509384,
+      "learning_rate": 5.582759867294385e-06,
+      "loss": 0.905,
+      "step": 27856
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7276796637715917,
+      "learning_rate": 5.5820753171128235e-06,
+      "loss": 0.6656,
+      "step": 27857
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.745223930834159,
+      "learning_rate": 5.581390792653815e-06,
+      "loss": 0.7934,
+      "step": 27858
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.523398839822236,
+      "learning_rate": 5.580706293921348e-06,
+      "loss": 0.8998,
+      "step": 27859
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5247120360589963,
+      "learning_rate": 5.580021820919411e-06,
+      "loss": 0.7687,
+      "step": 27860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.776516071399948,
+      "learning_rate": 5.5793373736519876e-06,
+      "loss": 0.6973,
+      "step": 27861
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6546589480727918,
+      "learning_rate": 5.578652952123056e-06,
+      "loss": 0.8064,
+      "step": 27862
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6507885740236605,
+      "learning_rate": 5.577968556336608e-06,
+      "loss": 0.8171,
+      "step": 27863
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7737028284224686,
+      "learning_rate": 5.577284186296631e-06,
+      "loss": 0.6855,
+      "step": 27864
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7538486649849164,
+      "learning_rate": 5.576599842007105e-06,
+      "loss": 0.7242,
+      "step": 27865
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1711410836987084,
+      "learning_rate": 5.575915523472012e-06,
+      "loss": 0.8444,
+      "step": 27866
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5004620408951848,
+      "learning_rate": 5.575231230695339e-06,
+      "loss": 0.867,
+      "step": 27867
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5590460202950305,
+      "learning_rate": 5.5745469636810746e-06,
+      "loss": 0.9149,
+      "step": 27868
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4820433773739252,
+      "learning_rate": 5.573862722433199e-06,
+      "loss": 0.8153,
+      "step": 27869
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5093830836994169,
+      "learning_rate": 5.573178506955692e-06,
+      "loss": 0.8216,
+      "step": 27870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9629132097965005,
+      "learning_rate": 5.572494317252545e-06,
+      "loss": 0.9661,
+      "step": 27871
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6872344049300738,
+      "learning_rate": 5.571810153327734e-06,
+      "loss": 0.8893,
+      "step": 27872
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.860368105797289,
+      "learning_rate": 5.5711260151852494e-06,
+      "loss": 0.7878,
+      "step": 27873
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.393749535642487,
+      "learning_rate": 5.570441902829069e-06,
+      "loss": 0.741,
+      "step": 27874
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.424893042007749,
+      "learning_rate": 5.569757816263181e-06,
+      "loss": 0.7267,
+      "step": 27875
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.660614659713147,
+      "learning_rate": 5.569073755491561e-06,
+      "loss": 0.9243,
+      "step": 27876
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6425911101146637,
+      "learning_rate": 5.568389720518201e-06,
+      "loss": 0.7762,
+      "step": 27877
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6484907127061164,
+      "learning_rate": 5.567705711347075e-06,
+      "loss": 0.7796,
+      "step": 27878
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6087760241239657,
+      "learning_rate": 5.5670217279821745e-06,
+      "loss": 0.7685,
+      "step": 27879
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4297775761252187,
+      "learning_rate": 5.566337770427472e-06,
+      "loss": 0.7985,
+      "step": 27880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.742549936053301,
+      "learning_rate": 5.565653838686959e-06,
+      "loss": 0.9583,
+      "step": 27881
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.620281584380173,
+      "learning_rate": 5.564969932764615e-06,
+      "loss": 0.8921,
+      "step": 27882
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7580928876718067,
+      "learning_rate": 5.5642860526644135e-06,
+      "loss": 0.8536,
+      "step": 27883
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9638868011388388,
+      "learning_rate": 5.563602198390343e-06,
+      "loss": 0.9375,
+      "step": 27884
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6437741019883472,
+      "learning_rate": 5.562918369946386e-06,
+      "loss": 0.9248,
+      "step": 27885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4574926226494604,
+      "learning_rate": 5.562234567336531e-06,
+      "loss": 0.8107,
+      "step": 27886
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9362529684578778,
+      "learning_rate": 5.561550790564744e-06,
+      "loss": 0.88,
+      "step": 27887
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7181803553317339,
+      "learning_rate": 5.560867039635012e-06,
+      "loss": 0.8705,
+      "step": 27888
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6357741140338644,
+      "learning_rate": 5.560183314551323e-06,
+      "loss": 0.8909,
+      "step": 27889
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6008000648097036,
+      "learning_rate": 5.559499615317652e-06,
+      "loss": 0.8602,
+      "step": 27890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9110773422328562,
+      "learning_rate": 5.558815941937975e-06,
+      "loss": 0.7795,
+      "step": 27891
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5826647890292325,
+      "learning_rate": 5.558132294416278e-06,
+      "loss": 0.8178,
+      "step": 27892
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.585176420452135,
+      "learning_rate": 5.557448672756546e-06,
+      "loss": 0.8495,
+      "step": 27893
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4937956377828354,
+      "learning_rate": 5.556765076962749e-06,
+      "loss": 0.8435,
+      "step": 27894
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5592335827593136,
+      "learning_rate": 5.556081507038877e-06,
+      "loss": 0.7966,
+      "step": 27895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7050351281728515,
+      "learning_rate": 5.555397962988905e-06,
+      "loss": 0.8612,
+      "step": 27896
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8579752481752814,
+      "learning_rate": 5.5547144448168115e-06,
+      "loss": 0.784,
+      "step": 27897
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5273207861852087,
+      "learning_rate": 5.554030952526574e-06,
+      "loss": 0.9417,
+      "step": 27898
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6140313697142392,
+      "learning_rate": 5.553347486122182e-06,
+      "loss": 0.8406,
+      "step": 27899
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.437023166441743,
+      "learning_rate": 5.55266404560761e-06,
+      "loss": 0.9309,
+      "step": 27900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8348702078527026,
+      "learning_rate": 5.5519806309868304e-06,
+      "loss": 0.7276,
+      "step": 27901
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6216573898355946,
+      "learning_rate": 5.551297242263827e-06,
+      "loss": 0.7461,
+      "step": 27902
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5383246133692423,
+      "learning_rate": 5.550613879442589e-06,
+      "loss": 0.7533,
+      "step": 27903
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6558374398068978,
+      "learning_rate": 5.5499305425270765e-06,
+      "loss": 0.8068,
+      "step": 27904
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7677741038891543,
+      "learning_rate": 5.549247231521278e-06,
+      "loss": 0.8132,
+      "step": 27905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.47255426622191,
+      "learning_rate": 5.54856394642917e-06,
+      "loss": 0.8574,
+      "step": 27906
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.043297244861119,
+      "learning_rate": 5.547880687254739e-06,
+      "loss": 0.6858,
+      "step": 27907
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6061380721384646,
+      "learning_rate": 5.5471974540019536e-06,
+      "loss": 0.8995,
+      "step": 27908
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7555376173607012,
+      "learning_rate": 5.54651424667479e-06,
+      "loss": 0.6927,
+      "step": 27909
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5116708119925863,
+      "learning_rate": 5.545831065277236e-06,
+      "loss": 0.7618,
+      "step": 27910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5442018364609564,
+      "learning_rate": 5.545147909813259e-06,
+      "loss": 0.8968,
+      "step": 27911
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6282865029774691,
+      "learning_rate": 5.544464780286845e-06,
+      "loss": 0.9141,
+      "step": 27912
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.59869153486565,
+      "learning_rate": 5.543781676701964e-06,
+      "loss": 0.886,
+      "step": 27913
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.657604382076302,
+      "learning_rate": 5.543098599062602e-06,
+      "loss": 0.6696,
+      "step": 27914
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.546655565303277,
+      "learning_rate": 5.542415547372724e-06,
+      "loss": 0.7805,
+      "step": 27915
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5052502726258195,
+      "learning_rate": 5.54173252163632e-06,
+      "loss": 0.8285,
+      "step": 27916
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.831308535829335,
+      "learning_rate": 5.541049521857361e-06,
+      "loss": 0.7568,
+      "step": 27917
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6538172544388314,
+      "learning_rate": 5.5403665480398196e-06,
+      "loss": 0.8213,
+      "step": 27918
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4410174960409348,
+      "learning_rate": 5.539683600187675e-06,
+      "loss": 0.7923,
+      "step": 27919
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6777441479475441,
+      "learning_rate": 5.539000678304909e-06,
+      "loss": 0.8182,
+      "step": 27920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8310737801192196,
+      "learning_rate": 5.538317782395493e-06,
+      "loss": 0.8291,
+      "step": 27921
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.796886958877356,
+      "learning_rate": 5.537634912463399e-06,
+      "loss": 0.7003,
+      "step": 27922
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.277828589582131,
+      "learning_rate": 5.536952068512609e-06,
+      "loss": 0.8821,
+      "step": 27923
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6100883458715998,
+      "learning_rate": 5.536269250547099e-06,
+      "loss": 0.6382,
+      "step": 27924
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7722845097953075,
+      "learning_rate": 5.535586458570844e-06,
+      "loss": 0.6989,
+      "step": 27925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.394986804932787,
+      "learning_rate": 5.534903692587814e-06,
+      "loss": 0.801,
+      "step": 27926
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6398289393438645,
+      "learning_rate": 5.534220952601987e-06,
+      "loss": 0.823,
+      "step": 27927
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7618251995141051,
+      "learning_rate": 5.533538238617345e-06,
+      "loss": 0.6775,
+      "step": 27928
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9819487492486327,
+      "learning_rate": 5.532855550637856e-06,
+      "loss": 0.732,
+      "step": 27929
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5275436600803225,
+      "learning_rate": 5.532172888667494e-06,
+      "loss": 0.8785,
+      "step": 27930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.773963396484454,
+      "learning_rate": 5.531490252710239e-06,
+      "loss": 0.8901,
+      "step": 27931
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4765982481034143,
+      "learning_rate": 5.53080764277006e-06,
+      "loss": 0.7775,
+      "step": 27932
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8356218203262322,
+      "learning_rate": 5.530125058850935e-06,
+      "loss": 0.8659,
+      "step": 27933
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7835549679013973,
+      "learning_rate": 5.529442500956834e-06,
+      "loss": 0.7539,
+      "step": 27934
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5077963160167593,
+      "learning_rate": 5.528759969091739e-06,
+      "loss": 0.8876,
+      "step": 27935
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.779923391838876,
+      "learning_rate": 5.5280774632596145e-06,
+      "loss": 0.6888,
+      "step": 27936
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7581785961562404,
+      "learning_rate": 5.5273949834644425e-06,
+      "loss": 0.9232,
+      "step": 27937
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5073473689621915,
+      "learning_rate": 5.526712529710194e-06,
+      "loss": 0.7483,
+      "step": 27938
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6137805380171741,
+      "learning_rate": 5.526030102000836e-06,
+      "loss": 0.8163,
+      "step": 27939
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.238321603970164,
+      "learning_rate": 5.525347700340349e-06,
+      "loss": 0.8281,
+      "step": 27940
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6298017206133983,
+      "learning_rate": 5.524665324732706e-06,
+      "loss": 0.7925,
+      "step": 27941
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1628359268422943,
+      "learning_rate": 5.52398297518188e-06,
+      "loss": 0.859,
+      "step": 27942
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5371181707118429,
+      "learning_rate": 5.523300651691837e-06,
+      "loss": 0.6877,
+      "step": 27943
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5381522786123212,
+      "learning_rate": 5.522618354266556e-06,
+      "loss": 0.8336,
+      "step": 27944
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6682880212116258,
+      "learning_rate": 5.521936082910012e-06,
+      "loss": 0.7315,
+      "step": 27945
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3657096221114995,
+      "learning_rate": 5.521253837626175e-06,
+      "loss": 0.8546,
+      "step": 27946
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6133982800525257,
+      "learning_rate": 5.5205716184190106e-06,
+      "loss": 0.8974,
+      "step": 27947
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5852827832355956,
+      "learning_rate": 5.519889425292496e-06,
+      "loss": 0.9215,
+      "step": 27948
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7820733985790093,
+      "learning_rate": 5.5192072582506086e-06,
+      "loss": 0.7208,
+      "step": 27949
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.114445619578178,
+      "learning_rate": 5.5185251172973155e-06,
+      "loss": 0.9633,
+      "step": 27950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8144799374926017,
+      "learning_rate": 5.517843002436583e-06,
+      "loss": 0.8336,
+      "step": 27951
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6890759675523868,
+      "learning_rate": 5.517160913672393e-06,
+      "loss": 0.7683,
+      "step": 27952
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2679616707331776,
+      "learning_rate": 5.516478851008705e-06,
+      "loss": 0.8364,
+      "step": 27953
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0016727475925897,
+      "learning_rate": 5.515796814449502e-06,
+      "loss": 0.7565,
+      "step": 27954
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4838422958161512,
+      "learning_rate": 5.515114803998745e-06,
+      "loss": 0.8654,
+      "step": 27955
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4971956739960988,
+      "learning_rate": 5.5144328196604145e-06,
+      "loss": 0.8646,
+      "step": 27956
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.681094596123314,
+      "learning_rate": 5.513750861438472e-06,
+      "loss": 0.7418,
+      "step": 27957
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.54813979365368,
+      "learning_rate": 5.5130689293368955e-06,
+      "loss": 0.9416,
+      "step": 27958
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.765008049536819,
+      "learning_rate": 5.5123870233596536e-06,
+      "loss": 0.8651,
+      "step": 27959
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4787198326504674,
+      "learning_rate": 5.511705143510711e-06,
+      "loss": 0.8949,
+      "step": 27960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.734797715103078,
+      "learning_rate": 5.51102328979404e-06,
+      "loss": 0.7969,
+      "step": 27961
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5924968265322417,
+      "learning_rate": 5.510341462213619e-06,
+      "loss": 0.7533,
+      "step": 27962
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9309707014987048,
+      "learning_rate": 5.50965966077341e-06,
+      "loss": 0.9271,
+      "step": 27963
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.706171604640526,
+      "learning_rate": 5.50897788547738e-06,
+      "loss": 0.7508,
+      "step": 27964
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4289938497024912,
+      "learning_rate": 5.508296136329504e-06,
+      "loss": 0.7638,
+      "step": 27965
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.48961443637344,
+      "learning_rate": 5.507614413333753e-06,
+      "loss": 0.8323,
+      "step": 27966
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7600489869459044,
+      "learning_rate": 5.506932716494093e-06,
+      "loss": 0.8681,
+      "step": 27967
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7373131655950695,
+      "learning_rate": 5.50625104581449e-06,
+      "loss": 0.6988,
+      "step": 27968
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0743852430561716,
+      "learning_rate": 5.505569401298916e-06,
+      "loss": 0.8293,
+      "step": 27969
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6850598824603056,
+      "learning_rate": 5.504887782951343e-06,
+      "loss": 0.7829,
+      "step": 27970
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7411980173179316,
+      "learning_rate": 5.504206190775736e-06,
+      "loss": 0.7952,
+      "step": 27971
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7115763093490972,
+      "learning_rate": 5.503524624776061e-06,
+      "loss": 0.8368,
+      "step": 27972
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0219231286312875,
+      "learning_rate": 5.502843084956293e-06,
+      "loss": 0.9345,
+      "step": 27973
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4498472527368795,
+      "learning_rate": 5.502161571320393e-06,
+      "loss": 0.8726,
+      "step": 27974
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5758777992779254,
+      "learning_rate": 5.501480083872335e-06,
+      "loss": 0.7896,
+      "step": 27975
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.709155823196796,
+      "learning_rate": 5.500798622616081e-06,
+      "loss": 0.9328,
+      "step": 27976
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.437062214237366,
+      "learning_rate": 5.500117187555604e-06,
+      "loss": 0.8283,
+      "step": 27977
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6644779601258068,
+      "learning_rate": 5.499435778694868e-06,
+      "loss": 0.8647,
+      "step": 27978
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.927585962357233,
+      "learning_rate": 5.4987543960378445e-06,
+      "loss": 0.7131,
+      "step": 27979
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.634962003370403,
+      "learning_rate": 5.498073039588498e-06,
+      "loss": 0.922,
+      "step": 27980
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.473697726986555,
+      "learning_rate": 5.497391709350792e-06,
+      "loss": 0.803,
+      "step": 27981
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.904275826863584,
+      "learning_rate": 5.496710405328698e-06,
+      "loss": 0.7559,
+      "step": 27982
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.617858162693163,
+      "learning_rate": 5.496029127526179e-06,
+      "loss": 0.887,
+      "step": 27983
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6662435317733342,
+      "learning_rate": 5.495347875947215e-06,
+      "loss": 0.8719,
+      "step": 27984
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8027877074043277,
+      "learning_rate": 5.494666650595752e-06,
+      "loss": 0.7884,
+      "step": 27985
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7516803832493832,
+      "learning_rate": 5.4939854514757675e-06,
+      "loss": 0.8272,
+      "step": 27986
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7123495526303139,
+      "learning_rate": 5.493304278591229e-06,
+      "loss": 0.9112,
+      "step": 27987
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9027756013485655,
+      "learning_rate": 5.4926231319461e-06,
+      "loss": 0.8285,
+      "step": 27988
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.636725485257656,
+      "learning_rate": 5.491942011544342e-06,
+      "loss": 0.8944,
+      "step": 27989
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5875170160661656,
+      "learning_rate": 5.4912609173899255e-06,
+      "loss": 0.829,
+      "step": 27990
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6177660932232283,
+      "learning_rate": 5.490579849486819e-06,
+      "loss": 0.8242,
+      "step": 27991
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5383521827663407,
+      "learning_rate": 5.48989880783898e-06,
+      "loss": 0.8191,
+      "step": 27992
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7666914980328898,
+      "learning_rate": 5.489217792450383e-06,
+      "loss": 0.7222,
+      "step": 27993
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6364409411987835,
+      "learning_rate": 5.488536803324989e-06,
+      "loss": 0.9193,
+      "step": 27994
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6879120914321704,
+      "learning_rate": 5.487855840466756e-06,
+      "loss": 0.8437,
+      "step": 27995
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7668455321106427,
+      "learning_rate": 5.4871749038796554e-06,
+      "loss": 0.9054,
+      "step": 27996
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4688075671500807,
+      "learning_rate": 5.486493993567654e-06,
+      "loss": 0.8071,
+      "step": 27997
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7581039615502732,
+      "learning_rate": 5.485813109534716e-06,
+      "loss": 0.7431,
+      "step": 27998
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.556781460310099,
+      "learning_rate": 5.485132251784797e-06,
+      "loss": 0.8506,
+      "step": 27999
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5678640312797867,
+      "learning_rate": 5.484451420321869e-06,
+      "loss": 0.8863,
+      "step": 28000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.54942322611338,
+      "learning_rate": 5.483770615149898e-06,
+      "loss": 0.8426,
+      "step": 28001
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8182076431574865,
+      "learning_rate": 5.483089836272844e-06,
+      "loss": 0.8412,
+      "step": 28002
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6621337346829896,
+      "learning_rate": 5.482409083694667e-06,
+      "loss": 0.8772,
+      "step": 28003
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8087785864170889,
+      "learning_rate": 5.481728357419336e-06,
+      "loss": 0.728,
+      "step": 28004
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5029078208405882,
+      "learning_rate": 5.481047657450816e-06,
+      "loss": 0.8279,
+      "step": 28005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9060106810769823,
+      "learning_rate": 5.480366983793066e-06,
+      "loss": 1.0058,
+      "step": 28006
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.615649335685592,
+      "learning_rate": 5.479686336450047e-06,
+      "loss": 0.8853,
+      "step": 28007
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.870884063482596,
+      "learning_rate": 5.479005715425725e-06,
+      "loss": 0.8761,
+      "step": 28008
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7955656371897366,
+      "learning_rate": 5.4783251207240685e-06,
+      "loss": 0.8207,
+      "step": 28009
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6825283842334344,
+      "learning_rate": 5.477644552349033e-06,
+      "loss": 0.9638,
+      "step": 28010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.563908102727641,
+      "learning_rate": 5.476964010304579e-06,
+      "loss": 0.8216,
+      "step": 28011
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6030475439690863,
+      "learning_rate": 5.476283494594677e-06,
+      "loss": 0.8683,
+      "step": 28012
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7453496662526142,
+      "learning_rate": 5.475603005223281e-06,
+      "loss": 0.8572,
+      "step": 28013
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5297961511081373,
+      "learning_rate": 5.474922542194361e-06,
+      "loss": 0.8323,
+      "step": 28014
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.538171041609204,
+      "learning_rate": 5.4742421055118686e-06,
+      "loss": 0.8296,
+      "step": 28015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.710737763961326,
+      "learning_rate": 5.473561695179777e-06,
+      "loss": 0.9412,
+      "step": 28016
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8841748840397685,
+      "learning_rate": 5.472881311202039e-06,
+      "loss": 0.8059,
+      "step": 28017
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8998573217802626,
+      "learning_rate": 5.472200953582622e-06,
+      "loss": 0.8732,
+      "step": 28018
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.983996945683618,
+      "learning_rate": 5.471520622325484e-06,
+      "loss": 0.9622,
+      "step": 28019
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.673818612328999,
+      "learning_rate": 5.470840317434584e-06,
+      "loss": 0.9047,
+      "step": 28020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5459612963045688,
+      "learning_rate": 5.470160038913884e-06,
+      "loss": 0.8321,
+      "step": 28021
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.56310544048582,
+      "learning_rate": 5.469479786767351e-06,
+      "loss": 0.9665,
+      "step": 28022
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6379882991360128,
+      "learning_rate": 5.468799560998942e-06,
+      "loss": 0.8275,
+      "step": 28023
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5552705885398348,
+      "learning_rate": 5.468119361612612e-06,
+      "loss": 0.8269,
+      "step": 28024
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7803986087715835,
+      "learning_rate": 5.467439188612326e-06,
+      "loss": 0.9374,
+      "step": 28025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.93141761664129,
+      "learning_rate": 5.4667590420020475e-06,
+      "loss": 0.8209,
+      "step": 28026
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.583365454847993,
+      "learning_rate": 5.466078921785733e-06,
+      "loss": 0.7993,
+      "step": 28027
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5019567421992364,
+      "learning_rate": 5.465398827967339e-06,
+      "loss": 0.7712,
+      "step": 28028
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9596506416123696,
+      "learning_rate": 5.464718760550829e-06,
+      "loss": 0.8684,
+      "step": 28029
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6233860586297204,
+      "learning_rate": 5.464038719540166e-06,
+      "loss": 0.9816,
+      "step": 28030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7563069018902684,
+      "learning_rate": 5.463358704939305e-06,
+      "loss": 0.7215,
+      "step": 28031
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8288264345334002,
+      "learning_rate": 5.462678716752201e-06,
+      "loss": 0.8666,
+      "step": 28032
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8145886364088406,
+      "learning_rate": 5.461998754982825e-06,
+      "loss": 0.8383,
+      "step": 28033
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6800984718660388,
+      "learning_rate": 5.461318819635122e-06,
+      "loss": 0.8249,
+      "step": 28034
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6363442768641032,
+      "learning_rate": 5.4606389107130634e-06,
+      "loss": 0.9377,
+      "step": 28035
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5595935685416795,
+      "learning_rate": 5.459959028220597e-06,
+      "loss": 0.6578,
+      "step": 28036
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7862490000961496,
+      "learning_rate": 5.459279172161691e-06,
+      "loss": 0.7379,
+      "step": 28037
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0775030866802506,
+      "learning_rate": 5.458599342540296e-06,
+      "loss": 0.8959,
+      "step": 28038
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4839959491162658,
+      "learning_rate": 5.457919539360377e-06,
+      "loss": 0.8453,
+      "step": 28039
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5879073450412329,
+      "learning_rate": 5.457239762625888e-06,
+      "loss": 0.7618,
+      "step": 28040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6070993289673043,
+      "learning_rate": 5.456560012340784e-06,
+      "loss": 0.7654,
+      "step": 28041
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4505208555557711,
+      "learning_rate": 5.4558802885090255e-06,
+      "loss": 0.8746,
+      "step": 28042
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.55712755706118,
+      "learning_rate": 5.455200591134574e-06,
+      "loss": 0.8909,
+      "step": 28043
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5819877828260343,
+      "learning_rate": 5.4545209202213845e-06,
+      "loss": 0.7752,
+      "step": 28044
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8024776865969225,
+      "learning_rate": 5.453841275773408e-06,
+      "loss": 0.8692,
+      "step": 28045
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4957281832964864,
+      "learning_rate": 5.453161657794609e-06,
+      "loss": 0.8524,
+      "step": 28046
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5954880821577926,
+      "learning_rate": 5.452482066288945e-06,
+      "loss": 1.0251,
+      "step": 28047
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5860470216983247,
+      "learning_rate": 5.45180250126037e-06,
+      "loss": 0.914,
+      "step": 28048
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6506212981822426,
+      "learning_rate": 5.451122962712838e-06,
+      "loss": 0.8743,
+      "step": 28049
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.818925237882966,
+      "learning_rate": 5.450443450650306e-06,
+      "loss": 0.7046,
+      "step": 28050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7616179753719272,
+      "learning_rate": 5.449763965076739e-06,
+      "loss": 0.6852,
+      "step": 28051
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5842976981597703,
+      "learning_rate": 5.449084505996085e-06,
+      "loss": 0.7801,
+      "step": 28052
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6411237970468973,
+      "learning_rate": 5.4484050734122995e-06,
+      "loss": 0.7325,
+      "step": 28053
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4759578985910373,
+      "learning_rate": 5.447725667329343e-06,
+      "loss": 0.8742,
+      "step": 28054
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.745423692111868,
+      "learning_rate": 5.447046287751165e-06,
+      "loss": 0.8562,
+      "step": 28055
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7607211962917486,
+      "learning_rate": 5.446366934681731e-06,
+      "loss": 0.7554,
+      "step": 28056
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.556288087436046,
+      "learning_rate": 5.445687608124984e-06,
+      "loss": 0.9248,
+      "step": 28057
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4854191957535605,
+      "learning_rate": 5.445008308084892e-06,
+      "loss": 0.8853,
+      "step": 28058
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6245896032236176,
+      "learning_rate": 5.444329034565397e-06,
+      "loss": 0.7832,
+      "step": 28059
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5266961216655424,
+      "learning_rate": 5.443649787570467e-06,
+      "loss": 0.8777,
+      "step": 28060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6985943241612314,
+      "learning_rate": 5.4429705671040486e-06,
+      "loss": 0.9105,
+      "step": 28061
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.721798612763468,
+      "learning_rate": 5.442291373170095e-06,
+      "loss": 0.888,
+      "step": 28062
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.525307967812844,
+      "learning_rate": 5.441612205772564e-06,
+      "loss": 0.853,
+      "step": 28063
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8339245039500434,
+      "learning_rate": 5.440933064915414e-06,
+      "loss": 0.8424,
+      "step": 28064
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7851712072142847,
+      "learning_rate": 5.440253950602594e-06,
+      "loss": 0.8521,
+      "step": 28065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4857348814538687,
+      "learning_rate": 5.439574862838057e-06,
+      "loss": 0.8572,
+      "step": 28066
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.090945501863575,
+      "learning_rate": 5.438895801625758e-06,
+      "loss": 0.8122,
+      "step": 28067
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3851490720836503,
+      "learning_rate": 5.438216766969655e-06,
+      "loss": 0.8384,
+      "step": 28068
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7290895492213926,
+      "learning_rate": 5.437537758873699e-06,
+      "loss": 0.7195,
+      "step": 28069
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.818199414249931,
+      "learning_rate": 5.436858777341839e-06,
+      "loss": 0.7386,
+      "step": 28070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8011492602980186,
+      "learning_rate": 5.436179822378031e-06,
+      "loss": 0.84,
+      "step": 28071
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6862598340123718,
+      "learning_rate": 5.435500893986234e-06,
+      "loss": 0.9041,
+      "step": 28072
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9200825610489423,
+      "learning_rate": 5.434821992170396e-06,
+      "loss": 0.8025,
+      "step": 28073
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4549730878557736,
+      "learning_rate": 5.434143116934465e-06,
+      "loss": 0.7837,
+      "step": 28074
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7121404391998671,
+      "learning_rate": 5.433464268282401e-06,
+      "loss": 0.7413,
+      "step": 28075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.567698048519791,
+      "learning_rate": 5.4327854462181515e-06,
+      "loss": 0.7237,
+      "step": 28076
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5082832429425688,
+      "learning_rate": 5.432106650745675e-06,
+      "loss": 0.8542,
+      "step": 28077
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5231607148825015,
+      "learning_rate": 5.431427881868914e-06,
+      "loss": 0.8744,
+      "step": 28078
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8907542661098575,
+      "learning_rate": 5.430749139591833e-06,
+      "loss": 0.9239,
+      "step": 28079
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6501557682110315,
+      "learning_rate": 5.430070423918371e-06,
+      "loss": 0.9156,
+      "step": 28080
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7788519792013298,
+      "learning_rate": 5.429391734852487e-06,
+      "loss": 0.8784,
+      "step": 28081
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0143280343983174,
+      "learning_rate": 5.428713072398139e-06,
+      "loss": 0.8563,
+      "step": 28082
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5829655478302078,
+      "learning_rate": 5.4280344365592615e-06,
+      "loss": 0.8267,
+      "step": 28083
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.023437852349145,
+      "learning_rate": 5.427355827339818e-06,
+      "loss": 0.7959,
+      "step": 28084
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8243245930682244,
+      "learning_rate": 5.4266772447437535e-06,
+      "loss": 0.7715,
+      "step": 28085
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5963384995056649,
+      "learning_rate": 5.4259986887750315e-06,
+      "loss": 0.7021,
+      "step": 28086
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6496094353075352,
+      "learning_rate": 5.4253201594375835e-06,
+      "loss": 0.7919,
+      "step": 28087
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7324834784298009,
+      "learning_rate": 5.424641656735372e-06,
+      "loss": 0.7804,
+      "step": 28088
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6587880189860738,
+      "learning_rate": 5.423963180672349e-06,
+      "loss": 0.9303,
+      "step": 28089
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.631068201654158,
+      "learning_rate": 5.4232847312524565e-06,
+      "loss": 0.9878,
+      "step": 28090
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0513565355457555,
+      "learning_rate": 5.4226063084796536e-06,
+      "loss": 0.881,
+      "step": 28091
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.694386358088044,
+      "learning_rate": 5.421927912357881e-06,
+      "loss": 0.7372,
+      "step": 28092
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6414995413353168,
+      "learning_rate": 5.421249542891098e-06,
+      "loss": 0.96,
+      "step": 28093
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5775036784511405,
+      "learning_rate": 5.420571200083246e-06,
+      "loss": 0.8234,
+      "step": 28094
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6211443222980453,
+      "learning_rate": 5.419892883938282e-06,
+      "loss": 0.7438,
+      "step": 28095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6456896834228603,
+      "learning_rate": 5.419214594460152e-06,
+      "loss": 0.896,
+      "step": 28096
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6661417511833272,
+      "learning_rate": 5.418536331652802e-06,
+      "loss": 0.7943,
+      "step": 28097
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7458394696097388,
+      "learning_rate": 5.417858095520182e-06,
+      "loss": 0.704,
+      "step": 28098
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5091895289561177,
+      "learning_rate": 5.417179886066249e-06,
+      "loss": 0.8675,
+      "step": 28099
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7753153803444888,
+      "learning_rate": 5.416501703294943e-06,
+      "loss": 0.7669,
+      "step": 28100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8115605112084326,
+      "learning_rate": 5.415823547210213e-06,
+      "loss": 0.7462,
+      "step": 28101
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8707370360788358,
+      "learning_rate": 5.415145417816008e-06,
+      "loss": 0.8331,
+      "step": 28102
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.651220656656043,
+      "learning_rate": 5.414467315116283e-06,
+      "loss": 0.9055,
+      "step": 28103
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.403509357176057,
+      "learning_rate": 5.4137892391149815e-06,
+      "loss": 0.6872,
+      "step": 28104
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6908300282866882,
+      "learning_rate": 5.413111189816047e-06,
+      "loss": 0.9561,
+      "step": 28105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8634429410295181,
+      "learning_rate": 5.41243316722343e-06,
+      "loss": 0.8746,
+      "step": 28106
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.508289964753048,
+      "learning_rate": 5.411755171341085e-06,
+      "loss": 0.8204,
+      "step": 28107
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.72329861334627,
+      "learning_rate": 5.411077202172953e-06,
+      "loss": 0.8815,
+      "step": 28108
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8140762992146866,
+      "learning_rate": 5.410399259722978e-06,
+      "loss": 0.9924,
+      "step": 28109
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7833989114853241,
+      "learning_rate": 5.409721343995116e-06,
+      "loss": 0.7569,
+      "step": 28110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9126967773622519,
+      "learning_rate": 5.409043454993307e-06,
+      "loss": 0.7574,
+      "step": 28111
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9301283235629987,
+      "learning_rate": 5.408365592721503e-06,
+      "loss": 0.896,
+      "step": 28112
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.732850906809442,
+      "learning_rate": 5.407687757183645e-06,
+      "loss": 0.9153,
+      "step": 28113
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8138844059305272,
+      "learning_rate": 5.407009948383686e-06,
+      "loss": 0.7972,
+      "step": 28114
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6550976611377202,
+      "learning_rate": 5.406332166325565e-06,
+      "loss": 0.8031,
+      "step": 28115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6100703921155899,
+      "learning_rate": 5.405654411013237e-06,
+      "loss": 0.8701,
+      "step": 28116
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5449800858641156,
+      "learning_rate": 5.404976682450644e-06,
+      "loss": 0.9158,
+      "step": 28117
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.659017622694868,
+      "learning_rate": 5.404298980641727e-06,
+      "loss": 0.7945,
+      "step": 28118
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5538186076869016,
+      "learning_rate": 5.403621305590436e-06,
+      "loss": 0.9092,
+      "step": 28119
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6722322415260926,
+      "learning_rate": 5.402943657300722e-06,
+      "loss": 0.7774,
+      "step": 28120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4664651343040502,
+      "learning_rate": 5.402266035776525e-06,
+      "loss": 0.8851,
+      "step": 28121
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5334836308407045,
+      "learning_rate": 5.401588441021786e-06,
+      "loss": 0.8233,
+      "step": 28122
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7746291744989975,
+      "learning_rate": 5.400910873040456e-06,
+      "loss": 0.8177,
+      "step": 28123
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5837928947682447,
+      "learning_rate": 5.400233331836482e-06,
+      "loss": 0.8517,
+      "step": 28124
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7024898135945676,
+      "learning_rate": 5.399555817413806e-06,
+      "loss": 0.8199,
+      "step": 28125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.624482766185234,
+      "learning_rate": 5.398878329776369e-06,
+      "loss": 0.7839,
+      "step": 28126
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6549558032877212,
+      "learning_rate": 5.398200868928118e-06,
+      "loss": 0.8221,
+      "step": 28127
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5559328561702055,
+      "learning_rate": 5.397523434873004e-06,
+      "loss": 0.8125,
+      "step": 28128
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4676173709774933,
+      "learning_rate": 5.3968460276149645e-06,
+      "loss": 0.7962,
+      "step": 28129
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8744194188326648,
+      "learning_rate": 5.396168647157942e-06,
+      "loss": 0.8711,
+      "step": 28130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5965891121734948,
+      "learning_rate": 5.395491293505881e-06,
+      "loss": 0.8081,
+      "step": 28131
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8068471108177602,
+      "learning_rate": 5.3948139666627335e-06,
+      "loss": 0.8633,
+      "step": 28132
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4438119102549942,
+      "learning_rate": 5.394136666632437e-06,
+      "loss": 0.7571,
+      "step": 28133
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.894337492951365,
+      "learning_rate": 5.3934593934189295e-06,
+      "loss": 0.883,
+      "step": 28134
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5064373033153435,
+      "learning_rate": 5.392782147026166e-06,
+      "loss": 0.6963,
+      "step": 28135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6355238349562449,
+      "learning_rate": 5.392104927458077e-06,
+      "loss": 0.8717,
+      "step": 28136
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5761484366840397,
+      "learning_rate": 5.3914277347186174e-06,
+      "loss": 0.9336,
+      "step": 28137
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.588480340442072,
+      "learning_rate": 5.39075056881172e-06,
+      "loss": 0.7472,
+      "step": 28138
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6764834147843968,
+      "learning_rate": 5.390073429741336e-06,
+      "loss": 0.7392,
+      "step": 28139
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7470232399823803,
+      "learning_rate": 5.389396317511401e-06,
+      "loss": 0.6886,
+      "step": 28140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5762025879143362,
+      "learning_rate": 5.388719232125863e-06,
+      "loss": 0.8857,
+      "step": 28141
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.181084282083003,
+      "learning_rate": 5.388042173588662e-06,
+      "loss": 0.7962,
+      "step": 28142
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.560915745401537,
+      "learning_rate": 5.387365141903735e-06,
+      "loss": 0.8086,
+      "step": 28143
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4787773229609418,
+      "learning_rate": 5.386688137075031e-06,
+      "loss": 0.8109,
+      "step": 28144
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8896033144431557,
+      "learning_rate": 5.38601115910649e-06,
+      "loss": 0.8681,
+      "step": 28145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.927117397392163,
+      "learning_rate": 5.385334208002055e-06,
+      "loss": 0.9119,
+      "step": 28146
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4751650835238321,
+      "learning_rate": 5.384657283765661e-06,
+      "loss": 0.8273,
+      "step": 28147
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9082636986772645,
+      "learning_rate": 5.383980386401253e-06,
+      "loss": 0.9212,
+      "step": 28148
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8752699996281257,
+      "learning_rate": 5.383303515912777e-06,
+      "loss": 0.7995,
+      "step": 28149
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2437236653397727,
+      "learning_rate": 5.382626672304168e-06,
+      "loss": 0.7929,
+      "step": 28150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7950451033892327,
+      "learning_rate": 5.381949855579365e-06,
+      "loss": 0.8666,
+      "step": 28151
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7583885436301132,
+      "learning_rate": 5.381273065742313e-06,
+      "loss": 0.8338,
+      "step": 28152
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7614591629775389,
+      "learning_rate": 5.380596302796954e-06,
+      "loss": 0.6992,
+      "step": 28153
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.908098624306378,
+      "learning_rate": 5.379919566747227e-06,
+      "loss": 0.8601,
+      "step": 28154
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4785204459213186,
+      "learning_rate": 5.379242857597067e-06,
+      "loss": 0.7136,
+      "step": 28155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5203461325833072,
+      "learning_rate": 5.378566175350421e-06,
+      "loss": 1.034,
+      "step": 28156
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1044242030865727,
+      "learning_rate": 5.377889520011224e-06,
+      "loss": 0.7774,
+      "step": 28157
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7975914049516508,
+      "learning_rate": 5.377212891583419e-06,
+      "loss": 0.9876,
+      "step": 28158
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6276906857224176,
+      "learning_rate": 5.376536290070942e-06,
+      "loss": 0.8401,
+      "step": 28159
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6985156970166984,
+      "learning_rate": 5.375859715477738e-06,
+      "loss": 0.9376,
+      "step": 28160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.636593281450099,
+      "learning_rate": 5.3751831678077385e-06,
+      "loss": 0.7854,
+      "step": 28161
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5130325178033668,
+      "learning_rate": 5.3745066470648895e-06,
+      "loss": 0.7286,
+      "step": 28162
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.509442325403182,
+      "learning_rate": 5.373830153253129e-06,
+      "loss": 0.8721,
+      "step": 28163
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7318218212173762,
+      "learning_rate": 5.373153686376389e-06,
+      "loss": 0.8539,
+      "step": 28164
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4863908816679943,
+      "learning_rate": 5.3724772464386145e-06,
+      "loss": 0.904,
+      "step": 28165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7062918424360984,
+      "learning_rate": 5.371800833443741e-06,
+      "loss": 0.762,
+      "step": 28166
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4265580421964161,
+      "learning_rate": 5.371124447395717e-06,
+      "loss": 0.9027,
+      "step": 28167
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5317476786334923,
+      "learning_rate": 5.370448088298463e-06,
+      "loss": 0.862,
+      "step": 28168
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6016036220794116,
+      "learning_rate": 5.369771756155926e-06,
+      "loss": 0.9572,
+      "step": 28169
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.526936026467171,
+      "learning_rate": 5.3690954509720484e-06,
+      "loss": 0.9189,
+      "step": 28170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6009307486136077,
+      "learning_rate": 5.368419172750763e-06,
+      "loss": 0.8678,
+      "step": 28171
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6285240464246968,
+      "learning_rate": 5.367742921496004e-06,
+      "loss": 1.0371,
+      "step": 28172
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8024747371990572,
+      "learning_rate": 5.367066697211711e-06,
+      "loss": 0.7414,
+      "step": 28173
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5412307561216625,
+      "learning_rate": 5.366390499901828e-06,
+      "loss": 0.899,
+      "step": 28174
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7983494508490417,
+      "learning_rate": 5.365714329570286e-06,
+      "loss": 0.7131,
+      "step": 28175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5818702087809378,
+      "learning_rate": 5.3650381862210185e-06,
+      "loss": 0.7328,
+      "step": 28176
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9608082734063887,
+      "learning_rate": 5.364362069857969e-06,
+      "loss": 0.8593,
+      "step": 28177
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5554752710862292,
+      "learning_rate": 5.363685980485069e-06,
+      "loss": 0.8251,
+      "step": 28178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.623055217261858,
+      "learning_rate": 5.363009918106256e-06,
+      "loss": 0.8989,
+      "step": 28179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5089945817910662,
+      "learning_rate": 5.362333882725471e-06,
+      "loss": 0.8742,
+      "step": 28180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7580078878136364,
+      "learning_rate": 5.361657874346647e-06,
+      "loss": 0.7013,
+      "step": 28181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.604868260960673,
+      "learning_rate": 5.360981892973716e-06,
+      "loss": 0.7423,
+      "step": 28182
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5793830596202005,
+      "learning_rate": 5.3603059386106145e-06,
+      "loss": 0.8256,
+      "step": 28183
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1536532009028813,
+      "learning_rate": 5.359630011261291e-06,
+      "loss": 0.8952,
+      "step": 28184
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.62645722169201,
+      "learning_rate": 5.3589541109296615e-06,
+      "loss": 0.8736,
+      "step": 28185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8089420166930132,
+      "learning_rate": 5.358278237619672e-06,
+      "loss": 0.8281,
+      "step": 28186
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7922941023064425,
+      "learning_rate": 5.357602391335255e-06,
+      "loss": 0.8731,
+      "step": 28187
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8228957139951134,
+      "learning_rate": 5.35692657208035e-06,
+      "loss": 0.8343,
+      "step": 28188
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7872919743905429,
+      "learning_rate": 5.356250779858889e-06,
+      "loss": 0.7792,
+      "step": 28189
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7040470291500622,
+      "learning_rate": 5.3555750146748034e-06,
+      "loss": 0.913,
+      "step": 28190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5032654680392878,
+      "learning_rate": 5.354899276532033e-06,
+      "loss": 0.8066,
+      "step": 28191
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5919904393893602,
+      "learning_rate": 5.354223565434507e-06,
+      "loss": 0.928,
+      "step": 28192
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4924385400441318,
+      "learning_rate": 5.353547881386165e-06,
+      "loss": 0.8906,
+      "step": 28193
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6412688411311864,
+      "learning_rate": 5.3528722243909346e-06,
+      "loss": 0.8361,
+      "step": 28194
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7225785315109552,
+      "learning_rate": 5.352196594452756e-06,
+      "loss": 0.9122,
+      "step": 28195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.717324481103857,
+      "learning_rate": 5.351520991575558e-06,
+      "loss": 0.9082,
+      "step": 28196
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5482378150370923,
+      "learning_rate": 5.3508454157632815e-06,
+      "loss": 1.0615,
+      "step": 28197
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7881019015967208,
+      "learning_rate": 5.350169867019853e-06,
+      "loss": 0.7308,
+      "step": 28198
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4825027569742852,
+      "learning_rate": 5.349494345349204e-06,
+      "loss": 0.9047,
+      "step": 28199
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5973699939432757,
+      "learning_rate": 5.348818850755272e-06,
+      "loss": 0.8758,
+      "step": 28200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6788755107212805,
+      "learning_rate": 5.348143383241992e-06,
+      "loss": 0.8487,
+      "step": 28201
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.635287942408266,
+      "learning_rate": 5.347467942813296e-06,
+      "loss": 0.8409,
+      "step": 28202
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7139257643476107,
+      "learning_rate": 5.346792529473108e-06,
+      "loss": 0.9024,
+      "step": 28203
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5861770788493825,
+      "learning_rate": 5.346117143225369e-06,
+      "loss": 0.8434,
+      "step": 28204
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6721233818699688,
+      "learning_rate": 5.3454417840740125e-06,
+      "loss": 0.8245,
+      "step": 28205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6392437158727229,
+      "learning_rate": 5.344766452022967e-06,
+      "loss": 0.957,
+      "step": 28206
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5732136211319154,
+      "learning_rate": 5.344091147076162e-06,
+      "loss": 0.8556,
+      "step": 28207
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.944918062270655,
+      "learning_rate": 5.343415869237533e-06,
+      "loss": 0.7227,
+      "step": 28208
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7672261292334532,
+      "learning_rate": 5.342740618511013e-06,
+      "loss": 0.8639,
+      "step": 28209
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.699538244912174,
+      "learning_rate": 5.342065394900533e-06,
+      "loss": 0.9627,
+      "step": 28210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9310079047800917,
+      "learning_rate": 5.3413901984100195e-06,
+      "loss": 0.5897,
+      "step": 28211
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.029067286620054,
+      "learning_rate": 5.340715029043411e-06,
+      "loss": 0.8811,
+      "step": 28212
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6667908636421656,
+      "learning_rate": 5.3400398868046285e-06,
+      "loss": 0.7343,
+      "step": 28213
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5558072753927286,
+      "learning_rate": 5.339364771697615e-06,
+      "loss": 0.9552,
+      "step": 28214
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.470444387151761,
+      "learning_rate": 5.338689683726291e-06,
+      "loss": 0.8683,
+      "step": 28215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7230369468617843,
+      "learning_rate": 5.338014622894596e-06,
+      "loss": 0.9964,
+      "step": 28216
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9780684816505674,
+      "learning_rate": 5.337339589206452e-06,
+      "loss": 0.8343,
+      "step": 28217
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.734585945276858,
+      "learning_rate": 5.3366645826657956e-06,
+      "loss": 0.8104,
+      "step": 28218
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5627606500193045,
+      "learning_rate": 5.3359896032765545e-06,
+      "loss": 0.9169,
+      "step": 28219
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5100155047366988,
+      "learning_rate": 5.335314651042656e-06,
+      "loss": 0.8918,
+      "step": 28220
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7068966081635502,
+      "learning_rate": 5.334639725968031e-06,
+      "loss": 0.908,
+      "step": 28221
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.406203504048687,
+      "learning_rate": 5.333964828056615e-06,
+      "loss": 0.7718,
+      "step": 28222
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6635536541632028,
+      "learning_rate": 5.333289957312333e-06,
+      "loss": 0.8661,
+      "step": 28223
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3548819210541174,
+      "learning_rate": 5.332615113739111e-06,
+      "loss": 0.8475,
+      "step": 28224
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0031090385064516,
+      "learning_rate": 5.3319402973408805e-06,
+      "loss": 0.7907,
+      "step": 28225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5522251915971848,
+      "learning_rate": 5.331265508121577e-06,
+      "loss": 0.8009,
+      "step": 28226
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5099973717399062,
+      "learning_rate": 5.330590746085122e-06,
+      "loss": 0.7968,
+      "step": 28227
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3790050371709544,
+      "learning_rate": 5.329916011235443e-06,
+      "loss": 0.7453,
+      "step": 28228
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7121194252791436,
+      "learning_rate": 5.329241303576471e-06,
+      "loss": 0.8392,
+      "step": 28229
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5884675769618897,
+      "learning_rate": 5.328566623112139e-06,
+      "loss": 0.8759,
+      "step": 28230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5005911978151898,
+      "learning_rate": 5.327891969846371e-06,
+      "loss": 0.7513,
+      "step": 28231
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6495773159364955,
+      "learning_rate": 5.327217343783091e-06,
+      "loss": 0.8489,
+      "step": 28232
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6589116014242933,
+      "learning_rate": 5.326542744926235e-06,
+      "loss": 0.8206,
+      "step": 28233
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3493375452044982,
+      "learning_rate": 5.325868173279725e-06,
+      "loss": 0.782,
+      "step": 28234
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.610856129412788,
+      "learning_rate": 5.325193628847492e-06,
+      "loss": 0.8853,
+      "step": 28235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.989080449275436,
+      "learning_rate": 5.324519111633457e-06,
+      "loss": 0.9731,
+      "step": 28236
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6671086587111459,
+      "learning_rate": 5.323844621641556e-06,
+      "loss": 0.9589,
+      "step": 28237
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6945009732495951,
+      "learning_rate": 5.32317015887571e-06,
+      "loss": 0.6912,
+      "step": 28238
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6855767765824137,
+      "learning_rate": 5.322495723339851e-06,
+      "loss": 0.8108,
+      "step": 28239
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.725444979154409,
+      "learning_rate": 5.321821315037903e-06,
+      "loss": 0.8881,
+      "step": 28240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5715021767908524,
+      "learning_rate": 5.3211469339737885e-06,
+      "loss": 0.8227,
+      "step": 28241
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8055463359911528,
+      "learning_rate": 5.320472580151437e-06,
+      "loss": 0.9197,
+      "step": 28242
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6465676738838633,
+      "learning_rate": 5.31979825357478e-06,
+      "loss": 0.9053,
+      "step": 28243
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.801170748360046,
+      "learning_rate": 5.31912395424774e-06,
+      "loss": 0.7816,
+      "step": 28244
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7980793708440224,
+      "learning_rate": 5.318449682174238e-06,
+      "loss": 0.8451,
+      "step": 28245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4411531677796705,
+      "learning_rate": 5.317775437358203e-06,
+      "loss": 0.845,
+      "step": 28246
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6531834869638915,
+      "learning_rate": 5.317101219803566e-06,
+      "loss": 0.7715,
+      "step": 28247
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1951325117036307,
+      "learning_rate": 5.316427029514251e-06,
+      "loss": 0.8682,
+      "step": 28248
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6473340505244762,
+      "learning_rate": 5.315752866494174e-06,
+      "loss": 0.7307,
+      "step": 28249
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7362371759368531,
+      "learning_rate": 5.315078730747268e-06,
+      "loss": 0.9804,
+      "step": 28250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5611280417788516,
+      "learning_rate": 5.3144046222774595e-06,
+      "loss": 0.7695,
+      "step": 28251
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5218837245543846,
+      "learning_rate": 5.313730541088671e-06,
+      "loss": 0.8585,
+      "step": 28252
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5806502872961763,
+      "learning_rate": 5.313056487184824e-06,
+      "loss": 0.8171,
+      "step": 28253
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6564170208733398,
+      "learning_rate": 5.3123824605698444e-06,
+      "loss": 0.9101,
+      "step": 28254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5252053026156285,
+      "learning_rate": 5.311708461247664e-06,
+      "loss": 0.7643,
+      "step": 28255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6563378950664631,
+      "learning_rate": 5.3110344892222e-06,
+      "loss": 0.9582,
+      "step": 28256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9436997171024666,
+      "learning_rate": 5.3103605444973735e-06,
+      "loss": 0.9322,
+      "step": 28257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.580560003663342,
+      "learning_rate": 5.309686627077117e-06,
+      "loss": 0.8447,
+      "step": 28258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.591252818316014,
+      "learning_rate": 5.3090127369653465e-06,
+      "loss": 0.8739,
+      "step": 28259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.45457572160238,
+      "learning_rate": 5.308338874165993e-06,
+      "loss": 0.9576,
+      "step": 28260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7772848697440775,
+      "learning_rate": 5.307665038682972e-06,
+      "loss": 0.67,
+      "step": 28261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.656645860115745,
+      "learning_rate": 5.306991230520213e-06,
+      "loss": 0.8619,
+      "step": 28262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5883163512719256,
+      "learning_rate": 5.306317449681635e-06,
+      "loss": 0.995,
+      "step": 28263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9550702805128313,
+      "learning_rate": 5.305643696171161e-06,
+      "loss": 0.7961,
+      "step": 28264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7719720057070644,
+      "learning_rate": 5.304969969992725e-06,
+      "loss": 0.6857,
+      "step": 28265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7307207019015913,
+      "learning_rate": 5.304296271150231e-06,
+      "loss": 0.7977,
+      "step": 28266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.617848630073132,
+      "learning_rate": 5.303622599647613e-06,
+      "loss": 0.9175,
+      "step": 28267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.389600357668503,
+      "learning_rate": 5.302948955488789e-06,
+      "loss": 0.8382,
+      "step": 28268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5842454163955935,
+      "learning_rate": 5.302275338677692e-06,
+      "loss": 0.8871,
+      "step": 28269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6228774915299637,
+      "learning_rate": 5.301601749218227e-06,
+      "loss": 0.93,
+      "step": 28270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6255489657743305,
+      "learning_rate": 5.300928187114323e-06,
+      "loss": 0.8204,
+      "step": 28271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.672082832934764,
+      "learning_rate": 5.30025465236991e-06,
+      "loss": 0.8456,
+      "step": 28272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8508333203915446,
+      "learning_rate": 5.299581144988899e-06,
+      "loss": 0.7869,
+      "step": 28273
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4156095575710488,
+      "learning_rate": 5.298907664975213e-06,
+      "loss": 0.8429,
+      "step": 28274
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3945165321455988,
+      "learning_rate": 5.298234212332773e-06,
+      "loss": 0.9028,
+      "step": 28275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.441569613358652,
+      "learning_rate": 5.297560787065506e-06,
+      "loss": 0.7315,
+      "step": 28276
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6432694931340008,
+      "learning_rate": 5.296887389177326e-06,
+      "loss": 0.8501,
+      "step": 28277
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4406254723408736,
+      "learning_rate": 5.296214018672161e-06,
+      "loss": 0.8769,
+      "step": 28278
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3651695763383476,
+      "learning_rate": 5.295540675553926e-06,
+      "loss": 0.7598,
+      "step": 28279
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.719499569511786,
+      "learning_rate": 5.294867359826539e-06,
+      "loss": 0.9903,
+      "step": 28280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8308336563536562,
+      "learning_rate": 5.294194071493925e-06,
+      "loss": 0.7805,
+      "step": 28281
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7540799834258107,
+      "learning_rate": 5.293520810560006e-06,
+      "loss": 0.8891,
+      "step": 28282
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8829666615198248,
+      "learning_rate": 5.292847577028698e-06,
+      "loss": 0.9057,
+      "step": 28283
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6276123557321849,
+      "learning_rate": 5.292174370903919e-06,
+      "loss": 0.8899,
+      "step": 28284
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6193135162614558,
+      "learning_rate": 5.29150119218959e-06,
+      "loss": 0.8695,
+      "step": 28285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0028434649680786,
+      "learning_rate": 5.290828040889637e-06,
+      "loss": 0.801,
+      "step": 28286
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.57685297820337,
+      "learning_rate": 5.2901549170079736e-06,
+      "loss": 0.7475,
+      "step": 28287
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5498883623128048,
+      "learning_rate": 5.2894818205485144e-06,
+      "loss": 0.87,
+      "step": 28288
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6301824559681977,
+      "learning_rate": 5.288808751515184e-06,
+      "loss": 0.9211,
+      "step": 28289
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6540162646928858,
+      "learning_rate": 5.288135709911906e-06,
+      "loss": 0.8533,
+      "step": 28290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5147311331131772,
+      "learning_rate": 5.287462695742593e-06,
+      "loss": 0.7716,
+      "step": 28291
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5025062863742784,
+      "learning_rate": 5.2867897090111594e-06,
+      "loss": 0.8179,
+      "step": 28292
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6331917664365114,
+      "learning_rate": 5.286116749721533e-06,
+      "loss": 0.8225,
+      "step": 28293
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5366176551244595,
+      "learning_rate": 5.285443817877622e-06,
+      "loss": 0.7707,
+      "step": 28294
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6355250340944427,
+      "learning_rate": 5.284770913483355e-06,
+      "loss": 0.9438,
+      "step": 28295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5849922253328663,
+      "learning_rate": 5.28409803654264e-06,
+      "loss": 0.8258,
+      "step": 28296
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.528214268444569,
+      "learning_rate": 5.283425187059403e-06,
+      "loss": 0.8417,
+      "step": 28297
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7975691489875874,
+      "learning_rate": 5.282752365037554e-06,
+      "loss": 0.8331,
+      "step": 28298
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9382313018268962,
+      "learning_rate": 5.282079570481018e-06,
+      "loss": 1.0034,
+      "step": 28299
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6576875393659605,
+      "learning_rate": 5.2814068033937096e-06,
+      "loss": 0.9425,
+      "step": 28300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4980930239624968,
+      "learning_rate": 5.28073406377954e-06,
+      "loss": 0.714,
+      "step": 28301
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8125143117701876,
+      "learning_rate": 5.280061351642429e-06,
+      "loss": 0.8315,
+      "step": 28302
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5833369380969922,
+      "learning_rate": 5.279388666986301e-06,
+      "loss": 0.9203,
+      "step": 28303
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7430416792829146,
+      "learning_rate": 5.278716009815067e-06,
+      "loss": 0.7864,
+      "step": 28304
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6819600292151748,
+      "learning_rate": 5.278043380132638e-06,
+      "loss": 0.7754,
+      "step": 28305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7576500459114774,
+      "learning_rate": 5.277370777942936e-06,
+      "loss": 0.71,
+      "step": 28306
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.376942761721233,
+      "learning_rate": 5.276698203249882e-06,
+      "loss": 0.7525,
+      "step": 28307
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.484963042471464,
+      "learning_rate": 5.276025656057384e-06,
+      "loss": 0.7592,
+      "step": 28308
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5154647892251887,
+      "learning_rate": 5.275353136369357e-06,
+      "loss": 0.7766,
+      "step": 28309
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8125189547559883,
+      "learning_rate": 5.27468064418972e-06,
+      "loss": 0.7715,
+      "step": 28310
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4410661227076311,
+      "learning_rate": 5.274008179522392e-06,
+      "loss": 0.83,
+      "step": 28311
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8821935560559881,
+      "learning_rate": 5.273335742371284e-06,
+      "loss": 0.9018,
+      "step": 28312
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9699128214295998,
+      "learning_rate": 5.272663332740308e-06,
+      "loss": 0.8579,
+      "step": 28313
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7680587649192643,
+      "learning_rate": 5.271990950633388e-06,
+      "loss": 0.6225,
+      "step": 28314
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5656995227765624,
+      "learning_rate": 5.271318596054429e-06,
+      "loss": 0.8207,
+      "step": 28315
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8058211972644824,
+      "learning_rate": 5.270646269007355e-06,
+      "loss": 0.8577,
+      "step": 28316
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7309880515245629,
+      "learning_rate": 5.269973969496069e-06,
+      "loss": 0.6763,
+      "step": 28317
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9025870857591403,
+      "learning_rate": 5.269301697524498e-06,
+      "loss": 0.7828,
+      "step": 28318
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.608490504476492,
+      "learning_rate": 5.268629453096547e-06,
+      "loss": 0.8236,
+      "step": 28319
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.573354627089363,
+      "learning_rate": 5.267957236216137e-06,
+      "loss": 0.9448,
+      "step": 28320
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7138995062601123,
+      "learning_rate": 5.267285046887176e-06,
+      "loss": 0.805,
+      "step": 28321
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7616947581411055,
+      "learning_rate": 5.2666128851135775e-06,
+      "loss": 0.8227,
+      "step": 28322
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5590894698844555,
+      "learning_rate": 5.265940750899256e-06,
+      "loss": 0.8738,
+      "step": 28323
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7836383975979204,
+      "learning_rate": 5.265268644248133e-06,
+      "loss": 0.7892,
+      "step": 28324
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1041340712982324,
+      "learning_rate": 5.264596565164113e-06,
+      "loss": 0.7169,
+      "step": 28325
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3519010195253391,
+      "learning_rate": 5.263924513651107e-06,
+      "loss": 0.7288,
+      "step": 28326
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6134618473231903,
+      "learning_rate": 5.263252489713032e-06,
+      "loss": 0.8582,
+      "step": 28327
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.612470844034176,
+      "learning_rate": 5.262580493353805e-06,
+      "loss": 1.0366,
+      "step": 28328
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8672916687429428,
+      "learning_rate": 5.261908524577334e-06,
+      "loss": 0.8394,
+      "step": 28329
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7831255673604068,
+      "learning_rate": 5.261236583387528e-06,
+      "loss": 0.7266,
+      "step": 28330
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6496292493214868,
+      "learning_rate": 5.260564669788302e-06,
+      "loss": 0.8888,
+      "step": 28331
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5451476926209926,
+      "learning_rate": 5.259892783783573e-06,
+      "loss": 0.9049,
+      "step": 28332
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5338613423348355,
+      "learning_rate": 5.259220925377249e-06,
+      "loss": 0.8348,
+      "step": 28333
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7541735795239726,
+      "learning_rate": 5.258549094573239e-06,
+      "loss": 0.8329,
+      "step": 28334
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.883678248242473,
+      "learning_rate": 5.257877291375459e-06,
+      "loss": 0.791,
+      "step": 28335
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.19656621230952,
+      "learning_rate": 5.257205515787817e-06,
+      "loss": 0.7551,
+      "step": 28336
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6404729528765976,
+      "learning_rate": 5.256533767814229e-06,
+      "loss": 0.8114,
+      "step": 28337
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6926232052928953,
+      "learning_rate": 5.2558620474586e-06,
+      "loss": 0.842,
+      "step": 28338
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4737094446219128,
+      "learning_rate": 5.255190354724846e-06,
+      "loss": 0.7971,
+      "step": 28339
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.638829578974806,
+      "learning_rate": 5.254518689616874e-06,
+      "loss": 0.9131,
+      "step": 28340
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6243439962729205,
+      "learning_rate": 5.2538470521385996e-06,
+      "loss": 0.8988,
+      "step": 28341
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8417477718905413,
+      "learning_rate": 5.253175442293929e-06,
+      "loss": 0.9451,
+      "step": 28342
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7003856331043572,
+      "learning_rate": 5.252503860086773e-06,
+      "loss": 0.8039,
+      "step": 28343
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7016860145153014,
+      "learning_rate": 5.2518323055210405e-06,
+      "loss": 0.8818,
+      "step": 28344
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5301115475522515,
+      "learning_rate": 5.251160778600648e-06,
+      "loss": 0.8194,
+      "step": 28345
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.582851853143583,
+      "learning_rate": 5.250489279329501e-06,
+      "loss": 0.8037,
+      "step": 28346
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5869049718302992,
+      "learning_rate": 5.249817807711505e-06,
+      "loss": 0.8929,
+      "step": 28347
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.850481203450852,
+      "learning_rate": 5.249146363750574e-06,
+      "loss": 0.8966,
+      "step": 28348
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6312525951507701,
+      "learning_rate": 5.24847494745062e-06,
+      "loss": 0.8306,
+      "step": 28349
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9561290781723977,
+      "learning_rate": 5.24780355881555e-06,
+      "loss": 0.8422,
+      "step": 28350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.664475552088982,
+      "learning_rate": 5.247132197849267e-06,
+      "loss": 0.7929,
+      "step": 28351
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.627695489349971,
+      "learning_rate": 5.246460864555687e-06,
+      "loss": 0.8917,
+      "step": 28352
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6129660543101032,
+      "learning_rate": 5.245789558938719e-06,
+      "loss": 0.8336,
+      "step": 28353
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6431087365337538,
+      "learning_rate": 5.245118281002271e-06,
+      "loss": 0.8024,
+      "step": 28354
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6213387917399935,
+      "learning_rate": 5.244447030750244e-06,
+      "loss": 0.8195,
+      "step": 28355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.559493732964648,
+      "learning_rate": 5.243775808186557e-06,
+      "loss": 0.8549,
+      "step": 28356
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5928800314682678,
+      "learning_rate": 5.2431046133151085e-06,
+      "loss": 0.6523,
+      "step": 28357
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9657591868828221,
+      "learning_rate": 5.242433446139816e-06,
+      "loss": 0.8914,
+      "step": 28358
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.909326734453257,
+      "learning_rate": 5.241762306664578e-06,
+      "loss": 0.9082,
+      "step": 28359
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9309280042456598,
+      "learning_rate": 5.241091194893309e-06,
+      "loss": 0.9095,
+      "step": 28360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6120348056774232,
+      "learning_rate": 5.24042011082991e-06,
+      "loss": 0.8862,
+      "step": 28361
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4513735425884975,
+      "learning_rate": 5.2397490544782935e-06,
+      "loss": 0.8124,
+      "step": 28362
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6416881943597126,
+      "learning_rate": 5.2390780258423715e-06,
+      "loss": 0.8693,
+      "step": 28363
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8101931947812964,
+      "learning_rate": 5.238407024926038e-06,
+      "loss": 0.7788,
+      "step": 28364
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6349193639820445,
+      "learning_rate": 5.237736051733205e-06,
+      "loss": 0.8536,
+      "step": 28365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5796360783502466,
+      "learning_rate": 5.23706510626778e-06,
+      "loss": 0.7765,
+      "step": 28366
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6024160286143545,
+      "learning_rate": 5.236394188533679e-06,
+      "loss": 0.8326,
+      "step": 28367
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4857587189282475,
+      "learning_rate": 5.2357232985347905e-06,
+      "loss": 0.8492,
+      "step": 28368
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6579516303514856,
+      "learning_rate": 5.23505243627503e-06,
+      "loss": 0.8822,
+      "step": 28369
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7409789371320771,
+      "learning_rate": 5.234381601758306e-06,
+      "loss": 0.6652,
+      "step": 28370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7608587406572883,
+      "learning_rate": 5.233710794988518e-06,
+      "loss": 0.768,
+      "step": 28371
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5377149614389127,
+      "learning_rate": 5.233040015969577e-06,
+      "loss": 0.9183,
+      "step": 28372
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6507235480452058,
+      "learning_rate": 5.232369264705384e-06,
+      "loss": 0.7558,
+      "step": 28373
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7728128738264491,
+      "learning_rate": 5.23169854119985e-06,
+      "loss": 0.6757,
+      "step": 28374
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7543098710697729,
+      "learning_rate": 5.231027845456873e-06,
+      "loss": 0.8261,
+      "step": 28375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8481279176994576,
+      "learning_rate": 5.230357177480365e-06,
+      "loss": 0.7635,
+      "step": 28376
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7356041812384416,
+      "learning_rate": 5.229686537274223e-06,
+      "loss": 0.6273,
+      "step": 28377
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.662790424263679,
+      "learning_rate": 5.229015924842362e-06,
+      "loss": 0.8557,
+      "step": 28378
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.508685353508678,
+      "learning_rate": 5.228345340188677e-06,
+      "loss": 0.8744,
+      "step": 28379
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.446794649994911,
+      "learning_rate": 5.227674783317078e-06,
+      "loss": 0.8217,
+      "step": 28380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6161910723923274,
+      "learning_rate": 5.227004254231468e-06,
+      "loss": 0.9109,
+      "step": 28381
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.568939478200577,
+      "learning_rate": 5.226333752935747e-06,
+      "loss": 0.7797,
+      "step": 28382
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8008497909986798,
+      "learning_rate": 5.2256632794338216e-06,
+      "loss": 0.8487,
+      "step": 28383
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9387205974864414,
+      "learning_rate": 5.224992833729601e-06,
+      "loss": 0.6761,
+      "step": 28384
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4194478517573417,
+      "learning_rate": 5.2243224158269835e-06,
+      "loss": 0.8933,
+      "step": 28385
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8060909287160727,
+      "learning_rate": 5.223652025729868e-06,
+      "loss": 0.8922,
+      "step": 28386
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6546305492119124,
+      "learning_rate": 5.222981663442164e-06,
+      "loss": 0.8527,
+      "step": 28387
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9565717361915034,
+      "learning_rate": 5.222311328967775e-06,
+      "loss": 0.8647,
+      "step": 28388
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.118695597296113,
+      "learning_rate": 5.221641022310604e-06,
+      "loss": 0.9401,
+      "step": 28389
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6458825023145272,
+      "learning_rate": 5.220970743474547e-06,
+      "loss": 0.9433,
+      "step": 28390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4919028305381445,
+      "learning_rate": 5.220300492463511e-06,
+      "loss": 0.7639,
+      "step": 28391
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6314151021230672,
+      "learning_rate": 5.219630269281404e-06,
+      "loss": 0.9154,
+      "step": 28392
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6392580076258385,
+      "learning_rate": 5.218960073932122e-06,
+      "loss": 0.8958,
+      "step": 28393
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6961390927632805,
+      "learning_rate": 5.218289906419564e-06,
+      "loss": 0.7894,
+      "step": 28394
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7130576648998608,
+      "learning_rate": 5.217619766747641e-06,
+      "loss": 0.6997,
+      "step": 28395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.58370389574339,
+      "learning_rate": 5.2169496549202444e-06,
+      "loss": 0.7897,
+      "step": 28396
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4319232980002181,
+      "learning_rate": 5.216279570941287e-06,
+      "loss": 0.7781,
+      "step": 28397
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6431145778488774,
+      "learning_rate": 5.2156095148146594e-06,
+      "loss": 0.734,
+      "step": 28398
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.422956219784478,
+      "learning_rate": 5.214939486544271e-06,
+      "loss": 0.7958,
+      "step": 28399
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4360297303727967,
+      "learning_rate": 5.214269486134017e-06,
+      "loss": 0.8996,
+      "step": 28400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4271109728615756,
+      "learning_rate": 5.213599513587805e-06,
+      "loss": 0.8485,
+      "step": 28401
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7718462483511783,
+      "learning_rate": 5.212929568909532e-06,
+      "loss": 0.7639,
+      "step": 28402
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6942687879813676,
+      "learning_rate": 5.212259652103095e-06,
+      "loss": 0.8751,
+      "step": 28403
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.437824834478716,
+      "learning_rate": 5.211589763172398e-06,
+      "loss": 0.7464,
+      "step": 28404
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6293649446432217,
+      "learning_rate": 5.210919902121344e-06,
+      "loss": 0.8142,
+      "step": 28405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7462190238519248,
+      "learning_rate": 5.2102500689538316e-06,
+      "loss": 0.8925,
+      "step": 28406
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.769309877117148,
+      "learning_rate": 5.2095802636737556e-06,
+      "loss": 0.9163,
+      "step": 28407
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7077713320303325,
+      "learning_rate": 5.20891048628502e-06,
+      "loss": 0.7925,
+      "step": 28408
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6582707760151054,
+      "learning_rate": 5.208240736791528e-06,
+      "loss": 0.8694,
+      "step": 28409
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5696454738404035,
+      "learning_rate": 5.207571015197177e-06,
+      "loss": 0.8625,
+      "step": 28410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5892008717964208,
+      "learning_rate": 5.206901321505859e-06,
+      "loss": 0.7863,
+      "step": 28411
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6256240030053362,
+      "learning_rate": 5.2062316557214785e-06,
+      "loss": 0.7663,
+      "step": 28412
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5465659941610008,
+      "learning_rate": 5.2055620178479405e-06,
+      "loss": 0.849,
+      "step": 28413
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5487791993179838,
+      "learning_rate": 5.204892407889138e-06,
+      "loss": 0.8775,
+      "step": 28414
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7951233806207296,
+      "learning_rate": 5.204222825848966e-06,
+      "loss": 1.0468,
+      "step": 28415
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5591674329959622,
+      "learning_rate": 5.203553271731332e-06,
+      "loss": 0.8965,
+      "step": 28416
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.693179986488374,
+      "learning_rate": 5.202883745540124e-06,
+      "loss": 0.8078,
+      "step": 28417
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5157745862203507,
+      "learning_rate": 5.2022142472792494e-06,
+      "loss": 0.8956,
+      "step": 28418
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4402059147667312,
+      "learning_rate": 5.2015447769525985e-06,
+      "loss": 0.8267,
+      "step": 28419
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.754159276006277,
+      "learning_rate": 5.2008753345640775e-06,
+      "loss": 0.7223,
+      "step": 28420
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.635455101106759,
+      "learning_rate": 5.200205920117576e-06,
+      "loss": 0.8407,
+      "step": 28421
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8565082646136895,
+      "learning_rate": 5.199536533616999e-06,
+      "loss": 0.8596,
+      "step": 28422
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5206837131269089,
+      "learning_rate": 5.19886717506624e-06,
+      "loss": 0.6767,
+      "step": 28423
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.688904437575465,
+      "learning_rate": 5.198197844469193e-06,
+      "loss": 0.8112,
+      "step": 28424
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7560349159101658,
+      "learning_rate": 5.197528541829758e-06,
+      "loss": 0.8067,
+      "step": 28425
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6600076066643616,
+      "learning_rate": 5.196859267151835e-06,
+      "loss": 0.87,
+      "step": 28426
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.590205566150923,
+      "learning_rate": 5.196190020439319e-06,
+      "loss": 0.7725,
+      "step": 28427
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.991579797239793,
+      "learning_rate": 5.195520801696101e-06,
+      "loss": 0.8573,
+      "step": 28428
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0566196808230828,
+      "learning_rate": 5.194851610926083e-06,
+      "loss": 0.8144,
+      "step": 28429
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.762243124005277,
+      "learning_rate": 5.194182448133163e-06,
+      "loss": 0.7123,
+      "step": 28430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.510798053241769,
+      "learning_rate": 5.1935133133212345e-06,
+      "loss": 0.8053,
+      "step": 28431
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.468057027781075,
+      "learning_rate": 5.19284420649419e-06,
+      "loss": 0.7214,
+      "step": 28432
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6789756459264213,
+      "learning_rate": 5.192175127655929e-06,
+      "loss": 0.8943,
+      "step": 28433
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5888901500692636,
+      "learning_rate": 5.191506076810349e-06,
+      "loss": 0.7459,
+      "step": 28434
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.562968990043909,
+      "learning_rate": 5.190837053961344e-06,
+      "loss": 0.8024,
+      "step": 28435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6781979835956349,
+      "learning_rate": 5.190168059112804e-06,
+      "loss": 0.8312,
+      "step": 28436
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7374757993681291,
+      "learning_rate": 5.189499092268632e-06,
+      "loss": 0.9026,
+      "step": 28437
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.699897722814449,
+      "learning_rate": 5.188830153432716e-06,
+      "loss": 0.8415,
+      "step": 28438
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6540624630560994,
+      "learning_rate": 5.188161242608957e-06,
+      "loss": 0.9032,
+      "step": 28439
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7964062497210381,
+      "learning_rate": 5.187492359801243e-06,
+      "loss": 0.7399,
+      "step": 28440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6678039804551807,
+      "learning_rate": 5.186823505013476e-06,
+      "loss": 0.8382,
+      "step": 28441
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.461965699675877,
+      "learning_rate": 5.186154678249542e-06,
+      "loss": 0.7984,
+      "step": 28442
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6259041237732828,
+      "learning_rate": 5.185485879513344e-06,
+      "loss": 0.7863,
+      "step": 28443
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6018647597226974,
+      "learning_rate": 5.184817108808773e-06,
+      "loss": 0.8888,
+      "step": 28444
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.466643998580152,
+      "learning_rate": 5.184148366139715e-06,
+      "loss": 0.7704,
+      "step": 28445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8129903953864128,
+      "learning_rate": 5.18347965151007e-06,
+      "loss": 0.7046,
+      "step": 28446
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4519960077566658,
+      "learning_rate": 5.182810964923733e-06,
+      "loss": 0.7767,
+      "step": 28447
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5293604210233402,
+      "learning_rate": 5.182142306384602e-06,
+      "loss": 0.8871,
+      "step": 28448
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6522199364742105,
+      "learning_rate": 5.181473675896557e-06,
+      "loss": 0.9559,
+      "step": 28449
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.143341155932967,
+      "learning_rate": 5.180805073463497e-06,
+      "loss": 0.9718,
+      "step": 28450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6861282731193068,
+      "learning_rate": 5.18013649908932e-06,
+      "loss": 0.8489,
+      "step": 28451
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9382014200643296,
+      "learning_rate": 5.179467952777915e-06,
+      "loss": 0.9496,
+      "step": 28452
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4912875033800064,
+      "learning_rate": 5.178799434533168e-06,
+      "loss": 0.7284,
+      "step": 28453
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4323111809770264,
+      "learning_rate": 5.178130944358978e-06,
+      "loss": 0.9351,
+      "step": 28454
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4593039292583616,
+      "learning_rate": 5.17746248225924e-06,
+      "loss": 0.8514,
+      "step": 28455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.46944337012384,
+      "learning_rate": 5.176794048237842e-06,
+      "loss": 0.8358,
+      "step": 28456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7249246920614749,
+      "learning_rate": 5.176125642298673e-06,
+      "loss": 0.674,
+      "step": 28457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5320842945947388,
+      "learning_rate": 5.1754572644456315e-06,
+      "loss": 0.848,
+      "step": 28458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3465267999961381,
+      "learning_rate": 5.1747889146826e-06,
+      "loss": 0.768,
+      "step": 28459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.504252299022209,
+      "learning_rate": 5.174120593013476e-06,
+      "loss": 0.8392,
+      "step": 28460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5548771576935212,
+      "learning_rate": 5.173452299442154e-06,
+      "loss": 0.7981,
+      "step": 28461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5047108892881673,
+      "learning_rate": 5.17278403397252e-06,
+      "loss": 0.8579,
+      "step": 28462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5819767007283072,
+      "learning_rate": 5.172115796608462e-06,
+      "loss": 0.9271,
+      "step": 28463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5931203393567623,
+      "learning_rate": 5.171447587353875e-06,
+      "loss": 0.8311,
+      "step": 28464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7902462443035634,
+      "learning_rate": 5.170779406212656e-06,
+      "loss": 0.8071,
+      "step": 28465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.601606366945851,
+      "learning_rate": 5.170111253188681e-06,
+      "loss": 0.8341,
+      "step": 28466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8967150193330955,
+      "learning_rate": 5.169443128285846e-06,
+      "loss": 0.7762,
+      "step": 28467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6014210353572238,
+      "learning_rate": 5.1687750315080446e-06,
+      "loss": 0.8362,
+      "step": 28468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4656202732868575,
+      "learning_rate": 5.168106962859166e-06,
+      "loss": 0.7926,
+      "step": 28469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7888288351123107,
+      "learning_rate": 5.1674389223431e-06,
+      "loss": 0.6972,
+      "step": 28470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5620591003142617,
+      "learning_rate": 5.16677090996373e-06,
+      "loss": 0.9072,
+      "step": 28471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7471759785111471,
+      "learning_rate": 5.166102925724954e-06,
+      "loss": 0.8244,
+      "step": 28472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7631272751786661,
+      "learning_rate": 5.165434969630654e-06,
+      "loss": 0.7161,
+      "step": 28473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8094604625017743,
+      "learning_rate": 5.164767041684724e-06,
+      "loss": 0.8788,
+      "step": 28474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.629177160581198,
+      "learning_rate": 5.1640991418910494e-06,
+      "loss": 0.8702,
+      "step": 28475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6839840990944872,
+      "learning_rate": 5.163431270253523e-06,
+      "loss": 0.7601,
+      "step": 28476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7017465033508947,
+      "learning_rate": 5.162763426776027e-06,
+      "loss": 0.9096,
+      "step": 28477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6251829486122933,
+      "learning_rate": 5.162095611462458e-06,
+      "loss": 0.7982,
+      "step": 28478
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7778222856237353,
+      "learning_rate": 5.161427824316703e-06,
+      "loss": 0.7092,
+      "step": 28479
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5331251581680727,
+      "learning_rate": 5.160760065342639e-06,
+      "loss": 0.8413,
+      "step": 28480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4136563965689746,
+      "learning_rate": 5.160092334544165e-06,
+      "loss": 0.9278,
+      "step": 28481
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.573368848912643,
+      "learning_rate": 5.159424631925168e-06,
+      "loss": 0.9373,
+      "step": 28482
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6612327606656172,
+      "learning_rate": 5.158756957489533e-06,
+      "loss": 0.8536,
+      "step": 28483
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5824493386493639,
+      "learning_rate": 5.158089311241143e-06,
+      "loss": 0.8604,
+      "step": 28484
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8068802827300804,
+      "learning_rate": 5.157421693183893e-06,
+      "loss": 0.7117,
+      "step": 28485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6724121779233856,
+      "learning_rate": 5.156754103321669e-06,
+      "loss": 0.8564,
+      "step": 28486
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6272574670984867,
+      "learning_rate": 5.156086541658356e-06,
+      "loss": 0.8044,
+      "step": 28487
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7571412772897712,
+      "learning_rate": 5.155419008197837e-06,
+      "loss": 0.8975,
+      "step": 28488
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.432292080327357,
+      "learning_rate": 5.154751502944002e-06,
+      "loss": 0.8282,
+      "step": 28489
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3272114687583425,
+      "learning_rate": 5.154084025900742e-06,
+      "loss": 0.9306,
+      "step": 28490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.539518943023221,
+      "learning_rate": 5.153416577071939e-06,
+      "loss": 0.9574,
+      "step": 28491
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.698649249166463,
+      "learning_rate": 5.152749156461475e-06,
+      "loss": 0.9232,
+      "step": 28492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.853466308754679,
+      "learning_rate": 5.152081764073244e-06,
+      "loss": 0.8157,
+      "step": 28493
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7673188946176603,
+      "learning_rate": 5.1514143999111234e-06,
+      "loss": 0.7348,
+      "step": 28494
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.581864435049705,
+      "learning_rate": 5.150747063979008e-06,
+      "loss": 0.7427,
+      "step": 28495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.586088958923914,
+      "learning_rate": 5.150079756280774e-06,
+      "loss": 0.7971,
+      "step": 28496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.412518592030089,
+      "learning_rate": 5.149412476820314e-06,
+      "loss": 0.817,
+      "step": 28497
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7574553445340826,
+      "learning_rate": 5.148745225601509e-06,
+      "loss": 0.7317,
+      "step": 28498
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5101764552836032,
+      "learning_rate": 5.148078002628247e-06,
+      "loss": 0.8466,
+      "step": 28499
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6710514517112223,
+      "learning_rate": 5.147410807904408e-06,
+      "loss": 0.7593,
+      "step": 28500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4315158522824276,
+      "learning_rate": 5.146743641433884e-06,
+      "loss": 0.7821,
+      "step": 28501
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6216423257461885,
+      "learning_rate": 5.146076503220549e-06,
+      "loss": 0.7282,
+      "step": 28502
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6385201108524416,
+      "learning_rate": 5.145409393268299e-06,
+      "loss": 0.838,
+      "step": 28503
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5242474352146345,
+      "learning_rate": 5.144742311581013e-06,
+      "loss": 0.7624,
+      "step": 28504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.599655842513696,
+      "learning_rate": 5.144075258162569e-06,
+      "loss": 0.9406,
+      "step": 28505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7392805952306067,
+      "learning_rate": 5.143408233016855e-06,
+      "loss": 0.7369,
+      "step": 28506
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7356410463022515,
+      "learning_rate": 5.142741236147762e-06,
+      "loss": 0.7008,
+      "step": 28507
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6097027922910156,
+      "learning_rate": 5.142074267559167e-06,
+      "loss": 0.8768,
+      "step": 28508
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.786165018336195,
+      "learning_rate": 5.141407327254951e-06,
+      "loss": 0.8736,
+      "step": 28509
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6379650166104902,
+      "learning_rate": 5.140740415238996e-06,
+      "loss": 0.9415,
+      "step": 28510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3827687090874208,
+      "learning_rate": 5.140073531515195e-06,
+      "loss": 0.8571,
+      "step": 28511
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.510281023956074,
+      "learning_rate": 5.139406676087424e-06,
+      "loss": 0.8199,
+      "step": 28512
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6818475343123382,
+      "learning_rate": 5.138739848959563e-06,
+      "loss": 0.9139,
+      "step": 28513
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7356590524194131,
+      "learning_rate": 5.138073050135497e-06,
+      "loss": 0.7128,
+      "step": 28514
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5345694081955417,
+      "learning_rate": 5.137406279619113e-06,
+      "loss": 0.7483,
+      "step": 28515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5776270200427707,
+      "learning_rate": 5.136739537414291e-06,
+      "loss": 0.8468,
+      "step": 28516
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5760143045719275,
+      "learning_rate": 5.136072823524905e-06,
+      "loss": 0.924,
+      "step": 28517
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7160528004918612,
+      "learning_rate": 5.1354061379548455e-06,
+      "loss": 0.9565,
+      "step": 28518
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9296926624611765,
+      "learning_rate": 5.134739480707991e-06,
+      "loss": 0.6948,
+      "step": 28519
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9226997475632899,
+      "learning_rate": 5.1340728517882256e-06,
+      "loss": 0.9278,
+      "step": 28520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4322321479344593,
+      "learning_rate": 5.133406251199424e-06,
+      "loss": 0.9239,
+      "step": 28521
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6814061092816033,
+      "learning_rate": 5.132739678945478e-06,
+      "loss": 0.8472,
+      "step": 28522
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.953820642923319,
+      "learning_rate": 5.132073135030255e-06,
+      "loss": 0.7838,
+      "step": 28523
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.642305741998046,
+      "learning_rate": 5.13140661945765e-06,
+      "loss": 0.8577,
+      "step": 28524
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.95505123037757,
+      "learning_rate": 5.1307401322315375e-06,
+      "loss": 0.7674,
+      "step": 28525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8556823028626421,
+      "learning_rate": 5.1300736733557936e-06,
+      "loss": 0.8387,
+      "step": 28526
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6799614720300784,
+      "learning_rate": 5.129407242834301e-06,
+      "loss": 0.8645,
+      "step": 28527
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6055903585928224,
+      "learning_rate": 5.128740840670946e-06,
+      "loss": 0.8691,
+      "step": 28528
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.710823481360823,
+      "learning_rate": 5.128074466869603e-06,
+      "loss": 0.6987,
+      "step": 28529
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6125185371661122,
+      "learning_rate": 5.127408121434152e-06,
+      "loss": 0.851,
+      "step": 28530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5339838714689245,
+      "learning_rate": 5.126741804368471e-06,
+      "loss": 0.8975,
+      "step": 28531
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.981043953454198,
+      "learning_rate": 5.1260755156764474e-06,
+      "loss": 0.8313,
+      "step": 28532
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6490273001023668,
+      "learning_rate": 5.125409255361954e-06,
+      "loss": 0.9224,
+      "step": 28533
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3752262956637122,
+      "learning_rate": 5.124743023428867e-06,
+      "loss": 0.8214,
+      "step": 28534
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8065322985907006,
+      "learning_rate": 5.124076819881071e-06,
+      "loss": 0.8837,
+      "step": 28535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8415372675092307,
+      "learning_rate": 5.123410644722446e-06,
+      "loss": 0.8534,
+      "step": 28536
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5303163638625985,
+      "learning_rate": 5.1227444979568694e-06,
+      "loss": 0.7414,
+      "step": 28537
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8412219088058208,
+      "learning_rate": 5.122078379588214e-06,
+      "loss": 0.9468,
+      "step": 28538
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7945275173291932,
+      "learning_rate": 5.121412289620367e-06,
+      "loss": 0.8996,
+      "step": 28539
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8837100982955008,
+      "learning_rate": 5.120746228057197e-06,
+      "loss": 0.7338,
+      "step": 28540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8887019805298775,
+      "learning_rate": 5.120080194902593e-06,
+      "loss": 0.8686,
+      "step": 28541
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7898329373966408,
+      "learning_rate": 5.119414190160421e-06,
+      "loss": 0.7807,
+      "step": 28542
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5869341777113521,
+      "learning_rate": 5.11874821383457e-06,
+      "loss": 0.6966,
+      "step": 28543
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.698508086311225,
+      "learning_rate": 5.118082265928908e-06,
+      "loss": 0.9075,
+      "step": 28544
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7756571592102637,
+      "learning_rate": 5.117416346447317e-06,
+      "loss": 0.8608,
+      "step": 28545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4824800647771268,
+      "learning_rate": 5.116750455393681e-06,
+      "loss": 0.7545,
+      "step": 28546
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4531592622125453,
+      "learning_rate": 5.116084592771861e-06,
+      "loss": 0.8997,
+      "step": 28547
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7090296894605306,
+      "learning_rate": 5.115418758585745e-06,
+      "loss": 0.8526,
+      "step": 28548
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8632380275872669,
+      "learning_rate": 5.114752952839207e-06,
+      "loss": 0.887,
+      "step": 28549
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7069457393690233,
+      "learning_rate": 5.114087175536131e-06,
+      "loss": 0.8341,
+      "step": 28550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5242285746384432,
+      "learning_rate": 5.113421426680378e-06,
+      "loss": 0.8849,
+      "step": 28551
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.669571844982561,
+      "learning_rate": 5.112755706275833e-06,
+      "loss": 0.915,
+      "step": 28552
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7532199365463488,
+      "learning_rate": 5.112090014326376e-06,
+      "loss": 0.9166,
+      "step": 28553
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0933987696838003,
+      "learning_rate": 5.111424350835879e-06,
+      "loss": 0.9199,
+      "step": 28554
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4756141508975638,
+      "learning_rate": 5.110758715808213e-06,
+      "loss": 0.8002,
+      "step": 28555
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7724941457771766,
+      "learning_rate": 5.1100931092472564e-06,
+      "loss": 0.8932,
+      "step": 28556
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5934064183008767,
+      "learning_rate": 5.10942753115689e-06,
+      "loss": 0.8043,
+      "step": 28557
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.002170997036697,
+      "learning_rate": 5.1087619815409815e-06,
+      "loss": 0.8297,
+      "step": 28558
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6282573754806706,
+      "learning_rate": 5.1080964604034136e-06,
+      "loss": 0.8143,
+      "step": 28559
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4528973756670918,
+      "learning_rate": 5.107430967748056e-06,
+      "loss": 0.9145,
+      "step": 28560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7739364754526143,
+      "learning_rate": 5.106765503578781e-06,
+      "loss": 0.8047,
+      "step": 28561
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.549628723832371,
+      "learning_rate": 5.1061000678994665e-06,
+      "loss": 0.8652,
+      "step": 28562
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6281594590253579,
+      "learning_rate": 5.10543466071399e-06,
+      "loss": 0.9123,
+      "step": 28563
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6426677795445397,
+      "learning_rate": 5.104769282026223e-06,
+      "loss": 0.818,
+      "step": 28564
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4922355939124605,
+      "learning_rate": 5.104103931840035e-06,
+      "loss": 0.8234,
+      "step": 28565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8342730668577418,
+      "learning_rate": 5.103438610159303e-06,
+      "loss": 0.7136,
+      "step": 28566
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5182127133227852,
+      "learning_rate": 5.102773316987905e-06,
+      "loss": 0.7652,
+      "step": 28567
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7618344483287096,
+      "learning_rate": 5.102108052329714e-06,
+      "loss": 0.6983,
+      "step": 28568
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1288146206648926,
+      "learning_rate": 5.1014428161885956e-06,
+      "loss": 0.8095,
+      "step": 28569
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7659345118742626,
+      "learning_rate": 5.1007776085684265e-06,
+      "loss": 0.7925,
+      "step": 28570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7051483019826685,
+      "learning_rate": 5.100112429473087e-06,
+      "loss": 0.7975,
+      "step": 28571
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1378433722400336,
+      "learning_rate": 5.0994472789064434e-06,
+      "loss": 0.9159,
+      "step": 28572
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1495418024798005,
+      "learning_rate": 5.098782156872365e-06,
+      "loss": 0.7726,
+      "step": 28573
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8376432620787189,
+      "learning_rate": 5.0981170633747326e-06,
+      "loss": 0.8096,
+      "step": 28574
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.047350031608153,
+      "learning_rate": 5.097451998417411e-06,
+      "loss": 0.7105,
+      "step": 28575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5758500443108163,
+      "learning_rate": 5.09678696200428e-06,
+      "loss": 0.8512,
+      "step": 28576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4613742339732732,
+      "learning_rate": 5.096121954139204e-06,
+      "loss": 0.9021,
+      "step": 28577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8526583352054924,
+      "learning_rate": 5.095456974826061e-06,
+      "loss": 0.8109,
+      "step": 28578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6729151998083034,
+      "learning_rate": 5.094792024068718e-06,
+      "loss": 0.8902,
+      "step": 28579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8260998071462978,
+      "learning_rate": 5.0941271018710514e-06,
+      "loss": 0.7139,
+      "step": 28580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.730135518810302,
+      "learning_rate": 5.093462208236931e-06,
+      "loss": 0.7498,
+      "step": 28581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5024707502385672,
+      "learning_rate": 5.092797343170223e-06,
+      "loss": 0.8883,
+      "step": 28582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7445968661988203,
+      "learning_rate": 5.092132506674801e-06,
+      "loss": 0.9252,
+      "step": 28583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.508982512076739,
+      "learning_rate": 5.091467698754543e-06,
+      "loss": 0.85,
+      "step": 28584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6021303866072159,
+      "learning_rate": 5.090802919413314e-06,
+      "loss": 0.849,
+      "step": 28585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7760237821440077,
+      "learning_rate": 5.09013816865498e-06,
+      "loss": 0.7348,
+      "step": 28586
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5751109316432748,
+      "learning_rate": 5.089473446483416e-06,
+      "loss": 0.917,
+      "step": 28587
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7688548170503883,
+      "learning_rate": 5.0888087529024965e-06,
+      "loss": 0.7762,
+      "step": 28588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.56696961666792,
+      "learning_rate": 5.088144087916087e-06,
+      "loss": 0.8064,
+      "step": 28589
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5150265498405013,
+      "learning_rate": 5.087479451528055e-06,
+      "loss": 0.783,
+      "step": 28590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6120153113790392,
+      "learning_rate": 5.086814843742272e-06,
+      "loss": 0.9464,
+      "step": 28591
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7948002294561827,
+      "learning_rate": 5.086150264562612e-06,
+      "loss": 0.9772,
+      "step": 28592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8983370954976995,
+      "learning_rate": 5.08548571399294e-06,
+      "loss": 0.8186,
+      "step": 28593
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5907035736358808,
+      "learning_rate": 5.0848211920371236e-06,
+      "loss": 0.9511,
+      "step": 28594
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4459020632251427,
+      "learning_rate": 5.084156698699038e-06,
+      "loss": 0.8356,
+      "step": 28595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.609389032456357,
+      "learning_rate": 5.0834922339825445e-06,
+      "loss": 0.7374,
+      "step": 28596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5251876649324108,
+      "learning_rate": 5.08282779789152e-06,
+      "loss": 0.8707,
+      "step": 28597
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4986195056934895,
+      "learning_rate": 5.082163390429824e-06,
+      "loss": 0.8275,
+      "step": 28598
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8981451388425463,
+      "learning_rate": 5.081499011601334e-06,
+      "loss": 0.84,
+      "step": 28599
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8662695976697825,
+      "learning_rate": 5.0808346614099106e-06,
+      "loss": 0.8205,
+      "step": 28600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4308765532272512,
+      "learning_rate": 5.080170339859429e-06,
+      "loss": 0.7849,
+      "step": 28601
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6185741921368577,
+      "learning_rate": 5.079506046953753e-06,
+      "loss": 0.8798,
+      "step": 28602
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6142286419648748,
+      "learning_rate": 5.078841782696746e-06,
+      "loss": 0.8305,
+      "step": 28603
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.670341533830161,
+      "learning_rate": 5.07817754709228e-06,
+      "loss": 0.9225,
+      "step": 28604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3935885674790045,
+      "learning_rate": 5.0775133401442275e-06,
+      "loss": 0.8056,
+      "step": 28605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.778805767149228,
+      "learning_rate": 5.0768491618564504e-06,
+      "loss": 0.8495,
+      "step": 28606
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4645862392936662,
+      "learning_rate": 5.076185012232813e-06,
+      "loss": 0.7186,
+      "step": 28607
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.507297898812066,
+      "learning_rate": 5.0755208912771835e-06,
+      "loss": 0.8722,
+      "step": 28608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.757345463909712,
+      "learning_rate": 5.074856798993436e-06,
+      "loss": 0.8729,
+      "step": 28609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6762327962080104,
+      "learning_rate": 5.0741927353854305e-06,
+      "loss": 0.88,
+      "step": 28610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2753198411193876,
+      "learning_rate": 5.073528700457031e-06,
+      "loss": 0.782,
+      "step": 28611
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7071430046416343,
+      "learning_rate": 5.0728646942121074e-06,
+      "loss": 0.9772,
+      "step": 28612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.559158204853203,
+      "learning_rate": 5.0722007166545294e-06,
+      "loss": 0.8922,
+      "step": 28613
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1509481865423457,
+      "learning_rate": 5.071536767788158e-06,
+      "loss": 0.7745,
+      "step": 28614
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.533721708033607,
+      "learning_rate": 5.070872847616857e-06,
+      "loss": 0.704,
+      "step": 28615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.447254326537531,
+      "learning_rate": 5.070208956144498e-06,
+      "loss": 0.8142,
+      "step": 28616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7878674857307244,
+      "learning_rate": 5.0695450933749395e-06,
+      "loss": 0.8318,
+      "step": 28617
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7336559448746234,
+      "learning_rate": 5.068881259312054e-06,
+      "loss": 0.8281,
+      "step": 28618
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5582583053643058,
+      "learning_rate": 5.068217453959698e-06,
+      "loss": 0.9371,
+      "step": 28619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4577071385033975,
+      "learning_rate": 5.067553677321748e-06,
+      "loss": 0.773,
+      "step": 28620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4292992563811346,
+      "learning_rate": 5.066889929402057e-06,
+      "loss": 0.8724,
+      "step": 28621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7366727513127602,
+      "learning_rate": 5.066226210204498e-06,
+      "loss": 0.7324,
+      "step": 28622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.800111033409798,
+      "learning_rate": 5.065562519732929e-06,
+      "loss": 0.7727,
+      "step": 28623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6035796397278077,
+      "learning_rate": 5.064898857991221e-06,
+      "loss": 0.8298,
+      "step": 28624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6351826004187608,
+      "learning_rate": 5.0642352249832295e-06,
+      "loss": 0.8896,
+      "step": 28625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5172051712654826,
+      "learning_rate": 5.0635716207128274e-06,
+      "loss": 0.8673,
+      "step": 28626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6873152375277798,
+      "learning_rate": 5.0629080451838745e-06,
+      "loss": 0.7932,
+      "step": 28627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.81803539508571,
+      "learning_rate": 5.062244498400228e-06,
+      "loss": 0.9171,
+      "step": 28628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.016404447695377,
+      "learning_rate": 5.0615809803657615e-06,
+      "loss": 0.8813,
+      "step": 28629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0000064822856527,
+      "learning_rate": 5.060917491084331e-06,
+      "loss": 0.7406,
+      "step": 28630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.533662354792436,
+      "learning_rate": 5.06025403055981e-06,
+      "loss": 0.8655,
+      "step": 28631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1925701803894384,
+      "learning_rate": 5.059590598796047e-06,
+      "loss": 0.8771,
+      "step": 28632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6814677425329783,
+      "learning_rate": 5.058927195796912e-06,
+      "loss": 0.8849,
+      "step": 28633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0106279278524095,
+      "learning_rate": 5.0582638215662715e-06,
+      "loss": 0.7853,
+      "step": 28634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6399896219975425,
+      "learning_rate": 5.057600476107982e-06,
+      "loss": 0.824,
+      "step": 28635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.914950946054046,
+      "learning_rate": 5.0569371594259056e-06,
+      "loss": 0.8273,
+      "step": 28636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.71464507992749,
+      "learning_rate": 5.056273871523904e-06,
+      "loss": 0.9122,
+      "step": 28637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5224406419067107,
+      "learning_rate": 5.055610612405846e-06,
+      "loss": 0.89,
+      "step": 28638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7490127814425794,
+      "learning_rate": 5.054947382075588e-06,
+      "loss": 0.7389,
+      "step": 28639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6187047018886787,
+      "learning_rate": 5.054284180536988e-06,
+      "loss": 0.858,
+      "step": 28640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6396074569840897,
+      "learning_rate": 5.053621007793916e-06,
+      "loss": 0.8051,
+      "step": 28641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.501094149229742,
+      "learning_rate": 5.052957863850223e-06,
+      "loss": 0.8215,
+      "step": 28642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6436589167639812,
+      "learning_rate": 5.052294748709776e-06,
+      "loss": 0.8313,
+      "step": 28643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7467183827455915,
+      "learning_rate": 5.051631662376441e-06,
+      "loss": 0.7259,
+      "step": 28644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8023442945472548,
+      "learning_rate": 5.050968604854071e-06,
+      "loss": 1.0112,
+      "step": 28645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5763387889512752,
+      "learning_rate": 5.050305576146526e-06,
+      "loss": 0.8157,
+      "step": 28646
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7591043632946288,
+      "learning_rate": 5.0496425762576685e-06,
+      "loss": 0.9713,
+      "step": 28647
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7423019754990151,
+      "learning_rate": 5.048979605191367e-06,
+      "loss": 0.6649,
+      "step": 28648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7759931858203334,
+      "learning_rate": 5.0483166629514654e-06,
+      "loss": 0.7104,
+      "step": 28649
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.560777596998241,
+      "learning_rate": 5.047653749541832e-06,
+      "loss": 0.7613,
+      "step": 28650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6279711129161616,
+      "learning_rate": 5.046990864966328e-06,
+      "loss": 0.862,
+      "step": 28651
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6430232000373806,
+      "learning_rate": 5.046328009228817e-06,
+      "loss": 0.8047,
+      "step": 28652
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.733582786441692,
+      "learning_rate": 5.045665182333146e-06,
+      "loss": 0.8494,
+      "step": 28653
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6202014942441505,
+      "learning_rate": 5.045002384283179e-06,
+      "loss": 0.865,
+      "step": 28654
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5871513519775904,
+      "learning_rate": 5.044339615082783e-06,
+      "loss": 0.859,
+      "step": 28655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.693840780756938,
+      "learning_rate": 5.043676874735807e-06,
+      "loss": 0.9025,
+      "step": 28656
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8175448105282468,
+      "learning_rate": 5.043014163246116e-06,
+      "loss": 0.7983,
+      "step": 28657
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4801657548755986,
+      "learning_rate": 5.042351480617562e-06,
+      "loss": 0.875,
+      "step": 28658
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.643416446934071,
+      "learning_rate": 5.041688826854012e-06,
+      "loss": 0.8991,
+      "step": 28659
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4882585032622742,
+      "learning_rate": 5.041026201959314e-06,
+      "loss": 0.8154,
+      "step": 28660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6900810105952717,
+      "learning_rate": 5.040363605937337e-06,
+      "loss": 0.8533,
+      "step": 28661
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.567234808790059,
+      "learning_rate": 5.039701038791932e-06,
+      "loss": 0.9079,
+      "step": 28662
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5829932123098647,
+      "learning_rate": 5.039038500526955e-06,
+      "loss": 0.9379,
+      "step": 28663
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.761481752305562,
+      "learning_rate": 5.038375991146266e-06,
+      "loss": 0.6979,
+      "step": 28664
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7571447055148808,
+      "learning_rate": 5.037713510653727e-06,
+      "loss": 0.8891,
+      "step": 28665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8215318646762861,
+      "learning_rate": 5.03705105905319e-06,
+      "loss": 0.8422,
+      "step": 28666
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8637975460188902,
+      "learning_rate": 5.0363886363485105e-06,
+      "loss": 0.9695,
+      "step": 28667
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6347872911804069,
+      "learning_rate": 5.035726242543546e-06,
+      "loss": 0.8378,
+      "step": 28668
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7670233395590567,
+      "learning_rate": 5.0350638776421615e-06,
+      "loss": 0.8035,
+      "step": 28669
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5612311833389503,
+      "learning_rate": 5.034401541648204e-06,
+      "loss": 0.7365,
+      "step": 28670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.244638967367016,
+      "learning_rate": 5.033739234565531e-06,
+      "loss": 0.8546,
+      "step": 28671
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.665134764253439,
+      "learning_rate": 5.033076956398e-06,
+      "loss": 0.8444,
+      "step": 28672
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.434482003090589,
+      "learning_rate": 5.032414707149471e-06,
+      "loss": 0.9367,
+      "step": 28673
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0577628716674607,
+      "learning_rate": 5.031752486823797e-06,
+      "loss": 0.8563,
+      "step": 28674
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.366943237104014,
+      "learning_rate": 5.031090295424828e-06,
+      "loss": 0.9197,
+      "step": 28675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5212281060489123,
+      "learning_rate": 5.030428132956428e-06,
+      "loss": 0.7411,
+      "step": 28676
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9467747412961505,
+      "learning_rate": 5.029765999422447e-06,
+      "loss": 0.8985,
+      "step": 28677
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5619013958608818,
+      "learning_rate": 5.029103894826745e-06,
+      "loss": 0.8654,
+      "step": 28678
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7995169377415603,
+      "learning_rate": 5.0284418191731684e-06,
+      "loss": 0.8352,
+      "step": 28679
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6349893468704237,
+      "learning_rate": 5.027779772465583e-06,
+      "loss": 0.7622,
+      "step": 28680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5065197059257516,
+      "learning_rate": 5.027117754707833e-06,
+      "loss": 0.8663,
+      "step": 28681
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5622490885185885,
+      "learning_rate": 5.026455765903783e-06,
+      "loss": 0.7801,
+      "step": 28682
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6298240427070683,
+      "learning_rate": 5.025793806057282e-06,
+      "loss": 0.8369,
+      "step": 28683
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.657288836838832,
+      "learning_rate": 5.025131875172179e-06,
+      "loss": 0.8648,
+      "step": 28684
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3106701044932896,
+      "learning_rate": 5.024469973252334e-06,
+      "loss": 0.8352,
+      "step": 28685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.213859211151051,
+      "learning_rate": 5.023808100301604e-06,
+      "loss": 0.7995,
+      "step": 28686
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7544771793069749,
+      "learning_rate": 5.023146256323839e-06,
+      "loss": 0.7058,
+      "step": 28687
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7437900729439328,
+      "learning_rate": 5.022484441322888e-06,
+      "loss": 0.671,
+      "step": 28688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8185503301884773,
+      "learning_rate": 5.021822655302608e-06,
+      "loss": 0.9344,
+      "step": 28689
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3789622177100478,
+      "learning_rate": 5.021160898266857e-06,
+      "loss": 0.9206,
+      "step": 28690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5564601979023767,
+      "learning_rate": 5.020499170219483e-06,
+      "loss": 0.8439,
+      "step": 28691
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8179010287076127,
+      "learning_rate": 5.019837471164335e-06,
+      "loss": 0.7952,
+      "step": 28692
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.049271679299936,
+      "learning_rate": 5.019175801105272e-06,
+      "loss": 0.7879,
+      "step": 28693
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.705040887393795,
+      "learning_rate": 5.018514160046146e-06,
+      "loss": 0.7209,
+      "step": 28694
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6232110339387764,
+      "learning_rate": 5.017852547990809e-06,
+      "loss": 0.8504,
+      "step": 28695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7418427430503541,
+      "learning_rate": 5.0171909649431075e-06,
+      "loss": 0.7074,
+      "step": 28696
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.257097191072643,
+      "learning_rate": 5.0165294109069e-06,
+      "loss": 0.9313,
+      "step": 28697
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1948034002804,
+      "learning_rate": 5.015867885886035e-06,
+      "loss": 0.8655,
+      "step": 28698
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7554692611316387,
+      "learning_rate": 5.0152063898843685e-06,
+      "loss": 0.9111,
+      "step": 28699
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.810689336342644,
+      "learning_rate": 5.014544922905744e-06,
+      "loss": 0.7035,
+      "step": 28700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.453424096189883,
+      "learning_rate": 5.013883484954021e-06,
+      "loss": 0.8154,
+      "step": 28701
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5353441342207608,
+      "learning_rate": 5.013222076033043e-06,
+      "loss": 0.8139,
+      "step": 28702
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7611855054860373,
+      "learning_rate": 5.012560696146669e-06,
+      "loss": 0.7498,
+      "step": 28703
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6062317468732579,
+      "learning_rate": 5.011899345298745e-06,
+      "loss": 0.8779,
+      "step": 28704
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9390423851105532,
+      "learning_rate": 5.011238023493119e-06,
+      "loss": 0.9108,
+      "step": 28705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5690784088004015,
+      "learning_rate": 5.010576730733645e-06,
+      "loss": 0.8135,
+      "step": 28706
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7446513262604524,
+      "learning_rate": 5.0099154670241756e-06,
+      "loss": 0.8733,
+      "step": 28707
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6617844776198476,
+      "learning_rate": 5.009254232368559e-06,
+      "loss": 0.9043,
+      "step": 28708
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7002450984250348,
+      "learning_rate": 5.0085930267706404e-06,
+      "loss": 0.8402,
+      "step": 28709
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5739099594091268,
+      "learning_rate": 5.007931850234273e-06,
+      "loss": 0.8951,
+      "step": 28710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5852117934144239,
+      "learning_rate": 5.0072707027633115e-06,
+      "loss": 0.8244,
+      "step": 28711
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8054359523295777,
+      "learning_rate": 5.006609584361601e-06,
+      "loss": 0.7288,
+      "step": 28712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.544103106018299,
+      "learning_rate": 5.005948495032985e-06,
+      "loss": 0.7974,
+      "step": 28713
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5033996766140023,
+      "learning_rate": 5.0052874347813186e-06,
+      "loss": 0.8266,
+      "step": 28714
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6105270233584386,
+      "learning_rate": 5.004626403610454e-06,
+      "loss": 0.7331,
+      "step": 28715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6036532585419392,
+      "learning_rate": 5.0039654015242355e-06,
+      "loss": 0.8854,
+      "step": 28716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.674587603376311,
+      "learning_rate": 5.00330442852651e-06,
+      "loss": 0.8309,
+      "step": 28717
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5997860792204504,
+      "learning_rate": 5.00264348462113e-06,
+      "loss": 0.7716,
+      "step": 28718
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5662231445684143,
+      "learning_rate": 5.001982569811938e-06,
+      "loss": 0.8962,
+      "step": 28719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5416447222202234,
+      "learning_rate": 5.001321684102791e-06,
+      "loss": 0.8296,
+      "step": 28720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5901062384632885,
+      "learning_rate": 5.000660827497527e-06,
+      "loss": 0.9402,
+      "step": 28721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.784986695621341,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.6742,
+      "step": 28722
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5120764285039556,
+      "learning_rate": 4.999339201614057e-06,
+      "loss": 0.8165,
+      "step": 28723
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.776531344953669,
+      "learning_rate": 4.998678432343545e-06,
+      "loss": 0.7298,
+      "step": 28724
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6320319413103022,
+      "learning_rate": 4.998017692192312e-06,
+      "loss": 0.889,
+      "step": 28725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7985675803377018,
+      "learning_rate": 4.9973569811641985e-06,
+      "loss": 0.7065,
+      "step": 28726
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7191524586441458,
+      "learning_rate": 4.996696299263057e-06,
+      "loss": 0.7509,
+      "step": 28727
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8084575354706647,
+      "learning_rate": 4.996035646492734e-06,
+      "loss": 0.7345,
+      "step": 28728
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7011918246637014,
+      "learning_rate": 4.995375022857083e-06,
+      "loss": 0.9338,
+      "step": 28729
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8105168098010491,
+      "learning_rate": 4.994714428359936e-06,
+      "loss": 0.7114,
+      "step": 28730
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5736698268183704,
+      "learning_rate": 4.994053863005145e-06,
+      "loss": 0.7514,
+      "step": 28731
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5670417169021926,
+      "learning_rate": 4.993393326796562e-06,
+      "loss": 0.8501,
+      "step": 28732
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.713071579951694,
+      "learning_rate": 4.992732819738028e-06,
+      "loss": 0.9344,
+      "step": 28733
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.64895349417641,
+      "learning_rate": 4.992072341833385e-06,
+      "loss": 0.8648,
+      "step": 28734
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5405328120461033,
+      "learning_rate": 4.991411893086482e-06,
+      "loss": 0.7914,
+      "step": 28735
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5965185825631507,
+      "learning_rate": 4.990751473501168e-06,
+      "loss": 0.7755,
+      "step": 28736
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.521633880683876,
+      "learning_rate": 4.990091083081285e-06,
+      "loss": 0.9051,
+      "step": 28737
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5126890347980344,
+      "learning_rate": 4.989430721830676e-06,
+      "loss": 0.7757,
+      "step": 28738
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6003707630880797,
+      "learning_rate": 4.988770389753189e-06,
+      "loss": 0.7969,
+      "step": 28739
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7214514117021428,
+      "learning_rate": 4.988110086852664e-06,
+      "loss": 0.7192,
+      "step": 28740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4929157680054437,
+      "learning_rate": 4.9874498131329495e-06,
+      "loss": 0.8331,
+      "step": 28741
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6781778271141121,
+      "learning_rate": 4.986789568597893e-06,
+      "loss": 0.7969,
+      "step": 28742
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6442633981134962,
+      "learning_rate": 4.986129353251335e-06,
+      "loss": 0.9532,
+      "step": 28743
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.611475586284772,
+      "learning_rate": 4.9854691670971145e-06,
+      "loss": 0.8367,
+      "step": 28744
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.571167119704665,
+      "learning_rate": 4.9848090101390815e-06,
+      "loss": 0.8744,
+      "step": 28745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5302318019671126,
+      "learning_rate": 4.9841488823810805e-06,
+      "loss": 0.8966,
+      "step": 28746
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8391249678136474,
+      "learning_rate": 4.983488783826953e-06,
+      "loss": 0.7876,
+      "step": 28747
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6066059875538352,
+      "learning_rate": 4.9828287144805385e-06,
+      "loss": 0.7682,
+      "step": 28748
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4929664076081401,
+      "learning_rate": 4.9821686743456835e-06,
+      "loss": 0.8182,
+      "step": 28749
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8006761526473263,
+      "learning_rate": 4.981508663426236e-06,
+      "loss": 1.013,
+      "step": 28750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.551247692832465,
+      "learning_rate": 4.9808486817260314e-06,
+      "loss": 0.816,
+      "step": 28751
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6410387002233575,
+      "learning_rate": 4.980188729248913e-06,
+      "loss": 0.8353,
+      "step": 28752
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6199014478125513,
+      "learning_rate": 4.979528805998724e-06,
+      "loss": 0.7877,
+      "step": 28753
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1699316355739393,
+      "learning_rate": 4.978868911979312e-06,
+      "loss": 0.6631,
+      "step": 28754
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7227753570303836,
+      "learning_rate": 4.9782090471945155e-06,
+      "loss": 0.851,
+      "step": 28755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4609245504948687,
+      "learning_rate": 4.977549211648172e-06,
+      "loss": 0.7265,
+      "step": 28756
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5075090659214996,
+      "learning_rate": 4.976889405344129e-06,
+      "loss": 0.779,
+      "step": 28757
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7273627981760316,
+      "learning_rate": 4.9762296282862245e-06,
+      "loss": 0.7248,
+      "step": 28758
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7554448392455706,
+      "learning_rate": 4.9755698804783045e-06,
+      "loss": 0.7592,
+      "step": 28759
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.820466296669582,
+      "learning_rate": 4.974910161924204e-06,
+      "loss": 0.8572,
+      "step": 28760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.67375720238576,
+      "learning_rate": 4.974250472627771e-06,
+      "loss": 0.8564,
+      "step": 28761
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.664657546442497,
+      "learning_rate": 4.973590812592839e-06,
+      "loss": 0.8498,
+      "step": 28762
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3398366595916842,
+      "learning_rate": 4.9729311818232565e-06,
+      "loss": 0.7626,
+      "step": 28763
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.515331036582327,
+      "learning_rate": 4.97227158032286e-06,
+      "loss": 0.8493,
+      "step": 28764
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4585502544294868,
+      "learning_rate": 4.971612008095486e-06,
+      "loss": 0.8278,
+      "step": 28765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.492014206596346,
+      "learning_rate": 4.970952465144979e-06,
+      "loss": 0.8792,
+      "step": 28766
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6339956670139122,
+      "learning_rate": 4.970292951475183e-06,
+      "loss": 0.8486,
+      "step": 28767
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.503968526081821,
+      "learning_rate": 4.969633467089934e-06,
+      "loss": 0.8539,
+      "step": 28768
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6135389493666084,
+      "learning_rate": 4.968974011993067e-06,
+      "loss": 0.9078,
+      "step": 28769
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6463877447627306,
+      "learning_rate": 4.968314586188427e-06,
+      "loss": 0.7316,
+      "step": 28770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7266036298261157,
+      "learning_rate": 4.9676551896798565e-06,
+      "loss": 0.6576,
+      "step": 28771
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5886153380072254,
+      "learning_rate": 4.96699582247119e-06,
+      "loss": 0.7569,
+      "step": 28772
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4989786554343327,
+      "learning_rate": 4.9663364845662635e-06,
+      "loss": 1.0019,
+      "step": 28773
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5204134858745966,
+      "learning_rate": 4.965677175968919e-06,
+      "loss": 0.8332,
+      "step": 28774
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5434877467395272,
+      "learning_rate": 4.965017896683001e-06,
+      "loss": 0.8463,
+      "step": 28775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.853214712762767,
+      "learning_rate": 4.9643586467123416e-06,
+      "loss": 0.8135,
+      "step": 28776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.811135170409245,
+      "learning_rate": 4.963699426060777e-06,
+      "loss": 0.6653,
+      "step": 28777
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6855510065706611,
+      "learning_rate": 4.963040234732154e-06,
+      "loss": 0.8268,
+      "step": 28778
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.637193413816319,
+      "learning_rate": 4.9623810727303e-06,
+      "loss": 0.8897,
+      "step": 28779
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6746551213682541,
+      "learning_rate": 4.961721940059062e-06,
+      "loss": 0.9433,
+      "step": 28780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9193159388844578,
+      "learning_rate": 4.961062836722271e-06,
+      "loss": 0.6667,
+      "step": 28781
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8444753365061841,
+      "learning_rate": 4.960403762723771e-06,
+      "loss": 0.8405,
+      "step": 28782
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.599452062401387,
+      "learning_rate": 4.9597447180673916e-06,
+      "loss": 0.9699,
+      "step": 28783
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7707915724718057,
+      "learning_rate": 4.959085702756979e-06,
+      "loss": 0.8084,
+      "step": 28784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5782632234619938,
+      "learning_rate": 4.958426716796364e-06,
+      "loss": 0.7983,
+      "step": 28785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7956233964591355,
+      "learning_rate": 4.957767760189382e-06,
+      "loss": 0.8978,
+      "step": 28786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7371578401299788,
+      "learning_rate": 4.957108832939872e-06,
+      "loss": 0.7509,
+      "step": 28787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8541152412439803,
+      "learning_rate": 4.956449935051675e-06,
+      "loss": 0.8993,
+      "step": 28788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.83282787229736,
+      "learning_rate": 4.9557910665286236e-06,
+      "loss": 0.7165,
+      "step": 28789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.065418983703974,
+      "learning_rate": 4.955132227374548e-06,
+      "loss": 0.8902,
+      "step": 28790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.614706744904742,
+      "learning_rate": 4.954473417593292e-06,
+      "loss": 0.8084,
+      "step": 28791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4843348193181056,
+      "learning_rate": 4.953814637188692e-06,
+      "loss": 0.8213,
+      "step": 28792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6644005959125783,
+      "learning_rate": 4.95315588616458e-06,
+      "loss": 0.8042,
+      "step": 28793
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9997692658125534,
+      "learning_rate": 4.95249716452479e-06,
+      "loss": 0.8032,
+      "step": 28794
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5358711319593275,
+      "learning_rate": 4.951838472273158e-06,
+      "loss": 0.8381,
+      "step": 28795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5724616243606482,
+      "learning_rate": 4.951179809413526e-06,
+      "loss": 0.7874,
+      "step": 28796
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.884682767070833,
+      "learning_rate": 4.950521175949722e-06,
+      "loss": 0.9828,
+      "step": 28797
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5086535622497235,
+      "learning_rate": 4.9498625718855786e-06,
+      "loss": 0.877,
+      "step": 28798
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5932529267209679,
+      "learning_rate": 4.94920399722494e-06,
+      "loss": 0.8434,
+      "step": 28799
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7888489686810478,
+      "learning_rate": 4.948545451971629e-06,
+      "loss": 0.962,
+      "step": 28800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.712444897181545,
+      "learning_rate": 4.94788693612949e-06,
+      "loss": 0.9351,
+      "step": 28801
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.585516762517265,
+      "learning_rate": 4.947228449702348e-06,
+      "loss": 0.914,
+      "step": 28802
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.946079951682795,
+      "learning_rate": 4.946569992694047e-06,
+      "loss": 0.8571,
+      "step": 28803
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8414797314380176,
+      "learning_rate": 4.945911565108411e-06,
+      "loss": 0.9877,
+      "step": 28804
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7025071462787449,
+      "learning_rate": 4.945253166949282e-06,
+      "loss": 0.8779,
+      "step": 28805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5208057694161023,
+      "learning_rate": 4.944594798220489e-06,
+      "loss": 0.8065,
+      "step": 28806
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7562987703767209,
+      "learning_rate": 4.943936458925862e-06,
+      "loss": 0.665,
+      "step": 28807
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8770988333029217,
+      "learning_rate": 4.943278149069237e-06,
+      "loss": 0.874,
+      "step": 28808
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.605733851004833,
+      "learning_rate": 4.942619868654453e-06,
+      "loss": 1.0068,
+      "step": 28809
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5821216901047503,
+      "learning_rate": 4.9419616176853355e-06,
+      "loss": 0.8902,
+      "step": 28810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4987273685525853,
+      "learning_rate": 4.941303396165716e-06,
+      "loss": 0.8616,
+      "step": 28811
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6448605244688104,
+      "learning_rate": 4.940645204099429e-06,
+      "loss": 0.7404,
+      "step": 28812
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9967733706439066,
+      "learning_rate": 4.939987041490312e-06,
+      "loss": 0.8178,
+      "step": 28813
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0706687936256936,
+      "learning_rate": 4.939328908342192e-06,
+      "loss": 0.7985,
+      "step": 28814
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8052384494401756,
+      "learning_rate": 4.938670804658897e-06,
+      "loss": 0.8878,
+      "step": 28815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6128228850416708,
+      "learning_rate": 4.938012730444264e-06,
+      "loss": 0.7998,
+      "step": 28816
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6596974548129018,
+      "learning_rate": 4.937354685702126e-06,
+      "loss": 0.9041,
+      "step": 28817
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.77780256493154,
+      "learning_rate": 4.936696670436313e-06,
+      "loss": 0.855,
+      "step": 28818
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.697349627231502,
+      "learning_rate": 4.93603868465065e-06,
+      "loss": 0.8098,
+      "step": 28819
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5604854836565913,
+      "learning_rate": 4.935380728348978e-06,
+      "loss": 0.8424,
+      "step": 28820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5532938283177076,
+      "learning_rate": 4.9347228015351185e-06,
+      "loss": 0.8158,
+      "step": 28821
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6232943846420482,
+      "learning_rate": 4.934064904212911e-06,
+      "loss": 0.8016,
+      "step": 28822
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7679300252830021,
+      "learning_rate": 4.9334070363861765e-06,
+      "loss": 0.8763,
+      "step": 28823
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6379762084778482,
+      "learning_rate": 4.932749198058755e-06,
+      "loss": 0.8266,
+      "step": 28824
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5134282032506974,
+      "learning_rate": 4.932091389234467e-06,
+      "loss": 0.7952,
+      "step": 28825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.613958077806911,
+      "learning_rate": 4.931433609917149e-06,
+      "loss": 0.7893,
+      "step": 28826
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.703432655797391,
+      "learning_rate": 4.930775860110636e-06,
+      "loss": 0.9322,
+      "step": 28827
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6246466545323888,
+      "learning_rate": 4.930118139818742e-06,
+      "loss": 0.8186,
+      "step": 28828
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6757182513433737,
+      "learning_rate": 4.929460449045307e-06,
+      "loss": 0.8194,
+      "step": 28829
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7709476197963295,
+      "learning_rate": 4.928802787794158e-06,
+      "loss": 0.7168,
+      "step": 28830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7047661364311575,
+      "learning_rate": 4.9281451560691325e-06,
+      "loss": 0.8126,
+      "step": 28831
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3899498142409448,
+      "learning_rate": 4.927487553874044e-06,
+      "loss": 0.831,
+      "step": 28832
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.58316901882401,
+      "learning_rate": 4.926829981212728e-06,
+      "loss": 0.8698,
+      "step": 28833
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6632516741313816,
+      "learning_rate": 4.9261724380890195e-06,
+      "loss": 0.8172,
+      "step": 28834
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7847877064039932,
+      "learning_rate": 4.92551492450674e-06,
+      "loss": 0.8646,
+      "step": 28835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5734153035694423,
+      "learning_rate": 4.924857440469716e-06,
+      "loss": 0.7141,
+      "step": 28836
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5763762178111853,
+      "learning_rate": 4.9241999859817794e-06,
+      "loss": 0.8008,
+      "step": 28837
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.872121652553732,
+      "learning_rate": 4.923542561046759e-06,
+      "loss": 1.0186,
+      "step": 28838
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3560784711878369,
+      "learning_rate": 4.9228851656684794e-06,
+      "loss": 0.8715,
+      "step": 28839
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.46128179976524,
+      "learning_rate": 4.922227799850772e-06,
+      "loss": 0.8381,
+      "step": 28840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7278969994424027,
+      "learning_rate": 4.921570463597464e-06,
+      "loss": 0.9937,
+      "step": 28841
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4672504684954533,
+      "learning_rate": 4.920913156912374e-06,
+      "loss": 0.7304,
+      "step": 28842
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3222562829682745,
+      "learning_rate": 4.920255879799338e-06,
+      "loss": 0.8713,
+      "step": 28843
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7375651875422973,
+      "learning_rate": 4.919598632262183e-06,
+      "loss": 0.8352,
+      "step": 28844
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9842823636017983,
+      "learning_rate": 4.918941414304733e-06,
+      "loss": 0.9451,
+      "step": 28845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5923184598187758,
+      "learning_rate": 4.918284225930812e-06,
+      "loss": 0.8569,
+      "step": 28846
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5495177544502525,
+      "learning_rate": 4.917627067144248e-06,
+      "loss": 0.8647,
+      "step": 28847
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6363582460872357,
+      "learning_rate": 4.916969937948874e-06,
+      "loss": 0.8107,
+      "step": 28848
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5537982973583313,
+      "learning_rate": 4.916312838348508e-06,
+      "loss": 0.8058,
+      "step": 28849
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6977471646375821,
+      "learning_rate": 4.915655768346975e-06,
+      "loss": 0.8231,
+      "step": 28850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5006012646316118,
+      "learning_rate": 4.914998727948105e-06,
+      "loss": 0.9653,
+      "step": 28851
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5613401828716738,
+      "learning_rate": 4.9143417171557255e-06,
+      "loss": 0.9548,
+      "step": 28852
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5980124974466525,
+      "learning_rate": 4.913684735973657e-06,
+      "loss": 0.8862,
+      "step": 28853
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5318353138352638,
+      "learning_rate": 4.913027784405724e-06,
+      "loss": 0.8936,
+      "step": 28854
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6485536787414352,
+      "learning_rate": 4.912370862455757e-06,
+      "loss": 0.8392,
+      "step": 28855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5516034118149264,
+      "learning_rate": 4.911713970127574e-06,
+      "loss": 0.877,
+      "step": 28856
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8698347976408172,
+      "learning_rate": 4.911057107425007e-06,
+      "loss": 0.8928,
+      "step": 28857
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.827751166929193,
+      "learning_rate": 4.910400274351873e-06,
+      "loss": 0.7608,
+      "step": 28858
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.447272383937774,
+      "learning_rate": 4.909743470912004e-06,
+      "loss": 0.844,
+      "step": 28859
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6905975981510526,
+      "learning_rate": 4.909086697109215e-06,
+      "loss": 0.7605,
+      "step": 28860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.430981693469883,
+      "learning_rate": 4.90842995294734e-06,
+      "loss": 0.6974,
+      "step": 28861
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4793797807120537,
+      "learning_rate": 4.907773238430198e-06,
+      "loss": 0.809,
+      "step": 28862
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5330432900786857,
+      "learning_rate": 4.907116553561608e-06,
+      "loss": 0.8591,
+      "step": 28863
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.34522003345475,
+      "learning_rate": 4.906459898345397e-06,
+      "loss": 0.7641,
+      "step": 28864
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6931884998333468,
+      "learning_rate": 4.905803272785394e-06,
+      "loss": 0.7888,
+      "step": 28865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0602473646381854,
+      "learning_rate": 4.905146676885417e-06,
+      "loss": 0.8891,
+      "step": 28866
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4822664513142312,
+      "learning_rate": 4.904490110649285e-06,
+      "loss": 0.8267,
+      "step": 28867
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.54968966098971,
+      "learning_rate": 4.903833574080825e-06,
+      "loss": 0.8931,
+      "step": 28868
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7385308119523672,
+      "learning_rate": 4.903177067183863e-06,
+      "loss": 0.8485,
+      "step": 28869
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8745624242432224,
+      "learning_rate": 4.902520589962218e-06,
+      "loss": 0.8306,
+      "step": 28870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7804342341514346,
+      "learning_rate": 4.9018641424197075e-06,
+      "loss": 0.6913,
+      "step": 28871
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6631588864378366,
+      "learning_rate": 4.901207724560158e-06,
+      "loss": 0.829,
+      "step": 28872
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5391979264599494,
+      "learning_rate": 4.900551336387395e-06,
+      "loss": 0.8033,
+      "step": 28873
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5338232545013493,
+      "learning_rate": 4.899894977905239e-06,
+      "loss": 0.8617,
+      "step": 28874
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5606371542348547,
+      "learning_rate": 4.8992386491175025e-06,
+      "loss": 0.9535,
+      "step": 28875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6566686924812406,
+      "learning_rate": 4.898582350028015e-06,
+      "loss": 0.8695,
+      "step": 28876
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.680511341049591,
+      "learning_rate": 4.897926080640599e-06,
+      "loss": 0.8556,
+      "step": 28877
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5793894301787457,
+      "learning_rate": 4.897269840959072e-06,
+      "loss": 0.9623,
+      "step": 28878
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4743070810258638,
+      "learning_rate": 4.896613630987254e-06,
+      "loss": 0.8147,
+      "step": 28879
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.99430747615842,
+      "learning_rate": 4.89595745072897e-06,
+      "loss": 0.8641,
+      "step": 28880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7640740141136325,
+      "learning_rate": 4.8953013001880324e-06,
+      "loss": 0.6889,
+      "step": 28881
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7343710192039166,
+      "learning_rate": 4.894645179368272e-06,
+      "loss": 0.7763,
+      "step": 28882
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6409396001918681,
+      "learning_rate": 4.8939890882735e-06,
+      "loss": 0.8277,
+      "step": 28883
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.719298249417394,
+      "learning_rate": 4.893333026907543e-06,
+      "loss": 0.8663,
+      "step": 28884
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7020369490545069,
+      "learning_rate": 4.8926769952742145e-06,
+      "loss": 0.7638,
+      "step": 28885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5584371890993869,
+      "learning_rate": 4.892020993377341e-06,
+      "loss": 0.908,
+      "step": 28886
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7998240725659383,
+      "learning_rate": 4.891365021220739e-06,
+      "loss": 0.7231,
+      "step": 28887
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7632344795857873,
+      "learning_rate": 4.8907090788082225e-06,
+      "loss": 0.7484,
+      "step": 28888
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7123843233032152,
+      "learning_rate": 4.890053166143617e-06,
+      "loss": 1.0078,
+      "step": 28889
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7092687905734312,
+      "learning_rate": 4.889397283230742e-06,
+      "loss": 0.7739,
+      "step": 28890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8052678208146004,
+      "learning_rate": 4.888741430073414e-06,
+      "loss": 0.6609,
+      "step": 28891
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7313077986866998,
+      "learning_rate": 4.888085606675449e-06,
+      "loss": 0.7125,
+      "step": 28892
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4393093254533307,
+      "learning_rate": 4.8874298130406665e-06,
+      "loss": 0.8952,
+      "step": 28893
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.524757328124971,
+      "learning_rate": 4.886774049172891e-06,
+      "loss": 0.8615,
+      "step": 28894
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3542250814522787,
+      "learning_rate": 4.886118315075935e-06,
+      "loss": 0.7477,
+      "step": 28895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7245646338434788,
+      "learning_rate": 4.885462610753614e-06,
+      "loss": 0.9581,
+      "step": 28896
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.721509258498823,
+      "learning_rate": 4.884806936209749e-06,
+      "loss": 0.8229,
+      "step": 28897
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4502371158057572,
+      "learning_rate": 4.884151291448161e-06,
+      "loss": 0.77,
+      "step": 28898
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5107850807167749,
+      "learning_rate": 4.883495676472665e-06,
+      "loss": 0.9031,
+      "step": 28899
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5738658219891422,
+      "learning_rate": 4.882840091287072e-06,
+      "loss": 0.8305,
+      "step": 28900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5105112534801617,
+      "learning_rate": 4.882184535895208e-06,
+      "loss": 0.8415,
+      "step": 28901
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6139942185363865,
+      "learning_rate": 4.881529010300882e-06,
+      "loss": 0.6673,
+      "step": 28902
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7768799161600813,
+      "learning_rate": 4.880873514507919e-06,
+      "loss": 0.7448,
+      "step": 28903
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.564213302119461,
+      "learning_rate": 4.880218048520127e-06,
+      "loss": 0.8607,
+      "step": 28904
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4873818552951188,
+      "learning_rate": 4.87956261234133e-06,
+      "loss": 0.7043,
+      "step": 28905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.736328539852301,
+      "learning_rate": 4.878907205975336e-06,
+      "loss": 0.7617,
+      "step": 28906
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.715058189267271,
+      "learning_rate": 4.87825182942597e-06,
+      "loss": 0.8857,
+      "step": 28907
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6737305357893135,
+      "learning_rate": 4.877596482697045e-06,
+      "loss": 0.8451,
+      "step": 28908
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6609733369277846,
+      "learning_rate": 4.87694116579237e-06,
+      "loss": 0.8591,
+      "step": 28909
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6718997526648236,
+      "learning_rate": 4.876285878715764e-06,
+      "loss": 0.8543,
+      "step": 28910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7568935197413529,
+      "learning_rate": 4.875630621471045e-06,
+      "loss": 0.7136,
+      "step": 28911
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.657969073090903,
+      "learning_rate": 4.874975394062034e-06,
+      "loss": 0.8889,
+      "step": 28912
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0100118884266878,
+      "learning_rate": 4.874320196492533e-06,
+      "loss": 0.994,
+      "step": 28913
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7402063801426974,
+      "learning_rate": 4.873665028766361e-06,
+      "loss": 0.6872,
+      "step": 28914
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7565384891726261,
+      "learning_rate": 4.873009890887338e-06,
+      "loss": 0.8049,
+      "step": 28915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.545953477681981,
+      "learning_rate": 4.872354782859274e-06,
+      "loss": 0.9,
+      "step": 28916
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6858379750578953,
+      "learning_rate": 4.8716997046859816e-06,
+      "loss": 0.8814,
+      "step": 28917
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4984164130917843,
+      "learning_rate": 4.871044656371276e-06,
+      "loss": 0.8285,
+      "step": 28918
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5428172618102578,
+      "learning_rate": 4.870389637918976e-06,
+      "loss": 0.9292,
+      "step": 28919
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5872450598714214,
+      "learning_rate": 4.869734649332892e-06,
+      "loss": 0.7159,
+      "step": 28920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3881158320348626,
+      "learning_rate": 4.8690796906168334e-06,
+      "loss": 0.6442,
+      "step": 28921
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6851951653417199,
+      "learning_rate": 4.868424761774622e-06,
+      "loss": 0.816,
+      "step": 28922
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6110631341577852,
+      "learning_rate": 4.867769862810061e-06,
+      "loss": 1.0191,
+      "step": 28923
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6824288799572888,
+      "learning_rate": 4.8671149937269694e-06,
+      "loss": 0.7476,
+      "step": 28924
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6745804882162236,
+      "learning_rate": 4.866460154529163e-06,
+      "loss": 0.8235,
+      "step": 28925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.642727834591334,
+      "learning_rate": 4.86580534522045e-06,
+      "loss": 0.8498,
+      "step": 28926
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5342724344371244,
+      "learning_rate": 4.86515056580464e-06,
+      "loss": 0.7622,
+      "step": 28927
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2288333020096154,
+      "learning_rate": 4.864495816285549e-06,
+      "loss": 0.8899,
+      "step": 28928
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7501092939102714,
+      "learning_rate": 4.863841096666998e-06,
+      "loss": 1.0005,
+      "step": 28929
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0875018240831364,
+      "learning_rate": 4.863186406952782e-06,
+      "loss": 0.9435,
+      "step": 28930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5838954598846644,
+      "learning_rate": 4.862531747146722e-06,
+      "loss": 0.7597,
+      "step": 28931
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4197383293674974,
+      "learning_rate": 4.861877117252627e-06,
+      "loss": 0.7899,
+      "step": 28932
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.082233932313073,
+      "learning_rate": 4.861222517274319e-06,
+      "loss": 0.8254,
+      "step": 28933
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4267311504731566,
+      "learning_rate": 4.860567947215592e-06,
+      "loss": 0.8114,
+      "step": 28934
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5106121129681236,
+      "learning_rate": 4.8599134070802656e-06,
+      "loss": 0.8141,
+      "step": 28935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.723899044198669,
+      "learning_rate": 4.859258896872155e-06,
+      "loss": 0.858,
+      "step": 28936
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.101680378907761,
+      "learning_rate": 4.858604416595063e-06,
+      "loss": 0.8644,
+      "step": 28937
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9236423818787616,
+      "learning_rate": 4.857949966252806e-06,
+      "loss": 0.7114,
+      "step": 28938
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5712767398489136,
+      "learning_rate": 4.85729554584919e-06,
+      "loss": 0.7066,
+      "step": 28939
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6585960770845654,
+      "learning_rate": 4.856641155388031e-06,
+      "loss": 0.7667,
+      "step": 28940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6375851729900686,
+      "learning_rate": 4.8559867948731325e-06,
+      "loss": 0.6906,
+      "step": 28941
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6041307708715618,
+      "learning_rate": 4.85533246430831e-06,
+      "loss": 0.9277,
+      "step": 28942
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6288770596858386,
+      "learning_rate": 4.854678163697371e-06,
+      "loss": 0.774,
+      "step": 28943
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.765737862963493,
+      "learning_rate": 4.854023893044121e-06,
+      "loss": 0.8236,
+      "step": 28944
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8321102722036837,
+      "learning_rate": 4.853369652352373e-06,
+      "loss": 0.8714,
+      "step": 28945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4608631665947565,
+      "learning_rate": 4.85271544162594e-06,
+      "loss": 0.7711,
+      "step": 28946
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.748823100071395,
+      "learning_rate": 4.8520612608686265e-06,
+      "loss": 0.6861,
+      "step": 28947
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7729433515296504,
+      "learning_rate": 4.8514071100842395e-06,
+      "loss": 0.6828,
+      "step": 28948
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4981831924692262,
+      "learning_rate": 4.85075298927659e-06,
+      "loss": 0.8064,
+      "step": 28949
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8857490935697978,
+      "learning_rate": 4.850098898449491e-06,
+      "loss": 0.8306,
+      "step": 28950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.575363375846999,
+      "learning_rate": 4.849444837606747e-06,
+      "loss": 0.8691,
+      "step": 28951
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1938090301765634,
+      "learning_rate": 4.848790806752162e-06,
+      "loss": 0.8485,
+      "step": 28952
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7711841158156318,
+      "learning_rate": 4.848136805889546e-06,
+      "loss": 0.841,
+      "step": 28953
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6335191509608242,
+      "learning_rate": 4.8474828350227135e-06,
+      "loss": 0.945,
+      "step": 28954
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5566471973143436,
+      "learning_rate": 4.846828894155468e-06,
+      "loss": 0.8326,
+      "step": 28955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4732715536042704,
+      "learning_rate": 4.8461749832916104e-06,
+      "loss": 0.8868,
+      "step": 28956
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7489416271726476,
+      "learning_rate": 4.845521102434959e-06,
+      "loss": 0.7289,
+      "step": 28957
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6909231868248096,
+      "learning_rate": 4.844867251589311e-06,
+      "loss": 0.7514,
+      "step": 28958
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6048439304067859,
+      "learning_rate": 4.844213430758483e-06,
+      "loss": 0.8637,
+      "step": 28959
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6950195384072277,
+      "learning_rate": 4.8435596399462725e-06,
+      "loss": 0.9248,
+      "step": 28960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.76720711961433,
+      "learning_rate": 4.842905879156493e-06,
+      "loss": 0.7826,
+      "step": 28961
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.610614636148591,
+      "learning_rate": 4.842252148392945e-06,
+      "loss": 0.85,
+      "step": 28962
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.343413823878255,
+      "learning_rate": 4.841598447659443e-06,
+      "loss": 0.7424,
+      "step": 28963
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0048760752476458,
+      "learning_rate": 4.840944776959786e-06,
+      "loss": 0.9513,
+      "step": 28964
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5608839163534887,
+      "learning_rate": 4.840291136297779e-06,
+      "loss": 0.929,
+      "step": 28965
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2502833570483993,
+      "learning_rate": 4.83963752567723e-06,
+      "loss": 0.8144,
+      "step": 28966
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.455302324939961,
+      "learning_rate": 4.838983945101949e-06,
+      "loss": 0.8347,
+      "step": 28967
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5506432079440686,
+      "learning_rate": 4.8383303945757374e-06,
+      "loss": 0.8651,
+      "step": 28968
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5648997824060957,
+      "learning_rate": 4.837676874102397e-06,
+      "loss": 0.9635,
+      "step": 28969
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.809811823217894,
+      "learning_rate": 4.837023383685736e-06,
+      "loss": 0.8329,
+      "step": 28970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5999480073441874,
+      "learning_rate": 4.836369923329564e-06,
+      "loss": 0.7614,
+      "step": 28971
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5801712346950143,
+      "learning_rate": 4.83571649303768e-06,
+      "loss": 0.8309,
+      "step": 28972
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5491044720329599,
+      "learning_rate": 4.835063092813886e-06,
+      "loss": 0.8536,
+      "step": 28973
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.457852290617037,
+      "learning_rate": 4.83440972266199e-06,
+      "loss": 0.7417,
+      "step": 28974
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3933345820970897,
+      "learning_rate": 4.833756382585799e-06,
+      "loss": 0.9112,
+      "step": 28975
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5312908934100589,
+      "learning_rate": 4.8331030725891145e-06,
+      "loss": 0.9328,
+      "step": 28976
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.709204255911185,
+      "learning_rate": 4.832449792675737e-06,
+      "loss": 0.8881,
+      "step": 28977
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7045511657018209,
+      "learning_rate": 4.831796542849476e-06,
+      "loss": 0.8275,
+      "step": 28978
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.400107345121314,
+      "learning_rate": 4.831143323114127e-06,
+      "loss": 0.81,
+      "step": 28979
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.567120747906462,
+      "learning_rate": 4.8304901334735035e-06,
+      "loss": 0.8301,
+      "step": 28980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7159680157306147,
+      "learning_rate": 4.8298369739313975e-06,
+      "loss": 0.7698,
+      "step": 28981
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.449934773208679,
+      "learning_rate": 4.829183844491623e-06,
+      "loss": 0.8785,
+      "step": 28982
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6886886244053396,
+      "learning_rate": 4.828530745157972e-06,
+      "loss": 0.824,
+      "step": 28983
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5239545844401654,
+      "learning_rate": 4.827877675934256e-06,
+      "loss": 0.7515,
+      "step": 28984
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7339914987535217,
+      "learning_rate": 4.827224636824275e-06,
+      "loss": 0.866,
+      "step": 28985
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.512232314763118,
+      "learning_rate": 4.826571627831824e-06,
+      "loss": 0.8838,
+      "step": 28986
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3469713136166126,
+      "learning_rate": 4.825918648960712e-06,
+      "loss": 0.8423,
+      "step": 28987
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5810061436312723,
+      "learning_rate": 4.825265700214744e-06,
+      "loss": 0.8694,
+      "step": 28988
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6316461818757753,
+      "learning_rate": 4.824612781597717e-06,
+      "loss": 0.8668,
+      "step": 28989
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.584129246393822,
+      "learning_rate": 4.823959893113429e-06,
+      "loss": 0.9171,
+      "step": 28990
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6149750075015301,
+      "learning_rate": 4.823307034765686e-06,
+      "loss": 0.8786,
+      "step": 28991
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.569966845722245,
+      "learning_rate": 4.822654206558292e-06,
+      "loss": 0.9632,
+      "step": 28992
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7353576448399245,
+      "learning_rate": 4.822001408495043e-06,
+      "loss": 0.665,
+      "step": 28993
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.725799235476976,
+      "learning_rate": 4.821348640579738e-06,
+      "loss": 0.8584,
+      "step": 28994
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.576713667613023,
+      "learning_rate": 4.820695902816181e-06,
+      "loss": 0.9724,
+      "step": 28995
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.942532947711272,
+      "learning_rate": 4.820043195208176e-06,
+      "loss": 0.881,
+      "step": 28996
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.466872815323988,
+      "learning_rate": 4.819390517759519e-06,
+      "loss": 0.8784,
+      "step": 28997
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7622858175690579,
+      "learning_rate": 4.818737870474007e-06,
+      "loss": 0.764,
+      "step": 28998
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.561577323660888,
+      "learning_rate": 4.818085253355443e-06,
+      "loss": 0.9146,
+      "step": 28999
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.585686146673489,
+      "learning_rate": 4.8174326664076324e-06,
+      "loss": 0.7601,
+      "step": 29000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5528957642197774,
+      "learning_rate": 4.816780109634368e-06,
+      "loss": 0.7596,
+      "step": 29001
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4266563858597603,
+      "learning_rate": 4.816127583039448e-06,
+      "loss": 0.7759,
+      "step": 29002
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7679306625045743,
+      "learning_rate": 4.8154750866266785e-06,
+      "loss": 0.7039,
+      "step": 29003
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6747198523195483,
+      "learning_rate": 4.814822620399849e-06,
+      "loss": 0.8209,
+      "step": 29004
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7724369834300608,
+      "learning_rate": 4.81417018436277e-06,
+      "loss": 0.7795,
+      "step": 29005
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7934148503975926,
+      "learning_rate": 4.8135177785192275e-06,
+      "loss": 0.7565,
+      "step": 29006
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.45280273996441,
+      "learning_rate": 4.812865402873032e-06,
+      "loss": 0.8042,
+      "step": 29007
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7412163204709449,
+      "learning_rate": 4.812213057427972e-06,
+      "loss": 0.7002,
+      "step": 29008
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8031613340275108,
+      "learning_rate": 4.811560742187851e-06,
+      "loss": 0.7698,
+      "step": 29009
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7713626436424181,
+      "learning_rate": 4.810908457156472e-06,
+      "loss": 0.9508,
+      "step": 29010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5629725798068765,
+      "learning_rate": 4.810256202337621e-06,
+      "loss": 0.8202,
+      "step": 29011
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6487322028414262,
+      "learning_rate": 4.8096039777351e-06,
+      "loss": 0.7566,
+      "step": 29012
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8143952848119065,
+      "learning_rate": 4.808951783352709e-06,
+      "loss": 0.902,
+      "step": 29013
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.649249838675814,
+      "learning_rate": 4.808299619194251e-06,
+      "loss": 0.7901,
+      "step": 29014
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6367498345939797,
+      "learning_rate": 4.807647485263508e-06,
+      "loss": 0.9948,
+      "step": 29015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.892971055988037,
+      "learning_rate": 4.8069953815642865e-06,
+      "loss": 0.8368,
+      "step": 29016
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5362787464040408,
+      "learning_rate": 4.806343308100385e-06,
+      "loss": 0.7756,
+      "step": 29017
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5972861382491832,
+      "learning_rate": 4.8056912648755975e-06,
+      "loss": 0.7892,
+      "step": 29018
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5757812489538898,
+      "learning_rate": 4.805039251893716e-06,
+      "loss": 0.9725,
+      "step": 29019
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.492694820045386,
+      "learning_rate": 4.80438726915854e-06,
+      "loss": 0.8984,
+      "step": 29020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7142932754802251,
+      "learning_rate": 4.803735316673871e-06,
+      "loss": 0.9714,
+      "step": 29021
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6135479157204973,
+      "learning_rate": 4.803083394443497e-06,
+      "loss": 0.8742,
+      "step": 29022
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7082007107961872,
+      "learning_rate": 4.802431502471219e-06,
+      "loss": 0.8306,
+      "step": 29023
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.476523617509706,
+      "learning_rate": 4.801779640760831e-06,
+      "loss": 0.8172,
+      "step": 29024
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.106038351490549,
+      "learning_rate": 4.801127809316123e-06,
+      "loss": 0.7609,
+      "step": 29025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6081201486522108,
+      "learning_rate": 4.8004760081408966e-06,
+      "loss": 0.8876,
+      "step": 29026
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5547017451078786,
+      "learning_rate": 4.799824237238947e-06,
+      "loss": 0.9531,
+      "step": 29027
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5461310838258233,
+      "learning_rate": 4.7991724966140674e-06,
+      "loss": 0.9399,
+      "step": 29028
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6573102970603737,
+      "learning_rate": 4.798520786270049e-06,
+      "loss": 0.7961,
+      "step": 29029
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0491134699804774,
+      "learning_rate": 4.79786910621069e-06,
+      "loss": 1.0246,
+      "step": 29030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.555851679948003,
+      "learning_rate": 4.79721745643979e-06,
+      "loss": 0.8308,
+      "step": 29031
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.550219704875522,
+      "learning_rate": 4.79656583696113e-06,
+      "loss": 0.821,
+      "step": 29032
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7259713150331857,
+      "learning_rate": 4.795914247778511e-06,
+      "loss": 0.9999,
+      "step": 29033
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0444745013167003,
+      "learning_rate": 4.795262688895728e-06,
+      "loss": 0.8898,
+      "step": 29034
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8399904554792255,
+      "learning_rate": 4.794611160316576e-06,
+      "loss": 0.9078,
+      "step": 29035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.614994191037303,
+      "learning_rate": 4.793959662044846e-06,
+      "loss": 0.7559,
+      "step": 29036
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.320613492663593,
+      "learning_rate": 4.793308194084327e-06,
+      "loss": 0.8268,
+      "step": 29037
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7046369024955783,
+      "learning_rate": 4.792656756438821e-06,
+      "loss": 0.7852,
+      "step": 29038
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5024422495479777,
+      "learning_rate": 4.7920053491121114e-06,
+      "loss": 0.9687,
+      "step": 29039
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4672117056409202,
+      "learning_rate": 4.791353972107999e-06,
+      "loss": 0.8098,
+      "step": 29040
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6557064986168135,
+      "learning_rate": 4.790702625430269e-06,
+      "loss": 0.8138,
+      "step": 29041
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6474176087876027,
+      "learning_rate": 4.7900513090827225e-06,
+      "loss": 0.9644,
+      "step": 29042
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.336694852750086,
+      "learning_rate": 4.789400023069142e-06,
+      "loss": 0.8388,
+      "step": 29043
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4742742218368454,
+      "learning_rate": 4.788748767393328e-06,
+      "loss": 0.9033,
+      "step": 29044
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5842146692466001,
+      "learning_rate": 4.788097542059069e-06,
+      "loss": 0.7711,
+      "step": 29045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.55244443003714,
+      "learning_rate": 4.787446347070152e-06,
+      "loss": 1.007,
+      "step": 29046
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5674981384727882,
+      "learning_rate": 4.7867951824303725e-06,
+      "loss": 0.869,
+      "step": 29047
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5140052254916991,
+      "learning_rate": 4.786144048143526e-06,
+      "loss": 0.7973,
+      "step": 29048
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6736394693712404,
+      "learning_rate": 4.785492944213399e-06,
+      "loss": 0.9603,
+      "step": 29049
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.468309502264924,
+      "learning_rate": 4.784841870643781e-06,
+      "loss": 0.7932,
+      "step": 29050
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8029167819447138,
+      "learning_rate": 4.784190827438462e-06,
+      "loss": 0.9344,
+      "step": 29051
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4436193106988933,
+      "learning_rate": 4.783539814601242e-06,
+      "loss": 0.9357,
+      "step": 29052
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.730565448753806,
+      "learning_rate": 4.782888832135903e-06,
+      "loss": 0.8989,
+      "step": 29053
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.587920746156155,
+      "learning_rate": 4.782237880046232e-06,
+      "loss": 0.8568,
+      "step": 29054
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8897096379113283,
+      "learning_rate": 4.781586958336025e-06,
+      "loss": 0.8425,
+      "step": 29055
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.706972184696521,
+      "learning_rate": 4.780936067009075e-06,
+      "loss": 0.961,
+      "step": 29056
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5679689930210665,
+      "learning_rate": 4.7802852060691676e-06,
+      "loss": 0.7234,
+      "step": 29057
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5559436438694807,
+      "learning_rate": 4.779634375520088e-06,
+      "loss": 0.8105,
+      "step": 29058
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4510327427262835,
+      "learning_rate": 4.778983575365634e-06,
+      "loss": 0.7664,
+      "step": 29059
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4180898451910535,
+      "learning_rate": 4.778332805609587e-06,
+      "loss": 0.779,
+      "step": 29060
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.738976212173984,
+      "learning_rate": 4.777682066255743e-06,
+      "loss": 0.9859,
+      "step": 29061
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8437642396935412,
+      "learning_rate": 4.777031357307884e-06,
+      "loss": 0.9061,
+      "step": 29062
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5411011117048952,
+      "learning_rate": 4.776380678769805e-06,
+      "loss": 0.8929,
+      "step": 29063
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4602308676441282,
+      "learning_rate": 4.775730030645288e-06,
+      "loss": 0.9097,
+      "step": 29064
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4941563595110041,
+      "learning_rate": 4.7750794129381285e-06,
+      "loss": 0.8821,
+      "step": 29065
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5997777888103089,
+      "learning_rate": 4.774428825652111e-06,
+      "loss": 0.8175,
+      "step": 29066
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8828014843461938,
+      "learning_rate": 4.77377826879102e-06,
+      "loss": 0.9228,
+      "step": 29067
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5687187212076292,
+      "learning_rate": 4.773127742358645e-06,
+      "loss": 0.8998,
+      "step": 29068
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.424061171307659,
+      "learning_rate": 4.772477246358781e-06,
+      "loss": 0.8015,
+      "step": 29069
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8070914322554037,
+      "learning_rate": 4.771826780795208e-06,
+      "loss": 0.914,
+      "step": 29070
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8679786060769041,
+      "learning_rate": 4.771176345671711e-06,
+      "loss": 0.9003,
+      "step": 29071
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5961207308435617,
+      "learning_rate": 4.770525940992081e-06,
+      "loss": 0.7871,
+      "step": 29072
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7063069420299701,
+      "learning_rate": 4.769875566760108e-06,
+      "loss": 0.7565,
+      "step": 29073
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.75138826923353,
+      "learning_rate": 4.769225222979575e-06,
+      "loss": 0.7871,
+      "step": 29074
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6751124854578978,
+      "learning_rate": 4.768574909654267e-06,
+      "loss": 0.8277,
+      "step": 29075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0103075531456827,
+      "learning_rate": 4.76792462678797e-06,
+      "loss": 0.9296,
+      "step": 29076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4596383943946176,
+      "learning_rate": 4.767274374384476e-06,
+      "loss": 0.8488,
+      "step": 29077
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0687195979919974,
+      "learning_rate": 4.766624152447569e-06,
+      "loss": 0.7045,
+      "step": 29078
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.228955838143697,
+      "learning_rate": 4.765973960981029e-06,
+      "loss": 0.7044,
+      "step": 29079
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6004041061303187,
+      "learning_rate": 4.765323799988648e-06,
+      "loss": 0.7418,
+      "step": 29080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6787003899759427,
+      "learning_rate": 4.764673669474206e-06,
+      "loss": 0.8098,
+      "step": 29081
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7669390912561675,
+      "learning_rate": 4.764023569441495e-06,
+      "loss": 0.7364,
+      "step": 29082
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.757756571177986,
+      "learning_rate": 4.763373499894293e-06,
+      "loss": 1.0008,
+      "step": 29083
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.878990114264489,
+      "learning_rate": 4.762723460836392e-06,
+      "loss": 0.8123,
+      "step": 29084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7256924942169515,
+      "learning_rate": 4.762073452271569e-06,
+      "loss": 0.6598,
+      "step": 29085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8998548481655735,
+      "learning_rate": 4.761423474203617e-06,
+      "loss": 0.8683,
+      "step": 29086
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5251660967627283,
+      "learning_rate": 4.760773526636315e-06,
+      "loss": 0.7789,
+      "step": 29087
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7374516843103758,
+      "learning_rate": 4.760123609573446e-06,
+      "loss": 0.821,
+      "step": 29088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7074126371407528,
+      "learning_rate": 4.759473723018796e-06,
+      "loss": 0.8141,
+      "step": 29089
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1002495425446193,
+      "learning_rate": 4.758823866976152e-06,
+      "loss": 0.8421,
+      "step": 29090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5352310760067296,
+      "learning_rate": 4.758174041449295e-06,
+      "loss": 0.8156,
+      "step": 29091
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6131033662586436,
+      "learning_rate": 4.7575242464420045e-06,
+      "loss": 1.0017,
+      "step": 29092
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5002897957690546,
+      "learning_rate": 4.756874481958068e-06,
+      "loss": 0.895,
+      "step": 29093
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.086831330170992,
+      "learning_rate": 4.756224748001272e-06,
+      "loss": 0.9379,
+      "step": 29094
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.594409914074594,
+      "learning_rate": 4.755575044575396e-06,
+      "loss": 0.7782,
+      "step": 29095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7862207978850326,
+      "learning_rate": 4.754925371684217e-06,
+      "loss": 0.7512,
+      "step": 29096
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296413166229243,
+      "learning_rate": 4.754275729331524e-06,
+      "loss": 0.6913,
+      "step": 29097
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.802623222003761,
+      "learning_rate": 4.753626117521103e-06,
+      "loss": 0.9532,
+      "step": 29098
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.527432274985477,
+      "learning_rate": 4.752976536256731e-06,
+      "loss": 0.7443,
+      "step": 29099
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5636056644044232,
+      "learning_rate": 4.752326985542187e-06,
+      "loss": 0.824,
+      "step": 29100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.565343048096619,
+      "learning_rate": 4.75167746538126e-06,
+      "loss": 0.9276,
+      "step": 29101
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4843709567212544,
+      "learning_rate": 4.751027975777726e-06,
+      "loss": 0.9601,
+      "step": 29102
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.663946100265482,
+      "learning_rate": 4.750378516735372e-06,
+      "loss": 0.8698,
+      "step": 29103
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6015615414916053,
+      "learning_rate": 4.7497290882579724e-06,
+      "loss": 0.7409,
+      "step": 29104
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5547688661856685,
+      "learning_rate": 4.749079690349316e-06,
+      "loss": 0.7707,
+      "step": 29105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5965638960475612,
+      "learning_rate": 4.748430323013176e-06,
+      "loss": 0.9398,
+      "step": 29106
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.725935051418735,
+      "learning_rate": 4.747780986253337e-06,
+      "loss": 0.9224,
+      "step": 29107
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5154967170607965,
+      "learning_rate": 4.747131680073589e-06,
+      "loss": 0.8681,
+      "step": 29108
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4582145625012881,
+      "learning_rate": 4.746482404477695e-06,
+      "loss": 0.7526,
+      "step": 29109
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.656591297053954,
+      "learning_rate": 4.745833159469443e-06,
+      "loss": 0.8407,
+      "step": 29110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4958658266534788,
+      "learning_rate": 4.745183945052615e-06,
+      "loss": 0.7809,
+      "step": 29111
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6689120886484252,
+      "learning_rate": 4.744534761230996e-06,
+      "loss": 0.7335,
+      "step": 29112
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4618986035933075,
+      "learning_rate": 4.7438856080083525e-06,
+      "loss": 0.7879,
+      "step": 29113
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6799778769427707,
+      "learning_rate": 4.743236485388471e-06,
+      "loss": 0.7866,
+      "step": 29114
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1533143126691763,
+      "learning_rate": 4.742587393375131e-06,
+      "loss": 0.8876,
+      "step": 29115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5832972606096751,
+      "learning_rate": 4.741938331972119e-06,
+      "loss": 0.7846,
+      "step": 29116
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6970741534151212,
+      "learning_rate": 4.7412893011831995e-06,
+      "loss": 0.8182,
+      "step": 29117
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3475991839527475,
+      "learning_rate": 4.7406403010121595e-06,
+      "loss": 0.7559,
+      "step": 29118
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5909675981625953,
+      "learning_rate": 4.7399913314627785e-06,
+      "loss": 0.9124,
+      "step": 29119
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.985597699780658,
+      "learning_rate": 4.739342392538831e-06,
+      "loss": 0.9968,
+      "step": 29120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5979815841311928,
+      "learning_rate": 4.738693484244101e-06,
+      "loss": 0.816,
+      "step": 29121
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6955765140273706,
+      "learning_rate": 4.738044606582359e-06,
+      "loss": 0.9669,
+      "step": 29122
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7041692829100006,
+      "learning_rate": 4.737395759557392e-06,
+      "loss": 0.87,
+      "step": 29123
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6340338624539392,
+      "learning_rate": 4.736746943172969e-06,
+      "loss": 0.821,
+      "step": 29124
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.485332155888559,
+      "learning_rate": 4.736098157432874e-06,
+      "loss": 0.8676,
+      "step": 29125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6223067977536645,
+      "learning_rate": 4.735449402340882e-06,
+      "loss": 0.9065,
+      "step": 29126
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3155967387779683,
+      "learning_rate": 4.7348006779007685e-06,
+      "loss": 0.658,
+      "step": 29127
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.749981871080214,
+      "learning_rate": 4.734151984116311e-06,
+      "loss": 0.8495,
+      "step": 29128
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.652124759190356,
+      "learning_rate": 4.733503320991292e-06,
+      "loss": 0.8183,
+      "step": 29129
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.585846536886384,
+      "learning_rate": 4.732854688529484e-06,
+      "loss": 0.736,
+      "step": 29130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.415218362993543,
+      "learning_rate": 4.73220608673466e-06,
+      "loss": 0.7282,
+      "step": 29131
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6772727397123652,
+      "learning_rate": 4.7315575156106e-06,
+      "loss": 0.8057,
+      "step": 29132
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4584334324374475,
+      "learning_rate": 4.730908975161083e-06,
+      "loss": 0.8138,
+      "step": 29133
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.354050455100984,
+      "learning_rate": 4.730260465389882e-06,
+      "loss": 0.8029,
+      "step": 29134
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7589247256560843,
+      "learning_rate": 4.729611986300771e-06,
+      "loss": 0.7714,
+      "step": 29135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.513880646713407,
+      "learning_rate": 4.728963537897526e-06,
+      "loss": 0.8859,
+      "step": 29136
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7648245341824564,
+      "learning_rate": 4.728315120183928e-06,
+      "loss": 0.7606,
+      "step": 29137
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7873609643256498,
+      "learning_rate": 4.727666733163748e-06,
+      "loss": 0.8558,
+      "step": 29138
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7122646175921685,
+      "learning_rate": 4.7270183768407585e-06,
+      "loss": 0.8829,
+      "step": 29139
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7423440818345933,
+      "learning_rate": 4.726370051218741e-06,
+      "loss": 0.7166,
+      "step": 29140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5202004611129591,
+      "learning_rate": 4.725721756301463e-06,
+      "loss": 0.9334,
+      "step": 29141
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6630070692699412,
+      "learning_rate": 4.725073492092707e-06,
+      "loss": 0.9498,
+      "step": 29142
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6992211189575508,
+      "learning_rate": 4.724425258596238e-06,
+      "loss": 0.8372,
+      "step": 29143
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6966532193381494,
+      "learning_rate": 4.72377705581584e-06,
+      "loss": 0.8727,
+      "step": 29144
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5681870357175183,
+      "learning_rate": 4.723128883755279e-06,
+      "loss": 0.7439,
+      "step": 29145
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9011948011239974,
+      "learning_rate": 4.722480742418335e-06,
+      "loss": 0.8475,
+      "step": 29146
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.938972980765102,
+      "learning_rate": 4.721832631808781e-06,
+      "loss": 0.8734,
+      "step": 29147
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5714005981690196,
+      "learning_rate": 4.721184551930382e-06,
+      "loss": 0.8911,
+      "step": 29148
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.752525303600208,
+      "learning_rate": 4.720536502786919e-06,
+      "loss": 0.809,
+      "step": 29149
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.964286793709146,
+      "learning_rate": 4.719888484382168e-06,
+      "loss": 0.8279,
+      "step": 29150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.599458351679643,
+      "learning_rate": 4.719240496719897e-06,
+      "loss": 0.822,
+      "step": 29151
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6091412438072177,
+      "learning_rate": 4.718592539803877e-06,
+      "loss": 0.7913,
+      "step": 29152
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6357919304849629,
+      "learning_rate": 4.717944613637883e-06,
+      "loss": 0.8131,
+      "step": 29153
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6595273849518102,
+      "learning_rate": 4.717296718225691e-06,
+      "loss": 0.8964,
+      "step": 29154
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4123349709476134,
+      "learning_rate": 4.7166488535710696e-06,
+      "loss": 0.8831,
+      "step": 29155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6326125961361342,
+      "learning_rate": 4.716001019677788e-06,
+      "loss": 0.8586,
+      "step": 29156
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4784446234942548,
+      "learning_rate": 4.715353216549622e-06,
+      "loss": 0.8602,
+      "step": 29157
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8007935843230142,
+      "learning_rate": 4.714705444190346e-06,
+      "loss": 0.7242,
+      "step": 29158
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6381346523823892,
+      "learning_rate": 4.714057702603728e-06,
+      "loss": 0.9099,
+      "step": 29159
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6429796426765872,
+      "learning_rate": 4.713409991793536e-06,
+      "loss": 0.9083,
+      "step": 29160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5893589140595392,
+      "learning_rate": 4.712762311763549e-06,
+      "loss": 0.8528,
+      "step": 29161
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5344028091925712,
+      "learning_rate": 4.71211466251753e-06,
+      "loss": 0.8146,
+      "step": 29162
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6089194284878081,
+      "learning_rate": 4.711467044059256e-06,
+      "loss": 0.8946,
+      "step": 29163
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6867977863063235,
+      "learning_rate": 4.710819456392492e-06,
+      "loss": 0.8036,
+      "step": 29164
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8923448113958676,
+      "learning_rate": 4.710171899521015e-06,
+      "loss": 0.7665,
+      "step": 29165
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9823372114300153,
+      "learning_rate": 4.709524373448589e-06,
+      "loss": 0.8935,
+      "step": 29166
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6874363102176084,
+      "learning_rate": 4.708876878178991e-06,
+      "loss": 0.9007,
+      "step": 29167
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.306397280630328,
+      "learning_rate": 4.708229413715987e-06,
+      "loss": 0.8129,
+      "step": 29168
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6179856514045423,
+      "learning_rate": 4.707581980063342e-06,
+      "loss": 0.7849,
+      "step": 29169
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5384064272924445,
+      "learning_rate": 4.706934577224831e-06,
+      "loss": 0.9717,
+      "step": 29170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8580417076256724,
+      "learning_rate": 4.706287205204225e-06,
+      "loss": 0.6995,
+      "step": 29171
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9044788889219662,
+      "learning_rate": 4.705639864005292e-06,
+      "loss": 0.9528,
+      "step": 29172
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4675451949570029,
+      "learning_rate": 4.704992553631796e-06,
+      "loss": 0.8388,
+      "step": 29173
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6449180109690182,
+      "learning_rate": 4.704345274087511e-06,
+      "loss": 0.9257,
+      "step": 29174
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6205563249719597,
+      "learning_rate": 4.703698025376207e-06,
+      "loss": 0.8013,
+      "step": 29175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5024731345308673,
+      "learning_rate": 4.703050807501649e-06,
+      "loss": 0.7879,
+      "step": 29176
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6029697428960903,
+      "learning_rate": 4.702403620467604e-06,
+      "loss": 0.8207,
+      "step": 29177
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.388156553628841,
+      "learning_rate": 4.701756464277841e-06,
+      "loss": 0.8767,
+      "step": 29178
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.542689246427221,
+      "learning_rate": 4.701109338936134e-06,
+      "loss": 0.7229,
+      "step": 29179
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6171200469754083,
+      "learning_rate": 4.700462244446246e-06,
+      "loss": 0.8083,
+      "step": 29180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5143783083350364,
+      "learning_rate": 4.699815180811939e-06,
+      "loss": 0.9122,
+      "step": 29181
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.638594448779738,
+      "learning_rate": 4.699168148036992e-06,
+      "loss": 0.8846,
+      "step": 29182
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8154165178396346,
+      "learning_rate": 4.698521146125161e-06,
+      "loss": 0.7763,
+      "step": 29183
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5105555805138124,
+      "learning_rate": 4.697874175080224e-06,
+      "loss": 0.8184,
+      "step": 29184
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6767593892799066,
+      "learning_rate": 4.6972272349059365e-06,
+      "loss": 0.7863,
+      "step": 29185
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.642689270152718,
+      "learning_rate": 4.696580325606075e-06,
+      "loss": 0.8736,
+      "step": 29186
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9007926280461496,
+      "learning_rate": 4.695933447184398e-06,
+      "loss": 0.7753,
+      "step": 29187
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5877017712620085,
+      "learning_rate": 4.695286599644681e-06,
+      "loss": 0.8442,
+      "step": 29188
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.522168134807835,
+      "learning_rate": 4.694639782990683e-06,
+      "loss": 0.7876,
+      "step": 29189
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6496508869187023,
+      "learning_rate": 4.69399299722617e-06,
+      "loss": 0.901,
+      "step": 29190
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8987546808840035,
+      "learning_rate": 4.693346242354908e-06,
+      "loss": 0.8404,
+      "step": 29191
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6891192869185758,
+      "learning_rate": 4.692699518380664e-06,
+      "loss": 0.8975,
+      "step": 29192
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6124023428900989,
+      "learning_rate": 4.692052825307213e-06,
+      "loss": 0.8515,
+      "step": 29193
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9524367883355556,
+      "learning_rate": 4.691406163138302e-06,
+      "loss": 0.8628,
+      "step": 29194
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6442583214159132,
+      "learning_rate": 4.690759531877706e-06,
+      "loss": 0.7202,
+      "step": 29195
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6212093358024704,
+      "learning_rate": 4.690112931529191e-06,
+      "loss": 0.83,
+      "step": 29196
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5726601543244212,
+      "learning_rate": 4.689466362096521e-06,
+      "loss": 0.7593,
+      "step": 29197
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7846842839425463,
+      "learning_rate": 4.688819823583454e-06,
+      "loss": 0.7544,
+      "step": 29198
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7306781783277354,
+      "learning_rate": 4.688173315993761e-06,
+      "loss": 0.8854,
+      "step": 29199
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.457982483739318,
+      "learning_rate": 4.687526839331208e-06,
+      "loss": 0.7356,
+      "step": 29200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.507422452035869,
+      "learning_rate": 4.686880393599556e-06,
+      "loss": 0.8135,
+      "step": 29201
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8512531031887174,
+      "learning_rate": 4.686233978802563e-06,
+      "loss": 0.9115,
+      "step": 29202
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7771246091018622,
+      "learning_rate": 4.685587594944003e-06,
+      "loss": 0.8876,
+      "step": 29203
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.537893921166631,
+      "learning_rate": 4.684941242027631e-06,
+      "loss": 0.8308,
+      "step": 29204
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8227052076306924,
+      "learning_rate": 4.684294920057214e-06,
+      "loss": 0.8938,
+      "step": 29205
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6710740251252556,
+      "learning_rate": 4.683648629036518e-06,
+      "loss": 0.8487,
+      "step": 29206
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7130517303416517,
+      "learning_rate": 4.683002368969303e-06,
+      "loss": 0.9459,
+      "step": 29207
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5154854029879876,
+      "learning_rate": 4.682356139859327e-06,
+      "loss": 0.7537,
+      "step": 29208
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.503669105467212,
+      "learning_rate": 4.681709941710358e-06,
+      "loss": 0.8794,
+      "step": 29209
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.60479581080767,
+      "learning_rate": 4.681063774526166e-06,
+      "loss": 0.7656,
+      "step": 29210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.556345947454919,
+      "learning_rate": 4.680417638310496e-06,
+      "loss": 0.939,
+      "step": 29211
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5857956775276154,
+      "learning_rate": 4.679771533067119e-06,
+      "loss": 0.9223,
+      "step": 29212
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8282957383197433,
+      "learning_rate": 4.679125458799796e-06,
+      "loss": 0.881,
+      "step": 29213
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5696538546847878,
+      "learning_rate": 4.678479415512297e-06,
+      "loss": 0.9022,
+      "step": 29214
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.641392231540471,
+      "learning_rate": 4.677833403208367e-06,
+      "loss": 0.9406,
+      "step": 29215
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.979604806096339,
+      "learning_rate": 4.6771874218917765e-06,
+      "loss": 0.7765,
+      "step": 29216
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5612170087320965,
+      "learning_rate": 4.676541471566291e-06,
+      "loss": 0.8765,
+      "step": 29217
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8095393879735152,
+      "learning_rate": 4.675895552235662e-06,
+      "loss": 0.655,
+      "step": 29218
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7673420763787613,
+      "learning_rate": 4.675249663903658e-06,
+      "loss": 0.7088,
+      "step": 29219
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6925848842749613,
+      "learning_rate": 4.674603806574033e-06,
+      "loss": 0.8786,
+      "step": 29220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7443335839628531,
+      "learning_rate": 4.673957980250553e-06,
+      "loss": 0.6425,
+      "step": 29221
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2534472686401217,
+      "learning_rate": 4.673312184936973e-06,
+      "loss": 0.8084,
+      "step": 29222
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4166988833344465,
+      "learning_rate": 4.672666420637061e-06,
+      "loss": 0.8465,
+      "step": 29223
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8702180180682781,
+      "learning_rate": 4.67202068735457e-06,
+      "loss": 0.8485,
+      "step": 29224
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5098323317262141,
+      "learning_rate": 4.6713749850932576e-06,
+      "loss": 0.7458,
+      "step": 29225
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.573676101569243,
+      "learning_rate": 4.670729313856887e-06,
+      "loss": 0.7681,
+      "step": 29226
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.542032200571671,
+      "learning_rate": 4.670083673649223e-06,
+      "loss": 0.845,
+      "step": 29227
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.543057763347849,
+      "learning_rate": 4.669438064474018e-06,
+      "loss": 0.7605,
+      "step": 29228
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5605011152278423,
+      "learning_rate": 4.668792486335029e-06,
+      "loss": 0.7392,
+      "step": 29229
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4709396501992977,
+      "learning_rate": 4.668146939236017e-06,
+      "loss": 0.7456,
+      "step": 29230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7342430176704051,
+      "learning_rate": 4.667501423180746e-06,
+      "loss": 0.6738,
+      "step": 29231
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8157267103954442,
+      "learning_rate": 4.666855938172969e-06,
+      "loss": 0.7223,
+      "step": 29232
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.636428681835598,
+      "learning_rate": 4.666210484216444e-06,
+      "loss": 0.7048,
+      "step": 29233
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.493650471180936,
+      "learning_rate": 4.665565061314927e-06,
+      "loss": 0.8564,
+      "step": 29234
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6303462379059783,
+      "learning_rate": 4.6649196694721845e-06,
+      "loss": 0.9466,
+      "step": 29235
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4965641352227586,
+      "learning_rate": 4.664274308691968e-06,
+      "loss": 0.8346,
+      "step": 29236
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9095638120278358,
+      "learning_rate": 4.663628978978033e-06,
+      "loss": 0.8625,
+      "step": 29237
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7633219464327131,
+      "learning_rate": 4.662983680334144e-06,
+      "loss": 0.6925,
+      "step": 29238
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6639737636327194,
+      "learning_rate": 4.662338412764048e-06,
+      "loss": 0.7973,
+      "step": 29239
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7853544467117233,
+      "learning_rate": 4.661693176271513e-06,
+      "loss": 0.9544,
+      "step": 29240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6288618586572507,
+      "learning_rate": 4.661047970860285e-06,
+      "loss": 0.8038,
+      "step": 29241
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5949097245008954,
+      "learning_rate": 4.66040279653413e-06,
+      "loss": 0.8119,
+      "step": 29242
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7727490665622896,
+      "learning_rate": 4.659757653296797e-06,
+      "loss": 0.7641,
+      "step": 29243
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8453042493871386,
+      "learning_rate": 4.65911254115205e-06,
+      "loss": 0.7677,
+      "step": 29244
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.708830737496752,
+      "learning_rate": 4.658467460103636e-06,
+      "loss": 0.7738,
+      "step": 29245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5934233295434417,
+      "learning_rate": 4.657822410155321e-06,
+      "loss": 0.8795,
+      "step": 29246
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.693329406866891,
+      "learning_rate": 4.65717739131085e-06,
+      "loss": 0.848,
+      "step": 29247
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5086634732378286,
+      "learning_rate": 4.656532403573987e-06,
+      "loss": 0.9034,
+      "step": 29248
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8477135078958107,
+      "learning_rate": 4.655887446948485e-06,
+      "loss": 0.7873,
+      "step": 29249
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4801646988390849,
+      "learning_rate": 4.655242521438095e-06,
+      "loss": 0.9073,
+      "step": 29250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6879028772372038,
+      "learning_rate": 4.654597627046574e-06,
+      "loss": 0.8464,
+      "step": 29251
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5689092703630663,
+      "learning_rate": 4.653952763777682e-06,
+      "loss": 0.9868,
+      "step": 29252
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4711652945943272,
+      "learning_rate": 4.653307931635169e-06,
+      "loss": 0.7716,
+      "step": 29253
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8270894382504672,
+      "learning_rate": 4.652663130622788e-06,
+      "loss": 0.6542,
+      "step": 29254
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7760580790866721,
+      "learning_rate": 4.652018360744294e-06,
+      "loss": 0.7746,
+      "step": 29255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9829274810033244,
+      "learning_rate": 4.651373622003445e-06,
+      "loss": 0.7711,
+      "step": 29256
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.498007963249774,
+      "learning_rate": 4.650728914403994e-06,
+      "loss": 0.703,
+      "step": 29257
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0815577543406016,
+      "learning_rate": 4.650084237949688e-06,
+      "loss": 0.7637,
+      "step": 29258
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6304654889943504,
+      "learning_rate": 4.649439592644286e-06,
+      "loss": 0.6797,
+      "step": 29259
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4384109044823208,
+      "learning_rate": 4.648794978491543e-06,
+      "loss": 0.7949,
+      "step": 29260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.919443539860783,
+      "learning_rate": 4.6481503954952114e-06,
+      "loss": 0.9358,
+      "step": 29261
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4230133958413227,
+      "learning_rate": 4.647505843659038e-06,
+      "loss": 0.8741,
+      "step": 29262
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9341218996200382,
+      "learning_rate": 4.6468613229867845e-06,
+      "loss": 0.773,
+      "step": 29263
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7436188240022086,
+      "learning_rate": 4.646216833482195e-06,
+      "loss": 0.757,
+      "step": 29264
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.697026570235951,
+      "learning_rate": 4.645572375149029e-06,
+      "loss": 0.8012,
+      "step": 29265
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.429009725139868,
+      "learning_rate": 4.644927947991034e-06,
+      "loss": 0.8401,
+      "step": 29266
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8081503298877082,
+      "learning_rate": 4.644283552011965e-06,
+      "loss": 0.7953,
+      "step": 29267
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8263107545656743,
+      "learning_rate": 4.6436391872155714e-06,
+      "loss": 0.8507,
+      "step": 29268
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.977593204861344,
+      "learning_rate": 4.642994853605609e-06,
+      "loss": 0.8819,
+      "step": 29269
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.684666991517555,
+      "learning_rate": 4.642350551185827e-06,
+      "loss": 0.838,
+      "step": 29270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.620059360046396,
+      "learning_rate": 4.6417062799599725e-06,
+      "loss": 0.8434,
+      "step": 29271
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6952792165048711,
+      "learning_rate": 4.641062039931799e-06,
+      "loss": 0.7707,
+      "step": 29272
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.876343741557211,
+      "learning_rate": 4.640417831105064e-06,
+      "loss": 0.8373,
+      "step": 29273
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.692194021936056,
+      "learning_rate": 4.6397736534835136e-06,
+      "loss": 0.7437,
+      "step": 29274
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7554932561738685,
+      "learning_rate": 4.639129507070893e-06,
+      "loss": 0.7892,
+      "step": 29275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6383177386284384,
+      "learning_rate": 4.638485391870958e-06,
+      "loss": 0.781,
+      "step": 29276
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6733661252612158,
+      "learning_rate": 4.637841307887463e-06,
+      "loss": 0.7869,
+      "step": 29277
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7868526384089756,
+      "learning_rate": 4.637197255124153e-06,
+      "loss": 0.913,
+      "step": 29278
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5770755097340055,
+      "learning_rate": 4.636553233584775e-06,
+      "loss": 0.87,
+      "step": 29279
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5363228853714916,
+      "learning_rate": 4.635909243273082e-06,
+      "loss": 0.8488,
+      "step": 29280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0055759071858152,
+      "learning_rate": 4.635265284192827e-06,
+      "loss": 0.7848,
+      "step": 29281
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7353417864223436,
+      "learning_rate": 4.634621356347756e-06,
+      "loss": 0.7355,
+      "step": 29282
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.011165953638918,
+      "learning_rate": 4.633977459741614e-06,
+      "loss": 0.926,
+      "step": 29283
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8315809074841478,
+      "learning_rate": 4.6333335943781586e-06,
+      "loss": 0.7596,
+      "step": 29284
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6146968776412354,
+      "learning_rate": 4.632689760261131e-06,
+      "loss": 0.8368,
+      "step": 29285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7123404152738164,
+      "learning_rate": 4.632045957394286e-06,
+      "loss": 0.7962,
+      "step": 29286
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6675509834634459,
+      "learning_rate": 4.631402185781364e-06,
+      "loss": 0.8873,
+      "step": 29287
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5918150272168115,
+      "learning_rate": 4.630758445426123e-06,
+      "loss": 0.8982,
+      "step": 29288
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.515068717947907,
+      "learning_rate": 4.630114736332302e-06,
+      "loss": 0.8243,
+      "step": 29289
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.515931005950349,
+      "learning_rate": 4.6294710585036536e-06,
+      "loss": 0.9333,
+      "step": 29290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.623740808612666,
+      "learning_rate": 4.628827411943933e-06,
+      "loss": 0.7909,
+      "step": 29291
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4124862899659858,
+      "learning_rate": 4.628183796656871e-06,
+      "loss": 0.8362,
+      "step": 29292
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.686103650814489,
+      "learning_rate": 4.627540212646223e-06,
+      "loss": 0.8351,
+      "step": 29293
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6947747103677289,
+      "learning_rate": 4.626896659915739e-06,
+      "loss": 0.7877,
+      "step": 29294
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6702961017693976,
+      "learning_rate": 4.626253138469169e-06,
+      "loss": 0.8921,
+      "step": 29295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8808726212036568,
+      "learning_rate": 4.625609648310248e-06,
+      "loss": 0.7764,
+      "step": 29296
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5660587977291578,
+      "learning_rate": 4.62496618944273e-06,
+      "loss": 0.8631,
+      "step": 29297
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7510595931590704,
+      "learning_rate": 4.624322761870363e-06,
+      "loss": 0.8859,
+      "step": 29298
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6125108006171318,
+      "learning_rate": 4.623679365596891e-06,
+      "loss": 0.9155,
+      "step": 29299
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.577970595152831,
+      "learning_rate": 4.623036000626056e-06,
+      "loss": 0.8048,
+      "step": 29300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4314580288785557,
+      "learning_rate": 4.62239266696161e-06,
+      "loss": 0.7927,
+      "step": 29301
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4653060343572433,
+      "learning_rate": 4.621749364607297e-06,
+      "loss": 0.847,
+      "step": 29302
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.577562439726002,
+      "learning_rate": 4.6211060935668595e-06,
+      "loss": 0.8453,
+      "step": 29303
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5581196793029521,
+      "learning_rate": 4.620462853844049e-06,
+      "loss": 0.8602,
+      "step": 29304
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6436028328627803,
+      "learning_rate": 4.619819645442607e-06,
+      "loss": 0.7489,
+      "step": 29305
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7780771821796189,
+      "learning_rate": 4.619176468366274e-06,
+      "loss": 0.7666,
+      "step": 29306
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8170978731778379,
+      "learning_rate": 4.6185333226188e-06,
+      "loss": 0.748,
+      "step": 29307
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6401549989259945,
+      "learning_rate": 4.617890208203932e-06,
+      "loss": 0.9041,
+      "step": 29308
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.767905929963206,
+      "learning_rate": 4.617247125125411e-06,
+      "loss": 0.9071,
+      "step": 29309
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4390709443985554,
+      "learning_rate": 4.616604073386978e-06,
+      "loss": 0.8005,
+      "step": 29310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.007130891235362,
+      "learning_rate": 4.6159610529923805e-06,
+      "loss": 0.8127,
+      "step": 29311
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.783057529684441,
+      "learning_rate": 4.61531806394537e-06,
+      "loss": 0.7449,
+      "step": 29312
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6282268818295276,
+      "learning_rate": 4.614675106249674e-06,
+      "loss": 0.8159,
+      "step": 29313
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.515026102705857,
+      "learning_rate": 4.614032179909045e-06,
+      "loss": 0.782,
+      "step": 29314
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.533209406652308,
+      "learning_rate": 4.613389284927225e-06,
+      "loss": 0.7545,
+      "step": 29315
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.563383168204524,
+      "learning_rate": 4.612746421307963e-06,
+      "loss": 0.8245,
+      "step": 29316
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5494574143540125,
+      "learning_rate": 4.612103589054997e-06,
+      "loss": 0.7425,
+      "step": 29317
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.785176475471089,
+      "learning_rate": 4.611460788172065e-06,
+      "loss": 0.6841,
+      "step": 29318
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.417303631948019,
+      "learning_rate": 4.610818018662919e-06,
+      "loss": 0.7772,
+      "step": 29319
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2046468044074485,
+      "learning_rate": 4.6101752805312926e-06,
+      "loss": 0.8326,
+      "step": 29320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6493601648339793,
+      "learning_rate": 4.609532573780936e-06,
+      "loss": 0.8409,
+      "step": 29321
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.520145383803483,
+      "learning_rate": 4.608889898415583e-06,
+      "loss": 0.9209,
+      "step": 29322
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6902891927735533,
+      "learning_rate": 4.608247254438983e-06,
+      "loss": 0.7721,
+      "step": 29323
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5225767595813906,
+      "learning_rate": 4.607604641854872e-06,
+      "loss": 0.7864,
+      "step": 29324
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.792633739116153,
+      "learning_rate": 4.606962060666997e-06,
+      "loss": 0.702,
+      "step": 29325
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6162605649878956,
+      "learning_rate": 4.606319510879097e-06,
+      "loss": 0.8459,
+      "step": 29326
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7956143177788668,
+      "learning_rate": 4.605676992494909e-06,
+      "loss": 0.6792,
+      "step": 29327
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4978449158958467,
+      "learning_rate": 4.605034505518176e-06,
+      "loss": 1.0047,
+      "step": 29328
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5329271750993287,
+      "learning_rate": 4.604392049952645e-06,
+      "loss": 0.8308,
+      "step": 29329
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5785199285269238,
+      "learning_rate": 4.603749625802051e-06,
+      "loss": 0.8751,
+      "step": 29330
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.793827447589726,
+      "learning_rate": 4.603107233070131e-06,
+      "loss": 0.7258,
+      "step": 29331
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.457924833237359,
+      "learning_rate": 4.60246487176063e-06,
+      "loss": 1.0186,
+      "step": 29332
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7301910799064575,
+      "learning_rate": 4.601822541877291e-06,
+      "loss": 0.7347,
+      "step": 29333
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.837090522712079,
+      "learning_rate": 4.601180243423851e-06,
+      "loss": 0.9182,
+      "step": 29334
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4297249835831782,
+      "learning_rate": 4.600537976404044e-06,
+      "loss": 0.8103,
+      "step": 29335
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.711346043684955,
+      "learning_rate": 4.5998957408216135e-06,
+      "loss": 0.698,
+      "step": 29336
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8400714810934014,
+      "learning_rate": 4.599253536680305e-06,
+      "loss": 0.795,
+      "step": 29337
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4556536286659396,
+      "learning_rate": 4.598611363983853e-06,
+      "loss": 0.8789,
+      "step": 29338
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.657573578321879,
+      "learning_rate": 4.59796922273599e-06,
+      "loss": 0.9073,
+      "step": 29339
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4949788691660262,
+      "learning_rate": 4.597327112940465e-06,
+      "loss": 0.7789,
+      "step": 29340
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5389965570157125,
+      "learning_rate": 4.596685034601008e-06,
+      "loss": 0.8829,
+      "step": 29341
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4459852077569395,
+      "learning_rate": 4.596042987721365e-06,
+      "loss": 0.9726,
+      "step": 29342
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7759739176420464,
+      "learning_rate": 4.595400972305267e-06,
+      "loss": 0.8617,
+      "step": 29343
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5342883135937782,
+      "learning_rate": 4.594758988356459e-06,
+      "loss": 0.8646,
+      "step": 29344
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7585779543711786,
+      "learning_rate": 4.594117035878672e-06,
+      "loss": 0.8577,
+      "step": 29345
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7646064180341328,
+      "learning_rate": 4.593475114875652e-06,
+      "loss": 0.8897,
+      "step": 29346
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7027416528182302,
+      "learning_rate": 4.592833225351131e-06,
+      "loss": 0.899,
+      "step": 29347
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.551537368925093,
+      "learning_rate": 4.592191367308842e-06,
+      "loss": 0.9462,
+      "step": 29348
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3336862862290926,
+      "learning_rate": 4.5915495407525275e-06,
+      "loss": 0.9611,
+      "step": 29349
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.659509877953551,
+      "learning_rate": 4.590907745685927e-06,
+      "loss": 0.9186,
+      "step": 29350
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8131014141905732,
+      "learning_rate": 4.590265982112776e-06,
+      "loss": 0.851,
+      "step": 29351
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0348576056716188,
+      "learning_rate": 4.589624250036804e-06,
+      "loss": 0.8602,
+      "step": 29352
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4340752762236657,
+      "learning_rate": 4.588982549461752e-06,
+      "loss": 0.7348,
+      "step": 29353
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7575634432889633,
+      "learning_rate": 4.588340880391362e-06,
+      "loss": 0.8728,
+      "step": 29354
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.799792146664475,
+      "learning_rate": 4.587699242829364e-06,
+      "loss": 0.7933,
+      "step": 29355
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.885003468566335,
+      "learning_rate": 4.58705763677949e-06,
+      "loss": 0.9164,
+      "step": 29356
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7060452483943163,
+      "learning_rate": 4.586416062245481e-06,
+      "loss": 0.9646,
+      "step": 29357
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.485760807514701,
+      "learning_rate": 4.5857745192310755e-06,
+      "loss": 0.7561,
+      "step": 29358
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4840880856262149,
+      "learning_rate": 4.585133007740005e-06,
+      "loss": 0.9057,
+      "step": 29359
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6670153674874213,
+      "learning_rate": 4.584491527775999e-06,
+      "loss": 0.8196,
+      "step": 29360
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.918551257492645,
+      "learning_rate": 4.583850079342803e-06,
+      "loss": 0.7999,
+      "step": 29361
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4991820968621465,
+      "learning_rate": 4.583208662444143e-06,
+      "loss": 0.8099,
+      "step": 29362
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6999499574566794,
+      "learning_rate": 4.58256727708376e-06,
+      "loss": 0.9248,
+      "step": 29363
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6725547639834473,
+      "learning_rate": 4.5819259232653824e-06,
+      "loss": 0.8458,
+      "step": 29364
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.882933255648038,
+      "learning_rate": 4.581284600992751e-06,
+      "loss": 0.7142,
+      "step": 29365
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6602040498527884,
+      "learning_rate": 4.580643310269592e-06,
+      "loss": 0.9163,
+      "step": 29366
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7863399181440953,
+      "learning_rate": 4.580002051099648e-06,
+      "loss": 0.7792,
+      "step": 29367
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5160102399572775,
+      "learning_rate": 4.579360823486644e-06,
+      "loss": 0.8475,
+      "step": 29368
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8400788179801264,
+      "learning_rate": 4.57871962743432e-06,
+      "loss": 0.8957,
+      "step": 29369
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4713664850995638,
+      "learning_rate": 4.578078462946404e-06,
+      "loss": 0.7701,
+      "step": 29370
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6133971502436186,
+      "learning_rate": 4.577437330026635e-06,
+      "loss": 0.7898,
+      "step": 29371
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7455815560697054,
+      "learning_rate": 4.576796228678744e-06,
+      "loss": 0.6557,
+      "step": 29372
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.590256244105654,
+      "learning_rate": 4.576155158906457e-06,
+      "loss": 1.0042,
+      "step": 29373
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9078260370400026,
+      "learning_rate": 4.575514120713513e-06,
+      "loss": 0.7512,
+      "step": 29374
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7810941814960158,
+      "learning_rate": 4.574873114103645e-06,
+      "loss": 0.6929,
+      "step": 29375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8288050419539248,
+      "learning_rate": 4.574232139080586e-06,
+      "loss": 0.9136,
+      "step": 29376
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5391873461875056,
+      "learning_rate": 4.573591195648059e-06,
+      "loss": 0.7764,
+      "step": 29377
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.759363483233666,
+      "learning_rate": 4.572950283809803e-06,
+      "loss": 0.7225,
+      "step": 29378
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6482663084601448,
+      "learning_rate": 4.572309403569552e-06,
+      "loss": 0.7983,
+      "step": 29379
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5863747233171392,
+      "learning_rate": 4.571668554931035e-06,
+      "loss": 0.7685,
+      "step": 29380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.660503038148919,
+      "learning_rate": 4.571027737897979e-06,
+      "loss": 0.8143,
+      "step": 29381
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8101157262237633,
+      "learning_rate": 4.570386952474116e-06,
+      "loss": 0.8843,
+      "step": 29382
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7828768344753596,
+      "learning_rate": 4.569746198663184e-06,
+      "loss": 0.7664,
+      "step": 29383
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8552253117511692,
+      "learning_rate": 4.569105476468909e-06,
+      "loss": 0.7871,
+      "step": 29384
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9013600298779136,
+      "learning_rate": 4.568464785895018e-06,
+      "loss": 0.8044,
+      "step": 29385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7521769996628194,
+      "learning_rate": 4.567824126945248e-06,
+      "loss": 0.8058,
+      "step": 29386
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8491563191614508,
+      "learning_rate": 4.567183499623322e-06,
+      "loss": 0.7878,
+      "step": 29387
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7303516144320514,
+      "learning_rate": 4.5665429039329735e-06,
+      "loss": 0.8084,
+      "step": 29388
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6917223216739297,
+      "learning_rate": 4.565902339877936e-06,
+      "loss": 0.7338,
+      "step": 29389
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.834657181490171,
+      "learning_rate": 4.5652618074619355e-06,
+      "loss": 0.7656,
+      "step": 29390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.507727839744958,
+      "learning_rate": 4.564621306688698e-06,
+      "loss": 0.9383,
+      "step": 29391
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8708879115514645,
+      "learning_rate": 4.563980837561955e-06,
+      "loss": 0.8253,
+      "step": 29392
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6661688985955403,
+      "learning_rate": 4.563340400085445e-06,
+      "loss": 0.8159,
+      "step": 29393
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8657139429072638,
+      "learning_rate": 4.562699994262881e-06,
+      "loss": 0.8172,
+      "step": 29394
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5457704604655902,
+      "learning_rate": 4.5620596200979996e-06,
+      "loss": 0.91,
+      "step": 29395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5328031171870373,
+      "learning_rate": 4.561419277594528e-06,
+      "loss": 0.964,
+      "step": 29396
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6958796285016828,
+      "learning_rate": 4.560778966756203e-06,
+      "loss": 0.7702,
+      "step": 29397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6041992627022112,
+      "learning_rate": 4.560138687586737e-06,
+      "loss": 0.9095,
+      "step": 29398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.570184852873709,
+      "learning_rate": 4.5594984400898655e-06,
+      "loss": 0.8542,
+      "step": 29399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.370824016164841,
+      "learning_rate": 4.5588582242693215e-06,
+      "loss": 0.8327,
+      "step": 29400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7782407689281094,
+      "learning_rate": 4.558218040128822e-06,
+      "loss": 0.817,
+      "step": 29401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7565466660616285,
+      "learning_rate": 4.5575778876721055e-06,
+      "loss": 0.673,
+      "step": 29402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6215973905099,
+      "learning_rate": 4.556937766902888e-06,
+      "loss": 0.8209,
+      "step": 29403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.772564967441748,
+      "learning_rate": 4.556297677824907e-06,
+      "loss": 0.7824,
+      "step": 29404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.564331965477835,
+      "learning_rate": 4.555657620441881e-06,
+      "loss": 0.8103,
+      "step": 29405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.528975158375195,
+      "learning_rate": 4.555017594757543e-06,
+      "loss": 0.9438,
+      "step": 29406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7016441657965153,
+      "learning_rate": 4.554377600775617e-06,
+      "loss": 0.924,
+      "step": 29407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7024224417304745,
+      "learning_rate": 4.553737638499824e-06,
+      "loss": 0.8909,
+      "step": 29408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7957543823794154,
+      "learning_rate": 4.553097707933894e-06,
+      "loss": 0.7277,
+      "step": 29409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8333008827337869,
+      "learning_rate": 4.55245780908156e-06,
+      "loss": 0.8474,
+      "step": 29410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5343044699471085,
+      "learning_rate": 4.551817941946539e-06,
+      "loss": 0.8958,
+      "step": 29411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4852304069840414,
+      "learning_rate": 4.5511781065325564e-06,
+      "loss": 0.8949,
+      "step": 29412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6923000753269166,
+      "learning_rate": 4.550538302843339e-06,
+      "loss": 0.8492,
+      "step": 29413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6252217454915932,
+      "learning_rate": 4.5498985308826174e-06,
+      "loss": 0.8606,
+      "step": 29414
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6748610003960758,
+      "learning_rate": 4.549258790654112e-06,
+      "loss": 0.953,
+      "step": 29415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.77705505764265,
+      "learning_rate": 4.548619082161543e-06,
+      "loss": 0.7219,
+      "step": 29416
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4916656978332432,
+      "learning_rate": 4.54797940540864e-06,
+      "loss": 0.8258,
+      "step": 29417
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6697112956448195,
+      "learning_rate": 4.547339760399131e-06,
+      "loss": 0.8099,
+      "step": 29418
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5414669055571668,
+      "learning_rate": 4.5467001471367355e-06,
+      "loss": 0.7843,
+      "step": 29419
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5612785480093334,
+      "learning_rate": 4.546060565625175e-06,
+      "loss": 0.806,
+      "step": 29420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9680715981608654,
+      "learning_rate": 4.545421015868181e-06,
+      "loss": 0.7361,
+      "step": 29421
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.688584353662975,
+      "learning_rate": 4.5447814978694685e-06,
+      "loss": 0.7316,
+      "step": 29422
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9688433447513853,
+      "learning_rate": 4.544142011632769e-06,
+      "loss": 0.8803,
+      "step": 29423
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4741157818787312,
+      "learning_rate": 4.543502557161797e-06,
+      "loss": 0.783,
+      "step": 29424
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7798147746798889,
+      "learning_rate": 4.542863134460286e-06,
+      "loss": 0.7306,
+      "step": 29425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7065365988349184,
+      "learning_rate": 4.542223743531949e-06,
+      "loss": 0.7763,
+      "step": 29426
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9517474780090152,
+      "learning_rate": 4.541584384380518e-06,
+      "loss": 0.7418,
+      "step": 29427
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7693626835718619,
+      "learning_rate": 4.540945057009709e-06,
+      "loss": 0.7494,
+      "step": 29428
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7619766726564414,
+      "learning_rate": 4.540305761423244e-06,
+      "loss": 0.7269,
+      "step": 29429
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4972504887344813,
+      "learning_rate": 4.5396664976248476e-06,
+      "loss": 0.8798,
+      "step": 29430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5012245341216566,
+      "learning_rate": 4.539027265618244e-06,
+      "loss": 0.8145,
+      "step": 29431
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4485821390460074,
+      "learning_rate": 4.5383880654071525e-06,
+      "loss": 0.893,
+      "step": 29432
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5495481617935811,
+      "learning_rate": 4.537748896995292e-06,
+      "loss": 0.8324,
+      "step": 29433
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4773182867897852,
+      "learning_rate": 4.537109760386387e-06,
+      "loss": 0.7835,
+      "step": 29434
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6887925902942413,
+      "learning_rate": 4.536470655584161e-06,
+      "loss": 0.8667,
+      "step": 29435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8085424060950565,
+      "learning_rate": 4.535831582592333e-06,
+      "loss": 0.6895,
+      "step": 29436
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.862161838536848,
+      "learning_rate": 4.53519254141462e-06,
+      "loss": 0.7666,
+      "step": 29437
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7712422722134139,
+      "learning_rate": 4.5345535320547465e-06,
+      "loss": 0.704,
+      "step": 29438
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4211966389293875,
+      "learning_rate": 4.533914554516435e-06,
+      "loss": 0.7229,
+      "step": 29439
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6382141490977562,
+      "learning_rate": 4.533275608803405e-06,
+      "loss": 1.0265,
+      "step": 29440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5658068415462434,
+      "learning_rate": 4.532636694919371e-06,
+      "loss": 0.917,
+      "step": 29441
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4702907160306031,
+      "learning_rate": 4.531997812868061e-06,
+      "loss": 0.8659,
+      "step": 29442
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5898624455203012,
+      "learning_rate": 4.5313589626531864e-06,
+      "loss": 0.9921,
+      "step": 29443
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.609058153359529,
+      "learning_rate": 4.5307201442784755e-06,
+      "loss": 0.7345,
+      "step": 29444
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7488021887365093,
+      "learning_rate": 4.53008135774764e-06,
+      "loss": 0.8109,
+      "step": 29445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.064348734764207,
+      "learning_rate": 4.529442603064406e-06,
+      "loss": 0.906,
+      "step": 29446
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7212779677924286,
+      "learning_rate": 4.528803880232485e-06,
+      "loss": 0.7763,
+      "step": 29447
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8594067245832846,
+      "learning_rate": 4.528165189255605e-06,
+      "loss": 0.8286,
+      "step": 29448
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5228825086648654,
+      "learning_rate": 4.527526530137478e-06,
+      "loss": 0.8858,
+      "step": 29449
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6564417166802154,
+      "learning_rate": 4.526887902881822e-06,
+      "loss": 0.7708,
+      "step": 29450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.672023067320013,
+      "learning_rate": 4.526249307492355e-06,
+      "loss": 0.9237,
+      "step": 29451
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6794084788668155,
+      "learning_rate": 4.525610743972804e-06,
+      "loss": 0.8184,
+      "step": 29452
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7078766982929918,
+      "learning_rate": 4.524972212326878e-06,
+      "loss": 0.8735,
+      "step": 29453
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6931032758962907,
+      "learning_rate": 4.524333712558294e-06,
+      "loss": 0.7025,
+      "step": 29454
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.456613371703249,
+      "learning_rate": 4.5236952446707735e-06,
+      "loss": 0.9501,
+      "step": 29455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7072738869303263,
+      "learning_rate": 4.523056808668037e-06,
+      "loss": 0.9636,
+      "step": 29456
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6814942956514174,
+      "learning_rate": 4.5224184045537964e-06,
+      "loss": 0.9414,
+      "step": 29457
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5192588029296752,
+      "learning_rate": 4.521780032331768e-06,
+      "loss": 0.8982,
+      "step": 29458
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8879773305757044,
+      "learning_rate": 4.521141692005669e-06,
+      "loss": 0.8336,
+      "step": 29459
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7024204202500857,
+      "learning_rate": 4.5205033835792226e-06,
+      "loss": 0.8767,
+      "step": 29460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5891766377597507,
+      "learning_rate": 4.519865107056139e-06,
+      "loss": 0.807,
+      "step": 29461
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4770792765863714,
+      "learning_rate": 4.519226862440134e-06,
+      "loss": 0.7928,
+      "step": 29462
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.822729420791259,
+      "learning_rate": 4.518588649734927e-06,
+      "loss": 0.8286,
+      "step": 29463
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5055252746603915,
+      "learning_rate": 4.51795046894423e-06,
+      "loss": 0.7585,
+      "step": 29464
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8155932297965651,
+      "learning_rate": 4.517312320071764e-06,
+      "loss": 0.7233,
+      "step": 29465
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7389498333504858,
+      "learning_rate": 4.5166742031212375e-06,
+      "loss": 0.876,
+      "step": 29466
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6091564753863128,
+      "learning_rate": 4.516036118096374e-06,
+      "loss": 0.8561,
+      "step": 29467
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6564254257203856,
+      "learning_rate": 4.515398065000881e-06,
+      "loss": 0.8293,
+      "step": 29468
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2405895175727997,
+      "learning_rate": 4.51476004383848e-06,
+      "loss": 0.9026,
+      "step": 29469
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.076915718499561,
+      "learning_rate": 4.514122054612883e-06,
+      "loss": 0.8567,
+      "step": 29470
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6689665350365432,
+      "learning_rate": 4.5134840973278e-06,
+      "loss": 1.0107,
+      "step": 29471
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4742223734091255,
+      "learning_rate": 4.51284617198695e-06,
+      "loss": 0.712,
+      "step": 29472
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8242620936016596,
+      "learning_rate": 4.51220827859405e-06,
+      "loss": 0.6981,
+      "step": 29473
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.530663043150715,
+      "learning_rate": 4.5115704171528105e-06,
+      "loss": 0.7728,
+      "step": 29474
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5389079138117423,
+      "learning_rate": 4.510932587666943e-06,
+      "loss": 1.0136,
+      "step": 29475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2071914139481454,
+      "learning_rate": 4.510294790140164e-06,
+      "loss": 0.8608,
+      "step": 29476
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4600262560115203,
+      "learning_rate": 4.509657024576189e-06,
+      "loss": 0.9511,
+      "step": 29477
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.371995885661374,
+      "learning_rate": 4.50901929097873e-06,
+      "loss": 0.794,
+      "step": 29478
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6935511151226774,
+      "learning_rate": 4.508381589351496e-06,
+      "loss": 0.7773,
+      "step": 29479
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.725584290478168,
+      "learning_rate": 4.507743919698202e-06,
+      "loss": 0.8177,
+      "step": 29480
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5180111920198096,
+      "learning_rate": 4.507106282022565e-06,
+      "loss": 0.8578,
+      "step": 29481
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4725148994178456,
+      "learning_rate": 4.506468676328295e-06,
+      "loss": 0.8307,
+      "step": 29482
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6803562484290409,
+      "learning_rate": 4.5058311026191e-06,
+      "loss": 0.7505,
+      "step": 29483
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5636246903556517,
+      "learning_rate": 4.5051935608986994e-06,
+      "loss": 1.0156,
+      "step": 29484
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.526951950677498,
+      "learning_rate": 4.504556051170797e-06,
+      "loss": 0.8828,
+      "step": 29485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6563371116898442,
+      "learning_rate": 4.50391857343911e-06,
+      "loss": 0.8686,
+      "step": 29486
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.810746655812847,
+      "learning_rate": 4.5032811277073525e-06,
+      "loss": 0.6737,
+      "step": 29487
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8613756673841497,
+      "learning_rate": 4.502643713979232e-06,
+      "loss": 0.9318,
+      "step": 29488
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.89408805414091,
+      "learning_rate": 4.5020063322584585e-06,
+      "loss": 0.9317,
+      "step": 29489
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8987403072744733,
+      "learning_rate": 4.501368982548743e-06,
+      "loss": 0.8563,
+      "step": 29490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8867242346528648,
+      "learning_rate": 4.500731664853801e-06,
+      "loss": 0.8443,
+      "step": 29491
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6685890989438752,
+      "learning_rate": 4.500094379177342e-06,
+      "loss": 0.8657,
+      "step": 29492
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6505390785393323,
+      "learning_rate": 4.499457125523071e-06,
+      "loss": 1.0049,
+      "step": 29493
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6865599856057012,
+      "learning_rate": 4.498819903894701e-06,
+      "loss": 0.742,
+      "step": 29494
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8676646872733242,
+      "learning_rate": 4.498182714295951e-06,
+      "loss": 0.7392,
+      "step": 29495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7223030362141356,
+      "learning_rate": 4.497545556730516e-06,
+      "loss": 0.9668,
+      "step": 29496
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.478172071864327,
+      "learning_rate": 4.496908431202113e-06,
+      "loss": 0.8235,
+      "step": 29497
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7316116200290128,
+      "learning_rate": 4.49627133771445e-06,
+      "loss": 0.7874,
+      "step": 29498
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.091353313705817,
+      "learning_rate": 4.495634276271244e-06,
+      "loss": 0.8826,
+      "step": 29499
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.377191470701526,
+      "learning_rate": 4.494997246876196e-06,
+      "loss": 0.7569,
+      "step": 29500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.922981186366824,
+      "learning_rate": 4.494360249533013e-06,
+      "loss": 0.9177,
+      "step": 29501
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.256175493538634,
+      "learning_rate": 4.493723284245413e-06,
+      "loss": 0.9694,
+      "step": 29502
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8598114862832482,
+      "learning_rate": 4.493086351017093e-06,
+      "loss": 0.9199,
+      "step": 29503
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7863465741177729,
+      "learning_rate": 4.4924494498517735e-06,
+      "loss": 0.7087,
+      "step": 29504
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7348438953459691,
+      "learning_rate": 4.491812580753152e-06,
+      "loss": 0.703,
+      "step": 29505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.716545789567592,
+      "learning_rate": 4.491175743724947e-06,
+      "loss": 0.7196,
+      "step": 29506
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5185053183550867,
+      "learning_rate": 4.490538938770855e-06,
+      "loss": 0.9619,
+      "step": 29507
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3856097547728237,
+      "learning_rate": 4.489902165894595e-06,
+      "loss": 0.8209,
+      "step": 29508
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6499728818608552,
+      "learning_rate": 4.489265425099868e-06,
+      "loss": 0.8026,
+      "step": 29509
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.626838070694095,
+      "learning_rate": 4.48862871639038e-06,
+      "loss": 0.9102,
+      "step": 29510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5444501913590194,
+      "learning_rate": 4.487992039769838e-06,
+      "loss": 0.9382,
+      "step": 29511
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6691894548685906,
+      "learning_rate": 4.487355395241958e-06,
+      "loss": 0.8739,
+      "step": 29512
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6034896076337373,
+      "learning_rate": 4.486718782810439e-06,
+      "loss": 0.8774,
+      "step": 29513
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4401301860596492,
+      "learning_rate": 4.486082202478984e-06,
+      "loss": 0.7476,
+      "step": 29514
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.644568589603091,
+      "learning_rate": 4.485445654251305e-06,
+      "loss": 0.76,
+      "step": 29515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6297518318513569,
+      "learning_rate": 4.484809138131111e-06,
+      "loss": 0.8147,
+      "step": 29516
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.40269442410795,
+      "learning_rate": 4.484172654122104e-06,
+      "loss": 0.8844,
+      "step": 29517
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.538800512389685,
+      "learning_rate": 4.483536202227987e-06,
+      "loss": 0.7973,
+      "step": 29518
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5765941584056613,
+      "learning_rate": 4.482899782452468e-06,
+      "loss": 0.8263,
+      "step": 29519
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.739873341115268,
+      "learning_rate": 4.482263394799258e-06,
+      "loss": 0.8724,
+      "step": 29520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6993371731945257,
+      "learning_rate": 4.481627039272056e-06,
+      "loss": 0.8701,
+      "step": 29521
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3975977119703837,
+      "learning_rate": 4.480990715874564e-06,
+      "loss": 0.6882,
+      "step": 29522
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7627129215579364,
+      "learning_rate": 4.480354424610498e-06,
+      "loss": 0.7013,
+      "step": 29523
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5957813643007586,
+      "learning_rate": 4.47971816548355e-06,
+      "loss": 0.8974,
+      "step": 29524
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5844919346810666,
+      "learning_rate": 4.479081938497435e-06,
+      "loss": 0.7843,
+      "step": 29525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.98316649472295,
+      "learning_rate": 4.478445743655849e-06,
+      "loss": 0.7788,
+      "step": 29526
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5829474346011703,
+      "learning_rate": 4.477809580962503e-06,
+      "loss": 0.8757,
+      "step": 29527
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7145287226678783,
+      "learning_rate": 4.477173450421096e-06,
+      "loss": 0.7382,
+      "step": 29528
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6251086672397317,
+      "learning_rate": 4.476537352035336e-06,
+      "loss": 0.7983,
+      "step": 29529
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5601172992469077,
+      "learning_rate": 4.475901285808924e-06,
+      "loss": 0.8255,
+      "step": 29530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7313596162805729,
+      "learning_rate": 4.4752652517455605e-06,
+      "loss": 0.6733,
+      "step": 29531
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5398266045506068,
+      "learning_rate": 4.4746292498489515e-06,
+      "loss": 0.8025,
+      "step": 29532
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4733435570755777,
+      "learning_rate": 4.4739932801228046e-06,
+      "loss": 0.7277,
+      "step": 29533
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4000667453698747,
+      "learning_rate": 4.473357342570819e-06,
+      "loss": 0.8204,
+      "step": 29534
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.515309231067651,
+      "learning_rate": 4.472721437196692e-06,
+      "loss": 0.8704,
+      "step": 29535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6069154585858496,
+      "learning_rate": 4.472085564004131e-06,
+      "loss": 0.7932,
+      "step": 29536
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4270489179161394,
+      "learning_rate": 4.47144972299684e-06,
+      "loss": 0.7946,
+      "step": 29537
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9246625421991301,
+      "learning_rate": 4.470813914178521e-06,
+      "loss": 0.8992,
+      "step": 29538
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.766573199950705,
+      "learning_rate": 4.47017813755287e-06,
+      "loss": 0.8923,
+      "step": 29539
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.788158767760824,
+      "learning_rate": 4.4695423931235915e-06,
+      "loss": 0.6588,
+      "step": 29540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.176538275637392,
+      "learning_rate": 4.4689066808943925e-06,
+      "loss": 0.8249,
+      "step": 29541
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7725910596620827,
+      "learning_rate": 4.46827100086897e-06,
+      "loss": 0.7759,
+      "step": 29542
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5814681578884162,
+      "learning_rate": 4.467635353051021e-06,
+      "loss": 0.9624,
+      "step": 29543
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.735594803021118,
+      "learning_rate": 4.466999737444254e-06,
+      "loss": 0.6726,
+      "step": 29544
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5165938692594756,
+      "learning_rate": 4.466364154052364e-06,
+      "loss": 0.7637,
+      "step": 29545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4933569193123415,
+      "learning_rate": 4.4657286028790555e-06,
+      "loss": 0.779,
+      "step": 29546
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.604743721895031,
+      "learning_rate": 4.465093083928025e-06,
+      "loss": 0.7679,
+      "step": 29547
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3800544501044907,
+      "learning_rate": 4.464457597202978e-06,
+      "loss": 0.7993,
+      "step": 29548
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7443657755402027,
+      "learning_rate": 4.4638221427076075e-06,
+      "loss": 0.7315,
+      "step": 29549
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.746759858962012,
+      "learning_rate": 4.4631867204456215e-06,
+      "loss": 0.819,
+      "step": 29550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.621212081816546,
+      "learning_rate": 4.462551330420715e-06,
+      "loss": 0.7515,
+      "step": 29551
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8011358877344714,
+      "learning_rate": 4.461915972636584e-06,
+      "loss": 0.7689,
+      "step": 29552
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6508621316618592,
+      "learning_rate": 4.461280647096932e-06,
+      "loss": 0.8375,
+      "step": 29553
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7593890491838166,
+      "learning_rate": 4.46064535380546e-06,
+      "loss": 0.714,
+      "step": 29554
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.452709683444677,
+      "learning_rate": 4.460010092765866e-06,
+      "loss": 0.8023,
+      "step": 29555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5317828760664371,
+      "learning_rate": 4.4593748639818425e-06,
+      "loss": 0.8297,
+      "step": 29556
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8955313287663755,
+      "learning_rate": 4.458739667457094e-06,
+      "loss": 0.725,
+      "step": 29557
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5639694261736854,
+      "learning_rate": 4.4581045031953195e-06,
+      "loss": 0.7579,
+      "step": 29558
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6709240568500032,
+      "learning_rate": 4.457469371200216e-06,
+      "loss": 0.8785,
+      "step": 29559
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4783666430266484,
+      "learning_rate": 4.456834271475477e-06,
+      "loss": 0.7683,
+      "step": 29560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.464139892912677,
+      "learning_rate": 4.456199204024803e-06,
+      "loss": 0.8696,
+      "step": 29561
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4730067042638741,
+      "learning_rate": 4.455564168851897e-06,
+      "loss": 0.7851,
+      "step": 29562
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9342397455865636,
+      "learning_rate": 4.454929165960451e-06,
+      "loss": 0.8646,
+      "step": 29563
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8210911952519744,
+      "learning_rate": 4.45429419535416e-06,
+      "loss": 0.809,
+      "step": 29564
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.615041868508214,
+      "learning_rate": 4.453659257036727e-06,
+      "loss": 0.862,
+      "step": 29565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6494629413957844,
+      "learning_rate": 4.453024351011842e-06,
+      "loss": 0.9598,
+      "step": 29566
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8423312859708687,
+      "learning_rate": 4.452389477283209e-06,
+      "loss": 0.9443,
+      "step": 29567
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6028995480405022,
+      "learning_rate": 4.4517546358545175e-06,
+      "loss": 0.881,
+      "step": 29568
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7361173699487848,
+      "learning_rate": 4.451119826729471e-06,
+      "loss": 0.7208,
+      "step": 29569
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8131292365930223,
+      "learning_rate": 4.450485049911757e-06,
+      "loss": 0.753,
+      "step": 29570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5021493604734946,
+      "learning_rate": 4.449850305405082e-06,
+      "loss": 0.9134,
+      "step": 29571
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3327389837060089,
+      "learning_rate": 4.449215593213133e-06,
+      "loss": 0.8137,
+      "step": 29572
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5791560561291182,
+      "learning_rate": 4.448580913339605e-06,
+      "loss": 0.9857,
+      "step": 29573
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3584402180851518,
+      "learning_rate": 4.447946265788197e-06,
+      "loss": 0.8407,
+      "step": 29574
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3628096450708256,
+      "learning_rate": 4.447311650562604e-06,
+      "loss": 0.7355,
+      "step": 29575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8079233645774225,
+      "learning_rate": 4.446677067666527e-06,
+      "loss": 0.7034,
+      "step": 29576
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.594314403071168,
+      "learning_rate": 4.446042517103647e-06,
+      "loss": 0.8128,
+      "step": 29577
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.505876585472251,
+      "learning_rate": 4.445407998877667e-06,
+      "loss": 0.8726,
+      "step": 29578
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5603928407122027,
+      "learning_rate": 4.444773512992282e-06,
+      "loss": 0.849,
+      "step": 29579
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7203412931225879,
+      "learning_rate": 4.4441390594511855e-06,
+      "loss": 0.8409,
+      "step": 29580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.580781801728292,
+      "learning_rate": 4.443504638258066e-06,
+      "loss": 0.9297,
+      "step": 29581
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5159836378698566,
+      "learning_rate": 4.442870249416622e-06,
+      "loss": 0.9145,
+      "step": 29582
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5782745490823173,
+      "learning_rate": 4.442235892930551e-06,
+      "loss": 0.8126,
+      "step": 29583
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6739553193481476,
+      "learning_rate": 4.441601568803538e-06,
+      "loss": 0.7916,
+      "step": 29584
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4925669286903422,
+      "learning_rate": 4.440967277039284e-06,
+      "loss": 0.8938,
+      "step": 29585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7012865256150245,
+      "learning_rate": 4.4403330176414795e-06,
+      "loss": 0.7886,
+      "step": 29586
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5113146860717146,
+      "learning_rate": 4.439698790613811e-06,
+      "loss": 0.9042,
+      "step": 29587
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4948145656798293,
+      "learning_rate": 4.439064595959978e-06,
+      "loss": 0.7324,
+      "step": 29588
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6347109642591038,
+      "learning_rate": 4.4384304336836745e-06,
+      "loss": 0.8719,
+      "step": 29589
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.866350024699216,
+      "learning_rate": 4.437796303788589e-06,
+      "loss": 0.7098,
+      "step": 29590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4995729679967835,
+      "learning_rate": 4.437162206278413e-06,
+      "loss": 0.829,
+      "step": 29591
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5992442476191604,
+      "learning_rate": 4.436528141156838e-06,
+      "loss": 0.8326,
+      "step": 29592
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8710710764548475,
+      "learning_rate": 4.435894108427564e-06,
+      "loss": 0.7925,
+      "step": 29593
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0366143051826344,
+      "learning_rate": 4.43526010809427e-06,
+      "loss": 0.9723,
+      "step": 29594
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7652023314265842,
+      "learning_rate": 4.434626140160654e-06,
+      "loss": 0.7316,
+      "step": 29595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7220317583642376,
+      "learning_rate": 4.4339922046304055e-06,
+      "loss": 0.653,
+      "step": 29596
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5855700031178093,
+      "learning_rate": 4.43335830150722e-06,
+      "loss": 0.7202,
+      "step": 29597
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.630611449944185,
+      "learning_rate": 4.432724430794786e-06,
+      "loss": 0.8457,
+      "step": 29598
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.475935453402998,
+      "learning_rate": 4.432090592496788e-06,
+      "loss": 0.9015,
+      "step": 29599
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5738336525658672,
+      "learning_rate": 4.431456786616927e-06,
+      "loss": 0.86,
+      "step": 29600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7097722072738255,
+      "learning_rate": 4.4308230131588826e-06,
+      "loss": 0.929,
+      "step": 29601
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.749020120087363,
+      "learning_rate": 4.430189272126353e-06,
+      "loss": 0.8058,
+      "step": 29602
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.21009358888762,
+      "learning_rate": 4.429555563523021e-06,
+      "loss": 0.8566,
+      "step": 29603
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.543607018871813,
+      "learning_rate": 4.4289218873525855e-06,
+      "loss": 0.8308,
+      "step": 29604
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3879174199490107,
+      "learning_rate": 4.428288243618725e-06,
+      "loss": 0.8319,
+      "step": 29605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.944253973443322,
+      "learning_rate": 4.427654632325139e-06,
+      "loss": 0.7807,
+      "step": 29606
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7501054966432636,
+      "learning_rate": 4.427021053475512e-06,
+      "loss": 0.7193,
+      "step": 29607
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.524597523988285,
+      "learning_rate": 4.426387507073529e-06,
+      "loss": 0.8292,
+      "step": 29608
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.610114623527964,
+      "learning_rate": 4.425753993122882e-06,
+      "loss": 0.8817,
+      "step": 29609
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5716348391195196,
+      "learning_rate": 4.425120511627263e-06,
+      "loss": 0.7647,
+      "step": 29610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.595674563686908,
+      "learning_rate": 4.424487062590359e-06,
+      "loss": 0.7686,
+      "step": 29611
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.113741045490256,
+      "learning_rate": 4.423853646015852e-06,
+      "loss": 0.9437,
+      "step": 29612
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5656672268199963,
+      "learning_rate": 4.423220261907435e-06,
+      "loss": 0.8458,
+      "step": 29613
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5959734051979513,
+      "learning_rate": 4.422586910268798e-06,
+      "loss": 0.9051,
+      "step": 29614
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7986189042217884,
+      "learning_rate": 4.421953591103627e-06,
+      "loss": 0.8342,
+      "step": 29615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5517537119275886,
+      "learning_rate": 4.421320304415604e-06,
+      "loss": 0.8072,
+      "step": 29616
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5355570124596238,
+      "learning_rate": 4.42068705020842e-06,
+      "loss": 0.9214,
+      "step": 29617
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5383709965430912,
+      "learning_rate": 4.420053828485765e-06,
+      "loss": 0.8771,
+      "step": 29618
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6002629738606962,
+      "learning_rate": 4.419420639251325e-06,
+      "loss": 0.8986,
+      "step": 29619
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4839939253144987,
+      "learning_rate": 4.4187874825087806e-06,
+      "loss": 0.7822,
+      "step": 29620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4663819951023258,
+      "learning_rate": 4.418154358261822e-06,
+      "loss": 0.8509,
+      "step": 29621
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4094429325179774,
+      "learning_rate": 4.417521266514141e-06,
+      "loss": 0.8689,
+      "step": 29622
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7283253977918784,
+      "learning_rate": 4.416888207269418e-06,
+      "loss": 0.873,
+      "step": 29623
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5556011787778046,
+      "learning_rate": 4.4162551805313345e-06,
+      "loss": 0.8352,
+      "step": 29624
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7056612071150226,
+      "learning_rate": 4.4156221863035854e-06,
+      "loss": 0.875,
+      "step": 29625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6128516520552507,
+      "learning_rate": 4.414989224589848e-06,
+      "loss": 0.775,
+      "step": 29626
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7855118984232526,
+      "learning_rate": 4.414356295393816e-06,
+      "loss": 0.7094,
+      "step": 29627
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6886621357246243,
+      "learning_rate": 4.413723398719167e-06,
+      "loss": 0.8897,
+      "step": 29628
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5569626040925673,
+      "learning_rate": 4.413090534569591e-06,
+      "loss": 0.9046,
+      "step": 29629
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5949367559067762,
+      "learning_rate": 4.412457702948768e-06,
+      "loss": 0.8908,
+      "step": 29630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5335489827923157,
+      "learning_rate": 4.411824903860387e-06,
+      "loss": 0.9347,
+      "step": 29631
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.768697636765615,
+      "learning_rate": 4.411192137308134e-06,
+      "loss": 0.8912,
+      "step": 29632
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6192589254188006,
+      "learning_rate": 4.410559403295683e-06,
+      "loss": 0.8889,
+      "step": 29633
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6606114320981178,
+      "learning_rate": 4.409926701826726e-06,
+      "loss": 0.882,
+      "step": 29634
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8084943675312022,
+      "learning_rate": 4.4092940329049495e-06,
+      "loss": 0.9116,
+      "step": 29635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8298919822664101,
+      "learning_rate": 4.408661396534033e-06,
+      "loss": 0.8418,
+      "step": 29636
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.609298133182667,
+      "learning_rate": 4.4080287927176565e-06,
+      "loss": 0.9146,
+      "step": 29637
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.87956468847724,
+      "learning_rate": 4.407396221459507e-06,
+      "loss": 0.8408,
+      "step": 29638
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8256406241686843,
+      "learning_rate": 4.40676368276327e-06,
+      "loss": 0.6912,
+      "step": 29639
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4607814874887235,
+      "learning_rate": 4.406131176632626e-06,
+      "loss": 0.7638,
+      "step": 29640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.814149189087361,
+      "learning_rate": 4.405498703071254e-06,
+      "loss": 0.6799,
+      "step": 29641
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8006652109191095,
+      "learning_rate": 4.404866262082839e-06,
+      "loss": 0.9552,
+      "step": 29642
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6212783614103778,
+      "learning_rate": 4.404233853671069e-06,
+      "loss": 0.9657,
+      "step": 29643
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.421422981437986,
+      "learning_rate": 4.403601477839619e-06,
+      "loss": 0.783,
+      "step": 29644
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5916530302793357,
+      "learning_rate": 4.402969134592171e-06,
+      "loss": 0.8692,
+      "step": 29645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.535427291146114,
+      "learning_rate": 4.402336823932412e-06,
+      "loss": 0.7926,
+      "step": 29646
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.800878889248083,
+      "learning_rate": 4.401704545864016e-06,
+      "loss": 0.8502,
+      "step": 29647
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.540765541390394,
+      "learning_rate": 4.401072300390672e-06,
+      "loss": 0.8646,
+      "step": 29648
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5260257369730894,
+      "learning_rate": 4.4004400875160545e-06,
+      "loss": 0.8096,
+      "step": 29649
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7254839996510059,
+      "learning_rate": 4.399807907243851e-06,
+      "loss": 0.7941,
+      "step": 29650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.665019758681016,
+      "learning_rate": 4.399175759577737e-06,
+      "loss": 0.8719,
+      "step": 29651
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.307282234999847,
+      "learning_rate": 4.398543644521396e-06,
+      "loss": 0.8724,
+      "step": 29652
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.490538054115947,
+      "learning_rate": 4.397911562078508e-06,
+      "loss": 0.9316,
+      "step": 29653
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8678792292576145,
+      "learning_rate": 4.39727951225275e-06,
+      "loss": 0.7396,
+      "step": 29654
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.007678168655352,
+      "learning_rate": 4.3966474950478036e-06,
+      "loss": 0.9668,
+      "step": 29655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7960413942799416,
+      "learning_rate": 4.396015510467353e-06,
+      "loss": 0.7696,
+      "step": 29656
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5320687044018573,
+      "learning_rate": 4.395383558515075e-06,
+      "loss": 0.7764,
+      "step": 29657
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7603319116085694,
+      "learning_rate": 4.394751639194644e-06,
+      "loss": 0.967,
+      "step": 29658
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.546170592393075,
+      "learning_rate": 4.394119752509743e-06,
+      "loss": 0.9557,
+      "step": 29659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4802787082871214,
+      "learning_rate": 4.393487898464056e-06,
+      "loss": 0.74,
+      "step": 29660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6889204228689512,
+      "learning_rate": 4.392856077061258e-06,
+      "loss": 0.8166,
+      "step": 29661
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5123259346083833,
+      "learning_rate": 4.3922242883050226e-06,
+      "loss": 0.8894,
+      "step": 29662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7692479616144823,
+      "learning_rate": 4.391592532199034e-06,
+      "loss": 0.7055,
+      "step": 29663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7913924492994135,
+      "learning_rate": 4.390960808746971e-06,
+      "loss": 0.7619,
+      "step": 29664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8029109719744378,
+      "learning_rate": 4.390329117952511e-06,
+      "loss": 0.9998,
+      "step": 29665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7684766150824869,
+      "learning_rate": 4.389697459819327e-06,
+      "loss": 0.7082,
+      "step": 29666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5414642131047709,
+      "learning_rate": 4.389065834351105e-06,
+      "loss": 0.9034,
+      "step": 29667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6009314259021656,
+      "learning_rate": 4.3884342415515135e-06,
+      "loss": 0.8355,
+      "step": 29668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5300441946729022,
+      "learning_rate": 4.387802681424234e-06,
+      "loss": 0.7677,
+      "step": 29669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6106727837924801,
+      "learning_rate": 4.387171153972949e-06,
+      "loss": 0.8005,
+      "step": 29670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5604360076347263,
+      "learning_rate": 4.38653965920133e-06,
+      "loss": 0.8348,
+      "step": 29671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7113633644065502,
+      "learning_rate": 4.3859081971130516e-06,
+      "loss": 0.8334,
+      "step": 29672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6795350278637895,
+      "learning_rate": 4.385276767711792e-06,
+      "loss": 0.7751,
+      "step": 29673
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6574948975351818,
+      "learning_rate": 4.384645371001236e-06,
+      "loss": 0.9326,
+      "step": 29674
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1169677993104705,
+      "learning_rate": 4.384014006985044e-06,
+      "loss": 0.915,
+      "step": 29675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7457421336179737,
+      "learning_rate": 4.383382675666903e-06,
+      "loss": 0.6643,
+      "step": 29676
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6058929965530666,
+      "learning_rate": 4.382751377050485e-06,
+      "loss": 0.8365,
+      "step": 29677
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7561520849412637,
+      "learning_rate": 4.382120111139474e-06,
+      "loss": 0.9285,
+      "step": 29678
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9601641117489819,
+      "learning_rate": 4.3814888779375305e-06,
+      "loss": 0.9096,
+      "step": 29679
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5723726306140615,
+      "learning_rate": 4.380857677448338e-06,
+      "loss": 0.9345,
+      "step": 29680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.789078930614551,
+      "learning_rate": 4.380226509675576e-06,
+      "loss": 0.7908,
+      "step": 29681
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7943263559794435,
+      "learning_rate": 4.379595374622909e-06,
+      "loss": 0.7359,
+      "step": 29682
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5904794616201314,
+      "learning_rate": 4.378964272294022e-06,
+      "loss": 0.7316,
+      "step": 29683
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6354925992202745,
+      "learning_rate": 4.37833320269258e-06,
+      "loss": 0.7722,
+      "step": 29684
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4853965398523419,
+      "learning_rate": 4.377702165822266e-06,
+      "loss": 0.8938,
+      "step": 29685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4881424863428339,
+      "learning_rate": 4.377071161686745e-06,
+      "loss": 0.8905,
+      "step": 29686
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7830085787382228,
+      "learning_rate": 4.376440190289703e-06,
+      "loss": 0.6509,
+      "step": 29687
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5710311841987106,
+      "learning_rate": 4.375809251634804e-06,
+      "loss": 0.9199,
+      "step": 29688
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5148000675467908,
+      "learning_rate": 4.37517834572572e-06,
+      "loss": 0.9092,
+      "step": 29689
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8380668164357061,
+      "learning_rate": 4.374547472566129e-06,
+      "loss": 0.8176,
+      "step": 29690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5279610953803693,
+      "learning_rate": 4.373916632159708e-06,
+      "loss": 0.8695,
+      "step": 29691
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5497936882819767,
+      "learning_rate": 4.373285824510125e-06,
+      "loss": 0.8674,
+      "step": 29692
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.787156030876152,
+      "learning_rate": 4.372655049621051e-06,
+      "loss": 0.8178,
+      "step": 29693
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.10803179271768,
+      "learning_rate": 4.372024307496159e-06,
+      "loss": 0.8864,
+      "step": 29694
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6546349476624242,
+      "learning_rate": 4.371393598139128e-06,
+      "loss": 0.743,
+      "step": 29695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7859951059535624,
+      "learning_rate": 4.370762921553625e-06,
+      "loss": 0.7538,
+      "step": 29696
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9875639623317163,
+      "learning_rate": 4.370132277743319e-06,
+      "loss": 0.8114,
+      "step": 29697
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4023776267374253,
+      "learning_rate": 4.369501666711885e-06,
+      "loss": 0.8031,
+      "step": 29698
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5781805413305403,
+      "learning_rate": 4.368871088462998e-06,
+      "loss": 0.8033,
+      "step": 29699
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6702246559470217,
+      "learning_rate": 4.368240543000327e-06,
+      "loss": 0.8657,
+      "step": 29700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5874110633483887,
+      "learning_rate": 4.36761003032754e-06,
+      "loss": 0.9573,
+      "step": 29701
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.602159468004549,
+      "learning_rate": 4.366979550448313e-06,
+      "loss": 0.7803,
+      "step": 29702
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5864133963141092,
+      "learning_rate": 4.366349103366311e-06,
+      "loss": 0.7586,
+      "step": 29703
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6742935797925607,
+      "learning_rate": 4.365718689085211e-06,
+      "loss": 0.8947,
+      "step": 29704
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6373641332722277,
+      "learning_rate": 4.365088307608678e-06,
+      "loss": 0.8018,
+      "step": 29705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.721157411483098,
+      "learning_rate": 4.364457958940388e-06,
+      "loss": 0.8397,
+      "step": 29706
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7689042092964992,
+      "learning_rate": 4.3638276430840045e-06,
+      "loss": 0.7052,
+      "step": 29707
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7530799675288031,
+      "learning_rate": 4.363197360043204e-06,
+      "loss": 0.7534,
+      "step": 29708
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8029265836934028,
+      "learning_rate": 4.362567109821654e-06,
+      "loss": 0.7031,
+      "step": 29709
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.739210981962636,
+      "learning_rate": 4.3619368924230185e-06,
+      "loss": 0.9335,
+      "step": 29710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.428363925150175,
+      "learning_rate": 4.361306707850972e-06,
+      "loss": 0.832,
+      "step": 29711
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6336400399945288,
+      "learning_rate": 4.360676556109186e-06,
+      "loss": 0.9033,
+      "step": 29712
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7895489905401745,
+      "learning_rate": 4.360046437201327e-06,
+      "loss": 0.8029,
+      "step": 29713
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.544150283018892,
+      "learning_rate": 4.3594163511310596e-06,
+      "loss": 0.9196,
+      "step": 29714
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4784769002134,
+      "learning_rate": 4.358786297902056e-06,
+      "loss": 0.8154,
+      "step": 29715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6010935613310748,
+      "learning_rate": 4.3581562775179875e-06,
+      "loss": 0.7524,
+      "step": 29716
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.814974879088237,
+      "learning_rate": 4.35752628998252e-06,
+      "loss": 0.686,
+      "step": 29717
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.648518992330305,
+      "learning_rate": 4.356896335299317e-06,
+      "loss": 0.9287,
+      "step": 29718
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7505200245671175,
+      "learning_rate": 4.35626641347205e-06,
+      "loss": 0.7599,
+      "step": 29719
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4841631708002327,
+      "learning_rate": 4.3556365245043905e-06,
+      "loss": 0.9115,
+      "step": 29720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1407243814770753,
+      "learning_rate": 4.355006668400003e-06,
+      "loss": 0.8015,
+      "step": 29721
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.835121148948915,
+      "learning_rate": 4.354376845162549e-06,
+      "loss": 0.8907,
+      "step": 29722
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5928581130208832,
+      "learning_rate": 4.3537470547957054e-06,
+      "loss": 0.9114,
+      "step": 29723
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5067246426334018,
+      "learning_rate": 4.353117297303129e-06,
+      "loss": 0.7449,
+      "step": 29724
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6103167613068998,
+      "learning_rate": 4.352487572688495e-06,
+      "loss": 0.8314,
+      "step": 29725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5109307566593864,
+      "learning_rate": 4.351857880955464e-06,
+      "loss": 0.8395,
+      "step": 29726
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7340449592792329,
+      "learning_rate": 4.351228222107706e-06,
+      "loss": 0.7512,
+      "step": 29727
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5226148218809943,
+      "learning_rate": 4.350598596148884e-06,
+      "loss": 0.8807,
+      "step": 29728
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7457721251261156,
+      "learning_rate": 4.349969003082669e-06,
+      "loss": 0.7186,
+      "step": 29729
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3384556304075375,
+      "learning_rate": 4.349339442912723e-06,
+      "loss": 0.7947,
+      "step": 29730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6360729562237326,
+      "learning_rate": 4.3487099156427075e-06,
+      "loss": 0.875,
+      "step": 29731
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8627571565915315,
+      "learning_rate": 4.348080421276293e-06,
+      "loss": 0.767,
+      "step": 29732
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5695881911593363,
+      "learning_rate": 4.347450959817146e-06,
+      "loss": 0.8458,
+      "step": 29733
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3968302122618985,
+      "learning_rate": 4.34682153126893e-06,
+      "loss": 0.8413,
+      "step": 29734
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3135563363057456,
+      "learning_rate": 4.346192135635305e-06,
+      "loss": 0.7685,
+      "step": 29735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6062981990548069,
+      "learning_rate": 4.345562772919939e-06,
+      "loss": 0.9018,
+      "step": 29736
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4773366678425033,
+      "learning_rate": 4.344933443126501e-06,
+      "loss": 0.923,
+      "step": 29737
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5150072096097293,
+      "learning_rate": 4.344304146258651e-06,
+      "loss": 0.7774,
+      "step": 29738
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8743608975293833,
+      "learning_rate": 4.343674882320048e-06,
+      "loss": 0.7659,
+      "step": 29739
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7962776370087891,
+      "learning_rate": 4.343045651314361e-06,
+      "loss": 0.6753,
+      "step": 29740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.654345728254618,
+      "learning_rate": 4.342416453245259e-06,
+      "loss": 0.8516,
+      "step": 29741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1572873868333717,
+      "learning_rate": 4.341787288116397e-06,
+      "loss": 0.8165,
+      "step": 29742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3807492820758847,
+      "learning_rate": 4.341158155931437e-06,
+      "loss": 0.8716,
+      "step": 29743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7061366419181279,
+      "learning_rate": 4.3405290566940475e-06,
+      "loss": 0.8435,
+      "step": 29744
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8886211325822317,
+      "learning_rate": 4.339899990407892e-06,
+      "loss": 0.8167,
+      "step": 29745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.576025228878997,
+      "learning_rate": 4.339270957076631e-06,
+      "loss": 0.9205,
+      "step": 29746
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5822375637464436,
+      "learning_rate": 4.338641956703923e-06,
+      "loss": 0.7853,
+      "step": 29747
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5091433968169632,
+      "learning_rate": 4.338012989293438e-06,
+      "loss": 0.8732,
+      "step": 29748
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7255411583477214,
+      "learning_rate": 4.33738405484883e-06,
+      "loss": 0.7994,
+      "step": 29749
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.692100035464094,
+      "learning_rate": 4.3367551533737685e-06,
+      "loss": 0.7991,
+      "step": 29750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5125502379388036,
+      "learning_rate": 4.336126284871908e-06,
+      "loss": 0.8673,
+      "step": 29751
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5858428066798458,
+      "learning_rate": 4.335497449346916e-06,
+      "loss": 0.7753,
+      "step": 29752
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6911519092315812,
+      "learning_rate": 4.334868646802448e-06,
+      "loss": 0.8743,
+      "step": 29753
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.623075959098393,
+      "learning_rate": 4.334239877242172e-06,
+      "loss": 0.8822,
+      "step": 29754
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7174487805421583,
+      "learning_rate": 4.333611140669746e-06,
+      "loss": 0.9214,
+      "step": 29755
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4654031558596905,
+      "learning_rate": 4.332982437088825e-06,
+      "loss": 0.8135,
+      "step": 29756
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4346937950462015,
+      "learning_rate": 4.332353766503075e-06,
+      "loss": 0.9028,
+      "step": 29757
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6376937892931758,
+      "learning_rate": 4.331725128916157e-06,
+      "loss": 0.7558,
+      "step": 29758
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8775168092310037,
+      "learning_rate": 4.331096524331735e-06,
+      "loss": 0.7722,
+      "step": 29759
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5593767293444942,
+      "learning_rate": 4.330467952753458e-06,
+      "loss": 0.8623,
+      "step": 29760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.658816487715013,
+      "learning_rate": 4.329839414184991e-06,
+      "loss": 0.7615,
+      "step": 29761
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6448445618426923,
+      "learning_rate": 4.329210908629997e-06,
+      "loss": 0.7946,
+      "step": 29762
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5127092252144474,
+      "learning_rate": 4.328582436092133e-06,
+      "loss": 0.8901,
+      "step": 29763
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6849286312683147,
+      "learning_rate": 4.3279539965750535e-06,
+      "loss": 0.7749,
+      "step": 29764
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7766440334849115,
+      "learning_rate": 4.327325590082423e-06,
+      "loss": 0.6821,
+      "step": 29765
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4604568896144798,
+      "learning_rate": 4.326697216617901e-06,
+      "loss": 0.7786,
+      "step": 29766
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4462774426264258,
+      "learning_rate": 4.326068876185141e-06,
+      "loss": 0.7743,
+      "step": 29767
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.543116505706865,
+      "learning_rate": 4.3254405687878075e-06,
+      "loss": 0.8314,
+      "step": 29768
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6235655901322266,
+      "learning_rate": 4.324812294429557e-06,
+      "loss": 0.8184,
+      "step": 29769
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6812071532312312,
+      "learning_rate": 4.324184053114041e-06,
+      "loss": 0.9545,
+      "step": 29770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9444396296630149,
+      "learning_rate": 4.323555844844922e-06,
+      "loss": 0.9449,
+      "step": 29771
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9453312338710698,
+      "learning_rate": 4.3229276696258625e-06,
+      "loss": 0.6687,
+      "step": 29772
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8561117772741957,
+      "learning_rate": 4.3222995274605155e-06,
+      "loss": 0.7707,
+      "step": 29773
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8787916543607528,
+      "learning_rate": 4.321671418352534e-06,
+      "loss": 0.8848,
+      "step": 29774
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.668565086558731,
+      "learning_rate": 4.3210433423055795e-06,
+      "loss": 0.7271,
+      "step": 29775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3150204310688318,
+      "learning_rate": 4.320415299323316e-06,
+      "loss": 0.8579,
+      "step": 29776
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5437167727722623,
+      "learning_rate": 4.319787289409385e-06,
+      "loss": 0.8124,
+      "step": 29777
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.49299924585921,
+      "learning_rate": 4.319159312567451e-06,
+      "loss": 0.8738,
+      "step": 29778
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.80180882186907,
+      "learning_rate": 4.318531368801171e-06,
+      "loss": 0.775,
+      "step": 29779
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.588281652660402,
+      "learning_rate": 4.317903458114202e-06,
+      "loss": 0.822,
+      "step": 29780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8332702170116433,
+      "learning_rate": 4.317275580510199e-06,
+      "loss": 0.9483,
+      "step": 29781
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.706493702472051,
+      "learning_rate": 4.316647735992814e-06,
+      "loss": 0.9846,
+      "step": 29782
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5127618528115472,
+      "learning_rate": 4.316019924565708e-06,
+      "loss": 0.8143,
+      "step": 29783
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.18849782125373,
+      "learning_rate": 4.315392146232529e-06,
+      "loss": 0.8206,
+      "step": 29784
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6228307059920295,
+      "learning_rate": 4.31476440099694e-06,
+      "loss": 0.8767,
+      "step": 29785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.676880357028844,
+      "learning_rate": 4.31413668886259e-06,
+      "loss": 0.8026,
+      "step": 29786
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7652756484778802,
+      "learning_rate": 4.31350900983314e-06,
+      "loss": 0.8369,
+      "step": 29787
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6206992914312977,
+      "learning_rate": 4.312881363912237e-06,
+      "loss": 0.9066,
+      "step": 29788
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5608139057291908,
+      "learning_rate": 4.312253751103544e-06,
+      "loss": 0.9153,
+      "step": 29789
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9355724929639295,
+      "learning_rate": 4.311626171410707e-06,
+      "loss": 0.7964,
+      "step": 29790
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6495063318441314,
+      "learning_rate": 4.310998624837383e-06,
+      "loss": 0.9584,
+      "step": 29791
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6303589791635837,
+      "learning_rate": 4.310371111387225e-06,
+      "loss": 0.8404,
+      "step": 29792
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7139212585203947,
+      "learning_rate": 4.309743631063892e-06,
+      "loss": 0.8498,
+      "step": 29793
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9094253259302245,
+      "learning_rate": 4.309116183871032e-06,
+      "loss": 0.7952,
+      "step": 29794
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4913976913978908,
+      "learning_rate": 4.308488769812296e-06,
+      "loss": 0.79,
+      "step": 29795
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.584140467041393,
+      "learning_rate": 4.30786138889134e-06,
+      "loss": 0.7359,
+      "step": 29796
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5722629160591153,
+      "learning_rate": 4.30723404111182e-06,
+      "loss": 0.9012,
+      "step": 29797
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6462506076543062,
+      "learning_rate": 4.306606726477388e-06,
+      "loss": 0.8577,
+      "step": 29798
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8207454082045091,
+      "learning_rate": 4.3059794449916896e-06,
+      "loss": 0.8834,
+      "step": 29799
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7409685303988751,
+      "learning_rate": 4.305352196658381e-06,
+      "loss": 0.8868,
+      "step": 29800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7532752668009695,
+      "learning_rate": 4.3047249814811185e-06,
+      "loss": 0.8551,
+      "step": 29801
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9711315639614495,
+      "learning_rate": 4.30409779946355e-06,
+      "loss": 0.8766,
+      "step": 29802
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5501616590563827,
+      "learning_rate": 4.303470650609325e-06,
+      "loss": 0.8488,
+      "step": 29803
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6636784360701495,
+      "learning_rate": 4.3028435349221e-06,
+      "loss": 0.6917,
+      "step": 29804
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7793164404085202,
+      "learning_rate": 4.302216452405521e-06,
+      "loss": 0.7401,
+      "step": 29805
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0457807176853158,
+      "learning_rate": 4.301589403063244e-06,
+      "loss": 0.7458,
+      "step": 29806
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7954121510089593,
+      "learning_rate": 4.300962386898915e-06,
+      "loss": 0.687,
+      "step": 29807
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6261304978834628,
+      "learning_rate": 4.3003354039161906e-06,
+      "loss": 0.8684,
+      "step": 29808
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5430064935572156,
+      "learning_rate": 4.299708454118714e-06,
+      "loss": 0.7477,
+      "step": 29809
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6267061790450421,
+      "learning_rate": 4.299081537510143e-06,
+      "loss": 0.8234,
+      "step": 29810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5682450706519473,
+      "learning_rate": 4.2984546540941245e-06,
+      "loss": 0.9065,
+      "step": 29811
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4029244275018038,
+      "learning_rate": 4.297827803874304e-06,
+      "loss": 0.7824,
+      "step": 29812
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.707743042595627,
+      "learning_rate": 4.297200986854335e-06,
+      "loss": 0.8572,
+      "step": 29813
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7668107314228472,
+      "learning_rate": 4.296574203037872e-06,
+      "loss": 0.7574,
+      "step": 29814
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4437416960084564,
+      "learning_rate": 4.295947452428559e-06,
+      "loss": 0.9068,
+      "step": 29815
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5625052085804552,
+      "learning_rate": 4.295320735030043e-06,
+      "loss": 0.8864,
+      "step": 29816
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5172542179904451,
+      "learning_rate": 4.294694050845976e-06,
+      "loss": 0.9326,
+      "step": 29817
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6250283796047094,
+      "learning_rate": 4.294067399880009e-06,
+      "loss": 0.8359,
+      "step": 29818
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5534717424320872,
+      "learning_rate": 4.293440782135788e-06,
+      "loss": 0.8113,
+      "step": 29819
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.205305538464568,
+      "learning_rate": 4.292814197616958e-06,
+      "loss": 0.8673,
+      "step": 29820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.803223447125344,
+      "learning_rate": 4.29218764632717e-06,
+      "loss": 0.7228,
+      "step": 29821
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7128928052236976,
+      "learning_rate": 4.2915611282700775e-06,
+      "loss": 0.8808,
+      "step": 29822
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4267126206930978,
+      "learning_rate": 4.290934643449324e-06,
+      "loss": 0.8202,
+      "step": 29823
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.794243955501344,
+      "learning_rate": 4.290308191868551e-06,
+      "loss": 0.7996,
+      "step": 29824
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8335660764898052,
+      "learning_rate": 4.289681773531415e-06,
+      "loss": 0.8669,
+      "step": 29825
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.400159688139081,
+      "learning_rate": 4.289055388441558e-06,
+      "loss": 0.7999,
+      "step": 29826
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5932579375496017,
+      "learning_rate": 4.28842903660263e-06,
+      "loss": 0.8359,
+      "step": 29827
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3733396536691092,
+      "learning_rate": 4.2878027180182745e-06,
+      "loss": 0.8481,
+      "step": 29828
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4734332641658117,
+      "learning_rate": 4.287176432692143e-06,
+      "loss": 0.9773,
+      "step": 29829
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7915448440955397,
+      "learning_rate": 4.286550180627874e-06,
+      "loss": 0.7185,
+      "step": 29830
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.89543704055118,
+      "learning_rate": 4.285923961829124e-06,
+      "loss": 0.8817,
+      "step": 29831
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4859940012177584,
+      "learning_rate": 4.285297776299533e-06,
+      "loss": 0.8394,
+      "step": 29832
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7461364177284224,
+      "learning_rate": 4.284671624042744e-06,
+      "loss": 0.6985,
+      "step": 29833
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7149534298668248,
+      "learning_rate": 4.284045505062406e-06,
+      "loss": 0.8837,
+      "step": 29834
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4755267926492979,
+      "learning_rate": 4.283419419362169e-06,
+      "loss": 0.8611,
+      "step": 29835
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7171109390855874,
+      "learning_rate": 4.282793366945675e-06,
+      "loss": 0.8682,
+      "step": 29836
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8984341775593805,
+      "learning_rate": 4.282167347816562e-06,
+      "loss": 0.7935,
+      "step": 29837
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.414300685662933,
+      "learning_rate": 4.281541361978483e-06,
+      "loss": 0.7774,
+      "step": 29838
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8156576228151903,
+      "learning_rate": 4.280915409435084e-06,
+      "loss": 0.7442,
+      "step": 29839
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5357622013774905,
+      "learning_rate": 4.280289490190005e-06,
+      "loss": 0.8527,
+      "step": 29840
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4624923820830844,
+      "learning_rate": 4.279663604246888e-06,
+      "loss": 0.8583,
+      "step": 29841
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8518069451399336,
+      "learning_rate": 4.27903775160938e-06,
+      "loss": 0.7968,
+      "step": 29842
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6101356793440889,
+      "learning_rate": 4.27841193228113e-06,
+      "loss": 0.8527,
+      "step": 29843
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7489656632046378,
+      "learning_rate": 4.277786146265778e-06,
+      "loss": 0.6958,
+      "step": 29844
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0001474371318038,
+      "learning_rate": 4.277160393566962e-06,
+      "loss": 1.0184,
+      "step": 29845
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7699631550980933,
+      "learning_rate": 4.2765346741883325e-06,
+      "loss": 0.6888,
+      "step": 29846
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.506626264655863,
+      "learning_rate": 4.275908988133528e-06,
+      "loss": 0.8478,
+      "step": 29847
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3711766424205076,
+      "learning_rate": 4.275283335406197e-06,
+      "loss": 0.7568,
+      "step": 29848
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.77709135517295,
+      "learning_rate": 4.274657716009976e-06,
+      "loss": 0.9293,
+      "step": 29849
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7468920401884211,
+      "learning_rate": 4.274032129948512e-06,
+      "loss": 0.8662,
+      "step": 29850
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7866333456279607,
+      "learning_rate": 4.273406577225444e-06,
+      "loss": 0.7442,
+      "step": 29851
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6314085011712995,
+      "learning_rate": 4.272781057844418e-06,
+      "loss": 0.8935,
+      "step": 29852
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.633973851601499,
+      "learning_rate": 4.272155571809076e-06,
+      "loss": 0.9481,
+      "step": 29853
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5718430595615625,
+      "learning_rate": 4.271530119123052e-06,
+      "loss": 0.8701,
+      "step": 29854
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4748419144888825,
+      "learning_rate": 4.270904699789994e-06,
+      "loss": 0.7219,
+      "step": 29855
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5711173492133614,
+      "learning_rate": 4.270279313813542e-06,
+      "loss": 0.9781,
+      "step": 29856
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5726114224332883,
+      "learning_rate": 4.269653961197346e-06,
+      "loss": 0.8119,
+      "step": 29857
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8110693259414898,
+      "learning_rate": 4.269028641945032e-06,
+      "loss": 0.7801,
+      "step": 29858
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7800902790257267,
+      "learning_rate": 4.2684033560602465e-06,
+      "loss": 0.7599,
+      "step": 29859
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7142664415283406,
+      "learning_rate": 4.267778103546632e-06,
+      "loss": 0.88,
+      "step": 29860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6759327299734441,
+      "learning_rate": 4.267152884407835e-06,
+      "loss": 0.8754,
+      "step": 29861
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5589530684952235,
+      "learning_rate": 4.266527698647482e-06,
+      "loss": 0.8926,
+      "step": 29862
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9112277865324028,
+      "learning_rate": 4.265902546269219e-06,
+      "loss": 0.8295,
+      "step": 29863
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.516634201845017,
+      "learning_rate": 4.265277427276692e-06,
+      "loss": 0.8456,
+      "step": 29864
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.568647466926971,
+      "learning_rate": 4.264652341673531e-06,
+      "loss": 0.9016,
+      "step": 29865
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8008462858002106,
+      "learning_rate": 4.2640272894633825e-06,
+      "loss": 0.8901,
+      "step": 29866
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6738617438207095,
+      "learning_rate": 4.2634022706498815e-06,
+      "loss": 0.7678,
+      "step": 29867
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6982297533223354,
+      "learning_rate": 4.26277728523667e-06,
+      "loss": 0.8354,
+      "step": 29868
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4289019556613873,
+      "learning_rate": 4.262152333227383e-06,
+      "loss": 0.8319,
+      "step": 29869
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7087905064871416,
+      "learning_rate": 4.2615274146256654e-06,
+      "loss": 0.9719,
+      "step": 29870
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6813950847946983,
+      "learning_rate": 4.2609025294351525e-06,
+      "loss": 0.8374,
+      "step": 29871
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7600798262369914,
+      "learning_rate": 4.260277677659477e-06,
+      "loss": 0.8035,
+      "step": 29872
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6319840998414559,
+      "learning_rate": 4.2596528593022835e-06,
+      "loss": 0.6729,
+      "step": 29873
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9556651391120972,
+      "learning_rate": 4.259028074367211e-06,
+      "loss": 0.9208,
+      "step": 29874
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8335861502635193,
+      "learning_rate": 4.258403322857895e-06,
+      "loss": 0.7444,
+      "step": 29875
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6809323321413407,
+      "learning_rate": 4.257778604777969e-06,
+      "loss": 0.8245,
+      "step": 29876
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7578222437595148,
+      "learning_rate": 4.257153920131074e-06,
+      "loss": 0.9285,
+      "step": 29877
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.506299984724778,
+      "learning_rate": 4.256529268920851e-06,
+      "loss": 0.7961,
+      "step": 29878
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7118394255552534,
+      "learning_rate": 4.255904651150931e-06,
+      "loss": 0.7921,
+      "step": 29879
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5059377968609722,
+      "learning_rate": 4.255280066824952e-06,
+      "loss": 0.7398,
+      "step": 29880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5900035360918756,
+      "learning_rate": 4.254655515946549e-06,
+      "loss": 0.8206,
+      "step": 29881
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.608656144696459,
+      "learning_rate": 4.254030998519365e-06,
+      "loss": 0.8073,
+      "step": 29882
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0081250644899336,
+      "learning_rate": 4.253406514547031e-06,
+      "loss": 0.8593,
+      "step": 29883
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5136624794330023,
+      "learning_rate": 4.25278206403318e-06,
+      "loss": 0.7209,
+      "step": 29884
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8036094052941392,
+      "learning_rate": 4.252157646981457e-06,
+      "loss": 0.8266,
+      "step": 29885
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4934465051027928,
+      "learning_rate": 4.251533263395486e-06,
+      "loss": 0.7725,
+      "step": 29886
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.188903207908654,
+      "learning_rate": 4.2509089132789126e-06,
+      "loss": 0.8224,
+      "step": 29887
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7502412829586629,
+      "learning_rate": 4.250284596635365e-06,
+      "loss": 0.8066,
+      "step": 29888
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.67934879496185,
+      "learning_rate": 4.249660313468483e-06,
+      "loss": 0.8284,
+      "step": 29889
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5648878443764285,
+      "learning_rate": 4.2490360637818965e-06,
+      "loss": 0.788,
+      "step": 29890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9854780401725647,
+      "learning_rate": 4.248411847579245e-06,
+      "loss": 0.8974,
+      "step": 29891
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3864763554257944,
+      "learning_rate": 4.247787664864163e-06,
+      "loss": 0.9285,
+      "step": 29892
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5190128217205245,
+      "learning_rate": 4.247163515640277e-06,
+      "loss": 0.7722,
+      "step": 29893
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.697795649816355,
+      "learning_rate": 4.2465393999112256e-06,
+      "loss": 0.881,
+      "step": 29894
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4440282495492043,
+      "learning_rate": 4.245915317680649e-06,
+      "loss": 0.7815,
+      "step": 29895
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6589090642717021,
+      "learning_rate": 4.245291268952174e-06,
+      "loss": 0.7515,
+      "step": 29896
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7543745114236369,
+      "learning_rate": 4.244667253729431e-06,
+      "loss": 0.9084,
+      "step": 29897
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6040043213595547,
+      "learning_rate": 4.244043272016059e-06,
+      "loss": 0.845,
+      "step": 29898
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4007946906865096,
+      "learning_rate": 4.243419323815692e-06,
+      "loss": 0.7601,
+      "step": 29899
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6366573838584462,
+      "learning_rate": 4.24279540913196e-06,
+      "loss": 0.8162,
+      "step": 29900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5348076585741164,
+      "learning_rate": 4.242171527968492e-06,
+      "loss": 0.857,
+      "step": 29901
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.612298370568599,
+      "learning_rate": 4.241547680328924e-06,
+      "loss": 0.8248,
+      "step": 29902
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7313907459243223,
+      "learning_rate": 4.240923866216894e-06,
+      "loss": 0.8398,
+      "step": 29903
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.23239760172385,
+      "learning_rate": 4.240300085636027e-06,
+      "loss": 0.8037,
+      "step": 29904
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6189092161862737,
+      "learning_rate": 4.239676338589953e-06,
+      "loss": 0.7971,
+      "step": 29905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4026638080387892,
+      "learning_rate": 4.239052625082311e-06,
+      "loss": 0.8145,
+      "step": 29906
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5535529042500587,
+      "learning_rate": 4.238428945116725e-06,
+      "loss": 0.8112,
+      "step": 29907
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9061185543557744,
+      "learning_rate": 4.237805298696832e-06,
+      "loss": 0.8313,
+      "step": 29908
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7599986830245516,
+      "learning_rate": 4.23718168582626e-06,
+      "loss": 0.6816,
+      "step": 29909
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4139877727772834,
+      "learning_rate": 4.236558106508641e-06,
+      "loss": 0.7042,
+      "step": 29910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7854233244345366,
+      "learning_rate": 4.235934560747604e-06,
+      "loss": 0.8438,
+      "step": 29911
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5672615391318991,
+      "learning_rate": 4.2353110485467845e-06,
+      "loss": 0.8828,
+      "step": 29912
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8743290895748816,
+      "learning_rate": 4.234687569909808e-06,
+      "loss": 0.6081,
+      "step": 29913
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7198237158513472,
+      "learning_rate": 4.234064124840302e-06,
+      "loss": 0.9288,
+      "step": 29914
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.431472355183374,
+      "learning_rate": 4.233440713341901e-06,
+      "loss": 0.9172,
+      "step": 29915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.114489913984776,
+      "learning_rate": 4.232817335418236e-06,
+      "loss": 0.769,
+      "step": 29916
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.633674362481389,
+      "learning_rate": 4.232193991072937e-06,
+      "loss": 0.8628,
+      "step": 29917
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5616657534124636,
+      "learning_rate": 4.231570680309626e-06,
+      "loss": 0.7939,
+      "step": 29918
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5499271912358972,
+      "learning_rate": 4.230947403131935e-06,
+      "loss": 0.8066,
+      "step": 29919
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7883475956413291,
+      "learning_rate": 4.2303241595435e-06,
+      "loss": 0.7291,
+      "step": 29920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4536734465190313,
+      "learning_rate": 4.229700949547944e-06,
+      "loss": 0.8161,
+      "step": 29921
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6252087181038066,
+      "learning_rate": 4.229077773148892e-06,
+      "loss": 0.7656,
+      "step": 29922
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5501548864791967,
+      "learning_rate": 4.228454630349977e-06,
+      "loss": 0.7416,
+      "step": 29923
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5861121906752074,
+      "learning_rate": 4.227831521154829e-06,
+      "loss": 0.8959,
+      "step": 29924
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5433577095192077,
+      "learning_rate": 4.227208445567075e-06,
+      "loss": 0.8317,
+      "step": 29925
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4639950381345943,
+      "learning_rate": 4.226585403590335e-06,
+      "loss": 0.8036,
+      "step": 29926
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7632463887974926,
+      "learning_rate": 4.225962395228248e-06,
+      "loss": 0.6921,
+      "step": 29927
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.56615175474569,
+      "learning_rate": 4.225339420484432e-06,
+      "loss": 0.9064,
+      "step": 29928
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6276309855655628,
+      "learning_rate": 4.224716479362521e-06,
+      "loss": 0.8162,
+      "step": 29929
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9714599181372443,
+      "learning_rate": 4.2240935718661365e-06,
+      "loss": 0.9137,
+      "step": 29930
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6913910944187531,
+      "learning_rate": 4.223470697998911e-06,
+      "loss": 0.6879,
+      "step": 29931
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.68497369989913,
+      "learning_rate": 4.222847857764464e-06,
+      "loss": 0.8837,
+      "step": 29932
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7725068883417239,
+      "learning_rate": 4.22222505116643e-06,
+      "loss": 0.9094,
+      "step": 29933
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8057888567423088,
+      "learning_rate": 4.221602278208431e-06,
+      "loss": 0.8264,
+      "step": 29934
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7796578156672993,
+      "learning_rate": 4.220979538894089e-06,
+      "loss": 0.7848,
+      "step": 29935
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7562539531368857,
+      "learning_rate": 4.220356833227034e-06,
+      "loss": 0.7356,
+      "step": 29936
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.06841381035024,
+      "learning_rate": 4.2197341612108944e-06,
+      "loss": 0.7999,
+      "step": 29937
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6844676316106901,
+      "learning_rate": 4.219111522849293e-06,
+      "loss": 0.8164,
+      "step": 29938
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.534068910601655,
+      "learning_rate": 4.21848891814585e-06,
+      "loss": 0.7444,
+      "step": 29939
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.729413094972385,
+      "learning_rate": 4.217866347104196e-06,
+      "loss": 0.8766,
+      "step": 29940
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5376974323994699,
+      "learning_rate": 4.217243809727958e-06,
+      "loss": 0.8642,
+      "step": 29941
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5373745378132537,
+      "learning_rate": 4.216621306020757e-06,
+      "loss": 0.8815,
+      "step": 29942
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.498389239170019,
+      "learning_rate": 4.215998835986215e-06,
+      "loss": 0.8667,
+      "step": 29943
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9470003958609363,
+      "learning_rate": 4.215376399627959e-06,
+      "loss": 0.9398,
+      "step": 29944
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7516165802671413,
+      "learning_rate": 4.214753996949616e-06,
+      "loss": 0.7938,
+      "step": 29945
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7844118960859484,
+      "learning_rate": 4.2141316279548065e-06,
+      "loss": 0.6992,
+      "step": 29946
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.727772245538109,
+      "learning_rate": 4.21350929264715e-06,
+      "loss": 0.8769,
+      "step": 29947
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.676688590135117,
+      "learning_rate": 4.21288699103028e-06,
+      "loss": 0.9046,
+      "step": 29948
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3995251027956495,
+      "learning_rate": 4.21226472310781e-06,
+      "loss": 0.8381,
+      "step": 29949
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7714820930836637,
+      "learning_rate": 4.211642488883369e-06,
+      "loss": 0.6914,
+      "step": 29950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0573702293352714,
+      "learning_rate": 4.211020288360576e-06,
+      "loss": 0.7965,
+      "step": 29951
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6058576039276897,
+      "learning_rate": 4.210398121543059e-06,
+      "loss": 0.696,
+      "step": 29952
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8514205806016775,
+      "learning_rate": 4.2097759884344335e-06,
+      "loss": 0.9367,
+      "step": 29953
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5412455382242585,
+      "learning_rate": 4.209153889038325e-06,
+      "loss": 0.8437,
+      "step": 29954
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5742839070663528,
+      "learning_rate": 4.208531823358364e-06,
+      "loss": 0.9416,
+      "step": 29955
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6970363896896228,
+      "learning_rate": 4.207909791398156e-06,
+      "loss": 0.741,
+      "step": 29956
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7735555886329728,
+      "learning_rate": 4.207287793161332e-06,
+      "loss": 0.7365,
+      "step": 29957
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7138171402302407,
+      "learning_rate": 4.2066658286515115e-06,
+      "loss": 0.7199,
+      "step": 29958
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5488323408202644,
+      "learning_rate": 4.206043897872324e-06,
+      "loss": 0.7665,
+      "step": 29959
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7208954181449816,
+      "learning_rate": 4.2054220008273765e-06,
+      "loss": 0.72,
+      "step": 29960
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5617412679640723,
+      "learning_rate": 4.204800137520296e-06,
+      "loss": 0.9006,
+      "step": 29961
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5291620531324723,
+      "learning_rate": 4.204178307954708e-06,
+      "loss": 0.9077,
+      "step": 29962
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6963043570275071,
+      "learning_rate": 4.203556512134225e-06,
+      "loss": 0.8578,
+      "step": 29963
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5586466597454958,
+      "learning_rate": 4.202934750062475e-06,
+      "loss": 0.8525,
+      "step": 29964
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7829483228026864,
+      "learning_rate": 4.202313021743071e-06,
+      "loss": 0.7723,
+      "step": 29965
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.039140084711358,
+      "learning_rate": 4.201691327179639e-06,
+      "loss": 0.8093,
+      "step": 29966
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5230858584183573,
+      "learning_rate": 4.201069666375792e-06,
+      "loss": 0.7934,
+      "step": 29967
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.765426536273967,
+      "learning_rate": 4.2004480393351565e-06,
+      "loss": 0.7241,
+      "step": 29968
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5684907566986883,
+      "learning_rate": 4.19982644606135e-06,
+      "loss": 0.852,
+      "step": 29969
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.166447709074528,
+      "learning_rate": 4.199204886557986e-06,
+      "loss": 0.8517,
+      "step": 29970
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5832592549666686,
+      "learning_rate": 4.198583360828688e-06,
+      "loss": 0.8463,
+      "step": 29971
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6504387359512342,
+      "learning_rate": 4.197961868877078e-06,
+      "loss": 0.8347,
+      "step": 29972
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7320185263556189,
+      "learning_rate": 4.19734041070677e-06,
+      "loss": 0.7042,
+      "step": 29973
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.604240744087504,
+      "learning_rate": 4.19671898632138e-06,
+      "loss": 0.9238,
+      "step": 29974
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8164784296902146,
+      "learning_rate": 4.1960975957245285e-06,
+      "loss": 0.8659,
+      "step": 29975
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6854994199179139,
+      "learning_rate": 4.19547623891984e-06,
+      "loss": 0.8495,
+      "step": 29976
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.170019591235002,
+      "learning_rate": 4.194854915910925e-06,
+      "loss": 0.8866,
+      "step": 29977
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5449571097973052,
+      "learning_rate": 4.194233626701399e-06,
+      "loss": 0.7959,
+      "step": 29978
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7918106359541408,
+      "learning_rate": 4.193612371294883e-06,
+      "loss": 0.8211,
+      "step": 29979
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5128446073798105,
+      "learning_rate": 4.1929911496949985e-06,
+      "loss": 0.8247,
+      "step": 29980
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5361769586729233,
+      "learning_rate": 4.192369961905357e-06,
+      "loss": 0.9444,
+      "step": 29981
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6584423952699408,
+      "learning_rate": 4.1917488079295735e-06,
+      "loss": 0.9505,
+      "step": 29982
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1013181736843976,
+      "learning_rate": 4.191127687771266e-06,
+      "loss": 0.9023,
+      "step": 29983
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4890208139973804,
+      "learning_rate": 4.190506601434057e-06,
+      "loss": 0.8166,
+      "step": 29984
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6968976844506833,
+      "learning_rate": 4.1898855489215574e-06,
+      "loss": 0.7584,
+      "step": 29985
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6287372480749973,
+      "learning_rate": 4.18926453023738e-06,
+      "loss": 0.8626,
+      "step": 29986
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5968390262058265,
+      "learning_rate": 4.188643545385148e-06,
+      "loss": 0.8968,
+      "step": 29987
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6408237492288433,
+      "learning_rate": 4.1880225943684695e-06,
+      "loss": 1.0125,
+      "step": 29988
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5067625396421613,
+      "learning_rate": 4.187401677190966e-06,
+      "loss": 0.8805,
+      "step": 29989
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4225814939558048,
+      "learning_rate": 4.186780793856247e-06,
+      "loss": 0.8434,
+      "step": 29990
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6128654165594982,
+      "learning_rate": 4.186159944367936e-06,
+      "loss": 0.7028,
+      "step": 29991
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6308309126593128,
+      "learning_rate": 4.185539128729636e-06,
+      "loss": 0.8519,
+      "step": 29992
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6427057608541118,
+      "learning_rate": 4.184918346944974e-06,
+      "loss": 0.7167,
+      "step": 29993
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7233410602276518,
+      "learning_rate": 4.1842975990175575e-06,
+      "loss": 1.0307,
+      "step": 29994
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7998711641120664,
+      "learning_rate": 4.183676884950998e-06,
+      "loss": 0.7754,
+      "step": 29995
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4754899820852516,
+      "learning_rate": 4.183056204748913e-06,
+      "loss": 0.8792,
+      "step": 29996
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5807880819777558,
+      "learning_rate": 4.18243555841492e-06,
+      "loss": 0.8711,
+      "step": 29997
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5546382863391026,
+      "learning_rate": 4.181814945952629e-06,
+      "loss": 0.8035,
+      "step": 29998
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.524321653369705,
+      "learning_rate": 4.18119436736565e-06,
+      "loss": 0.8296,
+      "step": 29999
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5821904957171629,
+      "learning_rate": 4.1805738226575995e-06,
+      "loss": 0.6999,
+      "step": 30000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5125549900008726,
+      "learning_rate": 4.179953311832094e-06,
+      "loss": 0.7841,
+      "step": 30001
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6514403449792345,
+      "learning_rate": 4.179332834892743e-06,
+      "loss": 0.9097,
+      "step": 30002
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7999939098453178,
+      "learning_rate": 4.178712391843156e-06,
+      "loss": 0.7362,
+      "step": 30003
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.569604852675638,
+      "learning_rate": 4.178091982686948e-06,
+      "loss": 0.7475,
+      "step": 30004
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6684520219404464,
+      "learning_rate": 4.177471607427736e-06,
+      "loss": 0.8686,
+      "step": 30005
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6543423229877234,
+      "learning_rate": 4.176851266069126e-06,
+      "loss": 0.8695,
+      "step": 30006
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6299967895949103,
+      "learning_rate": 4.176230958614729e-06,
+      "loss": 0.9057,
+      "step": 30007
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7675148521046133,
+      "learning_rate": 4.1756106850681624e-06,
+      "loss": 0.9441,
+      "step": 30008
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5284543370363677,
+      "learning_rate": 4.174990445433032e-06,
+      "loss": 0.8168,
+      "step": 30009
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.135847647847564,
+      "learning_rate": 4.174370239712952e-06,
+      "loss": 0.7318,
+      "step": 30010
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.676269302293838,
+      "learning_rate": 4.173750067911532e-06,
+      "loss": 0.793,
+      "step": 30011
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5775490941312817,
+      "learning_rate": 4.173129930032387e-06,
+      "loss": 0.9016,
+      "step": 30012
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.728484765834051,
+      "learning_rate": 4.17250982607912e-06,
+      "loss": 0.836,
+      "step": 30013
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4220012732396285,
+      "learning_rate": 4.17188975605535e-06,
+      "loss": 0.8588,
+      "step": 30014
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.481594548884287,
+      "learning_rate": 4.171269719964681e-06,
+      "loss": 0.8317,
+      "step": 30015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5977827272522693,
+      "learning_rate": 4.170649717810723e-06,
+      "loss": 0.8672,
+      "step": 30016
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8056972665271458,
+      "learning_rate": 4.170029749597087e-06,
+      "loss": 0.8917,
+      "step": 30017
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.656239476056616,
+      "learning_rate": 4.169409815327388e-06,
+      "loss": 0.8409,
+      "step": 30018
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8011710749028,
+      "learning_rate": 4.168789915005233e-06,
+      "loss": 0.7138,
+      "step": 30019
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5695775764338293,
+      "learning_rate": 4.168170048634223e-06,
+      "loss": 0.8281,
+      "step": 30020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7456019120968426,
+      "learning_rate": 4.167550216217974e-06,
+      "loss": 0.6856,
+      "step": 30021
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5322864681187311,
+      "learning_rate": 4.1669304177600964e-06,
+      "loss": 0.7796,
+      "step": 30022
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4278179964408704,
+      "learning_rate": 4.166310653264199e-06,
+      "loss": 0.9206,
+      "step": 30023
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6774827005946973,
+      "learning_rate": 4.1656909227338826e-06,
+      "loss": 0.8741,
+      "step": 30024
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6289991234134915,
+      "learning_rate": 4.1650712261727614e-06,
+      "loss": 0.8841,
+      "step": 30025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9304080037423907,
+      "learning_rate": 4.164451563584446e-06,
+      "loss": 0.9299,
+      "step": 30026
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.435873164184879,
+      "learning_rate": 4.163831934972541e-06,
+      "loss": 0.7335,
+      "step": 30027
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5040445247548448,
+      "learning_rate": 4.163212340340651e-06,
+      "loss": 0.8809,
+      "step": 30028
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7768254893099176,
+      "learning_rate": 4.16259277969239e-06,
+      "loss": 0.9434,
+      "step": 30029
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.600084421253995,
+      "learning_rate": 4.161973253031359e-06,
+      "loss": 0.8398,
+      "step": 30030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7826266368806942,
+      "learning_rate": 4.161353760361171e-06,
+      "loss": 0.7231,
+      "step": 30031
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5153300506482423,
+      "learning_rate": 4.160734301685427e-06,
+      "loss": 0.7533,
+      "step": 30032
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5942266075326108,
+      "learning_rate": 4.160114877007739e-06,
+      "loss": 0.7227,
+      "step": 30033
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.117130485341364,
+      "learning_rate": 4.159495486331708e-06,
+      "loss": 0.7927,
+      "step": 30034
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7006619888713663,
+      "learning_rate": 4.1588761296609485e-06,
+      "loss": 0.8394,
+      "step": 30035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6378359466065622,
+      "learning_rate": 4.158256806999059e-06,
+      "loss": 0.8096,
+      "step": 30036
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5069194877810812,
+      "learning_rate": 4.157637518349646e-06,
+      "loss": 0.775,
+      "step": 30037
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6736509602046412,
+      "learning_rate": 4.1570182637163155e-06,
+      "loss": 0.7765,
+      "step": 30038
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.662353430065327,
+      "learning_rate": 4.156399043102675e-06,
+      "loss": 0.7793,
+      "step": 30039
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8762764763127768,
+      "learning_rate": 4.155779856512338e-06,
+      "loss": 0.8293,
+      "step": 30040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8258099611390801,
+      "learning_rate": 4.155160703948892e-06,
+      "loss": 0.8538,
+      "step": 30041
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5104747067604747,
+      "learning_rate": 4.154541585415951e-06,
+      "loss": 0.8623,
+      "step": 30042
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5226955205184072,
+      "learning_rate": 4.153922500917123e-06,
+      "loss": 0.8917,
+      "step": 30043
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6634050989177185,
+      "learning_rate": 4.15330345045601e-06,
+      "loss": 0.8288,
+      "step": 30044
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7281212101792067,
+      "learning_rate": 4.152684434036212e-06,
+      "loss": 0.6798,
+      "step": 30045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5183158501935596,
+      "learning_rate": 4.152065451661334e-06,
+      "loss": 0.893,
+      "step": 30046
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6314342064278518,
+      "learning_rate": 4.151446503334988e-06,
+      "loss": 0.7812,
+      "step": 30047
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5603360268185547,
+      "learning_rate": 4.150827589060769e-06,
+      "loss": 0.8087,
+      "step": 30048
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7582608267971865,
+      "learning_rate": 4.150208708842286e-06,
+      "loss": 0.6954,
+      "step": 30049
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4666021314753985,
+      "learning_rate": 4.149589862683141e-06,
+      "loss": 0.6623,
+      "step": 30050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7891518459003762,
+      "learning_rate": 4.1489710505869305e-06,
+      "loss": 0.6982,
+      "step": 30051
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.519101430719181,
+      "learning_rate": 4.148352272557264e-06,
+      "loss": 0.8496,
+      "step": 30052
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.820150298110518,
+      "learning_rate": 4.147733528597748e-06,
+      "loss": 0.8777,
+      "step": 30053
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0505816250525792,
+      "learning_rate": 4.147114818711979e-06,
+      "loss": 0.6296,
+      "step": 30054
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7502656454127218,
+      "learning_rate": 4.146496142903557e-06,
+      "loss": 0.7344,
+      "step": 30055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5229250194159434,
+      "learning_rate": 4.145877501176088e-06,
+      "loss": 0.7525,
+      "step": 30056
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.172119465251554,
+      "learning_rate": 4.145258893533179e-06,
+      "loss": 0.9363,
+      "step": 30057
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5948891022332776,
+      "learning_rate": 4.144640319978422e-06,
+      "loss": 0.7954,
+      "step": 30058
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4842224717985815,
+      "learning_rate": 4.144021780515421e-06,
+      "loss": 0.7829,
+      "step": 30059
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6838578271674265,
+      "learning_rate": 4.1434032751477794e-06,
+      "loss": 0.7862,
+      "step": 30060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6833873931262666,
+      "learning_rate": 4.142784803879102e-06,
+      "loss": 0.8505,
+      "step": 30061
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5484889703220512,
+      "learning_rate": 4.142166366712985e-06,
+      "loss": 0.8913,
+      "step": 30062
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.038680847314602,
+      "learning_rate": 4.141547963653027e-06,
+      "loss": 0.8722,
+      "step": 30063
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5352446844226362,
+      "learning_rate": 4.140929594702836e-06,
+      "loss": 0.8688,
+      "step": 30064
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.640298479092134,
+      "learning_rate": 4.140311259866003e-06,
+      "loss": 0.9447,
+      "step": 30065
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6749175964219183,
+      "learning_rate": 4.139692959146138e-06,
+      "loss": 0.7886,
+      "step": 30066
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5491365385962943,
+      "learning_rate": 4.139074692546831e-06,
+      "loss": 0.8658,
+      "step": 30067
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7786274484816225,
+      "learning_rate": 4.138456460071691e-06,
+      "loss": 0.7287,
+      "step": 30068
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8242916467471104,
+      "learning_rate": 4.137838261724308e-06,
+      "loss": 0.7962,
+      "step": 30069
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7109806094177875,
+      "learning_rate": 4.137220097508292e-06,
+      "loss": 0.7946,
+      "step": 30070
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.527608878860428,
+      "learning_rate": 4.136601967427236e-06,
+      "loss": 0.9036,
+      "step": 30071
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7522410417476506,
+      "learning_rate": 4.135983871484735e-06,
+      "loss": 0.7453,
+      "step": 30072
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.585003230819823,
+      "learning_rate": 4.135365809684394e-06,
+      "loss": 0.8068,
+      "step": 30073
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6191993537251412,
+      "learning_rate": 4.134747782029814e-06,
+      "loss": 0.8208,
+      "step": 30074
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7221190972501101,
+      "learning_rate": 4.134129788524588e-06,
+      "loss": 0.9002,
+      "step": 30075
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5779922484362987,
+      "learning_rate": 4.133511829172313e-06,
+      "loss": 0.8739,
+      "step": 30076
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6885152046729182,
+      "learning_rate": 4.132893903976588e-06,
+      "loss": 0.9407,
+      "step": 30077
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.51429259624567,
+      "learning_rate": 4.132276012941016e-06,
+      "loss": 0.8253,
+      "step": 30078
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4641588125675604,
+      "learning_rate": 4.1316581560691924e-06,
+      "loss": 0.9123,
+      "step": 30079
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.686314751583417,
+      "learning_rate": 4.131040333364708e-06,
+      "loss": 0.827,
+      "step": 30080
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4586799991413453,
+      "learning_rate": 4.130422544831165e-06,
+      "loss": 0.9168,
+      "step": 30081
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7282539178816643,
+      "learning_rate": 4.129804790472164e-06,
+      "loss": 0.9444,
+      "step": 30082
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6329366955283835,
+      "learning_rate": 4.129187070291298e-06,
+      "loss": 0.9418,
+      "step": 30083
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6466219465301106,
+      "learning_rate": 4.128569384292161e-06,
+      "loss": 0.7597,
+      "step": 30084
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5685366342995886,
+      "learning_rate": 4.127951732478353e-06,
+      "loss": 0.8025,
+      "step": 30085
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.60904980293855,
+      "learning_rate": 4.127334114853467e-06,
+      "loss": 0.8443,
+      "step": 30086
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5765964772306114,
+      "learning_rate": 4.126716531421105e-06,
+      "loss": 0.812,
+      "step": 30087
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8000969070207922,
+      "learning_rate": 4.126098982184854e-06,
+      "loss": 0.6955,
+      "step": 30088
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2194687628690275,
+      "learning_rate": 4.125481467148319e-06,
+      "loss": 0.7871,
+      "step": 30089
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.638836182222275,
+      "learning_rate": 4.124863986315087e-06,
+      "loss": 0.8019,
+      "step": 30090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6044632347269812,
+      "learning_rate": 4.1242465396887595e-06,
+      "loss": 0.8251,
+      "step": 30091
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5589187524549664,
+      "learning_rate": 4.12362912727293e-06,
+      "loss": 0.7548,
+      "step": 30092
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7448363518006453,
+      "learning_rate": 4.123011749071188e-06,
+      "loss": 0.7089,
+      "step": 30093
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7604610686894782,
+      "learning_rate": 4.122394405087132e-06,
+      "loss": 0.7167,
+      "step": 30094
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4471625126715497,
+      "learning_rate": 4.121777095324361e-06,
+      "loss": 0.7909,
+      "step": 30095
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5776498917089021,
+      "learning_rate": 4.121159819786465e-06,
+      "loss": 0.9393,
+      "step": 30096
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4639176513460366,
+      "learning_rate": 4.120542578477033e-06,
+      "loss": 0.852,
+      "step": 30097
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9381820071468048,
+      "learning_rate": 4.119925371399665e-06,
+      "loss": 0.6532,
+      "step": 30098
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7667163625486955,
+      "learning_rate": 4.119308198557955e-06,
+      "loss": 0.7521,
+      "step": 30099
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4753521235038474,
+      "learning_rate": 4.1186910599554954e-06,
+      "loss": 0.8304,
+      "step": 30100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7865655060235757,
+      "learning_rate": 4.118073955595873e-06,
+      "loss": 0.8188,
+      "step": 30101
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.537609075844262,
+      "learning_rate": 4.117456885482689e-06,
+      "loss": 0.8409,
+      "step": 30102
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6062568889572935,
+      "learning_rate": 4.116839849619536e-06,
+      "loss": 0.9293,
+      "step": 30103
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5752513081654262,
+      "learning_rate": 4.116222848010004e-06,
+      "loss": 0.8944,
+      "step": 30104
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8085747580485023,
+      "learning_rate": 4.1156058806576815e-06,
+      "loss": 0.7627,
+      "step": 30105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8614330808949844,
+      "learning_rate": 4.114988947566167e-06,
+      "loss": 0.8992,
+      "step": 30106
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5603850905440644,
+      "learning_rate": 4.114372048739048e-06,
+      "loss": 0.9103,
+      "step": 30107
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.031485211699221,
+      "learning_rate": 4.113755184179922e-06,
+      "loss": 0.8599,
+      "step": 30108
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6113409412661843,
+      "learning_rate": 4.113138353892373e-06,
+      "loss": 0.7369,
+      "step": 30109
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6085636409855446,
+      "learning_rate": 4.11252155788e-06,
+      "loss": 0.8025,
+      "step": 30110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.661710869266473,
+      "learning_rate": 4.1119047961463855e-06,
+      "loss": 0.8907,
+      "step": 30111
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4617744880376862,
+      "learning_rate": 4.1112880686951306e-06,
+      "loss": 0.8321,
+      "step": 30112
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6160137168189164,
+      "learning_rate": 4.110671375529817e-06,
+      "loss": 0.8578,
+      "step": 30113
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8016934424898381,
+      "learning_rate": 4.110054716654042e-06,
+      "loss": 0.7463,
+      "step": 30114
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.092906368157351,
+      "learning_rate": 4.109438092071389e-06,
+      "loss": 0.8503,
+      "step": 30115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5440349251811942,
+      "learning_rate": 4.108821501785458e-06,
+      "loss": 0.9601,
+      "step": 30116
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7965138969347476,
+      "learning_rate": 4.1082049457998316e-06,
+      "loss": 0.9087,
+      "step": 30117
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6106007927410413,
+      "learning_rate": 4.107588424118098e-06,
+      "loss": 0.7843,
+      "step": 30118
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4843460317031196,
+      "learning_rate": 4.106971936743851e-06,
+      "loss": 0.8947,
+      "step": 30119
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5548700788420289,
+      "learning_rate": 4.106355483680683e-06,
+      "loss": 0.7856,
+      "step": 30120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5788342905318802,
+      "learning_rate": 4.105739064932178e-06,
+      "loss": 0.7309,
+      "step": 30121
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.86901272118571,
+      "learning_rate": 4.105122680501923e-06,
+      "loss": 0.8377,
+      "step": 30122
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6854196911218498,
+      "learning_rate": 4.1045063303935105e-06,
+      "loss": 0.8882,
+      "step": 30123
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6641096313556114,
+      "learning_rate": 4.103890014610532e-06,
+      "loss": 0.7816,
+      "step": 30124
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.914080196430791,
+      "learning_rate": 4.103273733156572e-06,
+      "loss": 0.7492,
+      "step": 30125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7546366973408067,
+      "learning_rate": 4.102657486035216e-06,
+      "loss": 0.6714,
+      "step": 30126
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7151367831303457,
+      "learning_rate": 4.102041273250055e-06,
+      "loss": 0.8356,
+      "step": 30127
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4494291022585388,
+      "learning_rate": 4.1014250948046805e-06,
+      "loss": 0.8131,
+      "step": 30128
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4987296217991006,
+      "learning_rate": 4.100808950702677e-06,
+      "loss": 0.7318,
+      "step": 30129
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.508335766507471,
+      "learning_rate": 4.100192840947628e-06,
+      "loss": 0.7799,
+      "step": 30130
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5059230803492027,
+      "learning_rate": 4.099576765543128e-06,
+      "loss": 0.9563,
+      "step": 30131
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.743108427154305,
+      "learning_rate": 4.098960724492756e-06,
+      "loss": 0.8188,
+      "step": 30132
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.621390963451724,
+      "learning_rate": 4.098344717800105e-06,
+      "loss": 0.8472,
+      "step": 30133
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6288787772085422,
+      "learning_rate": 4.0977287454687575e-06,
+      "loss": 0.8855,
+      "step": 30134
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6023453757812494,
+      "learning_rate": 4.097112807502306e-06,
+      "loss": 0.932,
+      "step": 30135
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7163254570844761,
+      "learning_rate": 4.096496903904328e-06,
+      "loss": 0.7704,
+      "step": 30136
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4816200127643127,
+      "learning_rate": 4.095881034678414e-06,
+      "loss": 0.9012,
+      "step": 30137
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7640281845885906,
+      "learning_rate": 4.095265199828157e-06,
+      "loss": 0.706,
+      "step": 30138
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.566895442649146,
+      "learning_rate": 4.094649399357129e-06,
+      "loss": 0.8904,
+      "step": 30139
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.526852953077378,
+      "learning_rate": 4.0940336332689216e-06,
+      "loss": 0.9522,
+      "step": 30140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5779943098926286,
+      "learning_rate": 4.093417901567119e-06,
+      "loss": 0.8409,
+      "step": 30141
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4917317818265357,
+      "learning_rate": 4.092802204255316e-06,
+      "loss": 0.9317,
+      "step": 30142
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.490381798761715,
+      "learning_rate": 4.092186541337081e-06,
+      "loss": 0.9651,
+      "step": 30143
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6909722623912715,
+      "learning_rate": 4.091570912816006e-06,
+      "loss": 0.9038,
+      "step": 30144
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.827215688093877,
+      "learning_rate": 4.090955318695679e-06,
+      "loss": 0.7504,
+      "step": 30145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7433194599421125,
+      "learning_rate": 4.090339758979679e-06,
+      "loss": 0.9045,
+      "step": 30146
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5997241322402271,
+      "learning_rate": 4.089724233671594e-06,
+      "loss": 0.8628,
+      "step": 30147
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7741839483129387,
+      "learning_rate": 4.089108742775003e-06,
+      "loss": 0.6995,
+      "step": 30148
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2985770015109086,
+      "learning_rate": 4.088493286293495e-06,
+      "loss": 0.8438,
+      "step": 30149
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3861537631970993,
+      "learning_rate": 4.0878778642306475e-06,
+      "loss": 0.7155,
+      "step": 30150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8430161143834851,
+      "learning_rate": 4.08726247659005e-06,
+      "loss": 0.7538,
+      "step": 30151
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6410374254661335,
+      "learning_rate": 4.086647123375282e-06,
+      "loss": 0.9003,
+      "step": 30152
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.106242454792911,
+      "learning_rate": 4.086031804589925e-06,
+      "loss": 0.8177,
+      "step": 30153
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8183142341168015,
+      "learning_rate": 4.085416520237562e-06,
+      "loss": 0.8215,
+      "step": 30154
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.552380913695353,
+      "learning_rate": 4.0848012703217785e-06,
+      "loss": 0.8135,
+      "step": 30155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3080863396525952,
+      "learning_rate": 4.0841860548461565e-06,
+      "loss": 0.7747,
+      "step": 30156
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7714603568024795,
+      "learning_rate": 4.083570873814273e-06,
+      "loss": 0.704,
+      "step": 30157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3801923632494435,
+      "learning_rate": 4.082955727229712e-06,
+      "loss": 0.8837,
+      "step": 30158
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3904570458389889,
+      "learning_rate": 4.08234061509606e-06,
+      "loss": 0.8086,
+      "step": 30159
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4865802147039275,
+      "learning_rate": 4.081725537416894e-06,
+      "loss": 0.8627,
+      "step": 30160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5550021621109889,
+      "learning_rate": 4.0811104941957924e-06,
+      "loss": 0.8,
+      "step": 30161
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6152645574974867,
+      "learning_rate": 4.0804954854363395e-06,
+      "loss": 0.7922,
+      "step": 30162
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7296845512688249,
+      "learning_rate": 4.079880511142119e-06,
+      "loss": 0.7012,
+      "step": 30163
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7499744107190219,
+      "learning_rate": 4.07926557131671e-06,
+      "loss": 0.7478,
+      "step": 30164
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7374282349081895,
+      "learning_rate": 4.078650665963686e-06,
+      "loss": 0.8974,
+      "step": 30165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6035910631400596,
+      "learning_rate": 4.078035795086636e-06,
+      "loss": 0.7997,
+      "step": 30166
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7843392554561253,
+      "learning_rate": 4.077420958689134e-06,
+      "loss": 0.8507,
+      "step": 30167
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7561637224066146,
+      "learning_rate": 4.076806156774765e-06,
+      "loss": 0.7109,
+      "step": 30168
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5472969893082977,
+      "learning_rate": 4.076191389347103e-06,
+      "loss": 0.7597,
+      "step": 30169
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5359513305694894,
+      "learning_rate": 4.075576656409733e-06,
+      "loss": 0.8595,
+      "step": 30170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.553058164484637,
+      "learning_rate": 4.074961957966228e-06,
+      "loss": 0.8293,
+      "step": 30171
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5094766970137874,
+      "learning_rate": 4.0743472940201746e-06,
+      "loss": 0.9026,
+      "step": 30172
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4739388270555212,
+      "learning_rate": 4.073732664575146e-06,
+      "loss": 0.865,
+      "step": 30173
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7447415031445838,
+      "learning_rate": 4.073118069634719e-06,
+      "loss": 0.6682,
+      "step": 30174
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4784433240185286,
+      "learning_rate": 4.072503509202476e-06,
+      "loss": 0.8075,
+      "step": 30175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.415470284856143,
+      "learning_rate": 4.071888983281996e-06,
+      "loss": 0.7316,
+      "step": 30176
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7326832476393674,
+      "learning_rate": 4.071274491876856e-06,
+      "loss": 0.7713,
+      "step": 30177
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21281696400679,
+      "learning_rate": 4.07066003499063e-06,
+      "loss": 0.8928,
+      "step": 30178
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7622023226584431,
+      "learning_rate": 4.070045612626898e-06,
+      "loss": 0.7473,
+      "step": 30179
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.640969732001809,
+      "learning_rate": 4.069431224789242e-06,
+      "loss": 0.8807,
+      "step": 30180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5317587932046852,
+      "learning_rate": 4.068816871481234e-06,
+      "loss": 0.727,
+      "step": 30181
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8980276076664364,
+      "learning_rate": 4.0682025527064486e-06,
+      "loss": 0.7337,
+      "step": 30182
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6920349503536332,
+      "learning_rate": 4.067588268468467e-06,
+      "loss": 0.7765,
+      "step": 30183
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8288897533255302,
+      "learning_rate": 4.066974018770866e-06,
+      "loss": 0.7595,
+      "step": 30184
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7968804628545711,
+      "learning_rate": 4.0663598036172225e-06,
+      "loss": 0.9075,
+      "step": 30185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7508574249180782,
+      "learning_rate": 4.065745623011106e-06,
+      "loss": 0.7015,
+      "step": 30186
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5779126948694286,
+      "learning_rate": 4.065131476956101e-06,
+      "loss": 0.8143,
+      "step": 30187
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6898262567102162,
+      "learning_rate": 4.0645173654557766e-06,
+      "loss": 0.8986,
+      "step": 30188
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5603580222946318,
+      "learning_rate": 4.063903288513714e-06,
+      "loss": 0.8015,
+      "step": 30189
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6319630731962997,
+      "learning_rate": 4.063289246133484e-06,
+      "loss": 0.8595,
+      "step": 30190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5223170103691626,
+      "learning_rate": 4.062675238318665e-06,
+      "loss": 0.7932,
+      "step": 30191
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5070726564063317,
+      "learning_rate": 4.062061265072829e-06,
+      "loss": 0.7971,
+      "step": 30192
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5986333135503288,
+      "learning_rate": 4.061447326399554e-06,
+      "loss": 0.9968,
+      "step": 30193
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2341672688208116,
+      "learning_rate": 4.060833422302412e-06,
+      "loss": 0.8299,
+      "step": 30194
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4882668169234132,
+      "learning_rate": 4.060219552784976e-06,
+      "loss": 0.9297,
+      "step": 30195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8081206098986334,
+      "learning_rate": 4.059605717850823e-06,
+      "loss": 0.7097,
+      "step": 30196
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8771452615303919,
+      "learning_rate": 4.058991917503529e-06,
+      "loss": 0.7618,
+      "step": 30197
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6204812611998638,
+      "learning_rate": 4.058378151746665e-06,
+      "loss": 0.8278,
+      "step": 30198
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6930461765292992,
+      "learning_rate": 4.057764420583801e-06,
+      "loss": 0.8459,
+      "step": 30199
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6819558277766335,
+      "learning_rate": 4.057150724018514e-06,
+      "loss": 0.8915,
+      "step": 30200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6084625035630387,
+      "learning_rate": 4.0565370620543795e-06,
+      "loss": 0.9768,
+      "step": 30201
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6393301109094116,
+      "learning_rate": 4.055923434694968e-06,
+      "loss": 0.9158,
+      "step": 30202
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3229513541503795,
+      "learning_rate": 4.05530984194385e-06,
+      "loss": 0.7389,
+      "step": 30203
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.711014899524393,
+      "learning_rate": 4.054696283804599e-06,
+      "loss": 0.9912,
+      "step": 30204
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5498310700113804,
+      "learning_rate": 4.054082760280792e-06,
+      "loss": 0.8915,
+      "step": 30205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7138956536947616,
+      "learning_rate": 4.053469271375998e-06,
+      "loss": 0.829,
+      "step": 30206
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9752896858600508,
+      "learning_rate": 4.052855817093785e-06,
+      "loss": 0.8298,
+      "step": 30207
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.875301037295736,
+      "learning_rate": 4.0522423974377325e-06,
+      "loss": 0.9371,
+      "step": 30208
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.595729658859461,
+      "learning_rate": 4.051629012411403e-06,
+      "loss": 0.9067,
+      "step": 30209
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8081161448230214,
+      "learning_rate": 4.051015662018376e-06,
+      "loss": 0.7475,
+      "step": 30210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6145495369938183,
+      "learning_rate": 4.050402346262217e-06,
+      "loss": 0.9229,
+      "step": 30211
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4844205617382493,
+      "learning_rate": 4.049789065146502e-06,
+      "loss": 0.8043,
+      "step": 30212
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4768419290106525,
+      "learning_rate": 4.049175818674795e-06,
+      "loss": 0.8577,
+      "step": 30213
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.50399921181398,
+      "learning_rate": 4.048562606850674e-06,
+      "loss": 0.7691,
+      "step": 30214
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8060470019484636,
+      "learning_rate": 4.047949429677706e-06,
+      "loss": 0.8654,
+      "step": 30215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5599268881552981,
+      "learning_rate": 4.0473362871594576e-06,
+      "loss": 0.7959,
+      "step": 30216
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4398318400129093,
+      "learning_rate": 4.0467231792995024e-06,
+      "loss": 0.856,
+      "step": 30217
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.888811393471805,
+      "learning_rate": 4.046110106101412e-06,
+      "loss": 0.7461,
+      "step": 30218
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6316913269603952,
+      "learning_rate": 4.045497067568753e-06,
+      "loss": 0.832,
+      "step": 30219
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5778446248683968,
+      "learning_rate": 4.0448840637050934e-06,
+      "loss": 0.7576,
+      "step": 30220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.167657775849669,
+      "learning_rate": 4.044271094514003e-06,
+      "loss": 1.001,
+      "step": 30221
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6265706503808306,
+      "learning_rate": 4.043658159999055e-06,
+      "loss": 0.8939,
+      "step": 30222
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7883148186849598,
+      "learning_rate": 4.043045260163816e-06,
+      "loss": 0.8586,
+      "step": 30223
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7259991950547369,
+      "learning_rate": 4.0424323950118495e-06,
+      "loss": 0.9062,
+      "step": 30224
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7494771966920809,
+      "learning_rate": 4.041819564546727e-06,
+      "loss": 0.6884,
+      "step": 30225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5072692044932043,
+      "learning_rate": 4.041206768772023e-06,
+      "loss": 0.85,
+      "step": 30226
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5927398127250918,
+      "learning_rate": 4.040594007691297e-06,
+      "loss": 0.8445,
+      "step": 30227
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5560898475393232,
+      "learning_rate": 4.039981281308117e-06,
+      "loss": 0.8358,
+      "step": 30228
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4011047821230367,
+      "learning_rate": 4.039368589626057e-06,
+      "loss": 0.7967,
+      "step": 30229
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5539304674579026,
+      "learning_rate": 4.038755932648677e-06,
+      "loss": 0.8838,
+      "step": 30230
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6520141094920588,
+      "learning_rate": 4.03814331037955e-06,
+      "loss": 0.8265,
+      "step": 30231
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6675646656639902,
+      "learning_rate": 4.037530722822236e-06,
+      "loss": 0.8315,
+      "step": 30232
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.853394043678712,
+      "learning_rate": 4.036918169980311e-06,
+      "loss": 0.8678,
+      "step": 30233
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5910501949752278,
+      "learning_rate": 4.03630565185733e-06,
+      "loss": 0.8126,
+      "step": 30234
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6270974619242833,
+      "learning_rate": 4.035693168456869e-06,
+      "loss": 0.8851,
+      "step": 30235
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3355727968886448,
+      "learning_rate": 4.035080719782491e-06,
+      "loss": 0.7057,
+      "step": 30236
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6553007428542816,
+      "learning_rate": 4.0344683058377645e-06,
+      "loss": 0.7653,
+      "step": 30237
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.570610301715308,
+      "learning_rate": 4.033855926626247e-06,
+      "loss": 0.8306,
+      "step": 30238
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5322082222423803,
+      "learning_rate": 4.033243582151509e-06,
+      "loss": 0.7733,
+      "step": 30239
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7227694412627832,
+      "learning_rate": 4.032631272417123e-06,
+      "loss": 0.7485,
+      "step": 30240
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0913185624059953,
+      "learning_rate": 4.03201899742664e-06,
+      "loss": 0.8275,
+      "step": 30241
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6084628106938361,
+      "learning_rate": 4.031406757183632e-06,
+      "loss": 0.9909,
+      "step": 30242
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4271576203005405,
+      "learning_rate": 4.030794551691663e-06,
+      "loss": 0.8047,
+      "step": 30243
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7395168761901534,
+      "learning_rate": 4.030182380954303e-06,
+      "loss": 0.8898,
+      "step": 30244
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5929122044983783,
+      "learning_rate": 4.02957024497511e-06,
+      "loss": 0.7538,
+      "step": 30245
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.508083907095951,
+      "learning_rate": 4.028958143757647e-06,
+      "loss": 0.9601,
+      "step": 30246
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7421967487854144,
+      "learning_rate": 4.028346077305482e-06,
+      "loss": 0.8196,
+      "step": 30247
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.867484702109382,
+      "learning_rate": 4.027734045622172e-06,
+      "loss": 0.9213,
+      "step": 30248
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7416914552903782,
+      "learning_rate": 4.027122048711292e-06,
+      "loss": 0.7739,
+      "step": 30249
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5810435473195101,
+      "learning_rate": 4.026510086576393e-06,
+      "loss": 0.7701,
+      "step": 30250
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4806148504851055,
+      "learning_rate": 4.0258981592210464e-06,
+      "loss": 0.8508,
+      "step": 30251
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6942716684231205,
+      "learning_rate": 4.0252862666488105e-06,
+      "loss": 0.8253,
+      "step": 30252
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6250253847715932,
+      "learning_rate": 4.024674408863252e-06,
+      "loss": 0.7903,
+      "step": 30253
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4678122276942136,
+      "learning_rate": 4.02406258586793e-06,
+      "loss": 0.695,
+      "step": 30254
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8986859413209611,
+      "learning_rate": 4.023450797666405e-06,
+      "loss": 0.8452,
+      "step": 30255
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.666063865177595,
+      "learning_rate": 4.022839044262241e-06,
+      "loss": 0.9293,
+      "step": 30256
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.625912310611145,
+      "learning_rate": 4.022227325659005e-06,
+      "loss": 0.8554,
+      "step": 30257
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5896581824421467,
+      "learning_rate": 4.0216156418602535e-06,
+      "loss": 0.8258,
+      "step": 30258
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.70234264492433,
+      "learning_rate": 4.021003992869545e-06,
+      "loss": 0.9562,
+      "step": 30259
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9009381727674912,
+      "learning_rate": 4.020392378690444e-06,
+      "loss": 0.9124,
+      "step": 30260
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7707535520080256,
+      "learning_rate": 4.019780799326515e-06,
+      "loss": 0.7164,
+      "step": 30261
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7873964800312349,
+      "learning_rate": 4.019169254781315e-06,
+      "loss": 0.7083,
+      "step": 30262
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7220819121067332,
+      "learning_rate": 4.018557745058403e-06,
+      "loss": 0.7153,
+      "step": 30263
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7418363828029813,
+      "learning_rate": 4.01794627016134e-06,
+      "loss": 0.7478,
+      "step": 30264
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6906885901471116,
+      "learning_rate": 4.017334830093693e-06,
+      "loss": 0.853,
+      "step": 30265
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6740712371946516,
+      "learning_rate": 4.016723424859016e-06,
+      "loss": 0.8958,
+      "step": 30266
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7093079843839916,
+      "learning_rate": 4.016112054460866e-06,
+      "loss": 0.9868,
+      "step": 30267
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9356004075759687,
+      "learning_rate": 4.0155007189028086e-06,
+      "loss": 0.7814,
+      "step": 30268
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5980500095834855,
+      "learning_rate": 4.014889418188398e-06,
+      "loss": 0.878,
+      "step": 30269
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7243531835446306,
+      "learning_rate": 4.014278152321199e-06,
+      "loss": 0.8954,
+      "step": 30270
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0048881204128843,
+      "learning_rate": 4.013666921304765e-06,
+      "loss": 0.8332,
+      "step": 30271
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6596044941484012,
+      "learning_rate": 4.013055725142661e-06,
+      "loss": 0.654,
+      "step": 30272
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7352669567427241,
+      "learning_rate": 4.012444563838437e-06,
+      "loss": 0.7955,
+      "step": 30273
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9546436119540087,
+      "learning_rate": 4.011833437395661e-06,
+      "loss": 0.813,
+      "step": 30274
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5203844554157546,
+      "learning_rate": 4.011222345817886e-06,
+      "loss": 0.9442,
+      "step": 30275
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6633218508264545,
+      "learning_rate": 4.0106112891086675e-06,
+      "loss": 0.8558,
+      "step": 30276
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.779068566515328,
+      "learning_rate": 4.010000267271566e-06,
+      "loss": 0.9682,
+      "step": 30277
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.46950903190442,
+      "learning_rate": 4.009389280310144e-06,
+      "loss": 0.8406,
+      "step": 30278
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7373254841356761,
+      "learning_rate": 4.008778328227953e-06,
+      "loss": 0.6982,
+      "step": 30279
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6892456995929421,
+      "learning_rate": 4.008167411028548e-06,
+      "loss": 1.0086,
+      "step": 30280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.803604671615575,
+      "learning_rate": 4.007556528715489e-06,
+      "loss": 0.8359,
+      "step": 30281
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.425829167695756,
+      "learning_rate": 4.006945681292338e-06,
+      "loss": 0.8179,
+      "step": 30282
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.512082447214055,
+      "learning_rate": 4.0063348687626455e-06,
+      "loss": 0.7512,
+      "step": 30283
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6010312672740783,
+      "learning_rate": 4.005724091129965e-06,
+      "loss": 0.8332,
+      "step": 30284
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6095555958108632,
+      "learning_rate": 4.005113348397858e-06,
+      "loss": 0.9207,
+      "step": 30285
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4701053983429273,
+      "learning_rate": 4.004502640569882e-06,
+      "loss": 0.782,
+      "step": 30286
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.759002828666681,
+      "learning_rate": 4.003891967649589e-06,
+      "loss": 0.7834,
+      "step": 30287
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6716235219026399,
+      "learning_rate": 4.003281329640533e-06,
+      "loss": 0.8943,
+      "step": 30288
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6424412368572383,
+      "learning_rate": 4.002670726546274e-06,
+      "loss": 0.8477,
+      "step": 30289
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5855374577494716,
+      "learning_rate": 4.002060158370361e-06,
+      "loss": 0.8384,
+      "step": 30290
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5503886535920155,
+      "learning_rate": 4.001449625116357e-06,
+      "loss": 0.8065,
+      "step": 30291
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5753364716836145,
+      "learning_rate": 4.000839126787809e-06,
+      "loss": 0.8068,
+      "step": 30292
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6487235965005411,
+      "learning_rate": 4.000228663388278e-06,
+      "loss": 0.8397,
+      "step": 30293
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.773438570618569,
+      "learning_rate": 3.999618234921312e-06,
+      "loss": 0.9977,
+      "step": 30294
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7394199321128839,
+      "learning_rate": 3.999007841390472e-06,
+      "loss": 0.6986,
+      "step": 30295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6228321635047165,
+      "learning_rate": 3.998397482799307e-06,
+      "loss": 0.8142,
+      "step": 30296
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8277682453847945,
+      "learning_rate": 3.9977871591513705e-06,
+      "loss": 0.824,
+      "step": 30297
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8161250921496759,
+      "learning_rate": 3.9971768704502156e-06,
+      "loss": 0.8286,
+      "step": 30298
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.551833132600255,
+      "learning_rate": 3.996566616699402e-06,
+      "loss": 0.842,
+      "step": 30299
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.509730641140475,
+      "learning_rate": 3.995956397902478e-06,
+      "loss": 0.8146,
+      "step": 30300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7498965034191016,
+      "learning_rate": 3.995346214062993e-06,
+      "loss": 0.8817,
+      "step": 30301
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4685318349188254,
+      "learning_rate": 3.994736065184503e-06,
+      "loss": 0.7692,
+      "step": 30302
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7864779448693083,
+      "learning_rate": 3.9941259512705654e-06,
+      "loss": 0.6845,
+      "step": 30303
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.582481788535632,
+      "learning_rate": 3.993515872324727e-06,
+      "loss": 0.8799,
+      "step": 30304
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6502861338328898,
+      "learning_rate": 3.992905828350538e-06,
+      "loss": 0.674,
+      "step": 30305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.590538763358762,
+      "learning_rate": 3.992295819351553e-06,
+      "loss": 0.8554,
+      "step": 30306
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.818099572412754,
+      "learning_rate": 3.991685845331326e-06,
+      "loss": 0.8105,
+      "step": 30307
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7895658869203652,
+      "learning_rate": 3.9910759062934065e-06,
+      "loss": 0.7136,
+      "step": 30308
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7210123946898155,
+      "learning_rate": 3.990466002241344e-06,
+      "loss": 0.8034,
+      "step": 30309
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6331873788082534,
+      "learning_rate": 3.989856133178692e-06,
+      "loss": 0.8236,
+      "step": 30310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5702182651574197,
+      "learning_rate": 3.989246299108999e-06,
+      "loss": 0.6984,
+      "step": 30311
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.543756432394832,
+      "learning_rate": 3.988636500035819e-06,
+      "loss": 0.969,
+      "step": 30312
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6210130977387998,
+      "learning_rate": 3.988026735962698e-06,
+      "loss": 0.7969,
+      "step": 30313
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6464486368726587,
+      "learning_rate": 3.987417006893191e-06,
+      "loss": 0.8942,
+      "step": 30314
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4696011601261438,
+      "learning_rate": 3.986807312830843e-06,
+      "loss": 0.8827,
+      "step": 30315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.62483843019309,
+      "learning_rate": 3.98619765377921e-06,
+      "loss": 0.8384,
+      "step": 30316
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5790754989836342,
+      "learning_rate": 3.985588029741838e-06,
+      "loss": 0.9129,
+      "step": 30317
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6856723044920894,
+      "learning_rate": 3.984978440722274e-06,
+      "loss": 0.8836,
+      "step": 30318
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8164606249012851,
+      "learning_rate": 3.9843688867240695e-06,
+      "loss": 0.84,
+      "step": 30319
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.146225493600363,
+      "learning_rate": 3.983759367750772e-06,
+      "loss": 0.9488,
+      "step": 30320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4756472936441272,
+      "learning_rate": 3.9831498838059415e-06,
+      "loss": 0.9012,
+      "step": 30321
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7635744087125242,
+      "learning_rate": 3.982540434893109e-06,
+      "loss": 0.774,
+      "step": 30322
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.61880642553493,
+      "learning_rate": 3.9819310210158324e-06,
+      "loss": 0.7668,
+      "step": 30323
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5658678707557898,
+      "learning_rate": 3.981321642177662e-06,
+      "loss": 0.8138,
+      "step": 30324
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8431719079621984,
+      "learning_rate": 3.980712298382141e-06,
+      "loss": 0.9227,
+      "step": 30325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7825623030769553,
+      "learning_rate": 3.980102989632816e-06,
+      "loss": 0.8308,
+      "step": 30326
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5900710429118006,
+      "learning_rate": 3.979493715933238e-06,
+      "loss": 0.7855,
+      "step": 30327
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.566806675738351,
+      "learning_rate": 3.978884477286956e-06,
+      "loss": 0.7651,
+      "step": 30328
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7512120083371419,
+      "learning_rate": 3.9782752736975165e-06,
+      "loss": 0.916,
+      "step": 30329
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.518960154968118,
+      "learning_rate": 3.97766610516846e-06,
+      "loss": 0.8871,
+      "step": 30330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0593394040212307,
+      "learning_rate": 3.977056971703342e-06,
+      "loss": 0.7948,
+      "step": 30331
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.801052364995032,
+      "learning_rate": 3.9764478733057e-06,
+      "loss": 0.8003,
+      "step": 30332
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5681678082053712,
+      "learning_rate": 3.975838809979087e-06,
+      "loss": 0.7441,
+      "step": 30333
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6179210297637008,
+      "learning_rate": 3.975229781727052e-06,
+      "loss": 0.9129,
+      "step": 30334
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.673284804051491,
+      "learning_rate": 3.974620788553135e-06,
+      "loss": 0.955,
+      "step": 30335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7467747284520472,
+      "learning_rate": 3.97401183046088e-06,
+      "loss": 0.7809,
+      "step": 30336
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.427483037394177,
+      "learning_rate": 3.973402907453837e-06,
+      "loss": 0.8943,
+      "step": 30337
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6964911102469853,
+      "learning_rate": 3.972794019535556e-06,
+      "loss": 0.7683,
+      "step": 30338
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.734430349882786,
+      "learning_rate": 3.97218516670957e-06,
+      "loss": 0.855,
+      "step": 30339
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9306682064655183,
+      "learning_rate": 3.971576348979431e-06,
+      "loss": 0.781,
+      "step": 30340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4719771560446293,
+      "learning_rate": 3.970967566348682e-06,
+      "loss": 0.8108,
+      "step": 30341
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.45834024444004,
+      "learning_rate": 3.970358818820873e-06,
+      "loss": 0.7335,
+      "step": 30342
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6029274721760431,
+      "learning_rate": 3.9697501063995435e-06,
+      "loss": 0.9501,
+      "step": 30343
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7395455184132382,
+      "learning_rate": 3.969141429088235e-06,
+      "loss": 0.8068,
+      "step": 30344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5348893513243027,
+      "learning_rate": 3.968532786890496e-06,
+      "loss": 0.8445,
+      "step": 30345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.514554570687896,
+      "learning_rate": 3.967924179809868e-06,
+      "loss": 0.8293,
+      "step": 30346
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6447285390259223,
+      "learning_rate": 3.9673156078498964e-06,
+      "loss": 0.735,
+      "step": 30347
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5735688393209917,
+      "learning_rate": 3.966707071014121e-06,
+      "loss": 0.7608,
+      "step": 30348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.65862773293195,
+      "learning_rate": 3.966098569306091e-06,
+      "loss": 0.9011,
+      "step": 30349
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1265222290685255,
+      "learning_rate": 3.9654901027293405e-06,
+      "loss": 0.8568,
+      "step": 30350
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7069221032067488,
+      "learning_rate": 3.964881671287422e-06,
+      "loss": 0.7132,
+      "step": 30351
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5722731337975153,
+      "learning_rate": 3.964273274983872e-06,
+      "loss": 0.7764,
+      "step": 30352
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6495020805255096,
+      "learning_rate": 3.963664913822231e-06,
+      "loss": 0.7779,
+      "step": 30353
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7235810346195048,
+      "learning_rate": 3.963056587806044e-06,
+      "loss": 0.8417,
+      "step": 30354
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4179070103536715,
+      "learning_rate": 3.962448296938854e-06,
+      "loss": 0.8245,
+      "step": 30355
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5376483269154666,
+      "learning_rate": 3.961840041224204e-06,
+      "loss": 0.7841,
+      "step": 30356
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7697437448720494,
+      "learning_rate": 3.96123182066563e-06,
+      "loss": 0.9044,
+      "step": 30357
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6465130058740027,
+      "learning_rate": 3.960623635266675e-06,
+      "loss": 0.8827,
+      "step": 30358
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8407395163219783,
+      "learning_rate": 3.960015485030885e-06,
+      "loss": 0.6679,
+      "step": 30359
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5744495618830967,
+      "learning_rate": 3.959407369961796e-06,
+      "loss": 0.7623,
+      "step": 30360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.558738510131151,
+      "learning_rate": 3.958799290062947e-06,
+      "loss": 0.7329,
+      "step": 30361
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0210732332312102,
+      "learning_rate": 3.958191245337881e-06,
+      "loss": 0.7333,
+      "step": 30362
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5109658435965272,
+      "learning_rate": 3.957583235790142e-06,
+      "loss": 0.9004,
+      "step": 30363
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.196550099403384,
+      "learning_rate": 3.956975261423267e-06,
+      "loss": 0.973,
+      "step": 30364
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.940276829147386,
+      "learning_rate": 3.956367322240792e-06,
+      "loss": 0.7753,
+      "step": 30365
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5038350487859355,
+      "learning_rate": 3.955759418246259e-06,
+      "loss": 0.8994,
+      "step": 30366
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4996172545021835,
+      "learning_rate": 3.955151549443211e-06,
+      "loss": 0.9064,
+      "step": 30367
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.726502686663732,
+      "learning_rate": 3.954543715835185e-06,
+      "loss": 0.8592,
+      "step": 30368
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.597292094489886,
+      "learning_rate": 3.953935917425716e-06,
+      "loss": 0.7242,
+      "step": 30369
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6924001960898922,
+      "learning_rate": 3.953328154218351e-06,
+      "loss": 0.7756,
+      "step": 30370
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7506882106548277,
+      "learning_rate": 3.952720426216618e-06,
+      "loss": 0.6945,
+      "step": 30371
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8033887389988679,
+      "learning_rate": 3.952112733424067e-06,
+      "loss": 0.6759,
+      "step": 30372
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7839757365458301,
+      "learning_rate": 3.951505075844225e-06,
+      "loss": 0.762,
+      "step": 30373
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9304178912302985,
+      "learning_rate": 3.95089745348064e-06,
+      "loss": 0.9257,
+      "step": 30374
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.448056575202179,
+      "learning_rate": 3.950289866336842e-06,
+      "loss": 0.836,
+      "step": 30375
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7444877208721916,
+      "learning_rate": 3.949682314416374e-06,
+      "loss": 0.8534,
+      "step": 30376
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.165217168800868,
+      "learning_rate": 3.9490747977227715e-06,
+      "loss": 0.8271,
+      "step": 30377
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9554491422114169,
+      "learning_rate": 3.948467316259569e-06,
+      "loss": 0.8861,
+      "step": 30378
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7949446770368453,
+      "learning_rate": 3.947859870030304e-06,
+      "loss": 0.8139,
+      "step": 30379
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4029583317111656,
+      "learning_rate": 3.947252459038519e-06,
+      "loss": 0.8062,
+      "step": 30380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6414664467546884,
+      "learning_rate": 3.946645083287747e-06,
+      "loss": 0.989,
+      "step": 30381
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6051530322010077,
+      "learning_rate": 3.946037742781519e-06,
+      "loss": 0.9595,
+      "step": 30382
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.423855062023762,
+      "learning_rate": 3.945430437523376e-06,
+      "loss": 0.8038,
+      "step": 30383
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8631424133015257,
+      "learning_rate": 3.9448231675168584e-06,
+      "loss": 0.7595,
+      "step": 30384
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9570930218649666,
+      "learning_rate": 3.944215932765498e-06,
+      "loss": 0.6858,
+      "step": 30385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5689253729197017,
+      "learning_rate": 3.943608733272824e-06,
+      "loss": 0.8192,
+      "step": 30386
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5131430393718026,
+      "learning_rate": 3.943001569042378e-06,
+      "loss": 0.6652,
+      "step": 30387
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5239020446660088,
+      "learning_rate": 3.942394440077699e-06,
+      "loss": 0.7973,
+      "step": 30388
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7661468757199326,
+      "learning_rate": 3.9417873463823165e-06,
+      "loss": 0.8955,
+      "step": 30389
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5427797883855985,
+      "learning_rate": 3.941180287959763e-06,
+      "loss": 0.9162,
+      "step": 30390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6403052604206185,
+      "learning_rate": 3.94057326481358e-06,
+      "loss": 0.856,
+      "step": 30391
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5539148387471098,
+      "learning_rate": 3.939966276947293e-06,
+      "loss": 0.7941,
+      "step": 30392
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6656348327761707,
+      "learning_rate": 3.939359324364445e-06,
+      "loss": 0.7401,
+      "step": 30393
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9468301756065571,
+      "learning_rate": 3.938752407068562e-06,
+      "loss": 0.8083,
+      "step": 30394
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6095181271192227,
+      "learning_rate": 3.9381455250631864e-06,
+      "loss": 0.8592,
+      "step": 30395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8530221821154428,
+      "learning_rate": 3.937538678351842e-06,
+      "loss": 0.625,
+      "step": 30396
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.52558921974745,
+      "learning_rate": 3.936931866938071e-06,
+      "loss": 0.9294,
+      "step": 30397
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6258845429078175,
+      "learning_rate": 3.9363250908254015e-06,
+      "loss": 0.8895,
+      "step": 30398
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.660779072743801,
+      "learning_rate": 3.9357183500173635e-06,
+      "loss": 0.9337,
+      "step": 30399
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0987112688021305,
+      "learning_rate": 3.935111644517493e-06,
+      "loss": 0.9093,
+      "step": 30400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.640932048649683,
+      "learning_rate": 3.934504974329326e-06,
+      "loss": 0.8698,
+      "step": 30401
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7340080854212385,
+      "learning_rate": 3.933898339456392e-06,
+      "loss": 0.8806,
+      "step": 30402
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.993420364026844,
+      "learning_rate": 3.9332917399022196e-06,
+      "loss": 0.8813,
+      "step": 30403
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.712534495693896,
+      "learning_rate": 3.9326851756703425e-06,
+      "loss": 0.9741,
+      "step": 30404
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.636118298212953,
+      "learning_rate": 3.932078646764297e-06,
+      "loss": 0.8228,
+      "step": 30405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7210546114007803,
+      "learning_rate": 3.93147215318761e-06,
+      "loss": 0.8178,
+      "step": 30406
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1641498072892613,
+      "learning_rate": 3.930865694943809e-06,
+      "loss": 0.8482,
+      "step": 30407
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.770154839490454,
+      "learning_rate": 3.930259272036432e-06,
+      "loss": 0.7899,
+      "step": 30408
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8363133254272133,
+      "learning_rate": 3.9296528844690085e-06,
+      "loss": 0.9805,
+      "step": 30409
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7214319908976126,
+      "learning_rate": 3.9290465322450685e-06,
+      "loss": 0.7249,
+      "step": 30410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.054899124061823,
+      "learning_rate": 3.9284402153681385e-06,
+      "loss": 0.9387,
+      "step": 30411
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8090163775052823,
+      "learning_rate": 3.927833933841755e-06,
+      "loss": 0.6867,
+      "step": 30412
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8051842297089844,
+      "learning_rate": 3.927227687669442e-06,
+      "loss": 0.9791,
+      "step": 30413
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.692351214778529,
+      "learning_rate": 3.9266214768547335e-06,
+      "loss": 0.8505,
+      "step": 30414
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.510880323444665,
+      "learning_rate": 3.926015301401156e-06,
+      "loss": 0.9181,
+      "step": 30415
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7324307050326715,
+      "learning_rate": 3.925409161312243e-06,
+      "loss": 0.7716,
+      "step": 30416
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4254424861935406,
+      "learning_rate": 3.924803056591517e-06,
+      "loss": 0.7966,
+      "step": 30417
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5895882231534946,
+      "learning_rate": 3.924196987242512e-06,
+      "loss": 0.7431,
+      "step": 30418
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.554267147123075,
+      "learning_rate": 3.923590953268762e-06,
+      "loss": 0.7088,
+      "step": 30419
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.121020563838409,
+      "learning_rate": 3.922984954673783e-06,
+      "loss": 0.8421,
+      "step": 30420
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8293202576591532,
+      "learning_rate": 3.92237899146111e-06,
+      "loss": 0.6999,
+      "step": 30421
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4999145640098885,
+      "learning_rate": 3.921773063634269e-06,
+      "loss": 0.8963,
+      "step": 30422
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4732565702083724,
+      "learning_rate": 3.921167171196798e-06,
+      "loss": 0.7278,
+      "step": 30423
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7691574897333937,
+      "learning_rate": 3.920561314152208e-06,
+      "loss": 0.824,
+      "step": 30424
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.750144897467453,
+      "learning_rate": 3.919955492504036e-06,
+      "loss": 0.7305,
+      "step": 30425
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6549339631515125,
+      "learning_rate": 3.91934970625581e-06,
+      "loss": 0.8716,
+      "step": 30426
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9118409231671776,
+      "learning_rate": 3.918743955411057e-06,
+      "loss": 0.816,
+      "step": 30427
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.46303649351047,
+      "learning_rate": 3.918138239973297e-06,
+      "loss": 0.7252,
+      "step": 30428
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8133962804722965,
+      "learning_rate": 3.917532559946062e-06,
+      "loss": 0.8372,
+      "step": 30429
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6822305152071664,
+      "learning_rate": 3.916926915332882e-06,
+      "loss": 0.8466,
+      "step": 30430
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8513303605973084,
+      "learning_rate": 3.916321306137276e-06,
+      "loss": 0.8258,
+      "step": 30431
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.753883312200533,
+      "learning_rate": 3.915715732362777e-06,
+      "loss": 0.8115,
+      "step": 30432
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9713743902174672,
+      "learning_rate": 3.915110194012907e-06,
+      "loss": 0.7857,
+      "step": 30433
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6388726429894782,
+      "learning_rate": 3.914504691091188e-06,
+      "loss": 0.8087,
+      "step": 30434
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5030346126054084,
+      "learning_rate": 3.913899223601149e-06,
+      "loss": 0.9635,
+      "step": 30435
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5113087143539499,
+      "learning_rate": 3.91329379154632e-06,
+      "loss": 0.7487,
+      "step": 30436
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.534370376043596,
+      "learning_rate": 3.912688394930221e-06,
+      "loss": 0.8866,
+      "step": 30437
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5016806693532567,
+      "learning_rate": 3.912083033756374e-06,
+      "loss": 0.8445,
+      "step": 30438
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.817526790523435,
+      "learning_rate": 3.911477708028308e-06,
+      "loss": 0.6553,
+      "step": 30439
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.675151870031959,
+      "learning_rate": 3.910872417749548e-06,
+      "loss": 0.833,
+      "step": 30440
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0242189513675304,
+      "learning_rate": 3.910267162923618e-06,
+      "loss": 0.7554,
+      "step": 30441
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8275183558577737,
+      "learning_rate": 3.909661943554036e-06,
+      "loss": 0.7955,
+      "step": 30442
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4604601570282223,
+      "learning_rate": 3.909056759644331e-06,
+      "loss": 0.7584,
+      "step": 30443
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9975042110926449,
+      "learning_rate": 3.9084516111980285e-06,
+      "loss": 0.8319,
+      "step": 30444
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8686109281242276,
+      "learning_rate": 3.9078464982186505e-06,
+      "loss": 0.9155,
+      "step": 30445
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.493329396533654,
+      "learning_rate": 3.907241420709716e-06,
+      "loss": 0.8285,
+      "step": 30446
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4416818747970717,
+      "learning_rate": 3.906636378674752e-06,
+      "loss": 0.8026,
+      "step": 30447
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7554772361669146,
+      "learning_rate": 3.906031372117278e-06,
+      "loss": 0.7121,
+      "step": 30448
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5688745494385796,
+      "learning_rate": 3.905426401040823e-06,
+      "loss": 0.7487,
+      "step": 30449
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6166294239128183,
+      "learning_rate": 3.9048214654489e-06,
+      "loss": 0.7753,
+      "step": 30450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5585855956823318,
+      "learning_rate": 3.90421656534504e-06,
+      "loss": 0.9404,
+      "step": 30451
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7610744866637682,
+      "learning_rate": 3.903611700732758e-06,
+      "loss": 0.6733,
+      "step": 30452
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1064639321129013,
+      "learning_rate": 3.903006871615582e-06,
+      "loss": 0.7944,
+      "step": 30453
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7776499894102131,
+      "learning_rate": 3.9024020779970305e-06,
+      "loss": 0.666,
+      "step": 30454
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8000000847008732,
+      "learning_rate": 3.901797319880622e-06,
+      "loss": 0.681,
+      "step": 30455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.449760879764098,
+      "learning_rate": 3.901192597269878e-06,
+      "loss": 0.8307,
+      "step": 30456
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.52906285603435,
+      "learning_rate": 3.9005879101683266e-06,
+      "loss": 0.8377,
+      "step": 30457
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.63019950660105,
+      "learning_rate": 3.899983258579483e-06,
+      "loss": 0.8433,
+      "step": 30458
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.777075600045917,
+      "learning_rate": 3.899378642506863e-06,
+      "loss": 0.8943,
+      "step": 30459
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8331758128134945,
+      "learning_rate": 3.898774061953995e-06,
+      "loss": 0.8907,
+      "step": 30460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5196677722514904,
+      "learning_rate": 3.8981695169243986e-06,
+      "loss": 0.9439,
+      "step": 30461
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9531778928628065,
+      "learning_rate": 3.89756500742159e-06,
+      "loss": 0.8656,
+      "step": 30462
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5727634727941489,
+      "learning_rate": 3.896960533449087e-06,
+      "loss": 0.7656,
+      "step": 30463
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4657060524948515,
+      "learning_rate": 3.896356095010412e-06,
+      "loss": 0.7758,
+      "step": 30464
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7263848849353065,
+      "learning_rate": 3.895751692109087e-06,
+      "loss": 0.8409,
+      "step": 30465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.715683292030655,
+      "learning_rate": 3.895147324748629e-06,
+      "loss": 0.8721,
+      "step": 30466
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8696848449560364,
+      "learning_rate": 3.894542992932553e-06,
+      "loss": 0.8156,
+      "step": 30467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8786033377802165,
+      "learning_rate": 3.893938696664383e-06,
+      "loss": 0.9335,
+      "step": 30468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9608210554851446,
+      "learning_rate": 3.893334435947632e-06,
+      "loss": 0.7463,
+      "step": 30469
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6682660431644771,
+      "learning_rate": 3.892730210785825e-06,
+      "loss": 0.8039,
+      "step": 30470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.533600935059688,
+      "learning_rate": 3.892126021182473e-06,
+      "loss": 0.6497,
+      "step": 30471
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4373841509309813,
+      "learning_rate": 3.891521867141099e-06,
+      "loss": 0.7785,
+      "step": 30472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4436329794483385,
+      "learning_rate": 3.890917748665217e-06,
+      "loss": 0.8304,
+      "step": 30473
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6713349292252682,
+      "learning_rate": 3.890313665758348e-06,
+      "loss": 0.9021,
+      "step": 30474
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0485115526184257,
+      "learning_rate": 3.889709618424008e-06,
+      "loss": 0.7705,
+      "step": 30475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.559146681516514,
+      "learning_rate": 3.889105606665709e-06,
+      "loss": 0.7247,
+      "step": 30476
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7660160766798224,
+      "learning_rate": 3.888501630486973e-06,
+      "loss": 0.6803,
+      "step": 30477
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.105055927776465,
+      "learning_rate": 3.887897689891318e-06,
+      "loss": 0.6813,
+      "step": 30478
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5424488246232757,
+      "learning_rate": 3.887293784882258e-06,
+      "loss": 0.8318,
+      "step": 30479
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6465860079641563,
+      "learning_rate": 3.886689915463305e-06,
+      "loss": 1.0229,
+      "step": 30480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.797916627773139,
+      "learning_rate": 3.886086081637979e-06,
+      "loss": 0.8418,
+      "step": 30481
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6089008504698432,
+      "learning_rate": 3.8854822834097985e-06,
+      "loss": 0.9056,
+      "step": 30482
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7583189191716949,
+      "learning_rate": 3.884878520782277e-06,
+      "loss": 0.7171,
+      "step": 30483
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6176380615589323,
+      "learning_rate": 3.884274793758924e-06,
+      "loss": 0.8516,
+      "step": 30484
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.600474926181718,
+      "learning_rate": 3.88367110234326e-06,
+      "loss": 0.9291,
+      "step": 30485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8769228225851036,
+      "learning_rate": 3.883067446538803e-06,
+      "loss": 1.0087,
+      "step": 30486
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9507621434513596,
+      "learning_rate": 3.8824638263490635e-06,
+      "loss": 0.9124,
+      "step": 30487
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5896476590736943,
+      "learning_rate": 3.881860241777553e-06,
+      "loss": 0.9104,
+      "step": 30488
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7395746857344967,
+      "learning_rate": 3.88125669282779e-06,
+      "loss": 0.853,
+      "step": 30489
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5351385050411128,
+      "learning_rate": 3.88065317950329e-06,
+      "loss": 0.8253,
+      "step": 30490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5124740921936217,
+      "learning_rate": 3.8800497018075645e-06,
+      "loss": 0.7356,
+      "step": 30491
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1205285475539655,
+      "learning_rate": 3.879446259744124e-06,
+      "loss": 0.9065,
+      "step": 30492
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6356261138868722,
+      "learning_rate": 3.878842853316489e-06,
+      "loss": 0.7831,
+      "step": 30493
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7759897779660565,
+      "learning_rate": 3.878239482528164e-06,
+      "loss": 0.8375,
+      "step": 30494
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5821952757052686,
+      "learning_rate": 3.877636147382672e-06,
+      "loss": 0.8341,
+      "step": 30495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6943345590862031,
+      "learning_rate": 3.877032847883516e-06,
+      "loss": 0.8907,
+      "step": 30496
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4392463653545429,
+      "learning_rate": 3.876429584034218e-06,
+      "loss": 0.8068,
+      "step": 30497
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.110195454690816,
+      "learning_rate": 3.875826355838282e-06,
+      "loss": 0.9326,
+      "step": 30498
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5399819344892864,
+      "learning_rate": 3.8752231632992254e-06,
+      "loss": 0.8179,
+      "step": 30499
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5648071769147034,
+      "learning_rate": 3.87462000642056e-06,
+      "loss": 0.8483,
+      "step": 30500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6711226771780998,
+      "learning_rate": 3.874016885205793e-06,
+      "loss": 0.979,
+      "step": 30501
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7653595898268437,
+      "learning_rate": 3.873413799658438e-06,
+      "loss": 0.9232,
+      "step": 30502
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7621742474985822,
+      "learning_rate": 3.872810749782009e-06,
+      "loss": 0.8991,
+      "step": 30503
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0423530988292167,
+      "learning_rate": 3.8722077355800225e-06,
+      "loss": 0.8789,
+      "step": 30504
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5152250440546682,
+      "learning_rate": 3.871604757055974e-06,
+      "loss": 0.8815,
+      "step": 30505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.450222588051428,
+      "learning_rate": 3.8710018142133844e-06,
+      "loss": 0.751,
+      "step": 30506
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4763624685385597,
+      "learning_rate": 3.870398907055766e-06,
+      "loss": 0.9225,
+      "step": 30507
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6529542980055936,
+      "learning_rate": 3.869796035586625e-06,
+      "loss": 0.8535,
+      "step": 30508
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7277897420546258,
+      "learning_rate": 3.869193199809469e-06,
+      "loss": 0.8846,
+      "step": 30509
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8929328248530126,
+      "learning_rate": 3.8685903997278115e-06,
+      "loss": 0.7867,
+      "step": 30510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7874502151093943,
+      "learning_rate": 3.867987635345165e-06,
+      "loss": 0.8795,
+      "step": 30511
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1584517683092885,
+      "learning_rate": 3.867384906665036e-06,
+      "loss": 0.7006,
+      "step": 30512
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5807451576706995,
+      "learning_rate": 3.8667822136909304e-06,
+      "loss": 0.7272,
+      "step": 30513
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6867488695993507,
+      "learning_rate": 3.866179556426364e-06,
+      "loss": 0.922,
+      "step": 30514
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4512593926967357,
+      "learning_rate": 3.865576934874839e-06,
+      "loss": 0.7882,
+      "step": 30515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5208314247374868,
+      "learning_rate": 3.864974349039866e-06,
+      "loss": 0.7402,
+      "step": 30516
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7090661755977987,
+      "learning_rate": 3.8643717989249585e-06,
+      "loss": 0.9515,
+      "step": 30517
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7538696964235254,
+      "learning_rate": 3.863769284533622e-06,
+      "loss": 0.926,
+      "step": 30518
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5839297805040153,
+      "learning_rate": 3.8631668058693585e-06,
+      "loss": 0.8414,
+      "step": 30519
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0854071674933716,
+      "learning_rate": 3.862564362935681e-06,
+      "loss": 0.984,
+      "step": 30520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6887035683184075,
+      "learning_rate": 3.8619619557361044e-06,
+      "loss": 0.7745,
+      "step": 30521
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7669885371113192,
+      "learning_rate": 3.861359584274122e-06,
+      "loss": 0.8417,
+      "step": 30522
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7827594812745193,
+      "learning_rate": 3.860757248553247e-06,
+      "loss": 1.0085,
+      "step": 30523
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5817389767926648,
+      "learning_rate": 3.8601549485769865e-06,
+      "loss": 0.8484,
+      "step": 30524
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5078494357044065,
+      "learning_rate": 3.8595526843488515e-06,
+      "loss": 0.8807,
+      "step": 30525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.521052707350212,
+      "learning_rate": 3.858950455872344e-06,
+      "loss": 0.9551,
+      "step": 30526
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7772595466902259,
+      "learning_rate": 3.8583482631509695e-06,
+      "loss": 0.7573,
+      "step": 30527
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.511725920290847,
+      "learning_rate": 3.857746106188238e-06,
+      "loss": 0.7781,
+      "step": 30528
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4425627295416679,
+      "learning_rate": 3.85714398498765e-06,
+      "loss": 0.8979,
+      "step": 30529
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.395419642411476,
+      "learning_rate": 3.8565418995527185e-06,
+      "loss": 0.8331,
+      "step": 30530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6734448184639217,
+      "learning_rate": 3.8559398498869416e-06,
+      "loss": 0.8785,
+      "step": 30531
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6029301117002526,
+      "learning_rate": 3.855337835993831e-06,
+      "loss": 0.9045,
+      "step": 30532
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6532428860489952,
+      "learning_rate": 3.8547358578768854e-06,
+      "loss": 0.9026,
+      "step": 30533
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9128336206521896,
+      "learning_rate": 3.854133915539617e-06,
+      "loss": 0.7036,
+      "step": 30534
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6219274234842413,
+      "learning_rate": 3.853532008985527e-06,
+      "loss": 0.8587,
+      "step": 30535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6632280180366066,
+      "learning_rate": 3.852930138218115e-06,
+      "loss": 0.8672,
+      "step": 30536
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7433971905624257,
+      "learning_rate": 3.852328303240891e-06,
+      "loss": 0.78,
+      "step": 30537
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7839682494293476,
+      "learning_rate": 3.8517265040573606e-06,
+      "loss": 0.7098,
+      "step": 30538
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7773236755171447,
+      "learning_rate": 3.851124740671026e-06,
+      "loss": 0.9651,
+      "step": 30539
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.729765402647794,
+      "learning_rate": 3.850523013085385e-06,
+      "loss": 0.8378,
+      "step": 30540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.654009477359405,
+      "learning_rate": 3.849921321303947e-06,
+      "loss": 0.8651,
+      "step": 30541
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5881393639815347,
+      "learning_rate": 3.849319665330218e-06,
+      "loss": 0.7887,
+      "step": 30542
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4642083916033384,
+      "learning_rate": 3.848718045167695e-06,
+      "loss": 0.8398,
+      "step": 30543
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8410509124111718,
+      "learning_rate": 3.848116460819882e-06,
+      "loss": 0.8155,
+      "step": 30544
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5123338300908307,
+      "learning_rate": 3.8475149122902815e-06,
+      "loss": 0.7825,
+      "step": 30545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4631699866031196,
+      "learning_rate": 3.846913399582401e-06,
+      "loss": 0.6559,
+      "step": 30546
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5345145711567985,
+      "learning_rate": 3.846311922699738e-06,
+      "loss": 0.8206,
+      "step": 30547
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4952562967121084,
+      "learning_rate": 3.845710481645793e-06,
+      "loss": 0.7343,
+      "step": 30548
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6161366583575645,
+      "learning_rate": 3.845109076424071e-06,
+      "loss": 0.8654,
+      "step": 30549
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4864378128842357,
+      "learning_rate": 3.844507707038071e-06,
+      "loss": 0.888,
+      "step": 30550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6285993486831387,
+      "learning_rate": 3.843906373491299e-06,
+      "loss": 0.8687,
+      "step": 30551
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5941451307812677,
+      "learning_rate": 3.843305075787249e-06,
+      "loss": 0.8962,
+      "step": 30552
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5411919070047428,
+      "learning_rate": 3.8427038139294304e-06,
+      "loss": 0.8914,
+      "step": 30553
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5400597308249107,
+      "learning_rate": 3.842102587921336e-06,
+      "loss": 0.872,
+      "step": 30554
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5536228329681818,
+      "learning_rate": 3.841501397766474e-06,
+      "loss": 0.6561,
+      "step": 30555
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7489017268110776,
+      "learning_rate": 3.84090024346834e-06,
+      "loss": 0.8982,
+      "step": 30556
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.684034332463015,
+      "learning_rate": 3.84029912503043e-06,
+      "loss": 1.0363,
+      "step": 30557
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8475340179377757,
+      "learning_rate": 3.83969804245625e-06,
+      "loss": 0.8034,
+      "step": 30558
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.579400149357194,
+      "learning_rate": 3.839096995749302e-06,
+      "loss": 0.9522,
+      "step": 30559
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6133401750787382,
+      "learning_rate": 3.838495984913081e-06,
+      "loss": 0.795,
+      "step": 30560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6968087951743485,
+      "learning_rate": 3.8378950099510835e-06,
+      "loss": 0.7136,
+      "step": 30561
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8110082955515713,
+      "learning_rate": 3.837294070866812e-06,
+      "loss": 0.7076,
+      "step": 30562
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.707410359594707,
+      "learning_rate": 3.83669316766377e-06,
+      "loss": 0.8222,
+      "step": 30563
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.679787665096715,
+      "learning_rate": 3.8360923003454505e-06,
+      "loss": 0.7445,
+      "step": 30564
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6504190959005336,
+      "learning_rate": 3.835491468915351e-06,
+      "loss": 0.8887,
+      "step": 30565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9139727235529727,
+      "learning_rate": 3.834890673376971e-06,
+      "loss": 0.8854,
+      "step": 30566
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4028192691590928,
+      "learning_rate": 3.834289913733813e-06,
+      "loss": 0.7507,
+      "step": 30567
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7203555977015954,
+      "learning_rate": 3.83368918998937e-06,
+      "loss": 0.8888,
+      "step": 30568
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5100633775719716,
+      "learning_rate": 3.833088502147139e-06,
+      "loss": 0.8896,
+      "step": 30569
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7342813928260759,
+      "learning_rate": 3.832487850210622e-06,
+      "loss": 0.7174,
+      "step": 30570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.73045886204219,
+      "learning_rate": 3.83188723418331e-06,
+      "loss": 0.8543,
+      "step": 30571
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4803348635786007,
+      "learning_rate": 3.831286654068706e-06,
+      "loss": 0.7542,
+      "step": 30572
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8106533208288272,
+      "learning_rate": 3.830686109870302e-06,
+      "loss": 0.6807,
+      "step": 30573
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6885287860652167,
+      "learning_rate": 3.830085601591599e-06,
+      "loss": 0.8225,
+      "step": 30574
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.725757308704205,
+      "learning_rate": 3.829485129236088e-06,
+      "loss": 0.7829,
+      "step": 30575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.712350190931149,
+      "learning_rate": 3.8288846928072705e-06,
+      "loss": 0.748,
+      "step": 30576
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.580181962088915,
+      "learning_rate": 3.8282842923086415e-06,
+      "loss": 0.8717,
+      "step": 30577
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.742689301464924,
+      "learning_rate": 3.827683927743692e-06,
+      "loss": 0.9028,
+      "step": 30578
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5271313290842745,
+      "learning_rate": 3.8270835991159195e-06,
+      "loss": 0.6725,
+      "step": 30579
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.608032345328359,
+      "learning_rate": 3.826483306428824e-06,
+      "loss": 0.9461,
+      "step": 30580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4364463812147572,
+      "learning_rate": 3.8258830496858975e-06,
+      "loss": 0.8655,
+      "step": 30581
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8199464946823214,
+      "learning_rate": 3.825282828890632e-06,
+      "loss": 0.9217,
+      "step": 30582
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5995284480032241,
+      "learning_rate": 3.824682644046524e-06,
+      "loss": 0.8155,
+      "step": 30583
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5535752021104348,
+      "learning_rate": 3.824082495157073e-06,
+      "loss": 0.8645,
+      "step": 30584
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4966790378644566,
+      "learning_rate": 3.823482382225768e-06,
+      "loss": 0.8867,
+      "step": 30585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4576320665451914,
+      "learning_rate": 3.822882305256101e-06,
+      "loss": 0.7575,
+      "step": 30586
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1143005641530936,
+      "learning_rate": 3.822282264251569e-06,
+      "loss": 0.8689,
+      "step": 30587
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4307539665483295,
+      "learning_rate": 3.8216822592156685e-06,
+      "loss": 0.7369,
+      "step": 30588
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.488194115392705,
+      "learning_rate": 3.8210822901518904e-06,
+      "loss": 0.7986,
+      "step": 30589
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5900723759352458,
+      "learning_rate": 3.820482357063724e-06,
+      "loss": 0.846,
+      "step": 30590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.659912499194065,
+      "learning_rate": 3.819882459954668e-06,
+      "loss": 0.8869,
+      "step": 30591
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5350377196352663,
+      "learning_rate": 3.819282598828211e-06,
+      "loss": 0.8481,
+      "step": 30592
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5579503594702937,
+      "learning_rate": 3.8186827736878505e-06,
+      "loss": 0.7803,
+      "step": 30593
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.59347349298449,
+      "learning_rate": 3.8180829845370714e-06,
+      "loss": 0.9345,
+      "step": 30594
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.48666144314756,
+      "learning_rate": 3.817483231379375e-06,
+      "loss": 0.8639,
+      "step": 30595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.576313968099381,
+      "learning_rate": 3.816883514218245e-06,
+      "loss": 0.8992,
+      "step": 30596
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4091241307866114,
+      "learning_rate": 3.816283833057179e-06,
+      "loss": 0.8687,
+      "step": 30597
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7379251513722993,
+      "learning_rate": 3.815684187899668e-06,
+      "loss": 0.7281,
+      "step": 30598
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8386171669398093,
+      "learning_rate": 3.815084578749197e-06,
+      "loss": 0.7715,
+      "step": 30599
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.407926450993732,
+      "learning_rate": 3.814485005609262e-06,
+      "loss": 0.8224,
+      "step": 30600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6530136330339285,
+      "learning_rate": 3.8138854684833537e-06,
+      "loss": 1.1203,
+      "step": 30601
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6354184740108206,
+      "learning_rate": 3.8132859673749688e-06,
+      "loss": 0.7992,
+      "step": 30602
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5795725821321314,
+      "learning_rate": 3.8126865022875858e-06,
+      "loss": 0.811,
+      "step": 30603
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.723953598216501,
+      "learning_rate": 3.812087073224701e-06,
+      "loss": 0.7513,
+      "step": 30604
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0804031401621765,
+      "learning_rate": 3.8114876801898036e-06,
+      "loss": 0.8262,
+      "step": 30605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5604183765863282,
+      "learning_rate": 3.810888323186391e-06,
+      "loss": 0.8355,
+      "step": 30606
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5544286682537982,
+      "learning_rate": 3.8102890022179393e-06,
+      "loss": 0.7199,
+      "step": 30607
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5342664165221822,
+      "learning_rate": 3.809689717287944e-06,
+      "loss": 0.8468,
+      "step": 30608
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7833125629642959,
+      "learning_rate": 3.8090904683998997e-06,
+      "loss": 0.719,
+      "step": 30609
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4521660695652185,
+      "learning_rate": 3.8084912555572907e-06,
+      "loss": 0.8119,
+      "step": 30610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7095950779000315,
+      "learning_rate": 3.807892078763602e-06,
+      "loss": 0.8612,
+      "step": 30611
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5512164293379385,
+      "learning_rate": 3.8072929380223256e-06,
+      "loss": 0.9347,
+      "step": 30612
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8722963809157023,
+      "learning_rate": 3.8066938333369542e-06,
+      "loss": 0.8501,
+      "step": 30613
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7578063921524267,
+      "learning_rate": 3.806094764710968e-06,
+      "loss": 0.8803,
+      "step": 30614
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5044356724017522,
+      "learning_rate": 3.8054957321478623e-06,
+      "loss": 0.8324,
+      "step": 30615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5682649279284908,
+      "learning_rate": 3.8048967356511234e-06,
+      "loss": 0.9034,
+      "step": 30616
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7115391115888172,
+      "learning_rate": 3.8042977752242315e-06,
+      "loss": 0.8896,
+      "step": 30617
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6487471788174217,
+      "learning_rate": 3.803698850870681e-06,
+      "loss": 0.9504,
+      "step": 30618
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.790445096364078,
+      "learning_rate": 3.8030999625939604e-06,
+      "loss": 0.6742,
+      "step": 30619
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9404849598067428,
+      "learning_rate": 3.802501110397553e-06,
+      "loss": 0.8791,
+      "step": 30620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.61397919169381,
+      "learning_rate": 3.8019022942849436e-06,
+      "loss": 0.7756,
+      "step": 30621
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.493205585855114,
+      "learning_rate": 3.8013035142596213e-06,
+      "loss": 0.8117,
+      "step": 30622
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9348148443721052,
+      "learning_rate": 3.8007047703250756e-06,
+      "loss": 0.8166,
+      "step": 30623
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6100401251526646,
+      "learning_rate": 3.8001060624847896e-06,
+      "loss": 0.8965,
+      "step": 30624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4138949076169733,
+      "learning_rate": 3.7995073907422454e-06,
+      "loss": 0.934,
+      "step": 30625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7179217326903617,
+      "learning_rate": 3.798908755100932e-06,
+      "loss": 0.7766,
+      "step": 30626
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5438025516068519,
+      "learning_rate": 3.798310155564339e-06,
+      "loss": 0.7339,
+      "step": 30627
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4951393974038667,
+      "learning_rate": 3.7977115921359477e-06,
+      "loss": 0.7939,
+      "step": 30628
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6603971616995172,
+      "learning_rate": 3.797113064819239e-06,
+      "loss": 0.8304,
+      "step": 30629
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.459346336177476,
+      "learning_rate": 3.7965145736177058e-06,
+      "loss": 0.889,
+      "step": 30630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.122891395032426,
+      "learning_rate": 3.795916118534826e-06,
+      "loss": 0.9187,
+      "step": 30631
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7762324716183266,
+      "learning_rate": 3.795317699574089e-06,
+      "loss": 0.7977,
+      "step": 30632
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.57049992801916,
+      "learning_rate": 3.7947193167389728e-06,
+      "loss": 0.837,
+      "step": 30633
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.596029011506916,
+      "learning_rate": 3.7941209700329706e-06,
+      "loss": 0.7766,
+      "step": 30634
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5632962900889018,
+      "learning_rate": 3.7935226594595564e-06,
+      "loss": 0.7928,
+      "step": 30635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6180811025784498,
+      "learning_rate": 3.7929243850222206e-06,
+      "loss": 0.8162,
+      "step": 30636
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8523811763074173,
+      "learning_rate": 3.7923261467244455e-06,
+      "loss": 0.7736,
+      "step": 30637
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5215512805461129,
+      "learning_rate": 3.79172794456971e-06,
+      "loss": 0.7301,
+      "step": 30638
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7584754137839353,
+      "learning_rate": 3.791129778561499e-06,
+      "loss": 0.9284,
+      "step": 30639
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6371867392357862,
+      "learning_rate": 3.7905316487032995e-06,
+      "loss": 0.8721,
+      "step": 30640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6168387676329654,
+      "learning_rate": 3.7899335549985903e-06,
+      "loss": 0.7527,
+      "step": 30641
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6715274448422068,
+      "learning_rate": 3.7893354974508513e-06,
+      "loss": 0.9782,
+      "step": 30642
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6792597160106948,
+      "learning_rate": 3.7887374760635675e-06,
+      "loss": 0.9105,
+      "step": 30643
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7858616272695267,
+      "learning_rate": 3.7881394908402237e-06,
+      "loss": 0.6175,
+      "step": 30644
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.634647843438813,
+      "learning_rate": 3.7875415417842974e-06,
+      "loss": 0.8152,
+      "step": 30645
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.629858160918165,
+      "learning_rate": 3.786943628899269e-06,
+      "loss": 0.8295,
+      "step": 30646
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7364288827226684,
+      "learning_rate": 3.7863457521886215e-06,
+      "loss": 0.6627,
+      "step": 30647
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9009126041484163,
+      "learning_rate": 3.7857479116558393e-06,
+      "loss": 0.7733,
+      "step": 30648
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5241118387486476,
+      "learning_rate": 3.7851501073044005e-06,
+      "loss": 0.8046,
+      "step": 30649
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.883316391436239,
+      "learning_rate": 3.7845523391377815e-06,
+      "loss": 0.8342,
+      "step": 30650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5578553466318867,
+      "learning_rate": 3.7839546071594703e-06,
+      "loss": 0.8329,
+      "step": 30651
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6543443984306911,
+      "learning_rate": 3.7833569113729394e-06,
+      "loss": 0.8722,
+      "step": 30652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5416736834027818,
+      "learning_rate": 3.782759251781677e-06,
+      "loss": 0.9246,
+      "step": 30653
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.729848500025528,
+      "learning_rate": 3.782161628389155e-06,
+      "loss": 0.7515,
+      "step": 30654
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5251226343922282,
+      "learning_rate": 3.78156404119886e-06,
+      "loss": 0.8973,
+      "step": 30655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8093574222134126,
+      "learning_rate": 3.780966490214264e-06,
+      "loss": 0.8838,
+      "step": 30656
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.430925240107706,
+      "learning_rate": 3.7803689754388538e-06,
+      "loss": 0.9382,
+      "step": 30657
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5540032062621536,
+      "learning_rate": 3.7797714968761045e-06,
+      "loss": 0.9029,
+      "step": 30658
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5323560065251067,
+      "learning_rate": 3.7791740545294907e-06,
+      "loss": 0.7426,
+      "step": 30659
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.795779164402007,
+      "learning_rate": 3.7785766484024954e-06,
+      "loss": 0.7738,
+      "step": 30660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5938519245547726,
+      "learning_rate": 3.7779792784986e-06,
+      "loss": 0.8068,
+      "step": 30661
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7644444242728718,
+      "learning_rate": 3.77738194482128e-06,
+      "loss": 0.7292,
+      "step": 30662
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7863519762500193,
+      "learning_rate": 3.7767846473740077e-06,
+      "loss": 0.781,
+      "step": 30663
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.551265736660205,
+      "learning_rate": 3.776187386160266e-06,
+      "loss": 0.7348,
+      "step": 30664
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6490918559961198,
+      "learning_rate": 3.7755901611835356e-06,
+      "loss": 0.8701,
+      "step": 30665
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.741950466768259,
+      "learning_rate": 3.774992972447289e-06,
+      "loss": 0.7527,
+      "step": 30666
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5126601254564689,
+      "learning_rate": 3.7743958199550013e-06,
+      "loss": 0.7429,
+      "step": 30667
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5012556582395304,
+      "learning_rate": 3.773798703710152e-06,
+      "loss": 0.8164,
+      "step": 30668
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9892171773864933,
+      "learning_rate": 3.7732016237162216e-06,
+      "loss": 0.8034,
+      "step": 30669
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8734559399868598,
+      "learning_rate": 3.772604579976682e-06,
+      "loss": 0.8522,
+      "step": 30670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.784629359884414,
+      "learning_rate": 3.772007572495007e-06,
+      "loss": 0.8772,
+      "step": 30671
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7841785241693797,
+      "learning_rate": 3.7714106012746786e-06,
+      "loss": 0.7184,
+      "step": 30672
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5128441532325536,
+      "learning_rate": 3.7708136663191673e-06,
+      "loss": 1.0276,
+      "step": 30673
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2473111951800244,
+      "learning_rate": 3.7702167676319534e-06,
+      "loss": 0.9356,
+      "step": 30674
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4910098594621701,
+      "learning_rate": 3.769619905216507e-06,
+      "loss": 0.8192,
+      "step": 30675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4677648093751379,
+      "learning_rate": 3.7690230790763094e-06,
+      "loss": 0.7709,
+      "step": 30676
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.552750493399314,
+      "learning_rate": 3.768426289214828e-06,
+      "loss": 0.8283,
+      "step": 30677
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8916181412577386,
+      "learning_rate": 3.7678295356355456e-06,
+      "loss": 0.7912,
+      "step": 30678
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5692135961266027,
+      "learning_rate": 3.7672328183419327e-06,
+      "loss": 0.778,
+      "step": 30679
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7352090449469074,
+      "learning_rate": 3.7666361373374607e-06,
+      "loss": 0.6881,
+      "step": 30680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8115463150320088,
+      "learning_rate": 3.766039492625605e-06,
+      "loss": 0.789,
+      "step": 30681
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7678116495786727,
+      "learning_rate": 3.7654428842098457e-06,
+      "loss": 0.7983,
+      "step": 30682
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7712273705039674,
+      "learning_rate": 3.764846312093652e-06,
+      "loss": 0.8062,
+      "step": 30683
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9583510112058353,
+      "learning_rate": 3.764249776280493e-06,
+      "loss": 0.7712,
+      "step": 30684
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7623644594967514,
+      "learning_rate": 3.7636532767738455e-06,
+      "loss": 0.7627,
+      "step": 30685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6319099950187,
+      "learning_rate": 3.7630568135771873e-06,
+      "loss": 0.8196,
+      "step": 30686
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6432481828155772,
+      "learning_rate": 3.762460386693987e-06,
+      "loss": 0.9024,
+      "step": 30687
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4972140650796257,
+      "learning_rate": 3.7618639961277124e-06,
+      "loss": 0.7492,
+      "step": 30688
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4496171453818156,
+      "learning_rate": 3.7612676418818415e-06,
+      "loss": 0.7737,
+      "step": 30689
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.547156799128067,
+      "learning_rate": 3.760671323959849e-06,
+      "loss": 0.7345,
+      "step": 30690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7669235364081415,
+      "learning_rate": 3.7600750423652043e-06,
+      "loss": 0.7485,
+      "step": 30691
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7054137951331567,
+      "learning_rate": 3.7594787971013736e-06,
+      "loss": 0.7307,
+      "step": 30692
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8202421902950245,
+      "learning_rate": 3.7588825881718372e-06,
+      "loss": 0.7218,
+      "step": 30693
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.676828818197339,
+      "learning_rate": 3.758286415580058e-06,
+      "loss": 0.8988,
+      "step": 30694
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.048535230393421,
+      "learning_rate": 3.7576902793295158e-06,
+      "loss": 0.8467,
+      "step": 30695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.591054866086399,
+      "learning_rate": 3.7570941794236725e-06,
+      "loss": 0.8719,
+      "step": 30696
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0640147202210404,
+      "learning_rate": 3.7564981158660073e-06,
+      "loss": 0.8457,
+      "step": 30697
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.728985526628335,
+      "learning_rate": 3.755902088659983e-06,
+      "loss": 0.8627,
+      "step": 30698
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6486253167423646,
+      "learning_rate": 3.755306097809074e-06,
+      "loss": 0.9594,
+      "step": 30699
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8809258527335859,
+      "learning_rate": 3.7547101433167567e-06,
+      "loss": 0.8285,
+      "step": 30700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5254382137012281,
+      "learning_rate": 3.7541142251864883e-06,
+      "loss": 0.8556,
+      "step": 30701
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9834492621784234,
+      "learning_rate": 3.7535183434217427e-06,
+      "loss": 0.8647,
+      "step": 30702
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6398556317914812,
+      "learning_rate": 3.7529224980259914e-06,
+      "loss": 0.7866,
+      "step": 30703
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7077380641812887,
+      "learning_rate": 3.75232668900271e-06,
+      "loss": 0.8035,
+      "step": 30704
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7650482376708335,
+      "learning_rate": 3.751730916355354e-06,
+      "loss": 0.7198,
+      "step": 30705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0268557981622006,
+      "learning_rate": 3.7511351800873987e-06,
+      "loss": 0.8228,
+      "step": 30706
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7189495605132465,
+      "learning_rate": 3.7505394802023163e-06,
+      "loss": 0.8245,
+      "step": 30707
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4348963391726774,
+      "learning_rate": 3.749943816703572e-06,
+      "loss": 0.7836,
+      "step": 30708
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3470340919623776,
+      "learning_rate": 3.7493481895946295e-06,
+      "loss": 0.8707,
+      "step": 30709
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.872679038927032,
+      "learning_rate": 3.7487525988789595e-06,
+      "loss": 0.7773,
+      "step": 30710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6948878993724927,
+      "learning_rate": 3.7481570445600357e-06,
+      "loss": 0.9217,
+      "step": 30711
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7560271753728294,
+      "learning_rate": 3.7475615266413178e-06,
+      "loss": 0.9068,
+      "step": 30712
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6869216064036225,
+      "learning_rate": 3.746966045126279e-06,
+      "loss": 0.7999,
+      "step": 30713
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6873328859472159,
+      "learning_rate": 3.7463706000183843e-06,
+      "loss": 0.9053,
+      "step": 30714
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.784803235377257,
+      "learning_rate": 3.745775191321096e-06,
+      "loss": 0.7372,
+      "step": 30715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6170279094354283,
+      "learning_rate": 3.745179819037884e-06,
+      "loss": 0.9366,
+      "step": 30716
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4602150418031392,
+      "learning_rate": 3.74458448317222e-06,
+      "loss": 0.8754,
+      "step": 30717
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7702211728503447,
+      "learning_rate": 3.7439891837275644e-06,
+      "loss": 0.8971,
+      "step": 30718
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9296656925250801,
+      "learning_rate": 3.743393920707381e-06,
+      "loss": 0.8381,
+      "step": 30719
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4765245384293735,
+      "learning_rate": 3.7427986941151394e-06,
+      "loss": 0.7399,
+      "step": 30720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6615588592400534,
+      "learning_rate": 3.742203503954308e-06,
+      "loss": 0.8987,
+      "step": 30721
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5493783065691298,
+      "learning_rate": 3.7416083502283494e-06,
+      "loss": 0.8116,
+      "step": 30722
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8451732310369777,
+      "learning_rate": 3.7410132329407258e-06,
+      "loss": 0.8342,
+      "step": 30723
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5100305041598265,
+      "learning_rate": 3.7404181520949033e-06,
+      "loss": 0.8411,
+      "step": 30724
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9936320438345363,
+      "learning_rate": 3.7398231076943513e-06,
+      "loss": 0.8684,
+      "step": 30725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7326690732316304,
+      "learning_rate": 3.7392280997425325e-06,
+      "loss": 0.6909,
+      "step": 30726
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7606490952257132,
+      "learning_rate": 3.7386331282429045e-06,
+      "loss": 0.7136,
+      "step": 30727
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8450490410367046,
+      "learning_rate": 3.7380381931989384e-06,
+      "loss": 0.7373,
+      "step": 30728
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5411476191966003,
+      "learning_rate": 3.7374432946140994e-06,
+      "loss": 0.7871,
+      "step": 30729
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4688220772089406,
+      "learning_rate": 3.7368484324918473e-06,
+      "loss": 0.7488,
+      "step": 30730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.510373422805044,
+      "learning_rate": 3.7362536068356448e-06,
+      "loss": 0.6788,
+      "step": 30731
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9207422788617685,
+      "learning_rate": 3.7356588176489595e-06,
+      "loss": 0.903,
+      "step": 30732
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.586942770276463,
+      "learning_rate": 3.7350640649352477e-06,
+      "loss": 0.794,
+      "step": 30733
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4768319540247616,
+      "learning_rate": 3.73446934869798e-06,
+      "loss": 0.9017,
+      "step": 30734
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7201286508619614,
+      "learning_rate": 3.7338746689406123e-06,
+      "loss": 0.6682,
+      "step": 30735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5630050964286644,
+      "learning_rate": 3.7332800256666134e-06,
+      "loss": 0.7853,
+      "step": 30736
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7361280833544014,
+      "learning_rate": 3.7326854188794393e-06,
+      "loss": 0.7136,
+      "step": 30737
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5797589577085231,
+      "learning_rate": 3.732090848582558e-06,
+      "loss": 0.8316,
+      "step": 30738
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6469188312099414,
+      "learning_rate": 3.731496314779428e-06,
+      "loss": 0.8886,
+      "step": 30739
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5786007608326615,
+      "learning_rate": 3.7309018174735078e-06,
+      "loss": 0.8649,
+      "step": 30740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.828875778517581,
+      "learning_rate": 3.7303073566682625e-06,
+      "loss": 0.8287,
+      "step": 30741
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.668453822538926,
+      "learning_rate": 3.7297129323671553e-06,
+      "loss": 0.8965,
+      "step": 30742
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.770676080268067,
+      "learning_rate": 3.7291185445736445e-06,
+      "loss": 0.8872,
+      "step": 30743
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5588623159223094,
+      "learning_rate": 3.7285241932911874e-06,
+      "loss": 0.7027,
+      "step": 30744
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8636302639869997,
+      "learning_rate": 3.7279298785232488e-06,
+      "loss": 0.8524,
+      "step": 30745
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7547823339211952,
+      "learning_rate": 3.7273356002732908e-06,
+      "loss": 0.8869,
+      "step": 30746
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7070274837870265,
+      "learning_rate": 3.726741358544771e-06,
+      "loss": 0.8782,
+      "step": 30747
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5778807954450285,
+      "learning_rate": 3.7261471533411464e-06,
+      "loss": 0.769,
+      "step": 30748
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.272231463540934,
+      "learning_rate": 3.7255529846658777e-06,
+      "loss": 0.8772,
+      "step": 30749
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7019058854872929,
+      "learning_rate": 3.724958852522431e-06,
+      "loss": 0.6507,
+      "step": 30750
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6775743572429138,
+      "learning_rate": 3.7243647569142606e-06,
+      "loss": 1.0165,
+      "step": 30751
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6463755251187555,
+      "learning_rate": 3.7237706978448206e-06,
+      "loss": 1.0035,
+      "step": 30752
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7066825729612927,
+      "learning_rate": 3.723176675317579e-06,
+      "loss": 0.7912,
+      "step": 30753
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.721425702232297,
+      "learning_rate": 3.722582689335986e-06,
+      "loss": 0.8465,
+      "step": 30754
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8120524890890417,
+      "learning_rate": 3.721988739903508e-06,
+      "loss": 0.8614,
+      "step": 30755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.348193009225784,
+      "learning_rate": 3.7213948270235954e-06,
+      "loss": 0.7818,
+      "step": 30756
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5691575066569061,
+      "learning_rate": 3.7208009506997135e-06,
+      "loss": 0.6757,
+      "step": 30757
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5756472721780928,
+      "learning_rate": 3.720207110935312e-06,
+      "loss": 0.7914,
+      "step": 30758
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4995837923285131,
+      "learning_rate": 3.719613307733858e-06,
+      "loss": 0.8721,
+      "step": 30759
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6817614884056196,
+      "learning_rate": 3.719019541098803e-06,
+      "loss": 0.9234,
+      "step": 30760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7070631167734855,
+      "learning_rate": 3.7184258110336e-06,
+      "loss": 0.813,
+      "step": 30761
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.68642396236957,
+      "learning_rate": 3.717832117541712e-06,
+      "loss": 0.7649,
+      "step": 30762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4841236503624913,
+      "learning_rate": 3.7172384606265976e-06,
+      "loss": 0.9084,
+      "step": 30763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1601626437473698,
+      "learning_rate": 3.71664484029171e-06,
+      "loss": 0.8212,
+      "step": 30764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8198065270554704,
+      "learning_rate": 3.7160512565405016e-06,
+      "loss": 0.8622,
+      "step": 30765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6510389583118052,
+      "learning_rate": 3.7154577093764334e-06,
+      "loss": 0.8391,
+      "step": 30766
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6667593469064876,
+      "learning_rate": 3.714864198802962e-06,
+      "loss": 0.8516,
+      "step": 30767
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5577198835340962,
+      "learning_rate": 3.714270724823542e-06,
+      "loss": 0.8229,
+      "step": 30768
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8198244399510074,
+      "learning_rate": 3.713677287441624e-06,
+      "loss": 0.7242,
+      "step": 30769
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.534149394534536,
+      "learning_rate": 3.7130838866606665e-06,
+      "loss": 0.9164,
+      "step": 30770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.54371694386404,
+      "learning_rate": 3.7124905224841303e-06,
+      "loss": 0.8271,
+      "step": 30771
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8556020097813632,
+      "learning_rate": 3.7118971949154624e-06,
+      "loss": 1.012,
+      "step": 30772
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5123610389421538,
+      "learning_rate": 3.7113039039581177e-06,
+      "loss": 0.7939,
+      "step": 30773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7505139980613222,
+      "learning_rate": 3.7107106496155565e-06,
+      "loss": 0.742,
+      "step": 30774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5944287749304271,
+      "learning_rate": 3.7101174318912257e-06,
+      "loss": 0.8762,
+      "step": 30775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.600207175172788,
+      "learning_rate": 3.7095242507885854e-06,
+      "loss": 0.7382,
+      "step": 30776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5631814393282597,
+      "learning_rate": 3.7089311063110823e-06,
+      "loss": 0.8489,
+      "step": 30777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6696432579046443,
+      "learning_rate": 3.708337998462178e-06,
+      "loss": 0.8708,
+      "step": 30778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.49560140918753,
+      "learning_rate": 3.707744927245319e-06,
+      "loss": 0.8484,
+      "step": 30779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5251902397584693,
+      "learning_rate": 3.707151892663964e-06,
+      "loss": 0.8749,
+      "step": 30780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0523884192006165,
+      "learning_rate": 3.7065588947215636e-06,
+      "loss": 0.7603,
+      "step": 30781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7359844389144108,
+      "learning_rate": 3.705965933421566e-06,
+      "loss": 0.8739,
+      "step": 30782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.779275001749078,
+      "learning_rate": 3.7053730087674267e-06,
+      "loss": 0.7182,
+      "step": 30783
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.763760277099858,
+      "learning_rate": 3.704780120762599e-06,
+      "loss": 0.8969,
+      "step": 30784
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7480337847220853,
+      "learning_rate": 3.704187269410542e-06,
+      "loss": 0.7554,
+      "step": 30785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5285470918051167,
+      "learning_rate": 3.7035944547146917e-06,
+      "loss": 0.9076,
+      "step": 30786
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0393466033031804,
+      "learning_rate": 3.7030016766785095e-06,
+      "loss": 0.8058,
+      "step": 30787
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5918272686540291,
+      "learning_rate": 3.702408935305447e-06,
+      "loss": 0.817,
+      "step": 30788
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7932349221192083,
+      "learning_rate": 3.701816230598955e-06,
+      "loss": 0.9371,
+      "step": 30789
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6807769220969597,
+      "learning_rate": 3.701223562562478e-06,
+      "loss": 0.8343,
+      "step": 30790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.650584497858534,
+      "learning_rate": 3.7006309311994727e-06,
+      "loss": 0.8412,
+      "step": 30791
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0547365484280387,
+      "learning_rate": 3.7000383365133916e-06,
+      "loss": 0.8343,
+      "step": 30792
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8525678607771512,
+      "learning_rate": 3.6994457785076823e-06,
+      "loss": 0.8818,
+      "step": 30793
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.615688362248305,
+      "learning_rate": 3.6988532571857895e-06,
+      "loss": 0.7672,
+      "step": 30794
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7168358954763856,
+      "learning_rate": 3.698260772551172e-06,
+      "loss": 0.8758,
+      "step": 30795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5593397828959863,
+      "learning_rate": 3.697668324607272e-06,
+      "loss": 0.6893,
+      "step": 30796
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4785623560155479,
+      "learning_rate": 3.697075913357543e-06,
+      "loss": 0.6861,
+      "step": 30797
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6081200671929092,
+      "learning_rate": 3.6964835388054367e-06,
+      "loss": 0.9214,
+      "step": 30798
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5542846408131457,
+      "learning_rate": 3.695891200954398e-06,
+      "loss": 0.8732,
+      "step": 30799
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9927132135938983,
+      "learning_rate": 3.695298899807874e-06,
+      "loss": 0.957,
+      "step": 30800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.55014246848381,
+      "learning_rate": 3.694706635369316e-06,
+      "loss": 0.8925,
+      "step": 30801
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5330641764993436,
+      "learning_rate": 3.6941144076421786e-06,
+      "loss": 0.8559,
+      "step": 30802
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.479749601934299,
+      "learning_rate": 3.693522216629897e-06,
+      "loss": 0.8066,
+      "step": 30803
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7753091474824934,
+      "learning_rate": 3.6929300623359255e-06,
+      "loss": 0.8769,
+      "step": 30804
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.845740744214319,
+      "learning_rate": 3.692337944763711e-06,
+      "loss": 0.9423,
+      "step": 30805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8603079562946325,
+      "learning_rate": 3.6917458639167105e-06,
+      "loss": 0.881,
+      "step": 30806
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9543361193745388,
+      "learning_rate": 3.691153819798354e-06,
+      "loss": 0.7835,
+      "step": 30807
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5351810884043637,
+      "learning_rate": 3.690561812412099e-06,
+      "loss": 0.9454,
+      "step": 30808
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7624848240903767,
+      "learning_rate": 3.689969841761393e-06,
+      "loss": 0.7052,
+      "step": 30809
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8513332801110927,
+      "learning_rate": 3.689377907849677e-06,
+      "loss": 0.7791,
+      "step": 30810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.524179440906587,
+      "learning_rate": 3.688786010680405e-06,
+      "loss": 0.8709,
+      "step": 30811
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.813997423340218,
+      "learning_rate": 3.688194150257015e-06,
+      "loss": 0.8064,
+      "step": 30812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5982531984780124,
+      "learning_rate": 3.687602326582961e-06,
+      "loss": 0.8803,
+      "step": 30813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4673122336067894,
+      "learning_rate": 3.6870105396616795e-06,
+      "loss": 0.7785,
+      "step": 30814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.545740115615673,
+      "learning_rate": 3.6864187894966263e-06,
+      "loss": 0.6793,
+      "step": 30815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7077624165044318,
+      "learning_rate": 3.685827076091242e-06,
+      "loss": 0.8114,
+      "step": 30816
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4325819701076694,
+      "learning_rate": 3.6852353994489665e-06,
+      "loss": 0.7191,
+      "step": 30817
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3467691196574412,
+      "learning_rate": 3.684643759573251e-06,
+      "loss": 0.7352,
+      "step": 30818
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5152921691782062,
+      "learning_rate": 3.6840521564675424e-06,
+      "loss": 0.8418,
+      "step": 30819
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7349815606811436,
+      "learning_rate": 3.6834605901352804e-06,
+      "loss": 0.7685,
+      "step": 30820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6034214831977671,
+      "learning_rate": 3.6828690605799087e-06,
+      "loss": 0.9802,
+      "step": 30821
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3022016088168407,
+      "learning_rate": 3.6822775678048727e-06,
+      "loss": 0.7491,
+      "step": 30822
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5952227567714958,
+      "learning_rate": 3.6816861118136204e-06,
+      "loss": 0.8296,
+      "step": 30823
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.646708445481272,
+      "learning_rate": 3.6810946926095914e-06,
+      "loss": 0.8054,
+      "step": 30824
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6357052057982207,
+      "learning_rate": 3.680503310196227e-06,
+      "loss": 0.8786,
+      "step": 30825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7748658065301586,
+      "learning_rate": 3.679911964576972e-06,
+      "loss": 0.8101,
+      "step": 30826
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.770431169443128,
+      "learning_rate": 3.6793206557552753e-06,
+      "loss": 0.8074,
+      "step": 30827
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9321703449821674,
+      "learning_rate": 3.678729383734574e-06,
+      "loss": 0.809,
+      "step": 30828
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7313689909776991,
+      "learning_rate": 3.678138148518308e-06,
+      "loss": 0.6878,
+      "step": 30829
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5813579416497083,
+      "learning_rate": 3.677546950109927e-06,
+      "loss": 1.002,
+      "step": 30830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8003281878476152,
+      "learning_rate": 3.676955788512866e-06,
+      "loss": 0.701,
+      "step": 30831
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6422354506262946,
+      "learning_rate": 3.6763646637305726e-06,
+      "loss": 0.84,
+      "step": 30832
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6272097833571184,
+      "learning_rate": 3.6757735757664835e-06,
+      "loss": 0.8382,
+      "step": 30833
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.750016294980045,
+      "learning_rate": 3.675182524624047e-06,
+      "loss": 0.683,
+      "step": 30834
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6535978649529797,
+      "learning_rate": 3.6745915103066952e-06,
+      "loss": 0.8412,
+      "step": 30835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7366398532627209,
+      "learning_rate": 3.6740005328178786e-06,
+      "loss": 0.6772,
+      "step": 30836
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7788389794900779,
+      "learning_rate": 3.6734095921610336e-06,
+      "loss": 0.8998,
+      "step": 30837
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6218607789871922,
+      "learning_rate": 3.6728186883395978e-06,
+      "loss": 0.8297,
+      "step": 30838
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6206276911631081,
+      "learning_rate": 3.672227821357014e-06,
+      "loss": 0.7834,
+      "step": 30839
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5710937100062017,
+      "learning_rate": 3.671636991216727e-06,
+      "loss": 0.9219,
+      "step": 30840
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.862286819882762,
+      "learning_rate": 3.6710461979221735e-06,
+      "loss": 0.8614,
+      "step": 30841
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4801089876172093,
+      "learning_rate": 3.670455441476788e-06,
+      "loss": 0.7043,
+      "step": 30842
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7559150808891868,
+      "learning_rate": 3.6698647218840157e-06,
+      "loss": 0.8151,
+      "step": 30843
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5012468174288278,
+      "learning_rate": 3.669274039147299e-06,
+      "loss": 0.7012,
+      "step": 30844
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.729621633140653,
+      "learning_rate": 3.6686833932700717e-06,
+      "loss": 0.8577,
+      "step": 30845
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.535372595615835,
+      "learning_rate": 3.6680927842557724e-06,
+      "loss": 0.7447,
+      "step": 30846
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.645866596090497,
+      "learning_rate": 3.667502212107841e-06,
+      "loss": 0.8325,
+      "step": 30847
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6477430403266418,
+      "learning_rate": 3.6669116768297187e-06,
+      "loss": 0.791,
+      "step": 30848
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4897973711649395,
+      "learning_rate": 3.6663211784248433e-06,
+      "loss": 0.9105,
+      "step": 30849
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6290662682830936,
+      "learning_rate": 3.665730716896647e-06,
+      "loss": 0.7132,
+      "step": 30850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7567966974090823,
+      "learning_rate": 3.6651402922485715e-06,
+      "loss": 0.7766,
+      "step": 30851
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4958129137879188,
+      "learning_rate": 3.6645499044840594e-06,
+      "loss": 0.7863,
+      "step": 30852
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5069595532353366,
+      "learning_rate": 3.6639595536065418e-06,
+      "loss": 0.8045,
+      "step": 30853
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.355745949600577,
+      "learning_rate": 3.6633692396194554e-06,
+      "loss": 0.804,
+      "step": 30854
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5013256865268108,
+      "learning_rate": 3.6627789625262434e-06,
+      "loss": 0.8437,
+      "step": 30855
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5738165495209042,
+      "learning_rate": 3.662188722330334e-06,
+      "loss": 0.7265,
+      "step": 30856
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5479551394221027,
+      "learning_rate": 3.6615985190351722e-06,
+      "loss": 0.94,
+      "step": 30857
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.660963716133527,
+      "learning_rate": 3.6610083526441876e-06,
+      "loss": 0.8087,
+      "step": 30858
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8968282848085376,
+      "learning_rate": 3.660418223160822e-06,
+      "loss": 0.809,
+      "step": 30859
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.563858997820463,
+      "learning_rate": 3.6598281305885052e-06,
+      "loss": 0.8884,
+      "step": 30860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5199459307326257,
+      "learning_rate": 3.65923807493068e-06,
+      "loss": 0.9282,
+      "step": 30861
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5206712434427225,
+      "learning_rate": 3.658648056190778e-06,
+      "loss": 0.7803,
+      "step": 30862
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5660556300248971,
+      "learning_rate": 3.6580580743722315e-06,
+      "loss": 0.7295,
+      "step": 30863
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.607120506522893,
+      "learning_rate": 3.6574681294784788e-06,
+      "loss": 0.9756,
+      "step": 30864
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6823470336361348,
+      "learning_rate": 3.6568782215129574e-06,
+      "loss": 0.8375,
+      "step": 30865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.582129010867351,
+      "learning_rate": 3.6562883504791004e-06,
+      "loss": 0.771,
+      "step": 30866
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4720567734004475,
+      "learning_rate": 3.655698516380336e-06,
+      "loss": 0.927,
+      "step": 30867
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.591904908519025,
+      "learning_rate": 3.655108719220104e-06,
+      "loss": 0.8199,
+      "step": 30868
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7040195935361704,
+      "learning_rate": 3.6545189590018416e-06,
+      "loss": 0.8706,
+      "step": 30869
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5497391982264503,
+      "learning_rate": 3.653929235728979e-06,
+      "loss": 0.8952,
+      "step": 30870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7380488092945252,
+      "learning_rate": 3.653339549404946e-06,
+      "loss": 0.8953,
+      "step": 30871
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8239437552575328,
+      "learning_rate": 3.652749900033179e-06,
+      "loss": 0.8154,
+      "step": 30872
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6145005446413023,
+      "learning_rate": 3.6521602876171147e-06,
+      "loss": 0.8507,
+      "step": 30873
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5997215539945924,
+      "learning_rate": 3.651570712160184e-06,
+      "loss": 0.6388,
+      "step": 30874
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5130033795786437,
+      "learning_rate": 3.650981173665814e-06,
+      "loss": 0.8164,
+      "step": 30875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7872551469566583,
+      "learning_rate": 3.6503916721374456e-06,
+      "loss": 0.6881,
+      "step": 30876
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6534041216708784,
+      "learning_rate": 3.6498022075785034e-06,
+      "loss": 0.8797,
+      "step": 30877
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5390594782571188,
+      "learning_rate": 3.649212779992427e-06,
+      "loss": 0.7401,
+      "step": 30878
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.745977159872378,
+      "learning_rate": 3.6486233893826405e-06,
+      "loss": 0.8835,
+      "step": 30879
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5910660186859091,
+      "learning_rate": 3.6480340357525832e-06,
+      "loss": 0.9851,
+      "step": 30880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7446223895772723,
+      "learning_rate": 3.6474447191056793e-06,
+      "loss": 0.732,
+      "step": 30881
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6143905645604806,
+      "learning_rate": 3.6468554394453636e-06,
+      "loss": 0.8535,
+      "step": 30882
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6323517557744653,
+      "learning_rate": 3.6462661967750744e-06,
+      "loss": 0.8129,
+      "step": 30883
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5993382471417235,
+      "learning_rate": 3.645676991098227e-06,
+      "loss": 0.8307,
+      "step": 30884
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9079477425779177,
+      "learning_rate": 3.6450878224182608e-06,
+      "loss": 0.8555,
+      "step": 30885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7963760797249265,
+      "learning_rate": 3.6444986907386035e-06,
+      "loss": 0.751,
+      "step": 30886
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7656153192791455,
+      "learning_rate": 3.6439095960626957e-06,
+      "loss": 0.7049,
+      "step": 30887
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.605737909989496,
+      "learning_rate": 3.643320538393951e-06,
+      "loss": 0.8089,
+      "step": 30888
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1434681023335225,
+      "learning_rate": 3.642731517735807e-06,
+      "loss": 0.8529,
+      "step": 30889
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7269215343497717,
+      "learning_rate": 3.642142534091695e-06,
+      "loss": 0.7997,
+      "step": 30890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7412392143778843,
+      "learning_rate": 3.641553587465043e-06,
+      "loss": 0.6711,
+      "step": 30891
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7892412268037988,
+      "learning_rate": 3.6409646778592756e-06,
+      "loss": 0.6421,
+      "step": 30892
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.591629182943013,
+      "learning_rate": 3.640375805277825e-06,
+      "loss": 0.8614,
+      "step": 30893
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7592340461039903,
+      "learning_rate": 3.639786969724123e-06,
+      "loss": 0.7004,
+      "step": 30894
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7078182416074839,
+      "learning_rate": 3.6391981712015922e-06,
+      "loss": 0.7459,
+      "step": 30895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.593896711650856,
+      "learning_rate": 3.638609409713666e-06,
+      "loss": 0.907,
+      "step": 30896
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5503067844631317,
+      "learning_rate": 3.63802068526377e-06,
+      "loss": 0.8485,
+      "step": 30897
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5396389824065067,
+      "learning_rate": 3.6374319978553284e-06,
+      "loss": 0.8451,
+      "step": 30898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5303835922570588,
+      "learning_rate": 3.636843347491772e-06,
+      "loss": 0.9534,
+      "step": 30899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6195866355829662,
+      "learning_rate": 3.636254734176532e-06,
+      "loss": 0.8101,
+      "step": 30900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3613707549298948,
+      "learning_rate": 3.6356661579130314e-06,
+      "loss": 0.8375,
+      "step": 30901
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5795591289107636,
+      "learning_rate": 3.635077618704694e-06,
+      "loss": 0.7219,
+      "step": 30902
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6697596058621074,
+      "learning_rate": 3.63448911655495e-06,
+      "loss": 0.7924,
+      "step": 30903
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6900701917561058,
+      "learning_rate": 3.633900651467228e-06,
+      "loss": 0.8883,
+      "step": 30904
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.408217946017876,
+      "learning_rate": 3.6333122234449523e-06,
+      "loss": 0.7849,
+      "step": 30905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4088458234010148,
+      "learning_rate": 3.632723832491546e-06,
+      "loss": 0.7916,
+      "step": 30906
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.9775352948681135,
+      "learning_rate": 3.632135478610437e-06,
+      "loss": 0.9581,
+      "step": 30907
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6307390072404844,
+      "learning_rate": 3.6315471618050546e-06,
+      "loss": 0.7567,
+      "step": 30908
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5187371350394931,
+      "learning_rate": 3.6309588820788213e-06,
+      "loss": 0.797,
+      "step": 30909
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6560647034302831,
+      "learning_rate": 3.6303706394351567e-06,
+      "loss": 0.9136,
+      "step": 30910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5318258419036146,
+      "learning_rate": 3.6297824338774957e-06,
+      "loss": 0.891,
+      "step": 30911
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8852664146314666,
+      "learning_rate": 3.6291942654092547e-06,
+      "loss": 0.9446,
+      "step": 30912
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8557708805136401,
+      "learning_rate": 3.628606134033865e-06,
+      "loss": 0.9196,
+      "step": 30913
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7309034627096078,
+      "learning_rate": 3.6280180397547438e-06,
+      "loss": 0.7867,
+      "step": 30914
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4958756677889635,
+      "learning_rate": 3.6274299825753225e-06,
+      "loss": 0.9296,
+      "step": 30915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0397346567723043,
+      "learning_rate": 3.626841962499018e-06,
+      "loss": 0.8876,
+      "step": 30916
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6048309647969015,
+      "learning_rate": 3.6262539795292604e-06,
+      "loss": 0.8687,
+      "step": 30917
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6198287604946686,
+      "learning_rate": 3.6256660336694705e-06,
+      "loss": 0.8348,
+      "step": 30918
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5454915377053933,
+      "learning_rate": 3.625078124923067e-06,
+      "loss": 0.9063,
+      "step": 30919
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9701765029661618,
+      "learning_rate": 3.624490253293477e-06,
+      "loss": 0.7741,
+      "step": 30920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5949415803563105,
+      "learning_rate": 3.6239024187841266e-06,
+      "loss": 0.9967,
+      "step": 30921
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4933802652107737,
+      "learning_rate": 3.623314621398436e-06,
+      "loss": 0.798,
+      "step": 30922
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.766351654937019,
+      "learning_rate": 3.6227268611398214e-06,
+      "loss": 0.8775,
+      "step": 30923
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6966954917513069,
+      "learning_rate": 3.622139138011711e-06,
+      "loss": 0.844,
+      "step": 30924
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7790770667366763,
+      "learning_rate": 3.62155145201753e-06,
+      "loss": 0.7489,
+      "step": 30925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.771671151461201,
+      "learning_rate": 3.6209638031606942e-06,
+      "loss": 0.7206,
+      "step": 30926
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4975922601038745,
+      "learning_rate": 3.6203761914446243e-06,
+      "loss": 0.9162,
+      "step": 30927
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8298976513663738,
+      "learning_rate": 3.6197886168727436e-06,
+      "loss": 0.8428,
+      "step": 30928
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4217882051281638,
+      "learning_rate": 3.6192010794484766e-06,
+      "loss": 0.7737,
+      "step": 30929
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9942767636373302,
+      "learning_rate": 3.618613579175242e-06,
+      "loss": 0.904,
+      "step": 30930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.087659906731851,
+      "learning_rate": 3.618026116056456e-06,
+      "loss": 0.8745,
+      "step": 30931
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5717514959897416,
+      "learning_rate": 3.617438690095545e-06,
+      "loss": 0.7368,
+      "step": 30932
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.638206542692173,
+      "learning_rate": 3.6168513012959227e-06,
+      "loss": 0.7013,
+      "step": 30933
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5338898568009867,
+      "learning_rate": 3.6162639496610173e-06,
+      "loss": 0.8403,
+      "step": 30934
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8821695588606495,
+      "learning_rate": 3.615676635194242e-06,
+      "loss": 0.8229,
+      "step": 30935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.445196045190325,
+      "learning_rate": 3.6150893578990198e-06,
+      "loss": 0.8389,
+      "step": 30936
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4929949475289388,
+      "learning_rate": 3.6145021177787666e-06,
+      "loss": 0.7118,
+      "step": 30937
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7713365912358577,
+      "learning_rate": 3.613914914836907e-06,
+      "loss": 0.7301,
+      "step": 30938
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4520057371292712,
+      "learning_rate": 3.613327749076856e-06,
+      "loss": 0.8239,
+      "step": 30939
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7203480913466422,
+      "learning_rate": 3.6127406205020298e-06,
+      "loss": 0.7032,
+      "step": 30940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9364598323931095,
+      "learning_rate": 3.6121535291158506e-06,
+      "loss": 1.0313,
+      "step": 30941
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.707020554308533,
+      "learning_rate": 3.6115664749217383e-06,
+      "loss": 0.8554,
+      "step": 30942
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7473955933916807,
+      "learning_rate": 3.6109794579231094e-06,
+      "loss": 0.8121,
+      "step": 30943
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7324856432029208,
+      "learning_rate": 3.610392478123377e-06,
+      "loss": 0.816,
+      "step": 30944
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.831814930352606,
+      "learning_rate": 3.609805535525962e-06,
+      "loss": 0.796,
+      "step": 30945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5621623470137087,
+      "learning_rate": 3.6092186301342858e-06,
+      "loss": 0.8111,
+      "step": 30946
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8742420552997827,
+      "learning_rate": 3.608631761951763e-06,
+      "loss": 0.9375,
+      "step": 30947
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.092694141071174,
+      "learning_rate": 3.6080449309818045e-06,
+      "loss": 0.9344,
+      "step": 30948
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.528429417040673,
+      "learning_rate": 3.607458137227834e-06,
+      "loss": 0.7744,
+      "step": 30949
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5985598826055971,
+      "learning_rate": 3.6068713806932678e-06,
+      "loss": 0.8304,
+      "step": 30950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5056915553812087,
+      "learning_rate": 3.6062846613815217e-06,
+      "loss": 0.7831,
+      "step": 30951
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1399315429517665,
+      "learning_rate": 3.6056979792960067e-06,
+      "loss": 0.8803,
+      "step": 30952
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.48802085015612,
+      "learning_rate": 3.6051113344401457e-06,
+      "loss": 0.6522,
+      "step": 30953
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7193905285897744,
+      "learning_rate": 3.604524726817348e-06,
+      "loss": 0.8677,
+      "step": 30954
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9491027790992383,
+      "learning_rate": 3.603938156431035e-06,
+      "loss": 0.918,
+      "step": 30955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.499283651259715,
+      "learning_rate": 3.6033516232846154e-06,
+      "loss": 0.805,
+      "step": 30956
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6668380836368595,
+      "learning_rate": 3.602765127381511e-06,
+      "loss": 0.7727,
+      "step": 30957
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5633587219101832,
+      "learning_rate": 3.6021786687251315e-06,
+      "loss": 0.9807,
+      "step": 30958
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5534758454388569,
+      "learning_rate": 3.601592247318896e-06,
+      "loss": 0.7089,
+      "step": 30959
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5653905118291387,
+      "learning_rate": 3.601005863166216e-06,
+      "loss": 0.8741,
+      "step": 30960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8116161973678464,
+      "learning_rate": 3.600419516270501e-06,
+      "loss": 0.8394,
+      "step": 30961
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8409972485901582,
+      "learning_rate": 3.5998332066351716e-06,
+      "loss": 0.8035,
+      "step": 30962
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.578245478776862,
+      "learning_rate": 3.599246934263643e-06,
+      "loss": 0.8176,
+      "step": 30963
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9464979471180959,
+      "learning_rate": 3.598660699159324e-06,
+      "loss": 0.852,
+      "step": 30964
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4227773282761629,
+      "learning_rate": 3.5980745013256258e-06,
+      "loss": 0.8618,
+      "step": 30965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5944820244011348,
+      "learning_rate": 3.5974883407659655e-06,
+      "loss": 0.965,
+      "step": 30966
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7703364349835577,
+      "learning_rate": 3.5969022174837576e-06,
+      "loss": 0.9136,
+      "step": 30967
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8763506572258217,
+      "learning_rate": 3.5963161314824123e-06,
+      "loss": 0.8448,
+      "step": 30968
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.580010094669336,
+      "learning_rate": 3.5957300827653385e-06,
+      "loss": 0.8791,
+      "step": 30969
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5093498870299509,
+      "learning_rate": 3.595144071335952e-06,
+      "loss": 0.7574,
+      "step": 30970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.765242254200734,
+      "learning_rate": 3.594558097197668e-06,
+      "loss": 0.8839,
+      "step": 30971
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8409917466218262,
+      "learning_rate": 3.5939721603538946e-06,
+      "loss": 0.8893,
+      "step": 30972
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7707181113493712,
+      "learning_rate": 3.5933862608080405e-06,
+      "loss": 0.713,
+      "step": 30973
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7645829586862287,
+      "learning_rate": 3.592800398563522e-06,
+      "loss": 0.7967,
+      "step": 30974
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4621157564380143,
+      "learning_rate": 3.5922145736237457e-06,
+      "loss": 0.8651,
+      "step": 30975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4854988435657148,
+      "learning_rate": 3.5916287859921273e-06,
+      "loss": 0.8039,
+      "step": 30976
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7069474568556449,
+      "learning_rate": 3.5910430356720727e-06,
+      "loss": 0.9578,
+      "step": 30977
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6610036175445313,
+      "learning_rate": 3.590457322666997e-06,
+      "loss": 0.7566,
+      "step": 30978
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.372396051811444,
+      "learning_rate": 3.589871646980305e-06,
+      "loss": 0.6955,
+      "step": 30979
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7710771820580291,
+      "learning_rate": 3.5892860086154103e-06,
+      "loss": 0.7309,
+      "step": 30980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8213879302754412,
+      "learning_rate": 3.588700407575725e-06,
+      "loss": 0.8295,
+      "step": 30981
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4486246679680952,
+      "learning_rate": 3.588114843864655e-06,
+      "loss": 0.7803,
+      "step": 30982
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5672221022841015,
+      "learning_rate": 3.5875293174856073e-06,
+      "loss": 0.7104,
+      "step": 30983
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6784127334037375,
+      "learning_rate": 3.5869438284419946e-06,
+      "loss": 0.8722,
+      "step": 30984
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4117319046821637,
+      "learning_rate": 3.586358376737231e-06,
+      "loss": 0.8642,
+      "step": 30985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5341311657442473,
+      "learning_rate": 3.585772962374714e-06,
+      "loss": 0.8509,
+      "step": 30986
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.917165299537371,
+      "learning_rate": 3.5851875853578568e-06,
+      "loss": 0.8908,
+      "step": 30987
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7026811883110096,
+      "learning_rate": 3.584602245690069e-06,
+      "loss": 0.9058,
+      "step": 30988
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5897984441678465,
+      "learning_rate": 3.584016943374764e-06,
+      "loss": 0.7902,
+      "step": 30989
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9276657380030056,
+      "learning_rate": 3.583431678415337e-06,
+      "loss": 0.8947,
+      "step": 30990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.496727131636126,
+      "learning_rate": 3.582846450815202e-06,
+      "loss": 0.8677,
+      "step": 30991
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.878398226958861,
+      "learning_rate": 3.5822612605777706e-06,
+      "loss": 0.9998,
+      "step": 30992
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8163694666879702,
+      "learning_rate": 3.581676107706442e-06,
+      "loss": 0.9885,
+      "step": 30993
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1871608762633414,
+      "learning_rate": 3.581090992204631e-06,
+      "loss": 0.8875,
+      "step": 30994
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5997343874369043,
+      "learning_rate": 3.580505914075737e-06,
+      "loss": 0.822,
+      "step": 30995
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5687081094670283,
+      "learning_rate": 3.5799208733231727e-06,
+      "loss": 0.907,
+      "step": 30996
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7756670944540305,
+      "learning_rate": 3.579335869950339e-06,
+      "loss": 0.8965,
+      "step": 30997
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4894702364143186,
+      "learning_rate": 3.578750903960647e-06,
+      "loss": 0.759,
+      "step": 30998
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.510909512879887,
+      "learning_rate": 3.5781659753575015e-06,
+      "loss": 0.8715,
+      "step": 30999
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6289281692819675,
+      "learning_rate": 3.5775810841443025e-06,
+      "loss": 0.7783,
+      "step": 31000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7248427540459498,
+      "learning_rate": 3.5769962303244597e-06,
+      "loss": 0.6373,
+      "step": 31001
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5754985542383029,
+      "learning_rate": 3.5764114139013828e-06,
+      "loss": 0.7298,
+      "step": 31002
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5153635517658908,
+      "learning_rate": 3.575826634878471e-06,
+      "loss": 0.7231,
+      "step": 31003
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.421367871333325,
+      "learning_rate": 3.575241893259128e-06,
+      "loss": 0.8924,
+      "step": 31004
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.06237800566534,
+      "learning_rate": 3.5746571890467607e-06,
+      "loss": 0.9096,
+      "step": 31005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7234971071541907,
+      "learning_rate": 3.574072522244776e-06,
+      "loss": 0.7707,
+      "step": 31006
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7496493735138663,
+      "learning_rate": 3.5734878928565763e-06,
+      "loss": 0.8658,
+      "step": 31007
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.774506967452363,
+      "learning_rate": 3.572903300885561e-06,
+      "loss": 0.7232,
+      "step": 31008
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6874804781241532,
+      "learning_rate": 3.572318746335137e-06,
+      "loss": 0.7826,
+      "step": 31009
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4463882004328426,
+      "learning_rate": 3.571734229208712e-06,
+      "loss": 0.7548,
+      "step": 31010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6365322915336897,
+      "learning_rate": 3.571149749509686e-06,
+      "loss": 0.7696,
+      "step": 31011
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8300069662055009,
+      "learning_rate": 3.570565307241457e-06,
+      "loss": 0.6852,
+      "step": 31012
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.561060059159068,
+      "learning_rate": 3.5699809024074373e-06,
+      "loss": 0.931,
+      "step": 31013
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8632855716597474,
+      "learning_rate": 3.5693965350110194e-06,
+      "loss": 0.8347,
+      "step": 31014
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6356585796183247,
+      "learning_rate": 3.568812205055615e-06,
+      "loss": 0.8684,
+      "step": 31015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0558957827280873,
+      "learning_rate": 3.5682279125446185e-06,
+      "loss": 0.8546,
+      "step": 31016
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.572609447842363,
+      "learning_rate": 3.567643657481439e-06,
+      "loss": 0.8243,
+      "step": 31017
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.54549904883278,
+      "learning_rate": 3.5670594398694713e-06,
+      "loss": 0.8131,
+      "step": 31018
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7163676340942033,
+      "learning_rate": 3.5664752597121233e-06,
+      "loss": 0.9543,
+      "step": 31019
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5025519428620928,
+      "learning_rate": 3.565891117012793e-06,
+      "loss": 0.8002,
+      "step": 31020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4846301713934227,
+      "learning_rate": 3.5653070117748777e-06,
+      "loss": 0.8179,
+      "step": 31021
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7593803567502018,
+      "learning_rate": 3.5647229440017827e-06,
+      "loss": 0.86,
+      "step": 31022
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6159495715001124,
+      "learning_rate": 3.564138913696912e-06,
+      "loss": 0.8682,
+      "step": 31023
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4595552796685645,
+      "learning_rate": 3.5635549208636623e-06,
+      "loss": 0.8942,
+      "step": 31024
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5421966532081217,
+      "learning_rate": 3.562970965505429e-06,
+      "loss": 0.7189,
+      "step": 31025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5293799129014227,
+      "learning_rate": 3.5623870476256174e-06,
+      "loss": 0.9093,
+      "step": 31026
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.559746187552289,
+      "learning_rate": 3.56180316722763e-06,
+      "loss": 0.8638,
+      "step": 31027
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.768547488349881,
+      "learning_rate": 3.5612193243148617e-06,
+      "loss": 0.6737,
+      "step": 31028
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6525515453928137,
+      "learning_rate": 3.5606355188907116e-06,
+      "loss": 0.8574,
+      "step": 31029
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.726424739303502,
+      "learning_rate": 3.5600517509585784e-06,
+      "loss": 0.9219,
+      "step": 31030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7073750520860989,
+      "learning_rate": 3.5594680205218667e-06,
+      "loss": 0.9288,
+      "step": 31031
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.415781808367532,
+      "learning_rate": 3.558884327583971e-06,
+      "loss": 0.7501,
+      "step": 31032
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1369787852104647,
+      "learning_rate": 3.5583006721482873e-06,
+      "loss": 0.7223,
+      "step": 31033
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.287027638786993,
+      "learning_rate": 3.5577170542182195e-06,
+      "loss": 0.9225,
+      "step": 31034
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.117818467200204,
+      "learning_rate": 3.5571334737971596e-06,
+      "loss": 0.9558,
+      "step": 31035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5461711581421407,
+      "learning_rate": 3.556549930888511e-06,
+      "loss": 0.8279,
+      "step": 31036
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.657747423890486,
+      "learning_rate": 3.555966425495666e-06,
+      "loss": 0.781,
+      "step": 31037
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6236227360027489,
+      "learning_rate": 3.5553829576220277e-06,
+      "loss": 0.9442,
+      "step": 31038
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9378727510597222,
+      "learning_rate": 3.554799527270987e-06,
+      "loss": 0.8146,
+      "step": 31039
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.798218683795611,
+      "learning_rate": 3.554216134445947e-06,
+      "loss": 0.6948,
+      "step": 31040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7321184721803593,
+      "learning_rate": 3.553632779150301e-06,
+      "loss": 0.7721,
+      "step": 31041
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6725277071195337,
+      "learning_rate": 3.5530494613874433e-06,
+      "loss": 0.6527,
+      "step": 31042
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5757198029432455,
+      "learning_rate": 3.552466181160772e-06,
+      "loss": 0.9493,
+      "step": 31043
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6966445732142943,
+      "learning_rate": 3.5518829384736888e-06,
+      "loss": 0.818,
+      "step": 31044
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6762120492326145,
+      "learning_rate": 3.551299733329583e-06,
+      "loss": 0.6924,
+      "step": 31045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6635694897698943,
+      "learning_rate": 3.5507165657318487e-06,
+      "loss": 0.8466,
+      "step": 31046
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5316374237164634,
+      "learning_rate": 3.5501334356838846e-06,
+      "loss": 0.8077,
+      "step": 31047
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6577952763092643,
+      "learning_rate": 3.549550343189089e-06,
+      "loss": 0.782,
+      "step": 31048
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.611200947601348,
+      "learning_rate": 3.548967288250853e-06,
+      "loss": 0.9408,
+      "step": 31049
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5729337031636903,
+      "learning_rate": 3.548384270872568e-06,
+      "loss": 0.8225,
+      "step": 31050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.506282567151283,
+      "learning_rate": 3.5478012910576332e-06,
+      "loss": 0.8992,
+      "step": 31051
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6645105084798042,
+      "learning_rate": 3.547218348809446e-06,
+      "loss": 0.8241,
+      "step": 31052
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5117132886182965,
+      "learning_rate": 3.5466354441313954e-06,
+      "loss": 0.8302,
+      "step": 31053
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.911776800408748,
+      "learning_rate": 3.5460525770268726e-06,
+      "loss": 0.8447,
+      "step": 31054
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.631078719172728,
+      "learning_rate": 3.5454697474992796e-06,
+      "loss": 0.8714,
+      "step": 31055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5645005079845042,
+      "learning_rate": 3.5448869555520016e-06,
+      "loss": 0.8417,
+      "step": 31056
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0871643554810113,
+      "learning_rate": 3.5443042011884377e-06,
+      "loss": 0.8006,
+      "step": 31057
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3211036934799307,
+      "learning_rate": 3.543721484411976e-06,
+      "loss": 0.7585,
+      "step": 31058
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5685236392111777,
+      "learning_rate": 3.543138805226015e-06,
+      "loss": 0.8358,
+      "step": 31059
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.716836500274421,
+      "learning_rate": 3.542556163633941e-06,
+      "loss": 0.6693,
+      "step": 31060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7160352620001327,
+      "learning_rate": 3.5419735596391537e-06,
+      "loss": 0.8352,
+      "step": 31061
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6494376121129193,
+      "learning_rate": 3.5413909932450398e-06,
+      "loss": 0.7968,
+      "step": 31062
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4473462482419939,
+      "learning_rate": 3.540808464454989e-06,
+      "loss": 0.8699,
+      "step": 31063
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9096491594675762,
+      "learning_rate": 3.5402259732723953e-06,
+      "loss": 1.0693,
+      "step": 31064
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4573295022298594,
+      "learning_rate": 3.539643519700653e-06,
+      "loss": 0.883,
+      "step": 31065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5888328962151361,
+      "learning_rate": 3.5390611037431564e-06,
+      "loss": 0.8349,
+      "step": 31066
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7596974906766278,
+      "learning_rate": 3.538478725403286e-06,
+      "loss": 0.8371,
+      "step": 31067
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8350068636548457,
+      "learning_rate": 3.5378963846844383e-06,
+      "loss": 0.7133,
+      "step": 31068
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5513017446088317,
+      "learning_rate": 3.5373140815900065e-06,
+      "loss": 0.8311,
+      "step": 31069
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.483336627159492,
+      "learning_rate": 3.5367318161233777e-06,
+      "loss": 0.7318,
+      "step": 31070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5724196629191325,
+      "learning_rate": 3.5361495882879406e-06,
+      "loss": 0.9337,
+      "step": 31071
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7603736659458658,
+      "learning_rate": 3.5355673980870853e-06,
+      "loss": 0.838,
+      "step": 31072
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4028802190245897,
+      "learning_rate": 3.5349852455242086e-06,
+      "loss": 0.7843,
+      "step": 31073
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6921588186065841,
+      "learning_rate": 3.534403130602693e-06,
+      "loss": 0.7292,
+      "step": 31074
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7644648281431727,
+      "learning_rate": 3.5338210533259264e-06,
+      "loss": 0.9987,
+      "step": 31075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5513907378788099,
+      "learning_rate": 3.5332390136973036e-06,
+      "loss": 0.9082,
+      "step": 31076
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.720107513669406,
+      "learning_rate": 3.5326570117202076e-06,
+      "loss": 0.926,
+      "step": 31077
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.490629222719129,
+      "learning_rate": 3.5320750473980304e-06,
+      "loss": 0.8386,
+      "step": 31078
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9544483751732071,
+      "learning_rate": 3.531493120734163e-06,
+      "loss": 0.7975,
+      "step": 31079
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6284374774210375,
+      "learning_rate": 3.5309112317319916e-06,
+      "loss": 0.7296,
+      "step": 31080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6738201301514808,
+      "learning_rate": 3.530329380394898e-06,
+      "loss": 1.0014,
+      "step": 31081
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.509370553029579,
+      "learning_rate": 3.529747566726276e-06,
+      "loss": 0.825,
+      "step": 31082
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4021883282116083,
+      "learning_rate": 3.5291657907295184e-06,
+      "loss": 0.8073,
+      "step": 31083
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.57634874411246,
+      "learning_rate": 3.5285840524079994e-06,
+      "loss": 0.7894,
+      "step": 31084
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7702112297343868,
+      "learning_rate": 3.528002351765114e-06,
+      "loss": 0.6865,
+      "step": 31085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9173311430095965,
+      "learning_rate": 3.527420688804247e-06,
+      "loss": 0.8417,
+      "step": 31086
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5441130566829022,
+      "learning_rate": 3.526839063528793e-06,
+      "loss": 0.8742,
+      "step": 31087
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6632959945923527,
+      "learning_rate": 3.5262574759421243e-06,
+      "loss": 1.0608,
+      "step": 31088
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6194348306859057,
+      "learning_rate": 3.525675926047634e-06,
+      "loss": 0.6931,
+      "step": 31089
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7384930743878888,
+      "learning_rate": 3.525094413848712e-06,
+      "loss": 0.6408,
+      "step": 31090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8838030306221363,
+      "learning_rate": 3.5245129393487365e-06,
+      "loss": 0.8464,
+      "step": 31091
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8070665452127385,
+      "learning_rate": 3.5239315025511e-06,
+      "loss": 0.7417,
+      "step": 31092
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8353617841239807,
+      "learning_rate": 3.523350103459182e-06,
+      "loss": 0.6943,
+      "step": 31093
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.584994730365866,
+      "learning_rate": 3.522768742076372e-06,
+      "loss": 0.9281,
+      "step": 31094
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7931933404546225,
+      "learning_rate": 3.5221874184060513e-06,
+      "loss": 0.858,
+      "step": 31095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7671633307797239,
+      "learning_rate": 3.5216061324516093e-06,
+      "loss": 0.8107,
+      "step": 31096
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6167415749731193,
+      "learning_rate": 3.5210248842164274e-06,
+      "loss": 0.9298,
+      "step": 31097
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8562592341225164,
+      "learning_rate": 3.520443673703886e-06,
+      "loss": 0.7677,
+      "step": 31098
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8299520540508598,
+      "learning_rate": 3.5198625009173726e-06,
+      "loss": 0.9517,
+      "step": 31099
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.641948683533227,
+      "learning_rate": 3.5192813658602754e-06,
+      "loss": 0.9575,
+      "step": 31100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6788738427514207,
+      "learning_rate": 3.518700268535974e-06,
+      "loss": 0.7937,
+      "step": 31101
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6312725564332724,
+      "learning_rate": 3.5181192089478477e-06,
+      "loss": 0.8125,
+      "step": 31102
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.526292263556944,
+      "learning_rate": 3.517538187099284e-06,
+      "loss": 0.7159,
+      "step": 31103
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.135415793920318,
+      "learning_rate": 3.5169572029936682e-06,
+      "loss": 0.9233,
+      "step": 31104
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6125965218020817,
+      "learning_rate": 3.516376256634381e-06,
+      "loss": 0.9036,
+      "step": 31105
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7125096608360385,
+      "learning_rate": 3.5157953480247985e-06,
+      "loss": 0.6922,
+      "step": 31106
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8891422992056084,
+      "learning_rate": 3.5152144771683107e-06,
+      "loss": 0.8173,
+      "step": 31107
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5355035544305862,
+      "learning_rate": 3.5146336440682995e-06,
+      "loss": 0.7361,
+      "step": 31108
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4670686297437254,
+      "learning_rate": 3.5140528487281443e-06,
+      "loss": 0.8857,
+      "step": 31109
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6194429871286657,
+      "learning_rate": 3.5134720911512233e-06,
+      "loss": 0.7771,
+      "step": 31110
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7121589324212028,
+      "learning_rate": 3.512891371340922e-06,
+      "loss": 0.8644,
+      "step": 31111
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4201231285901728,
+      "learning_rate": 3.5123106893006242e-06,
+      "loss": 0.8833,
+      "step": 31112
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.61775434538853,
+      "learning_rate": 3.5117300450337066e-06,
+      "loss": 0.6585,
+      "step": 31113
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6573366745419533,
+      "learning_rate": 3.5111494385435484e-06,
+      "loss": 0.8853,
+      "step": 31114
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6681166599057555,
+      "learning_rate": 3.5105688698335363e-06,
+      "loss": 0.8396,
+      "step": 31115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.46063333840646,
+      "learning_rate": 3.509988338907042e-06,
+      "loss": 0.9214,
+      "step": 31116
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078891735611307,
+      "learning_rate": 3.5094078457674553e-06,
+      "loss": 0.8564,
+      "step": 31117
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5644358177543343,
+      "learning_rate": 3.5088273904181457e-06,
+      "loss": 0.7376,
+      "step": 31118
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4818838369997016,
+      "learning_rate": 3.5082469728625034e-06,
+      "loss": 0.8647,
+      "step": 31119
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.762503043005377,
+      "learning_rate": 3.5076665931038976e-06,
+      "loss": 0.9061,
+      "step": 31120
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.501613826222138,
+      "learning_rate": 3.5070862511457172e-06,
+      "loss": 0.8754,
+      "step": 31121
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.942927141006117,
+      "learning_rate": 3.5065059469913356e-06,
+      "loss": 0.7899,
+      "step": 31122
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5618080367903013,
+      "learning_rate": 3.5059256806441276e-06,
+      "loss": 0.7093,
+      "step": 31123
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6718496268726482,
+      "learning_rate": 3.505345452107478e-06,
+      "loss": 0.8401,
+      "step": 31124
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5320954770854904,
+      "learning_rate": 3.504765261384766e-06,
+      "loss": 0.7837,
+      "step": 31125
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9729592305294372,
+      "learning_rate": 3.5041851084793665e-06,
+      "loss": 0.7951,
+      "step": 31126
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5032958862406909,
+      "learning_rate": 3.503604993394656e-06,
+      "loss": 0.7875,
+      "step": 31127
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4859141917992609,
+      "learning_rate": 3.5030249161340124e-06,
+      "loss": 0.7911,
+      "step": 31128
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6125185277447014,
+      "learning_rate": 3.502444876700819e-06,
+      "loss": 0.8736,
+      "step": 31129
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5527007552171985,
+      "learning_rate": 3.5018648750984473e-06,
+      "loss": 0.8622,
+      "step": 31130
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6513636653234884,
+      "learning_rate": 3.5012849113302737e-06,
+      "loss": 0.9117,
+      "step": 31131
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4669760706964319,
+      "learning_rate": 3.5007049853996767e-06,
+      "loss": 0.8858,
+      "step": 31132
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.914237931314017,
+      "learning_rate": 3.5001250973100353e-06,
+      "loss": 0.7592,
+      "step": 31133
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5221118380711864,
+      "learning_rate": 3.499545247064724e-06,
+      "loss": 0.712,
+      "step": 31134
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6334568192910146,
+      "learning_rate": 3.4989654346671143e-06,
+      "loss": 0.8473,
+      "step": 31135
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4554911838444247,
+      "learning_rate": 3.4983856601205903e-06,
+      "loss": 0.8665,
+      "step": 31136
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.508258580416763,
+      "learning_rate": 3.49780592342852e-06,
+      "loss": 0.8246,
+      "step": 31137
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6558177609501792,
+      "learning_rate": 3.4972262245942847e-06,
+      "loss": 0.8179,
+      "step": 31138
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7857980760736442,
+      "learning_rate": 3.4966465636212543e-06,
+      "loss": 0.7457,
+      "step": 31139
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7515333080296321,
+      "learning_rate": 3.49606694051281e-06,
+      "loss": 0.7293,
+      "step": 31140
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4898645232959788,
+      "learning_rate": 3.4954873552723188e-06,
+      "loss": 0.7204,
+      "step": 31141
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5579948728240933,
+      "learning_rate": 3.494907807903164e-06,
+      "loss": 0.7444,
+      "step": 31142
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9014422975733787,
+      "learning_rate": 3.4943282984087147e-06,
+      "loss": 0.8098,
+      "step": 31143
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4558844488978515,
+      "learning_rate": 3.4937488267923426e-06,
+      "loss": 0.7981,
+      "step": 31144
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5113842512508255,
+      "learning_rate": 3.493169393057425e-06,
+      "loss": 0.8134,
+      "step": 31145
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4990714415356647,
+      "learning_rate": 3.4925899972073387e-06,
+      "loss": 0.8266,
+      "step": 31146
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7167539743896848,
+      "learning_rate": 3.4920106392454523e-06,
+      "loss": 0.7319,
+      "step": 31147
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5230074345201,
+      "learning_rate": 3.491431319175138e-06,
+      "loss": 0.7602,
+      "step": 31148
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5367416761472776,
+      "learning_rate": 3.490852036999771e-06,
+      "loss": 0.8441,
+      "step": 31149
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8718819503874555,
+      "learning_rate": 3.4902727927227277e-06,
+      "loss": 0.8238,
+      "step": 31150
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4442280830797323,
+      "learning_rate": 3.489693586347378e-06,
+      "loss": 0.8495,
+      "step": 31151
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6425076441138673,
+      "learning_rate": 3.4891144178770885e-06,
+      "loss": 0.8397,
+      "step": 31152
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.556229178019339,
+      "learning_rate": 3.488535287315237e-06,
+      "loss": 0.8605,
+      "step": 31153
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125061333006547,
+      "learning_rate": 3.4879561946651974e-06,
+      "loss": 0.8527,
+      "step": 31154
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.619691922659208,
+      "learning_rate": 3.487377139930339e-06,
+      "loss": 0.7494,
+      "step": 31155
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7194805511699771,
+      "learning_rate": 3.486798123114029e-06,
+      "loss": 0.8872,
+      "step": 31156
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8143339166706567,
+      "learning_rate": 3.486219144219646e-06,
+      "loss": 0.8968,
+      "step": 31157
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.659112089158367,
+      "learning_rate": 3.485640203250553e-06,
+      "loss": 0.7201,
+      "step": 31158
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6151050872791397,
+      "learning_rate": 3.4850613002101295e-06,
+      "loss": 0.7642,
+      "step": 31159
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4807965380792514,
+      "learning_rate": 3.4844824351017394e-06,
+      "loss": 0.7186,
+      "step": 31160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.572209090255249,
+      "learning_rate": 3.483903607928757e-06,
+      "loss": 0.8262,
+      "step": 31161
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.764160479943143,
+      "learning_rate": 3.4833248186945477e-06,
+      "loss": 0.6874,
+      "step": 31162
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7310361126206382,
+      "learning_rate": 3.4827460674024847e-06,
+      "loss": 0.7253,
+      "step": 31163
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6204411522338777,
+      "learning_rate": 3.4821673540559442e-06,
+      "loss": 0.8363,
+      "step": 31164
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3905417860566598,
+      "learning_rate": 3.481588678658282e-06,
+      "loss": 0.7954,
+      "step": 31165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6366571535930705,
+      "learning_rate": 3.4810100412128743e-06,
+      "loss": 0.7889,
+      "step": 31166
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7155220170747794,
+      "learning_rate": 3.48043144172309e-06,
+      "loss": 0.8169,
+      "step": 31167
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.426860938128788,
+      "learning_rate": 3.479852880192305e-06,
+      "loss": 0.8178,
+      "step": 31168
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5723645659303993,
+      "learning_rate": 3.479274356623874e-06,
+      "loss": 0.8893,
+      "step": 31169
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5425125285646664,
+      "learning_rate": 3.4786958710211714e-06,
+      "loss": 0.8479,
+      "step": 31170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4743373471124266,
+      "learning_rate": 3.4781174233875705e-06,
+      "loss": 0.7477,
+      "step": 31171
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5812952639343922,
+      "learning_rate": 3.477539013726433e-06,
+      "loss": 0.8166,
+      "step": 31172
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5170559466875084,
+      "learning_rate": 3.4769606420411273e-06,
+      "loss": 0.8543,
+      "step": 31173
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5326199438184975,
+      "learning_rate": 3.4763823083350202e-06,
+      "loss": 0.9759,
+      "step": 31174
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0276446838623903,
+      "learning_rate": 3.4758040126114845e-06,
+      "loss": 0.9136,
+      "step": 31175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7746177114706374,
+      "learning_rate": 3.47522575487388e-06,
+      "loss": 0.8986,
+      "step": 31176
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6714364829770805,
+      "learning_rate": 3.4746475351255806e-06,
+      "loss": 0.8181,
+      "step": 31177
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7393032930940089,
+      "learning_rate": 3.474069353369949e-06,
+      "loss": 0.8966,
+      "step": 31178
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6063549120846865,
+      "learning_rate": 3.473491209610347e-06,
+      "loss": 0.7762,
+      "step": 31179
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5686384992743154,
+      "learning_rate": 3.4729131038501475e-06,
+      "loss": 0.8309,
+      "step": 31180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7366821125086267,
+      "learning_rate": 3.4723350360927165e-06,
+      "loss": 0.6174,
+      "step": 31181
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5281473856432823,
+      "learning_rate": 3.4717570063414174e-06,
+      "loss": 0.7829,
+      "step": 31182
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.101135261219831,
+      "learning_rate": 3.4711790145996127e-06,
+      "loss": 0.7535,
+      "step": 31183
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0585867969848604,
+      "learning_rate": 3.470601060870671e-06,
+      "loss": 0.7876,
+      "step": 31184
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.095024263878455,
+      "learning_rate": 3.4700231451579634e-06,
+      "loss": 0.888,
+      "step": 31185
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8586349179297807,
+      "learning_rate": 3.4694452674648425e-06,
+      "loss": 0.8292,
+      "step": 31186
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5500221022571368,
+      "learning_rate": 3.4688674277946775e-06,
+      "loss": 0.9246,
+      "step": 31187
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6725698471755683,
+      "learning_rate": 3.468289626150836e-06,
+      "loss": 0.869,
+      "step": 31188
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5805073938616447,
+      "learning_rate": 3.4677118625366825e-06,
+      "loss": 0.9234,
+      "step": 31189
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7806429311479244,
+      "learning_rate": 3.4671341369555776e-06,
+      "loss": 0.7795,
+      "step": 31190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6276742330927236,
+      "learning_rate": 3.4665564494108837e-06,
+      "loss": 0.9276,
+      "step": 31191
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.705175028423474,
+      "learning_rate": 3.46597879990597e-06,
+      "loss": 0.814,
+      "step": 31192
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4910752499599191,
+      "learning_rate": 3.465401188444193e-06,
+      "loss": 0.9492,
+      "step": 31193
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.755487975214288,
+      "learning_rate": 3.464823615028923e-06,
+      "loss": 0.6902,
+      "step": 31194
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6260625419852666,
+      "learning_rate": 3.4642460796635146e-06,
+      "loss": 0.7903,
+      "step": 31195
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3987136114091356,
+      "learning_rate": 3.4636685823513384e-06,
+      "loss": 0.9035,
+      "step": 31196
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6040580735647982,
+      "learning_rate": 3.4630911230957486e-06,
+      "loss": 0.8058,
+      "step": 31197
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8251279544047871,
+      "learning_rate": 3.4625137019001164e-06,
+      "loss": 0.8838,
+      "step": 31198
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4779702755033766,
+      "learning_rate": 3.461936318767799e-06,
+      "loss": 0.6385,
+      "step": 31199
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9185822604505856,
+      "learning_rate": 3.4613589737021557e-06,
+      "loss": 0.8843,
+      "step": 31200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1247614712445437,
+      "learning_rate": 3.4607816667065497e-06,
+      "loss": 0.7451,
+      "step": 31201
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.701953550581488,
+      "learning_rate": 3.460204397784347e-06,
+      "loss": 0.7661,
+      "step": 31202
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6573617220584604,
+      "learning_rate": 3.459627166938904e-06,
+      "loss": 0.7663,
+      "step": 31203
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.498664919226866,
+      "learning_rate": 3.4590499741735796e-06,
+      "loss": 0.7956,
+      "step": 31204
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1185337439739143,
+      "learning_rate": 3.4584728194917363e-06,
+      "loss": 0.8569,
+      "step": 31205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5109648388460761,
+      "learning_rate": 3.4578957028967388e-06,
+      "loss": 0.8857,
+      "step": 31206
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5826534430785022,
+      "learning_rate": 3.4573186243919443e-06,
+      "loss": 0.9227,
+      "step": 31207
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8070382117305026,
+      "learning_rate": 3.4567415839807083e-06,
+      "loss": 0.9337,
+      "step": 31208
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.407938943714929,
+      "learning_rate": 3.4561645816663945e-06,
+      "loss": 0.8584,
+      "step": 31209
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6679394921683748,
+      "learning_rate": 3.455587617452365e-06,
+      "loss": 0.7052,
+      "step": 31210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.451438622543245,
+      "learning_rate": 3.455010691341978e-06,
+      "loss": 0.7701,
+      "step": 31211
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5819590661717822,
+      "learning_rate": 3.454433803338586e-06,
+      "loss": 0.6095,
+      "step": 31212
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6603929699922793,
+      "learning_rate": 3.453856953445557e-06,
+      "loss": 0.969,
+      "step": 31213
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9619861287696783,
+      "learning_rate": 3.453280141666241e-06,
+      "loss": 0.9299,
+      "step": 31214
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7501263453783982,
+      "learning_rate": 3.4527033680040047e-06,
+      "loss": 0.7422,
+      "step": 31215
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6166232586856286,
+      "learning_rate": 3.4521266324621983e-06,
+      "loss": 0.7492,
+      "step": 31216
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4233982869388784,
+      "learning_rate": 3.451549935044187e-06,
+      "loss": 0.9888,
+      "step": 31217
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.475635479765036,
+      "learning_rate": 3.450973275753322e-06,
+      "loss": 0.8517,
+      "step": 31218
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.496515024133514,
+      "learning_rate": 3.450396654592968e-06,
+      "loss": 0.7962,
+      "step": 31219
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4436359703471107,
+      "learning_rate": 3.4498200715664775e-06,
+      "loss": 0.8701,
+      "step": 31220
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9128903787221607,
+      "learning_rate": 3.449243526677205e-06,
+      "loss": 0.8949,
+      "step": 31221
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6219957139030354,
+      "learning_rate": 3.448667019928511e-06,
+      "loss": 0.8357,
+      "step": 31222
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.514135068176155,
+      "learning_rate": 3.448090551323755e-06,
+      "loss": 0.7581,
+      "step": 31223
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.620561261542445,
+      "learning_rate": 3.44751412086629e-06,
+      "loss": 0.7552,
+      "step": 31224
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.376109475633645,
+      "learning_rate": 3.4469377285594687e-06,
+      "loss": 0.7835,
+      "step": 31225
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.574130214941863,
+      "learning_rate": 3.4463613744066516e-06,
+      "loss": 0.8831,
+      "step": 31226
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6321567249382414,
+      "learning_rate": 3.445785058411195e-06,
+      "loss": 0.7617,
+      "step": 31227
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.673933386015155,
+      "learning_rate": 3.4452087805764544e-06,
+      "loss": 0.8959,
+      "step": 31228
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6804942334117272,
+      "learning_rate": 3.4446325409057792e-06,
+      "loss": 0.8763,
+      "step": 31229
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.584565997092863,
+      "learning_rate": 3.4440563394025294e-06,
+      "loss": 0.7656,
+      "step": 31230
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9172676934977735,
+      "learning_rate": 3.443480176070062e-06,
+      "loss": 0.8167,
+      "step": 31231
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.713757748434657,
+      "learning_rate": 3.4429040509117295e-06,
+      "loss": 0.8488,
+      "step": 31232
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5249897333928686,
+      "learning_rate": 3.4423279639308815e-06,
+      "loss": 0.8683,
+      "step": 31233
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.623023767312557,
+      "learning_rate": 3.4417519151308763e-06,
+      "loss": 0.8296,
+      "step": 31234
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5556337116909265,
+      "learning_rate": 3.441175904515072e-06,
+      "loss": 0.8036,
+      "step": 31235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6860411881671193,
+      "learning_rate": 3.4405999320868165e-06,
+      "loss": 0.944,
+      "step": 31236
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6562137921447815,
+      "learning_rate": 3.440023997849462e-06,
+      "loss": 0.8906,
+      "step": 31237
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5415690202965273,
+      "learning_rate": 3.4394481018063685e-06,
+      "loss": 0.8683,
+      "step": 31238
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5418240942212276,
+      "learning_rate": 3.4388722439608822e-06,
+      "loss": 0.8824,
+      "step": 31239
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5441120000495896,
+      "learning_rate": 3.438296424316362e-06,
+      "loss": 0.7434,
+      "step": 31240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.640946571093667,
+      "learning_rate": 3.4377206428761534e-06,
+      "loss": 0.7747,
+      "step": 31241
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7734559169950823,
+      "learning_rate": 3.4371448996436163e-06,
+      "loss": 0.7137,
+      "step": 31242
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7863953166926804,
+      "learning_rate": 3.436569194622096e-06,
+      "loss": 0.7117,
+      "step": 31243
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.64941963169593,
+      "learning_rate": 3.4359935278149513e-06,
+      "loss": 0.8637,
+      "step": 31244
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6334527567755859,
+      "learning_rate": 3.4354178992255306e-06,
+      "loss": 0.8427,
+      "step": 31245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4151611308793208,
+      "learning_rate": 3.4348423088571815e-06,
+      "loss": 0.8551,
+      "step": 31246
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.568412706982141,
+      "learning_rate": 3.4342667567132593e-06,
+      "loss": 0.8927,
+      "step": 31247
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6793781179777136,
+      "learning_rate": 3.433691242797118e-06,
+      "loss": 0.8276,
+      "step": 31248
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7199062869953519,
+      "learning_rate": 3.433115767112105e-06,
+      "loss": 0.709,
+      "step": 31249
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.995481714148886,
+      "learning_rate": 3.4325403296615677e-06,
+      "loss": 0.7755,
+      "step": 31250
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.675713216169196,
+      "learning_rate": 3.43196493044886e-06,
+      "loss": 0.8246,
+      "step": 31251
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6522548871953697,
+      "learning_rate": 3.4313895694773357e-06,
+      "loss": 0.8319,
+      "step": 31252
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4372767018149404,
+      "learning_rate": 3.4308142467503404e-06,
+      "loss": 0.7961,
+      "step": 31253
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6377073585295272,
+      "learning_rate": 3.4302389622712207e-06,
+      "loss": 0.7989,
+      "step": 31254
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6247914250907074,
+      "learning_rate": 3.42966371604333e-06,
+      "loss": 0.7519,
+      "step": 31255
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4557812735931266,
+      "learning_rate": 3.429088508070021e-06,
+      "loss": 0.7974,
+      "step": 31256
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6389981399954774,
+      "learning_rate": 3.428513338354639e-06,
+      "loss": 0.8701,
+      "step": 31257
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9888708755589992,
+      "learning_rate": 3.4279382069005283e-06,
+      "loss": 0.91,
+      "step": 31258
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.431649155330322,
+      "learning_rate": 3.4273631137110464e-06,
+      "loss": 0.9134,
+      "step": 31259
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0115927711757653,
+      "learning_rate": 3.4267880587895343e-06,
+      "loss": 0.9024,
+      "step": 31260
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1318684933329655,
+      "learning_rate": 3.4262130421393424e-06,
+      "loss": 0.8266,
+      "step": 31261
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.474897528357021,
+      "learning_rate": 3.425638063763823e-06,
+      "loss": 0.9599,
+      "step": 31262
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.619770553971361,
+      "learning_rate": 3.42506312366632e-06,
+      "loss": 0.7985,
+      "step": 31263
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8109839160523549,
+      "learning_rate": 3.4244882218501775e-06,
+      "loss": 0.8766,
+      "step": 31264
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5767337386174318,
+      "learning_rate": 3.423913358318747e-06,
+      "loss": 0.8931,
+      "step": 31265
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6656406481160286,
+      "learning_rate": 3.423338533075381e-06,
+      "loss": 0.9288,
+      "step": 31266
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.487432178112543,
+      "learning_rate": 3.422763746123414e-06,
+      "loss": 0.8089,
+      "step": 31267
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8694492747661164,
+      "learning_rate": 3.422188997466199e-06,
+      "loss": 0.7965,
+      "step": 31268
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8117106599918071,
+      "learning_rate": 3.421614287107082e-06,
+      "loss": 0.8222,
+      "step": 31269
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5509366713446733,
+      "learning_rate": 3.4210396150494152e-06,
+      "loss": 0.8456,
+      "step": 31270
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4638464458711025,
+      "learning_rate": 3.4204649812965327e-06,
+      "loss": 0.8955,
+      "step": 31271
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7900476100067728,
+      "learning_rate": 3.4198903858517863e-06,
+      "loss": 0.9053,
+      "step": 31272
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5488428908153238,
+      "learning_rate": 3.4193158287185248e-06,
+      "loss": 0.7518,
+      "step": 31273
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6785178068039244,
+      "learning_rate": 3.4187413099000854e-06,
+      "loss": 0.8419,
+      "step": 31274
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8407634500870929,
+      "learning_rate": 3.4181668293998216e-06,
+      "loss": 0.8552,
+      "step": 31275
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7326517592472894,
+      "learning_rate": 3.4175923872210715e-06,
+      "loss": 0.8552,
+      "step": 31276
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7330906231700662,
+      "learning_rate": 3.4170179833671847e-06,
+      "loss": 0.6955,
+      "step": 31277
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7559146000394292,
+      "learning_rate": 3.4164436178415007e-06,
+      "loss": 0.8208,
+      "step": 31278
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.557181502657161,
+      "learning_rate": 3.41586929064737e-06,
+      "loss": 0.8064,
+      "step": 31279
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7196376276333951,
+      "learning_rate": 3.4152950017881325e-06,
+      "loss": 0.8018,
+      "step": 31280
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7145171661998637,
+      "learning_rate": 3.414720751267129e-06,
+      "loss": 0.7225,
+      "step": 31281
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.930195175475632,
+      "learning_rate": 3.414146539087706e-06,
+      "loss": 0.9221,
+      "step": 31282
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.647942284322207,
+      "learning_rate": 3.4135723652532093e-06,
+      "loss": 0.9309,
+      "step": 31283
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4322087081022654,
+      "learning_rate": 3.4129982297669807e-06,
+      "loss": 0.8684,
+      "step": 31284
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5144968502685443,
+      "learning_rate": 3.412424132632358e-06,
+      "loss": 0.8827,
+      "step": 31285
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5137342234814575,
+      "learning_rate": 3.411850073852688e-06,
+      "loss": 0.9526,
+      "step": 31286
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.943241910568344,
+      "learning_rate": 3.411276053431315e-06,
+      "loss": 0.9152,
+      "step": 31287
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7227616858187775,
+      "learning_rate": 3.41070207137158e-06,
+      "loss": 0.7628,
+      "step": 31288
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6215456201387555,
+      "learning_rate": 3.4101281276768194e-06,
+      "loss": 0.826,
+      "step": 31289
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6189803351970702,
+      "learning_rate": 3.4095542223503795e-06,
+      "loss": 0.7926,
+      "step": 31290
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6881865619642267,
+      "learning_rate": 3.4089803553956058e-06,
+      "loss": 0.7586,
+      "step": 31291
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6944235234887177,
+      "learning_rate": 3.4084065268158338e-06,
+      "loss": 0.76,
+      "step": 31292
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5332961270650616,
+      "learning_rate": 3.4078327366144025e-06,
+      "loss": 0.8605,
+      "step": 31293
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2589739062573213,
+      "learning_rate": 3.4072589847946603e-06,
+      "loss": 0.7977,
+      "step": 31294
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5558341018536948,
+      "learning_rate": 3.40668527135994e-06,
+      "loss": 0.7725,
+      "step": 31295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.555602083397845,
+      "learning_rate": 3.406111596313588e-06,
+      "loss": 0.8379,
+      "step": 31296
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1637350737750225,
+      "learning_rate": 3.4055379596589387e-06,
+      "loss": 0.8501,
+      "step": 31297
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6078860439359972,
+      "learning_rate": 3.404964361399339e-06,
+      "loss": 0.987,
+      "step": 31298
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.829308926888376,
+      "learning_rate": 3.4043908015381212e-06,
+      "loss": 0.9059,
+      "step": 31299
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8679675477395263,
+      "learning_rate": 3.4038172800786317e-06,
+      "loss": 0.9717,
+      "step": 31300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.609550367256968,
+      "learning_rate": 3.403243797024206e-06,
+      "loss": 0.8403,
+      "step": 31301
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7565167949352097,
+      "learning_rate": 3.40267035237818e-06,
+      "loss": 0.721,
+      "step": 31302
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7912077896531724,
+      "learning_rate": 3.4020969461438966e-06,
+      "loss": 1.0281,
+      "step": 31303
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.49691171482178,
+      "learning_rate": 3.401523578324697e-06,
+      "loss": 0.8703,
+      "step": 31304
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4505400589453779,
+      "learning_rate": 3.4009502489239157e-06,
+      "loss": 0.6545,
+      "step": 31305
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7765467216584765,
+      "learning_rate": 3.4003769579448875e-06,
+      "loss": 0.8363,
+      "step": 31306
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7693398857496374,
+      "learning_rate": 3.399803705390955e-06,
+      "loss": 0.8344,
+      "step": 31307
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.630596663321736,
+      "learning_rate": 3.3992304912654573e-06,
+      "loss": 0.8697,
+      "step": 31308
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5741947144653579,
+      "learning_rate": 3.3986573155717307e-06,
+      "loss": 0.7138,
+      "step": 31309
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3579015310168547,
+      "learning_rate": 3.3980841783131067e-06,
+      "loss": 0.863,
+      "step": 31310
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5525905545336265,
+      "learning_rate": 3.3975110794929267e-06,
+      "loss": 0.8188,
+      "step": 31311
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6605141303754989,
+      "learning_rate": 3.396938019114532e-06,
+      "loss": 0.8147,
+      "step": 31312
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7291667148648259,
+      "learning_rate": 3.396364997181254e-06,
+      "loss": 0.7114,
+      "step": 31313
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.647296854736395,
+      "learning_rate": 3.395792013696426e-06,
+      "loss": 0.9408,
+      "step": 31314
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.468120856338312,
+      "learning_rate": 3.3952190686633914e-06,
+      "loss": 0.8172,
+      "step": 31315
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6811381435888673,
+      "learning_rate": 3.3946461620854788e-06,
+      "loss": 0.9437,
+      "step": 31316
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5953532774762702,
+      "learning_rate": 3.3940732939660303e-06,
+      "loss": 0.7021,
+      "step": 31317
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6693708949017823,
+      "learning_rate": 3.393500464308377e-06,
+      "loss": 0.9024,
+      "step": 31318
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4211695403441411,
+      "learning_rate": 3.392927673115857e-06,
+      "loss": 0.8279,
+      "step": 31319
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7907400079215352,
+      "learning_rate": 3.3923549203918017e-06,
+      "loss": 0.8777,
+      "step": 31320
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6731805309309267,
+      "learning_rate": 3.391782206139551e-06,
+      "loss": 0.7674,
+      "step": 31321
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6099537747014199,
+      "learning_rate": 3.391209530362436e-06,
+      "loss": 0.8842,
+      "step": 31322
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5660925551287261,
+      "learning_rate": 3.390636893063789e-06,
+      "loss": 0.7704,
+      "step": 31323
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.672403596301627,
+      "learning_rate": 3.390064294246945e-06,
+      "loss": 0.7952,
+      "step": 31324
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9288594178496357,
+      "learning_rate": 3.3894917339152444e-06,
+      "loss": 0.8471,
+      "step": 31325
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5315619228040964,
+      "learning_rate": 3.388919212072015e-06,
+      "loss": 0.8695,
+      "step": 31326
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5852995251638282,
+      "learning_rate": 3.3883467287205874e-06,
+      "loss": 0.761,
+      "step": 31327
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4679489860191062,
+      "learning_rate": 3.387774283864298e-06,
+      "loss": 0.9087,
+      "step": 31328
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6709938774520001,
+      "learning_rate": 3.387201877506484e-06,
+      "loss": 0.8519,
+      "step": 31329
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.465875408195359,
+      "learning_rate": 3.3866295096504744e-06,
+      "loss": 0.7785,
+      "step": 31330
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5052598468016143,
+      "learning_rate": 3.386057180299599e-06,
+      "loss": 0.8001,
+      "step": 31331
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.779067680542931,
+      "learning_rate": 3.3854848894571913e-06,
+      "loss": 0.9056,
+      "step": 31332
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4832791137960277,
+      "learning_rate": 3.384912637126588e-06,
+      "loss": 0.8771,
+      "step": 31333
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8565345461590985,
+      "learning_rate": 3.3843404233111187e-06,
+      "loss": 0.7258,
+      "step": 31334
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8401667553526725,
+      "learning_rate": 3.38376824801411e-06,
+      "loss": 0.7205,
+      "step": 31335
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7043487934791626,
+      "learning_rate": 3.3831961112389e-06,
+      "loss": 0.6893,
+      "step": 31336
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.702373130504811,
+      "learning_rate": 3.3826240129888143e-06,
+      "loss": 0.7936,
+      "step": 31337
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.510205663565103,
+      "learning_rate": 3.3820519532671905e-06,
+      "loss": 0.9367,
+      "step": 31338
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6012328168732264,
+      "learning_rate": 3.3814799320773517e-06,
+      "loss": 0.7546,
+      "step": 31339
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6431960339588851,
+      "learning_rate": 3.3809079494226347e-06,
+      "loss": 0.8408,
+      "step": 31340
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3960611924950492,
+      "learning_rate": 3.380336005306365e-06,
+      "loss": 0.8507,
+      "step": 31341
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7674809508358842,
+      "learning_rate": 3.379764099731877e-06,
+      "loss": 0.9545,
+      "step": 31342
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8946222879931898,
+      "learning_rate": 3.3791922327024994e-06,
+      "loss": 0.8419,
+      "step": 31343
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7316989729849661,
+      "learning_rate": 3.3786204042215576e-06,
+      "loss": 0.8374,
+      "step": 31344
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5114429312009547,
+      "learning_rate": 3.378048614292383e-06,
+      "loss": 0.7579,
+      "step": 31345
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.87983762237803,
+      "learning_rate": 3.3774768629183065e-06,
+      "loss": 0.7275,
+      "step": 31346
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6942818827828674,
+      "learning_rate": 3.376905150102663e-06,
+      "loss": 0.8457,
+      "step": 31347
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5855024879414574,
+      "learning_rate": 3.3763334758487674e-06,
+      "loss": 0.8414,
+      "step": 31348
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203751351962395,
+      "learning_rate": 3.375761840159956e-06,
+      "loss": 0.8328,
+      "step": 31349
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.922050606629176,
+      "learning_rate": 3.3751902430395558e-06,
+      "loss": 0.8478,
+      "step": 31350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4949487497685336,
+      "learning_rate": 3.3746186844909025e-06,
+      "loss": 0.8642,
+      "step": 31351
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4974944827189183,
+      "learning_rate": 3.374047164517309e-06,
+      "loss": 0.9646,
+      "step": 31352
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.623379667695261,
+      "learning_rate": 3.3734756831221116e-06,
+      "loss": 0.7756,
+      "step": 31353
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7658461028716053,
+      "learning_rate": 3.37290424030864e-06,
+      "loss": 0.9216,
+      "step": 31354
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7347171622197082,
+      "learning_rate": 3.3723328360802187e-06,
+      "loss": 0.7447,
+      "step": 31355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6307094839699674,
+      "learning_rate": 3.371761470440169e-06,
+      "loss": 0.8056,
+      "step": 31356
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7731251329923072,
+      "learning_rate": 3.371190143391824e-06,
+      "loss": 0.8001,
+      "step": 31357
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4209625328615645,
+      "learning_rate": 3.370618854938511e-06,
+      "loss": 0.8042,
+      "step": 31358
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.377201573068278,
+      "learning_rate": 3.3700476050835506e-06,
+      "loss": 0.8343,
+      "step": 31359
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6400541385968728,
+      "learning_rate": 3.3694763938302764e-06,
+      "loss": 0.7873,
+      "step": 31360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.725952040719986,
+      "learning_rate": 3.368905221182008e-06,
+      "loss": 0.8335,
+      "step": 31361
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4651244844202862,
+      "learning_rate": 3.3683340871420712e-06,
+      "loss": 0.9062,
+      "step": 31362
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.429391466435904,
+      "learning_rate": 3.367762991713792e-06,
+      "loss": 0.8635,
+      "step": 31363
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.697232590930448,
+      "learning_rate": 3.3671919349005e-06,
+      "loss": 0.8916,
+      "step": 31364
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6189379410993538,
+      "learning_rate": 3.3666209167055163e-06,
+      "loss": 0.8215,
+      "step": 31365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5850278991862525,
+      "learning_rate": 3.366049937132162e-06,
+      "loss": 0.8227,
+      "step": 31366
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6310896691958077,
+      "learning_rate": 3.3654789961837666e-06,
+      "loss": 0.7403,
+      "step": 31367
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5934706428412522,
+      "learning_rate": 3.364908093863658e-06,
+      "loss": 0.8774,
+      "step": 31368
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6386256642846297,
+      "learning_rate": 3.3643372301751498e-06,
+      "loss": 0.8561,
+      "step": 31369
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4881711350864202,
+      "learning_rate": 3.3637664051215703e-06,
+      "loss": 0.7219,
+      "step": 31370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4059808371038367,
+      "learning_rate": 3.363195618706243e-06,
+      "loss": 0.8824,
+      "step": 31371
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5580895068768552,
+      "learning_rate": 3.3626248709324973e-06,
+      "loss": 0.8585,
+      "step": 31372
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7350294430905449,
+      "learning_rate": 3.3620541618036496e-06,
+      "loss": 0.7934,
+      "step": 31373
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5568412662261346,
+      "learning_rate": 3.361483491323021e-06,
+      "loss": 0.7454,
+      "step": 31374
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.804538472816006,
+      "learning_rate": 3.36091285949394e-06,
+      "loss": 0.6822,
+      "step": 31375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5922252714101988,
+      "learning_rate": 3.3603422663197236e-06,
+      "loss": 0.8032,
+      "step": 31376
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4421413387137019,
+      "learning_rate": 3.3597717118037e-06,
+      "loss": 0.9038,
+      "step": 31377
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.765984856140628,
+      "learning_rate": 3.3592011959491844e-06,
+      "loss": 0.9307,
+      "step": 31378
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5898842955059838,
+      "learning_rate": 3.3586307187595045e-06,
+      "loss": 0.7437,
+      "step": 31379
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6186529071441522,
+      "learning_rate": 3.358060280237977e-06,
+      "loss": 0.9352,
+      "step": 31380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6017745792736804,
+      "learning_rate": 3.3574898803879286e-06,
+      "loss": 0.8277,
+      "step": 31381
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7592568671606834,
+      "learning_rate": 3.3569195192126758e-06,
+      "loss": 0.8464,
+      "step": 31382
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4174968780826964,
+      "learning_rate": 3.356349196715538e-06,
+      "loss": 0.6691,
+      "step": 31383
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6039235754669476,
+      "learning_rate": 3.35577891289984e-06,
+      "loss": 0.8367,
+      "step": 31384
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4838157357067614,
+      "learning_rate": 3.355208667768902e-06,
+      "loss": 0.8494,
+      "step": 31385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6763000320391235,
+      "learning_rate": 3.354638461326043e-06,
+      "loss": 0.8282,
+      "step": 31386
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6212639033903091,
+      "learning_rate": 3.3540682935745804e-06,
+      "loss": 0.8892,
+      "step": 31387
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.72917833211955,
+      "learning_rate": 3.3534981645178356e-06,
+      "loss": 0.8584,
+      "step": 31388
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.830869585853778,
+      "learning_rate": 3.352928074159133e-06,
+      "loss": 1.0399,
+      "step": 31389
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8962336517454004,
+      "learning_rate": 3.3523580225017872e-06,
+      "loss": 0.804,
+      "step": 31390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7277401956108294,
+      "learning_rate": 3.351788009549114e-06,
+      "loss": 0.8656,
+      "step": 31391
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4353957756799567,
+      "learning_rate": 3.3512180353044356e-06,
+      "loss": 0.8224,
+      "step": 31392
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7146589037768536,
+      "learning_rate": 3.3506480997710743e-06,
+      "loss": 0.9261,
+      "step": 31393
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6413590978258963,
+      "learning_rate": 3.350078202952345e-06,
+      "loss": 0.8367,
+      "step": 31394
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7352978840647768,
+      "learning_rate": 3.349508344851562e-06,
+      "loss": 0.6983,
+      "step": 31395
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.579972537774891,
+      "learning_rate": 3.3489385254720507e-06,
+      "loss": 0.709,
+      "step": 31396
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5839576523807406,
+      "learning_rate": 3.348368744817121e-06,
+      "loss": 0.9049,
+      "step": 31397
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.918763852988626,
+      "learning_rate": 3.3477990028900986e-06,
+      "loss": 0.9754,
+      "step": 31398
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5711324348951272,
+      "learning_rate": 3.347229299694292e-06,
+      "loss": 0.8393,
+      "step": 31399
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4333570852459936,
+      "learning_rate": 3.346659635233027e-06,
+      "loss": 0.8767,
+      "step": 31400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9125305694903119,
+      "learning_rate": 3.346090009509613e-06,
+      "loss": 0.9374,
+      "step": 31401
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6523581621077534,
+      "learning_rate": 3.3455204225273718e-06,
+      "loss": 0.8483,
+      "step": 31402
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.255687385769303,
+      "learning_rate": 3.3449508742896176e-06,
+      "loss": 0.7176,
+      "step": 31403
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7514125378409942,
+      "learning_rate": 3.3443813647996638e-06,
+      "loss": 0.8385,
+      "step": 31404
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7664226240173058,
+      "learning_rate": 3.343811894060829e-06,
+      "loss": 0.8759,
+      "step": 31405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5089172382715728,
+      "learning_rate": 3.3432424620764304e-06,
+      "loss": 0.8542,
+      "step": 31406
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5508006312599911,
+      "learning_rate": 3.342673068849783e-06,
+      "loss": 0.9244,
+      "step": 31407
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.835663252401099,
+      "learning_rate": 3.342103714384197e-06,
+      "loss": 0.8457,
+      "step": 31408
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7860572790871404,
+      "learning_rate": 3.3415343986829908e-06,
+      "loss": 0.6945,
+      "step": 31409
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6095874948672844,
+      "learning_rate": 3.340965121749483e-06,
+      "loss": 0.924,
+      "step": 31410
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6881071747944358,
+      "learning_rate": 3.3403958835869843e-06,
+      "loss": 0.8905,
+      "step": 31411
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.706887794960307,
+      "learning_rate": 3.339826684198806e-06,
+      "loss": 0.8234,
+      "step": 31412
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7427949482767795,
+      "learning_rate": 3.3392575235882654e-06,
+      "loss": 0.7316,
+      "step": 31413
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.601855224241939,
+      "learning_rate": 3.3386884017586794e-06,
+      "loss": 0.8622,
+      "step": 31414
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5515306589463218,
+      "learning_rate": 3.3381193187133587e-06,
+      "loss": 0.8072,
+      "step": 31415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8174309134572674,
+      "learning_rate": 3.3375502744556134e-06,
+      "loss": 0.7661,
+      "step": 31416
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4290365753204555,
+      "learning_rate": 3.3369812689887617e-06,
+      "loss": 0.7273,
+      "step": 31417
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7203864810328726,
+      "learning_rate": 3.336412302316112e-06,
+      "loss": 0.9712,
+      "step": 31418
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5926872934122667,
+      "learning_rate": 3.3358433744409835e-06,
+      "loss": 0.8779,
+      "step": 31419
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6089274021954179,
+      "learning_rate": 3.3352744853666807e-06,
+      "loss": 0.8564,
+      "step": 31420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8113602605163082,
+      "learning_rate": 3.3347056350965235e-06,
+      "loss": 0.7365,
+      "step": 31421
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9270323745880962,
+      "learning_rate": 3.334136823633817e-06,
+      "loss": 0.7721,
+      "step": 31422
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.717518087606718,
+      "learning_rate": 3.333568050981879e-06,
+      "loss": 0.7661,
+      "step": 31423
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.573879372712607,
+      "learning_rate": 3.332999317144019e-06,
+      "loss": 0.7965,
+      "step": 31424
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.950316475398071,
+      "learning_rate": 3.3324306221235435e-06,
+      "loss": 0.783,
+      "step": 31425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.542352066321408,
+      "learning_rate": 3.3318619659237684e-06,
+      "loss": 0.7082,
+      "step": 31426
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7985950402056858,
+      "learning_rate": 3.331293348548007e-06,
+      "loss": 0.7686,
+      "step": 31427
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.050150566718566,
+      "learning_rate": 3.330724769999568e-06,
+      "loss": 0.9089,
+      "step": 31428
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6655404046473936,
+      "learning_rate": 3.3301562302817568e-06,
+      "loss": 0.7843,
+      "step": 31429
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5352456030279276,
+      "learning_rate": 3.3295877293978875e-06,
+      "loss": 0.7705,
+      "step": 31430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4989116078120224,
+      "learning_rate": 3.3290192673512747e-06,
+      "loss": 0.7781,
+      "step": 31431
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7497986364411686,
+      "learning_rate": 3.3284508441452225e-06,
+      "loss": 0.8403,
+      "step": 31432
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5171802111002106,
+      "learning_rate": 3.3278824597830383e-06,
+      "loss": 0.9113,
+      "step": 31433
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6741900500647036,
+      "learning_rate": 3.3273141142680356e-06,
+      "loss": 0.8704,
+      "step": 31434
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6975540981116337,
+      "learning_rate": 3.3267458076035255e-06,
+      "loss": 0.8038,
+      "step": 31435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6091722861588675,
+      "learning_rate": 3.3261775397928143e-06,
+      "loss": 0.8805,
+      "step": 31436
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9025985518797683,
+      "learning_rate": 3.3256093108392064e-06,
+      "loss": 0.8736,
+      "step": 31437
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8278302551033299,
+      "learning_rate": 3.3250411207460188e-06,
+      "loss": 0.852,
+      "step": 31438
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4776526767972493,
+      "learning_rate": 3.32447296951655e-06,
+      "loss": 0.8696,
+      "step": 31439
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5326943243481561,
+      "learning_rate": 3.323904857154118e-06,
+      "loss": 0.8102,
+      "step": 31440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8779657352048855,
+      "learning_rate": 3.323336783662021e-06,
+      "loss": 0.7295,
+      "step": 31441
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7990434528430697,
+      "learning_rate": 3.3227687490435746e-06,
+      "loss": 0.7054,
+      "step": 31442
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5482277969484632,
+      "learning_rate": 3.3222007533020793e-06,
+      "loss": 0.9044,
+      "step": 31443
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8707493711383096,
+      "learning_rate": 3.3216327964408456e-06,
+      "loss": 0.7829,
+      "step": 31444
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.657385729007701,
+      "learning_rate": 3.321064878463187e-06,
+      "loss": 0.8929,
+      "step": 31445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7566664453846401,
+      "learning_rate": 3.3204969993723955e-06,
+      "loss": 0.6819,
+      "step": 31446
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5785966552460018,
+      "learning_rate": 3.3199291591717865e-06,
+      "loss": 0.8934,
+      "step": 31447
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7954012123538212,
+      "learning_rate": 3.3193613578646633e-06,
+      "loss": 0.9258,
+      "step": 31448
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6303345906107571,
+      "learning_rate": 3.3187935954543417e-06,
+      "loss": 0.8519,
+      "step": 31449
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.561098167808049,
+      "learning_rate": 3.3182258719441107e-06,
+      "loss": 0.9945,
+      "step": 31450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2165589685564577,
+      "learning_rate": 3.3176581873372848e-06,
+      "loss": 0.8745,
+      "step": 31451
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4654238760109843,
+      "learning_rate": 3.3170905416371723e-06,
+      "loss": 0.8081,
+      "step": 31452
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6147960991458779,
+      "learning_rate": 3.316522934847074e-06,
+      "loss": 0.824,
+      "step": 31453
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.717977305072124,
+      "learning_rate": 3.3159553669702917e-06,
+      "loss": 0.8081,
+      "step": 31454
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5621878059499321,
+      "learning_rate": 3.3153878380101332e-06,
+      "loss": 0.9051,
+      "step": 31455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7280268110302472,
+      "learning_rate": 3.314820347969907e-06,
+      "loss": 0.8949,
+      "step": 31456
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5053775904682616,
+      "learning_rate": 3.3142528968529097e-06,
+      "loss": 0.7514,
+      "step": 31457
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.895045229462465,
+      "learning_rate": 3.313685484662452e-06,
+      "loss": 0.7599,
+      "step": 31458
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.91928901615495,
+      "learning_rate": 3.313118111401834e-06,
+      "loss": 0.7804,
+      "step": 31459
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6299395663118228,
+      "learning_rate": 3.312550777074356e-06,
+      "loss": 0.8773,
+      "step": 31460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7114340654593048,
+      "learning_rate": 3.311983481683325e-06,
+      "loss": 0.884,
+      "step": 31461
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.098533356825804,
+      "learning_rate": 3.311416225232047e-06,
+      "loss": 0.8856,
+      "step": 31462
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5335678787994265,
+      "learning_rate": 3.3108490077238207e-06,
+      "loss": 0.7649,
+      "step": 31463
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9159110579643586,
+      "learning_rate": 3.3102818291619467e-06,
+      "loss": 0.8851,
+      "step": 31464
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4724889177653602,
+      "learning_rate": 3.309714689549729e-06,
+      "loss": 0.73,
+      "step": 31465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5678938286640853,
+      "learning_rate": 3.3091475888904777e-06,
+      "loss": 0.9386,
+      "step": 31466
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5805269022209916,
+      "learning_rate": 3.3085805271874803e-06,
+      "loss": 0.7708,
+      "step": 31467
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3813119388267794,
+      "learning_rate": 3.3080135044440454e-06,
+      "loss": 0.7749,
+      "step": 31468
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5906291370619674,
+      "learning_rate": 3.3074465206634755e-06,
+      "loss": 0.9011,
+      "step": 31469
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7949148931712188,
+      "learning_rate": 3.3068795758490736e-06,
+      "loss": 0.6913,
+      "step": 31470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5695554376736793,
+      "learning_rate": 3.306312670004137e-06,
+      "loss": 0.8562,
+      "step": 31471
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.485242174426283,
+      "learning_rate": 3.3057458031319654e-06,
+      "loss": 0.8314,
+      "step": 31472
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1059894413889735,
+      "learning_rate": 3.30517897523586e-06,
+      "loss": 0.7492,
+      "step": 31473
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2504281524605063,
+      "learning_rate": 3.304612186319126e-06,
+      "loss": 0.7747,
+      "step": 31474
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.236861978552538,
+      "learning_rate": 3.304045436385059e-06,
+      "loss": 0.7178,
+      "step": 31475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.639652496610335,
+      "learning_rate": 3.303478725436956e-06,
+      "loss": 0.8441,
+      "step": 31476
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.539847715054025,
+      "learning_rate": 3.3029120534781244e-06,
+      "loss": 0.8,
+      "step": 31477
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6143410425255205,
+      "learning_rate": 3.3023454205118545e-06,
+      "loss": 0.825,
+      "step": 31478
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7957466347856023,
+      "learning_rate": 3.301778826541454e-06,
+      "loss": 0.7788,
+      "step": 31479
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4363087993412773,
+      "learning_rate": 3.3012122715702145e-06,
+      "loss": 0.7862,
+      "step": 31480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6563983762527197,
+      "learning_rate": 3.3006457556014415e-06,
+      "loss": 0.9955,
+      "step": 31481
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5363398872079688,
+      "learning_rate": 3.300079278638426e-06,
+      "loss": 0.8887,
+      "step": 31482
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5302725951419756,
+      "learning_rate": 3.2995128406844733e-06,
+      "loss": 0.7523,
+      "step": 31483
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7524620378387306,
+      "learning_rate": 3.2989464417428785e-06,
+      "loss": 0.7167,
+      "step": 31484
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6727479810172445,
+      "learning_rate": 3.2983800818169363e-06,
+      "loss": 0.8914,
+      "step": 31485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5780676130357016,
+      "learning_rate": 3.297813760909946e-06,
+      "loss": 0.7272,
+      "step": 31486
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7632856485895223,
+      "learning_rate": 3.2972474790252097e-06,
+      "loss": 0.9602,
+      "step": 31487
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9018352859439815,
+      "learning_rate": 3.2966812361660204e-06,
+      "loss": 0.8736,
+      "step": 31488
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6610382632116545,
+      "learning_rate": 3.2961150323356715e-06,
+      "loss": 0.7721,
+      "step": 31489
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7579564506809282,
+      "learning_rate": 3.2955488675374635e-06,
+      "loss": 0.7426,
+      "step": 31490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5593122589857775,
+      "learning_rate": 3.294982741774697e-06,
+      "loss": 0.8686,
+      "step": 31491
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8197546883856422,
+      "learning_rate": 3.294416655050663e-06,
+      "loss": 0.7052,
+      "step": 31492
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.449698054690433,
+      "learning_rate": 3.293850607368655e-06,
+      "loss": 0.7786,
+      "step": 31493
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6161558185424303,
+      "learning_rate": 3.2932845987319704e-06,
+      "loss": 0.8344,
+      "step": 31494
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5977779819653706,
+      "learning_rate": 3.2927186291439117e-06,
+      "loss": 0.8768,
+      "step": 31495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6938680947810332,
+      "learning_rate": 3.292152698607768e-06,
+      "loss": 0.7426,
+      "step": 31496
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4509111382929611,
+      "learning_rate": 3.291586807126831e-06,
+      "loss": 0.9104,
+      "step": 31497
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5137823257218745,
+      "learning_rate": 3.2910209547044027e-06,
+      "loss": 0.8138,
+      "step": 31498
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7337887735507316,
+      "learning_rate": 3.290455141343771e-06,
+      "loss": 0.8347,
+      "step": 31499
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6964680834791543,
+      "learning_rate": 3.289889367048238e-06,
+      "loss": 0.8589,
+      "step": 31500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.508663115213615,
+      "learning_rate": 3.28932363182109e-06,
+      "loss": 0.8342,
+      "step": 31501
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4648015389321118,
+      "learning_rate": 3.2887579356656274e-06,
+      "loss": 0.822,
+      "step": 31502
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5758927539367842,
+      "learning_rate": 3.2881922785851374e-06,
+      "loss": 0.7941,
+      "step": 31503
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9020175826946952,
+      "learning_rate": 3.28762666058292e-06,
+      "loss": 0.7849,
+      "step": 31504
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.011074986650604,
+      "learning_rate": 3.2870610816622663e-06,
+      "loss": 0.7919,
+      "step": 31505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4596375679101072,
+      "learning_rate": 3.286495541826463e-06,
+      "loss": 0.8089,
+      "step": 31506
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7706099696947948,
+      "learning_rate": 3.28593004107881e-06,
+      "loss": 0.9135,
+      "step": 31507
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6157925412836973,
+      "learning_rate": 3.2853645794226007e-06,
+      "loss": 0.7483,
+      "step": 31508
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6201023067314941,
+      "learning_rate": 3.284799156861125e-06,
+      "loss": 0.6765,
+      "step": 31509
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8828239298824512,
+      "learning_rate": 3.2842337733976704e-06,
+      "loss": 0.8533,
+      "step": 31510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4621233796246598,
+      "learning_rate": 3.2836684290355336e-06,
+      "loss": 0.7755,
+      "step": 31511
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7378828017481698,
+      "learning_rate": 3.2831031237780086e-06,
+      "loss": 0.8997,
+      "step": 31512
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6159418223808761,
+      "learning_rate": 3.2825378576283827e-06,
+      "loss": 0.7808,
+      "step": 31513
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7554440338455533,
+      "learning_rate": 3.2819726305899467e-06,
+      "loss": 0.8242,
+      "step": 31514
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6059353036967774,
+      "learning_rate": 3.2814074426659925e-06,
+      "loss": 0.7984,
+      "step": 31515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.727532557916094,
+      "learning_rate": 3.280842293859814e-06,
+      "loss": 0.9015,
+      "step": 31516
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3454845248699683,
+      "learning_rate": 3.2802771841746984e-06,
+      "loss": 0.8387,
+      "step": 31517
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6745967214547648,
+      "learning_rate": 3.2797121136139345e-06,
+      "loss": 0.863,
+      "step": 31518
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9571080295227459,
+      "learning_rate": 3.279147082180817e-06,
+      "loss": 0.8723,
+      "step": 31519
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7063158094882267,
+      "learning_rate": 3.278582089878629e-06,
+      "loss": 0.7871,
+      "step": 31520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7786322736867581,
+      "learning_rate": 3.2780171367106684e-06,
+      "loss": 0.764,
+      "step": 31521
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8236669115647681,
+      "learning_rate": 3.2774522226802162e-06,
+      "loss": 0.7499,
+      "step": 31522
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7982430778925587,
+      "learning_rate": 3.2768873477905696e-06,
+      "loss": 0.6817,
+      "step": 31523
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6069191116939823,
+      "learning_rate": 3.2763225120450094e-06,
+      "loss": 0.924,
+      "step": 31524
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5180986000279535,
+      "learning_rate": 3.275757715446831e-06,
+      "loss": 0.6659,
+      "step": 31525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.935823134737256,
+      "learning_rate": 3.2751929579993203e-06,
+      "loss": 0.8432,
+      "step": 31526
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6208775451552258,
+      "learning_rate": 3.2746282397057627e-06,
+      "loss": 0.7869,
+      "step": 31527
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5320497066038576,
+      "learning_rate": 3.274063560569448e-06,
+      "loss": 0.8908,
+      "step": 31528
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8716141237452788,
+      "learning_rate": 3.2734989205936684e-06,
+      "loss": 0.8761,
+      "step": 31529
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6117738820634284,
+      "learning_rate": 3.2729343197817078e-06,
+      "loss": 0.8007,
+      "step": 31530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.820780350920104,
+      "learning_rate": 3.2723697581368496e-06,
+      "loss": 0.9147,
+      "step": 31531
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.825234190658563,
+      "learning_rate": 3.271805235662384e-06,
+      "loss": 0.8155,
+      "step": 31532
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8378213488927482,
+      "learning_rate": 3.271240752361603e-06,
+      "loss": 0.7423,
+      "step": 31533
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8287031348760132,
+      "learning_rate": 3.270676308237788e-06,
+      "loss": 0.7283,
+      "step": 31534
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6317605899249463,
+      "learning_rate": 3.270111903294222e-06,
+      "loss": 0.7068,
+      "step": 31535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2780741517167664,
+      "learning_rate": 3.2695475375341967e-06,
+      "loss": 0.785,
+      "step": 31536
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5207196777237746,
+      "learning_rate": 3.2689832109609985e-06,
+      "loss": 0.8075,
+      "step": 31537
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7199880071504619,
+      "learning_rate": 3.268418923577912e-06,
+      "loss": 0.8087,
+      "step": 31538
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6191158612736563,
+      "learning_rate": 3.2678546753882177e-06,
+      "loss": 0.9533,
+      "step": 31539
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6899609225682946,
+      "learning_rate": 3.2672904663952078e-06,
+      "loss": 0.8555,
+      "step": 31540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7427650147799196,
+      "learning_rate": 3.2667262966021617e-06,
+      "loss": 0.6923,
+      "step": 31541
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7118013041223343,
+      "learning_rate": 3.2661621660123666e-06,
+      "loss": 0.8218,
+      "step": 31542
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7489040261037209,
+      "learning_rate": 3.2655980746291114e-06,
+      "loss": 0.7489,
+      "step": 31543
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0035882806719574,
+      "learning_rate": 3.2650340224556754e-06,
+      "loss": 0.822,
+      "step": 31544
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5869035158357079,
+      "learning_rate": 3.264470009495341e-06,
+      "loss": 0.714,
+      "step": 31545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9840175164717928,
+      "learning_rate": 3.263906035751393e-06,
+      "loss": 0.7593,
+      "step": 31546
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5108711358367115,
+      "learning_rate": 3.263342101227125e-06,
+      "loss": 0.7308,
+      "step": 31547
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9232502195489916,
+      "learning_rate": 3.2627782059258053e-06,
+      "loss": 0.7584,
+      "step": 31548
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4581475659791585,
+      "learning_rate": 3.262214349850723e-06,
+      "loss": 0.7706,
+      "step": 31549
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4926195655242664,
+      "learning_rate": 3.261650533005162e-06,
+      "loss": 0.7509,
+      "step": 31550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5910996492910967,
+      "learning_rate": 3.261086755392412e-06,
+      "loss": 0.8197,
+      "step": 31551
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.22102719469024,
+      "learning_rate": 3.2605230170157407e-06,
+      "loss": 0.8945,
+      "step": 31552
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4687857410306562,
+      "learning_rate": 3.2599593178784395e-06,
+      "loss": 0.8401,
+      "step": 31553
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.493714782080714,
+      "learning_rate": 3.2593956579837914e-06,
+      "loss": 0.9034,
+      "step": 31554
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5486096440606556,
+      "learning_rate": 3.2588320373350736e-06,
+      "loss": 0.8739,
+      "step": 31555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4902653829197137,
+      "learning_rate": 3.2582684559355727e-06,
+      "loss": 0.9311,
+      "step": 31556
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4169035697723722,
+      "learning_rate": 3.257704913788564e-06,
+      "loss": 0.8915,
+      "step": 31557
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5169540155276044,
+      "learning_rate": 3.257141410897334e-06,
+      "loss": 0.8902,
+      "step": 31558
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5841324454563204,
+      "learning_rate": 3.2565779472651595e-06,
+      "loss": 0.8763,
+      "step": 31559
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5153509497245174,
+      "learning_rate": 3.2560145228953267e-06,
+      "loss": 0.7421,
+      "step": 31560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7579783240390259,
+      "learning_rate": 3.2554511377911115e-06,
+      "loss": 0.9223,
+      "step": 31561
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7763793721307518,
+      "learning_rate": 3.254887791955792e-06,
+      "loss": 0.7605,
+      "step": 31562
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5572120403238885,
+      "learning_rate": 3.2543244853926515e-06,
+      "loss": 0.8907,
+      "step": 31563
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.06316289384257,
+      "learning_rate": 3.2537612181049726e-06,
+      "loss": 0.8112,
+      "step": 31564
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.53162190217052,
+      "learning_rate": 3.253197990096031e-06,
+      "loss": 0.9343,
+      "step": 31565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6473239700488995,
+      "learning_rate": 3.2526348013691043e-06,
+      "loss": 0.8665,
+      "step": 31566
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5796008682676954,
+      "learning_rate": 3.2520716519274743e-06,
+      "loss": 0.7563,
+      "step": 31567
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5732972100421163,
+      "learning_rate": 3.251508541774422e-06,
+      "loss": 0.8768,
+      "step": 31568
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7852419609540655,
+      "learning_rate": 3.250945470913224e-06,
+      "loss": 0.7055,
+      "step": 31569
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7783853652465678,
+      "learning_rate": 3.250382439347154e-06,
+      "loss": 0.7758,
+      "step": 31570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4632405709145304,
+      "learning_rate": 3.2498194470794954e-06,
+      "loss": 0.8164,
+      "step": 31571
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4963756941964812,
+      "learning_rate": 3.2492564941135273e-06,
+      "loss": 0.8179,
+      "step": 31572
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.553917123165472,
+      "learning_rate": 3.2486935804525264e-06,
+      "loss": 0.6973,
+      "step": 31573
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.612144547463062,
+      "learning_rate": 3.2481307060997645e-06,
+      "loss": 0.8974,
+      "step": 31574
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.474602489581675,
+      "learning_rate": 3.2475678710585267e-06,
+      "loss": 0.7244,
+      "step": 31575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6879345858698316,
+      "learning_rate": 3.247005075332084e-06,
+      "loss": 0.9493,
+      "step": 31576
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8867370613557872,
+      "learning_rate": 3.246442318923718e-06,
+      "loss": 0.9508,
+      "step": 31577
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.56123963239272,
+      "learning_rate": 3.245879601836701e-06,
+      "loss": 0.8361,
+      "step": 31578
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4623961382067148,
+      "learning_rate": 3.2453169240743132e-06,
+      "loss": 0.7891,
+      "step": 31579
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.556109338769261,
+      "learning_rate": 3.244754285639826e-06,
+      "loss": 0.8732,
+      "step": 31580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.639653295786103,
+      "learning_rate": 3.244191686536522e-06,
+      "loss": 0.8679,
+      "step": 31581
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7104214672284261,
+      "learning_rate": 3.2436291267676725e-06,
+      "loss": 0.749,
+      "step": 31582
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4620767767329639,
+      "learning_rate": 3.243066606336549e-06,
+      "loss": 0.879,
+      "step": 31583
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.458904742597276,
+      "learning_rate": 3.242504125246432e-06,
+      "loss": 0.7547,
+      "step": 31584
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.643236046127952,
+      "learning_rate": 3.241941683500598e-06,
+      "loss": 0.7426,
+      "step": 31585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6551109489509843,
+      "learning_rate": 3.241379281102319e-06,
+      "loss": 0.919,
+      "step": 31586
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7284952658529174,
+      "learning_rate": 3.2408169180548675e-06,
+      "loss": 0.8099,
+      "step": 31587
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7423634677549471,
+      "learning_rate": 3.2402545943615195e-06,
+      "loss": 0.7284,
+      "step": 31588
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5520757521782143,
+      "learning_rate": 3.2396923100255515e-06,
+      "loss": 0.8495,
+      "step": 31589
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.609696487442035,
+      "learning_rate": 3.2391300650502356e-06,
+      "loss": 0.8116,
+      "step": 31590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7630601493152889,
+      "learning_rate": 3.2385678594388416e-06,
+      "loss": 0.9006,
+      "step": 31591
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.579591083894623,
+      "learning_rate": 3.238005693194647e-06,
+      "loss": 0.8032,
+      "step": 31592
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5430783366229532,
+      "learning_rate": 3.2374435663209268e-06,
+      "loss": 0.8456,
+      "step": 31593
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5664499242858394,
+      "learning_rate": 3.23688147882095e-06,
+      "loss": 0.9078,
+      "step": 31594
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5665134254064916,
+      "learning_rate": 3.2363194306979884e-06,
+      "loss": 0.9003,
+      "step": 31595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7811104407679842,
+      "learning_rate": 3.235757421955317e-06,
+      "loss": 0.6945,
+      "step": 31596
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5846135483862511,
+      "learning_rate": 3.235195452596209e-06,
+      "loss": 0.8092,
+      "step": 31597
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4349357881832145,
+      "learning_rate": 3.2346335226239357e-06,
+      "loss": 0.796,
+      "step": 31598
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8952830599544148,
+      "learning_rate": 3.234071632041765e-06,
+      "loss": 0.8746,
+      "step": 31599
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4397341372916155,
+      "learning_rate": 3.2335097808529736e-06,
+      "loss": 0.7578,
+      "step": 31600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.599878006606787,
+      "learning_rate": 3.232947969060828e-06,
+      "loss": 0.9062,
+      "step": 31601
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.770015242538518,
+      "learning_rate": 3.2323861966686055e-06,
+      "loss": 1.0486,
+      "step": 31602
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8644690048972425,
+      "learning_rate": 3.231824463679569e-06,
+      "loss": 0.8112,
+      "step": 31603
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.491992235612082,
+      "learning_rate": 3.231262770096997e-06,
+      "loss": 0.8933,
+      "step": 31604
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5623669446791044,
+      "learning_rate": 3.230701115924152e-06,
+      "loss": 0.8021,
+      "step": 31605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5032198286150735,
+      "learning_rate": 3.2301395011643124e-06,
+      "loss": 0.7347,
+      "step": 31606
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.574850370680992,
+      "learning_rate": 3.2295779258207426e-06,
+      "loss": 0.8487,
+      "step": 31607
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4238978652029033,
+      "learning_rate": 3.229016389896712e-06,
+      "loss": 0.7366,
+      "step": 31608
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7762551199616028,
+      "learning_rate": 3.2284548933954905e-06,
+      "loss": 0.6184,
+      "step": 31609
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7282584354762767,
+      "learning_rate": 3.227893436320353e-06,
+      "loss": 0.9228,
+      "step": 31610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8827641105313742,
+      "learning_rate": 3.2273320186745625e-06,
+      "loss": 0.7614,
+      "step": 31611
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4446625268158109,
+      "learning_rate": 3.2267706404613864e-06,
+      "loss": 0.7165,
+      "step": 31612
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4207312626090363,
+      "learning_rate": 3.2262093016840955e-06,
+      "loss": 0.7463,
+      "step": 31613
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8014176602350489,
+      "learning_rate": 3.2256480023459633e-06,
+      "loss": 0.6737,
+      "step": 31614
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4885415667508495,
+      "learning_rate": 3.225086742450252e-06,
+      "loss": 0.889,
+      "step": 31615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9510422091696318,
+      "learning_rate": 3.224525522000228e-06,
+      "loss": 0.7452,
+      "step": 31616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.689057390013291,
+      "learning_rate": 3.2239643409991607e-06,
+      "loss": 0.7151,
+      "step": 31617
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.562007510757025,
+      "learning_rate": 3.2234031994503223e-06,
+      "loss": 0.766,
+      "step": 31618
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4711719094276563,
+      "learning_rate": 3.2228420973569753e-06,
+      "loss": 0.7956,
+      "step": 31619
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6548053517949624,
+      "learning_rate": 3.2222810347223845e-06,
+      "loss": 0.8173,
+      "step": 31620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0401264398681636,
+      "learning_rate": 3.2217200115498216e-06,
+      "loss": 0.7986,
+      "step": 31621
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6602730463820805,
+      "learning_rate": 3.2211590278425486e-06,
+      "loss": 0.8347,
+      "step": 31622
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.521115049292104,
+      "learning_rate": 3.2205980836038363e-06,
+      "loss": 0.8555,
+      "step": 31623
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9424277833692358,
+      "learning_rate": 3.2200371788369457e-06,
+      "loss": 0.912,
+      "step": 31624
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8070444094846567,
+      "learning_rate": 3.219476313545148e-06,
+      "loss": 0.7231,
+      "step": 31625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6535799879954896,
+      "learning_rate": 3.2189154877317032e-06,
+      "loss": 0.7005,
+      "step": 31626
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5809575102402018,
+      "learning_rate": 3.2183547013998816e-06,
+      "loss": 0.806,
+      "step": 31627
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.809772943051283,
+      "learning_rate": 3.217793954552946e-06,
+      "loss": 0.6572,
+      "step": 31628
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.535558269591766,
+      "learning_rate": 3.217233247194158e-06,
+      "loss": 0.7925,
+      "step": 31629
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8040909193723622,
+      "learning_rate": 3.216672579326786e-06,
+      "loss": 0.7028,
+      "step": 31630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7623973930761169,
+      "learning_rate": 3.216111950954094e-06,
+      "loss": 0.7246,
+      "step": 31631
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5778938830145264,
+      "learning_rate": 3.2155513620793523e-06,
+      "loss": 0.8419,
+      "step": 31632
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.738119816399671,
+      "learning_rate": 3.2149908127058116e-06,
+      "loss": 0.852,
+      "step": 31633
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5211857578344514,
+      "learning_rate": 3.214430302836743e-06,
+      "loss": 0.7725,
+      "step": 31634
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.639890325871838,
+      "learning_rate": 3.2138698324754126e-06,
+      "loss": 0.9731,
+      "step": 31635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7552681233992942,
+      "learning_rate": 3.2133094016250797e-06,
+      "loss": 0.8666,
+      "step": 31636
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.893029816070089,
+      "learning_rate": 3.212749010289006e-06,
+      "loss": 0.8681,
+      "step": 31637
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3813974038473928,
+      "learning_rate": 3.2121886584704553e-06,
+      "loss": 0.8976,
+      "step": 31638
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.637011778469138,
+      "learning_rate": 3.2116283461726948e-06,
+      "loss": 0.9076,
+      "step": 31639
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6115115610743675,
+      "learning_rate": 3.21106807339898e-06,
+      "loss": 0.8281,
+      "step": 31640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3701700512203403,
+      "learning_rate": 3.210507840152579e-06,
+      "loss": 0.8469,
+      "step": 31641
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6750651676409365,
+      "learning_rate": 3.209947646436752e-06,
+      "loss": 0.7568,
+      "step": 31642
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.571855155197024,
+      "learning_rate": 3.2093874922547563e-06,
+      "loss": 0.8377,
+      "step": 31643
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.529084935714119,
+      "learning_rate": 3.2088273776098554e-06,
+      "loss": 0.7644,
+      "step": 31644
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.61210874467462,
+      "learning_rate": 3.2082673025053157e-06,
+      "loss": 0.8081,
+      "step": 31645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4570741222072332,
+      "learning_rate": 3.207707266944393e-06,
+      "loss": 0.8894,
+      "step": 31646
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9497281123121948,
+      "learning_rate": 3.2071472709303466e-06,
+      "loss": 0.8363,
+      "step": 31647
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9681100603259878,
+      "learning_rate": 3.2065873144664393e-06,
+      "loss": 0.7864,
+      "step": 31648
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.467429720105088,
+      "learning_rate": 3.2060273975559374e-06,
+      "loss": 0.7238,
+      "step": 31649
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.664062406115686,
+      "learning_rate": 3.205467520202089e-06,
+      "loss": 0.7357,
+      "step": 31650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6047062408546513,
+      "learning_rate": 3.2049076824081595e-06,
+      "loss": 0.7401,
+      "step": 31651
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5677768638455363,
+      "learning_rate": 3.204347884177409e-06,
+      "loss": 0.786,
+      "step": 31652
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.120587462224276,
+      "learning_rate": 3.203788125513099e-06,
+      "loss": 0.8607,
+      "step": 31653
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5902098378573584,
+      "learning_rate": 3.2032284064184872e-06,
+      "loss": 0.8083,
+      "step": 31654
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5284293040284278,
+      "learning_rate": 3.202668726896828e-06,
+      "loss": 0.7784,
+      "step": 31655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5550290496629053,
+      "learning_rate": 3.202109086951386e-06,
+      "loss": 0.8241,
+      "step": 31656
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5603634966158615,
+      "learning_rate": 3.2015494865854144e-06,
+      "loss": 0.9854,
+      "step": 31657
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7908883824220783,
+      "learning_rate": 3.200989925802177e-06,
+      "loss": 0.882,
+      "step": 31658
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7112936368002607,
+      "learning_rate": 3.200430404604926e-06,
+      "loss": 0.7783,
+      "step": 31659
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5967730115928032,
+      "learning_rate": 3.199870922996925e-06,
+      "loss": 0.8876,
+      "step": 31660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9734370289592547,
+      "learning_rate": 3.199311480981425e-06,
+      "loss": 0.822,
+      "step": 31661
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5168346991831363,
+      "learning_rate": 3.1987520785616898e-06,
+      "loss": 0.7903,
+      "step": 31662
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5998647660826628,
+      "learning_rate": 3.198192715740973e-06,
+      "loss": 0.9166,
+      "step": 31663
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0755160066540217,
+      "learning_rate": 3.197633392522529e-06,
+      "loss": 0.7752,
+      "step": 31664
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9252211138813657,
+      "learning_rate": 3.1970741089096168e-06,
+      "loss": 0.8076,
+      "step": 31665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7717550963539026,
+      "learning_rate": 3.196514864905497e-06,
+      "loss": 0.9708,
+      "step": 31666
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5355525890627097,
+      "learning_rate": 3.19595566051342e-06,
+      "loss": 0.8042,
+      "step": 31667
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8926959729078348,
+      "learning_rate": 3.1953964957366413e-06,
+      "loss": 0.8396,
+      "step": 31668
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7791486484312709,
+      "learning_rate": 3.1948373705784173e-06,
+      "loss": 0.7617,
+      "step": 31669
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8518342722991934,
+      "learning_rate": 3.1942782850420086e-06,
+      "loss": 0.7503,
+      "step": 31670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5614296552928648,
+      "learning_rate": 3.193719239130666e-06,
+      "loss": 0.9199,
+      "step": 31671
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6715919331975038,
+      "learning_rate": 3.193160232847642e-06,
+      "loss": 0.8322,
+      "step": 31672
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7700367364468585,
+      "learning_rate": 3.1926012661961935e-06,
+      "loss": 0.7112,
+      "step": 31673
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5517855877908222,
+      "learning_rate": 3.1920423391795795e-06,
+      "loss": 0.802,
+      "step": 31674
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.856664233378138,
+      "learning_rate": 3.1914834518010506e-06,
+      "loss": 0.7738,
+      "step": 31675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7960736098768338,
+      "learning_rate": 3.1909246040638562e-06,
+      "loss": 0.6706,
+      "step": 31676
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0290977803990318,
+      "learning_rate": 3.1903657959712585e-06,
+      "loss": 0.9007,
+      "step": 31677
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.948776320864532,
+      "learning_rate": 3.1898070275265035e-06,
+      "loss": 0.914,
+      "step": 31678
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.611078595217313,
+      "learning_rate": 3.189248298732852e-06,
+      "loss": 0.7848,
+      "step": 31679
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5782612323568161,
+      "learning_rate": 3.1886896095935494e-06,
+      "loss": 0.7902,
+      "step": 31680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5005049479777322,
+      "learning_rate": 3.188130960111856e-06,
+      "loss": 0.7675,
+      "step": 31681
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7488466737059931,
+      "learning_rate": 3.1875723502910173e-06,
+      "loss": 0.9119,
+      "step": 31682
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5200174822886938,
+      "learning_rate": 3.1870137801342917e-06,
+      "loss": 0.8094,
+      "step": 31683
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.684198420738211,
+      "learning_rate": 3.1864552496449297e-06,
+      "loss": 0.8371,
+      "step": 31684
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.760567936285682,
+      "learning_rate": 3.185896758826178e-06,
+      "loss": 0.9339,
+      "step": 31685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5548325498975815,
+      "learning_rate": 3.185338307681294e-06,
+      "loss": 0.9054,
+      "step": 31686
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4959892337287435,
+      "learning_rate": 3.18477989621353e-06,
+      "loss": 0.8511,
+      "step": 31687
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8070094948108482,
+      "learning_rate": 3.1842215244261356e-06,
+      "loss": 0.9175,
+      "step": 31688
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4762851177259444,
+      "learning_rate": 3.1836631923223583e-06,
+      "loss": 0.7901,
+      "step": 31689
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.377217503625342,
+      "learning_rate": 3.1831048999054513e-06,
+      "loss": 0.8112,
+      "step": 31690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3820700132528783,
+      "learning_rate": 3.1825466471786693e-06,
+      "loss": 0.8422,
+      "step": 31691
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.659858043057765,
+      "learning_rate": 3.18198843414526e-06,
+      "loss": 0.7148,
+      "step": 31692
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.799557179059821,
+      "learning_rate": 3.1814302608084678e-06,
+      "loss": 0.8056,
+      "step": 31693
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7977538974672712,
+      "learning_rate": 3.180872127171548e-06,
+      "loss": 0.6521,
+      "step": 31694
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4639171302507172,
+      "learning_rate": 3.1803140332377535e-06,
+      "loss": 0.9023,
+      "step": 31695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.037461052794137,
+      "learning_rate": 3.1797559790103293e-06,
+      "loss": 0.8272,
+      "step": 31696
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8358168374647263,
+      "learning_rate": 3.179197964492522e-06,
+      "loss": 0.8627,
+      "step": 31697
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4511731621148034,
+      "learning_rate": 3.1786399896875862e-06,
+      "loss": 0.8649,
+      "step": 31698
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6550547274554497,
+      "learning_rate": 3.1780820545987647e-06,
+      "loss": 0.8927,
+      "step": 31699
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4332705964537207,
+      "learning_rate": 3.1775241592293127e-06,
+      "loss": 0.9775,
+      "step": 31700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.599883044746905,
+      "learning_rate": 3.176966303582473e-06,
+      "loss": 0.7668,
+      "step": 31701
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5633053044154062,
+      "learning_rate": 3.1764084876614965e-06,
+      "loss": 0.8855,
+      "step": 31702
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8165624124774006,
+      "learning_rate": 3.175850711469629e-06,
+      "loss": 0.7293,
+      "step": 31703
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7641011303154195,
+      "learning_rate": 3.1752929750101216e-06,
+      "loss": 0.8476,
+      "step": 31704
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5502159781749885,
+      "learning_rate": 3.1747352782862185e-06,
+      "loss": 0.7621,
+      "step": 31705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.77407535812382,
+      "learning_rate": 3.1741776213011643e-06,
+      "loss": 0.8456,
+      "step": 31706
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4332733649353446,
+      "learning_rate": 3.17362000405821e-06,
+      "loss": 0.836,
+      "step": 31707
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5982495343211822,
+      "learning_rate": 3.1730624265606037e-06,
+      "loss": 0.7136,
+      "step": 31708
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.526350795295701,
+      "learning_rate": 3.1725048888115893e-06,
+      "loss": 0.8673,
+      "step": 31709
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.682207252352055,
+      "learning_rate": 3.1719473908144106e-06,
+      "loss": 0.8021,
+      "step": 31710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9379233749999523,
+      "learning_rate": 3.1713899325723153e-06,
+      "loss": 0.875,
+      "step": 31711
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8684489544533303,
+      "learning_rate": 3.1708325140885533e-06,
+      "loss": 0.6969,
+      "step": 31712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6711125388556038,
+      "learning_rate": 3.1702751353663664e-06,
+      "loss": 0.8811,
+      "step": 31713
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6479711583937628,
+      "learning_rate": 3.1697177964089963e-06,
+      "loss": 0.7863,
+      "step": 31714
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7133085945636282,
+      "learning_rate": 3.1691604972196922e-06,
+      "loss": 0.9162,
+      "step": 31715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5502299231565726,
+      "learning_rate": 3.168603237801701e-06,
+      "loss": 0.7751,
+      "step": 31716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7265758362517716,
+      "learning_rate": 3.168046018158264e-06,
+      "loss": 0.8714,
+      "step": 31717
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8965255389826727,
+      "learning_rate": 3.1674888382926237e-06,
+      "loss": 0.8078,
+      "step": 31718
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6426288538290452,
+      "learning_rate": 3.1669316982080268e-06,
+      "loss": 0.8088,
+      "step": 31719
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5173665309815794,
+      "learning_rate": 3.1663745979077186e-06,
+      "loss": 0.8771,
+      "step": 31720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7120009074607683,
+      "learning_rate": 3.165817537394943e-06,
+      "loss": 0.851,
+      "step": 31721
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.658219879068419,
+      "learning_rate": 3.1652605166729367e-06,
+      "loss": 0.8676,
+      "step": 31722
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7628708147745387,
+      "learning_rate": 3.164703535744951e-06,
+      "loss": 0.8648,
+      "step": 31723
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.586629601567323,
+      "learning_rate": 3.1641465946142223e-06,
+      "loss": 0.7334,
+      "step": 31724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7392111096734123,
+      "learning_rate": 3.1635896932839994e-06,
+      "loss": 0.6956,
+      "step": 31725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.85671404150445,
+      "learning_rate": 3.1630328317575187e-06,
+      "loss": 0.7695,
+      "step": 31726
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7504934397680245,
+      "learning_rate": 3.162476010038028e-06,
+      "loss": 0.7173,
+      "step": 31727
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6269366291341576,
+      "learning_rate": 3.1619192281287647e-06,
+      "loss": 0.8058,
+      "step": 31728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8220446662892533,
+      "learning_rate": 3.1613624860329716e-06,
+      "loss": 0.6906,
+      "step": 31729
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1568910880578303,
+      "learning_rate": 3.1608057837538976e-06,
+      "loss": 0.7166,
+      "step": 31730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6428908210690267,
+      "learning_rate": 3.1602491212947717e-06,
+      "loss": 0.8426,
+      "step": 31731
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7927012163271356,
+      "learning_rate": 3.1596924986588397e-06,
+      "loss": 0.8353,
+      "step": 31732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8965919029050036,
+      "learning_rate": 3.1591359158493453e-06,
+      "loss": 0.9467,
+      "step": 31733
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.694511127930911,
+      "learning_rate": 3.1585793728695325e-06,
+      "loss": 0.9734,
+      "step": 31734
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.756770156396967,
+      "learning_rate": 3.15802286972263e-06,
+      "loss": 0.8662,
+      "step": 31735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4881231701453863,
+      "learning_rate": 3.1574664064118856e-06,
+      "loss": 0.7622,
+      "step": 31736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.541867354504313,
+      "learning_rate": 3.15690998294054e-06,
+      "loss": 0.801,
+      "step": 31737
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5770564235686064,
+      "learning_rate": 3.1563535993118287e-06,
+      "loss": 0.6933,
+      "step": 31738
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8185365943730214,
+      "learning_rate": 3.155797255528996e-06,
+      "loss": 0.9401,
+      "step": 31739
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4050817252816385,
+      "learning_rate": 3.1552409515952764e-06,
+      "loss": 0.7559,
+      "step": 31740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5533411484622777,
+      "learning_rate": 3.1546846875139136e-06,
+      "loss": 0.8216,
+      "step": 31741
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6509036117055949,
+      "learning_rate": 3.15412846328814e-06,
+      "loss": 0.8625,
+      "step": 31742
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.771330309137844,
+      "learning_rate": 3.153572278921202e-06,
+      "loss": 0.7837,
+      "step": 31743
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6564794357732218,
+      "learning_rate": 3.1530161344163324e-06,
+      "loss": 0.7776,
+      "step": 31744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7831580683523253,
+      "learning_rate": 3.1524600297767682e-06,
+      "loss": 0.7135,
+      "step": 31745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.965128039500734,
+      "learning_rate": 3.1519039650057502e-06,
+      "loss": 0.7667,
+      "step": 31746
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6239494130985064,
+      "learning_rate": 3.1513479401065174e-06,
+      "loss": 0.7994,
+      "step": 31747
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.669562426232145,
+      "learning_rate": 3.1507919550823063e-06,
+      "loss": 0.7229,
+      "step": 31748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5975519573764179,
+      "learning_rate": 3.150236009936349e-06,
+      "loss": 0.8637,
+      "step": 31749
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.537048754574436,
+      "learning_rate": 3.149680104671886e-06,
+      "loss": 0.8426,
+      "step": 31750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6431364359463696,
+      "learning_rate": 3.149124239292158e-06,
+      "loss": 0.8552,
+      "step": 31751
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.721162684958507,
+      "learning_rate": 3.148568413800397e-06,
+      "loss": 0.8475,
+      "step": 31752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7575339608553482,
+      "learning_rate": 3.148012628199837e-06,
+      "loss": 0.7466,
+      "step": 31753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8681709122388876,
+      "learning_rate": 3.147456882493717e-06,
+      "loss": 0.8064,
+      "step": 31754
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6071264974672028,
+      "learning_rate": 3.1469011766852754e-06,
+      "loss": 0.8978,
+      "step": 31755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6621038226359592,
+      "learning_rate": 3.146345510777745e-06,
+      "loss": 0.9277,
+      "step": 31756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.814723619619706,
+      "learning_rate": 3.1457898847743573e-06,
+      "loss": 0.9314,
+      "step": 31757
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5655145556693992,
+      "learning_rate": 3.1452342986783545e-06,
+      "loss": 0.862,
+      "step": 31758
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6859683768041933,
+      "learning_rate": 3.1446787524929644e-06,
+      "loss": 0.8027,
+      "step": 31759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7964452824542185,
+      "learning_rate": 3.144123246221428e-06,
+      "loss": 0.7149,
+      "step": 31760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.613996021500693,
+      "learning_rate": 3.1435677798669738e-06,
+      "loss": 0.7622,
+      "step": 31761
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.671899113107976,
+      "learning_rate": 3.1430123534328415e-06,
+      "loss": 0.8506,
+      "step": 31762
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5570680653164226,
+      "learning_rate": 3.142456966922259e-06,
+      "loss": 0.8862,
+      "step": 31763
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7647220017743925,
+      "learning_rate": 3.1419016203384666e-06,
+      "loss": 0.659,
+      "step": 31764
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6700916203488403,
+      "learning_rate": 3.1413463136846945e-06,
+      "loss": 0.8463,
+      "step": 31765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8873145674414875,
+      "learning_rate": 3.140791046964172e-06,
+      "loss": 0.8859,
+      "step": 31766
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9433873113913767,
+      "learning_rate": 3.1402358201801355e-06,
+      "loss": 0.769,
+      "step": 31767
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.73620758031094,
+      "learning_rate": 3.1396806333358208e-06,
+      "loss": 0.8272,
+      "step": 31768
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5937952020862522,
+      "learning_rate": 3.1391254864344577e-06,
+      "loss": 0.8693,
+      "step": 31769
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8314054270039148,
+      "learning_rate": 3.1385703794792742e-06,
+      "loss": 0.7287,
+      "step": 31770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6832805261827932,
+      "learning_rate": 3.138015312473508e-06,
+      "loss": 0.7655,
+      "step": 31771
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7007454197095587,
+      "learning_rate": 3.1374602854203908e-06,
+      "loss": 0.6504,
+      "step": 31772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5952918794793314,
+      "learning_rate": 3.1369052983231527e-06,
+      "loss": 1.0013,
+      "step": 31773
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9199266907521444,
+      "learning_rate": 3.1363503511850222e-06,
+      "loss": 0.9177,
+      "step": 31774
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8128079460296206,
+      "learning_rate": 3.135795444009232e-06,
+      "loss": 0.7262,
+      "step": 31775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6505809502432616,
+      "learning_rate": 3.135240576799018e-06,
+      "loss": 0.7856,
+      "step": 31776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5887366754070906,
+      "learning_rate": 3.1346857495576066e-06,
+      "loss": 0.934,
+      "step": 31777
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7728257937698891,
+      "learning_rate": 3.134130962288224e-06,
+      "loss": 0.7333,
+      "step": 31778
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4561386193742891,
+      "learning_rate": 3.1335762149941095e-06,
+      "loss": 0.7222,
+      "step": 31779
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6643976844089194,
+      "learning_rate": 3.133021507678484e-06,
+      "loss": 0.8231,
+      "step": 31780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6537515795958524,
+      "learning_rate": 3.1324668403445845e-06,
+      "loss": 0.8636,
+      "step": 31781
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6180813579526407,
+      "learning_rate": 3.1319122129956337e-06,
+      "loss": 0.9036,
+      "step": 31782
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0801659274906568,
+      "learning_rate": 3.1313576256348687e-06,
+      "loss": 0.8873,
+      "step": 31783
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6364186325335113,
+      "learning_rate": 3.1308030782655106e-06,
+      "loss": 0.9074,
+      "step": 31784
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7430341904609394,
+      "learning_rate": 3.130248570890795e-06,
+      "loss": 0.7822,
+      "step": 31785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.929233629497407,
+      "learning_rate": 3.129694103513946e-06,
+      "loss": 0.9252,
+      "step": 31786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6366404537277273,
+      "learning_rate": 3.129139676138191e-06,
+      "loss": 0.7963,
+      "step": 31787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.88495612088209,
+      "learning_rate": 3.12858528876676e-06,
+      "loss": 0.8087,
+      "step": 31788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5473562368451979,
+      "learning_rate": 3.128030941402883e-06,
+      "loss": 0.8315,
+      "step": 31789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6779913318668531,
+      "learning_rate": 3.1274766340497876e-06,
+      "loss": 0.9157,
+      "step": 31790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6000168672899642,
+      "learning_rate": 3.126922366710694e-06,
+      "loss": 0.6989,
+      "step": 31791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6714869251888407,
+      "learning_rate": 3.1263681393888347e-06,
+      "loss": 0.7663,
+      "step": 31792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5038869753986501,
+      "learning_rate": 3.1258139520874396e-06,
+      "loss": 0.8659,
+      "step": 31793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5963928544112453,
+      "learning_rate": 3.125259804809733e-06,
+      "loss": 0.8315,
+      "step": 31794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6092538434834591,
+      "learning_rate": 3.1247056975589364e-06,
+      "loss": 0.8741,
+      "step": 31795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1711024872270777,
+      "learning_rate": 3.124151630338279e-06,
+      "loss": 0.824,
+      "step": 31796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6720056916362949,
+      "learning_rate": 3.123597603150992e-06,
+      "loss": 0.8362,
+      "step": 31797
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7410755007183352,
+      "learning_rate": 3.1230436160002963e-06,
+      "loss": 0.8508,
+      "step": 31798
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6410223695361699,
+      "learning_rate": 3.1224896688894146e-06,
+      "loss": 0.8872,
+      "step": 31799
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6636069482534588,
+      "learning_rate": 3.1219357618215785e-06,
+      "loss": 0.8212,
+      "step": 31800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6745737382981267,
+      "learning_rate": 3.1213818948000074e-06,
+      "loss": 0.8729,
+      "step": 31801
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6869432276204186,
+      "learning_rate": 3.120828067827931e-06,
+      "loss": 0.8457,
+      "step": 31802
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9444520391066928,
+      "learning_rate": 3.1202742809085694e-06,
+      "loss": 0.777,
+      "step": 31803
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7381970673171825,
+      "learning_rate": 3.119720534045152e-06,
+      "loss": 0.9775,
+      "step": 31804
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7323571467416512,
+      "learning_rate": 3.119166827240896e-06,
+      "loss": 0.7791,
+      "step": 31805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5648674242762222,
+      "learning_rate": 3.118613160499032e-06,
+      "loss": 0.8067,
+      "step": 31806
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5455564302369535,
+      "learning_rate": 3.1180595338227827e-06,
+      "loss": 0.841,
+      "step": 31807
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6802237598616252,
+      "learning_rate": 3.1175059472153648e-06,
+      "loss": 0.8666,
+      "step": 31808
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6062473440187466,
+      "learning_rate": 3.1169524006800066e-06,
+      "loss": 1.015,
+      "step": 31809
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5745548769165354,
+      "learning_rate": 3.116398894219934e-06,
+      "loss": 0.8876,
+      "step": 31810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.642276893350407,
+      "learning_rate": 3.1158454278383666e-06,
+      "loss": 0.7519,
+      "step": 31811
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5029430391522296,
+      "learning_rate": 3.1152920015385225e-06,
+      "loss": 0.7776,
+      "step": 31812
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.570542060888744,
+      "learning_rate": 3.114738615323628e-06,
+      "loss": 1.0162,
+      "step": 31813
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7106377250413816,
+      "learning_rate": 3.1141852691969098e-06,
+      "loss": 0.783,
+      "step": 31814
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6929378374988688,
+      "learning_rate": 3.113631963161584e-06,
+      "loss": 0.7186,
+      "step": 31815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5056954097176853,
+      "learning_rate": 3.1130786972208694e-06,
+      "loss": 0.7647,
+      "step": 31816
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4486883136506605,
+      "learning_rate": 3.112525471377992e-06,
+      "loss": 0.858,
+      "step": 31817
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5739477991678987,
+      "learning_rate": 3.1119722856361744e-06,
+      "loss": 0.9306,
+      "step": 31818
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9576607010319926,
+      "learning_rate": 3.111419139998635e-06,
+      "loss": 0.8711,
+      "step": 31819
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5803084503074123,
+      "learning_rate": 3.1108660344685913e-06,
+      "loss": 0.7698,
+      "step": 31820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9031165769262706,
+      "learning_rate": 3.1103129690492694e-06,
+      "loss": 0.868,
+      "step": 31821
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5456713744575652,
+      "learning_rate": 3.109759943743884e-06,
+      "loss": 0.8826,
+      "step": 31822
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7637974627881585,
+      "learning_rate": 3.1092069585556573e-06,
+      "loss": 0.7179,
+      "step": 31823
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4779754157517306,
+      "learning_rate": 3.1086540134878117e-06,
+      "loss": 0.779,
+      "step": 31824
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7751357755415382,
+      "learning_rate": 3.1081011085435653e-06,
+      "loss": 0.8972,
+      "step": 31825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6610314480120434,
+      "learning_rate": 3.1075482437261326e-06,
+      "loss": 0.9208,
+      "step": 31826
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7604753396347184,
+      "learning_rate": 3.106995419038736e-06,
+      "loss": 0.6948,
+      "step": 31827
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6555136119897407,
+      "learning_rate": 3.106442634484601e-06,
+      "loss": 0.7674,
+      "step": 31828
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8855760530698316,
+      "learning_rate": 3.105889890066933e-06,
+      "loss": 0.9015,
+      "step": 31829
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6808960232850927,
+      "learning_rate": 3.105337185788957e-06,
+      "loss": 0.7977,
+      "step": 31830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.824868223145559,
+      "learning_rate": 3.1047845216538895e-06,
+      "loss": 0.8409,
+      "step": 31831
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7009614699785132,
+      "learning_rate": 3.1042318976649575e-06,
+      "loss": 0.7394,
+      "step": 31832
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.718269005813117,
+      "learning_rate": 3.1036793138253627e-06,
+      "loss": 0.7957,
+      "step": 31833
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.641241703202032,
+      "learning_rate": 3.103126770138332e-06,
+      "loss": 0.5673,
+      "step": 31834
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5515623093093933,
+      "learning_rate": 3.1025742666070823e-06,
+      "loss": 0.9336,
+      "step": 31835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5359766590369355,
+      "learning_rate": 3.102021803234827e-06,
+      "loss": 0.8614,
+      "step": 31836
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5419841676911772,
+      "learning_rate": 3.1014693800247865e-06,
+      "loss": 0.8255,
+      "step": 31837
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2681024516319015,
+      "learning_rate": 3.1009169969801735e-06,
+      "loss": 0.9113,
+      "step": 31838
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4897824482520465,
+      "learning_rate": 3.100364654104208e-06,
+      "loss": 0.8086,
+      "step": 31839
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8926862795507957,
+      "learning_rate": 3.099812351400101e-06,
+      "loss": 0.8867,
+      "step": 31840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4888186083074024,
+      "learning_rate": 3.0992600888710745e-06,
+      "loss": 0.8975,
+      "step": 31841
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4713725824921404,
+      "learning_rate": 3.098707866520341e-06,
+      "loss": 0.6353,
+      "step": 31842
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8405464921752583,
+      "learning_rate": 3.098155684351111e-06,
+      "loss": 0.8218,
+      "step": 31843
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.527212338494849,
+      "learning_rate": 3.097603542366604e-06,
+      "loss": 0.8696,
+      "step": 31844
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8600626871554198,
+      "learning_rate": 3.097051440570038e-06,
+      "loss": 0.7939,
+      "step": 31845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.658366982411253,
+      "learning_rate": 3.0964993789646236e-06,
+      "loss": 0.7705,
+      "step": 31846
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6412336390973312,
+      "learning_rate": 3.0959473575535725e-06,
+      "loss": 0.8575,
+      "step": 31847
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6343014670159606,
+      "learning_rate": 3.0953953763401013e-06,
+      "loss": 0.8621,
+      "step": 31848
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4713437163836711,
+      "learning_rate": 3.094843435327427e-06,
+      "loss": 0.8085,
+      "step": 31849
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.593302589804747,
+      "learning_rate": 3.0942915345187617e-06,
+      "loss": 0.7848,
+      "step": 31850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8214640002273663,
+      "learning_rate": 3.093739673917313e-06,
+      "loss": 0.7144,
+      "step": 31851
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6356808661062026,
+      "learning_rate": 3.0931878535262994e-06,
+      "loss": 0.7462,
+      "step": 31852
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6774965537046165,
+      "learning_rate": 3.0926360733489345e-06,
+      "loss": 0.9813,
+      "step": 31853
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7854070822878626,
+      "learning_rate": 3.092084333388431e-06,
+      "loss": 0.7152,
+      "step": 31854
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8801193408820631,
+      "learning_rate": 3.0915326336479947e-06,
+      "loss": 0.7783,
+      "step": 31855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.527260038965716,
+      "learning_rate": 3.0909809741308437e-06,
+      "loss": 0.8477,
+      "step": 31856
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.616369144073778,
+      "learning_rate": 3.090429354840192e-06,
+      "loss": 0.861,
+      "step": 31857
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9799569228497347,
+      "learning_rate": 3.0898777757792477e-06,
+      "loss": 0.8712,
+      "step": 31858
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6534450358485344,
+      "learning_rate": 3.089326236951219e-06,
+      "loss": 0.7954,
+      "step": 31859
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.451182873298829,
+      "learning_rate": 3.0887747383593245e-06,
+      "loss": 0.8464,
+      "step": 31860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1218220832022903,
+      "learning_rate": 3.088223280006769e-06,
+      "loss": 0.9178,
+      "step": 31861
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8784513350315442,
+      "learning_rate": 3.087671861896768e-06,
+      "loss": 0.7943,
+      "step": 31862
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4867550839421861,
+      "learning_rate": 3.0871204840325276e-06,
+      "loss": 0.859,
+      "step": 31863
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9171637162396453,
+      "learning_rate": 3.086569146417263e-06,
+      "loss": 0.9124,
+      "step": 31864
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0140002249598212,
+      "learning_rate": 3.086017849054178e-06,
+      "loss": 0.8529,
+      "step": 31865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.621786096457605,
+      "learning_rate": 3.08546659194649e-06,
+      "loss": 0.8646,
+      "step": 31866
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6651514665111795,
+      "learning_rate": 3.0849153750974036e-06,
+      "loss": 0.9509,
+      "step": 31867
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4486013973485115,
+      "learning_rate": 3.084364198510127e-06,
+      "loss": 0.9145,
+      "step": 31868
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.508956505610423,
+      "learning_rate": 3.0838130621878704e-06,
+      "loss": 0.7302,
+      "step": 31869
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7348306534143143,
+      "learning_rate": 3.0832619661338482e-06,
+      "loss": 0.9577,
+      "step": 31870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7575892053660982,
+      "learning_rate": 3.0827109103512643e-06,
+      "loss": 0.7375,
+      "step": 31871
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8196839525482886,
+      "learning_rate": 3.082159894843324e-06,
+      "loss": 0.7506,
+      "step": 31872
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4907529041972793,
+      "learning_rate": 3.081608919613238e-06,
+      "loss": 0.7795,
+      "step": 31873
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7601101495184757,
+      "learning_rate": 3.0810579846642196e-06,
+      "loss": 0.6988,
+      "step": 31874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6367993125144717,
+      "learning_rate": 3.0805070899994715e-06,
+      "loss": 0.8615,
+      "step": 31875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7616471902960928,
+      "learning_rate": 3.079956235622199e-06,
+      "loss": 0.8426,
+      "step": 31876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7784512276141187,
+      "learning_rate": 3.079405421535613e-06,
+      "loss": 0.8239,
+      "step": 31877
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5833391792200457,
+      "learning_rate": 3.0788546477429214e-06,
+      "loss": 0.9832,
+      "step": 31878
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5829656882596992,
+      "learning_rate": 3.0783039142473305e-06,
+      "loss": 0.8658,
+      "step": 31879
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5585430280868187,
+      "learning_rate": 3.0777532210520413e-06,
+      "loss": 0.7919,
+      "step": 31880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7363533707706315,
+      "learning_rate": 3.077202568160268e-06,
+      "loss": 0.6625,
+      "step": 31881
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5927468639489475,
+      "learning_rate": 3.0766519555752095e-06,
+      "loss": 0.829,
+      "step": 31882
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5667093333361461,
+      "learning_rate": 3.0761013833000785e-06,
+      "loss": 0.8419,
+      "step": 31883
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6743619940568069,
+      "learning_rate": 3.075550851338075e-06,
+      "loss": 0.8345,
+      "step": 31884
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3846688357102503,
+      "learning_rate": 3.0750003596924095e-06,
+      "loss": 0.7679,
+      "step": 31885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3658499360059766,
+      "learning_rate": 3.0744499083662803e-06,
+      "loss": 0.8279,
+      "step": 31886
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7646950118789422,
+      "learning_rate": 3.0738994973629e-06,
+      "loss": 0.8829,
+      "step": 31887
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5375158533711795,
+      "learning_rate": 3.07334912668547e-06,
+      "loss": 0.7669,
+      "step": 31888
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4382417572008614,
+      "learning_rate": 3.0727987963371906e-06,
+      "loss": 0.7741,
+      "step": 31889
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4443749782161914,
+      "learning_rate": 3.0722485063212703e-06,
+      "loss": 0.7332,
+      "step": 31890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6888945058406717,
+      "learning_rate": 3.071698256640915e-06,
+      "loss": 0.8249,
+      "step": 31891
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7812789547142234,
+      "learning_rate": 3.0711480472993262e-06,
+      "loss": 0.8813,
+      "step": 31892
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5121003952810783,
+      "learning_rate": 3.070597878299705e-06,
+      "loss": 0.7322,
+      "step": 31893
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7073604672870435,
+      "learning_rate": 3.070047749645255e-06,
+      "loss": 0.8191,
+      "step": 31894
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.303485275845378,
+      "learning_rate": 3.0694976613391845e-06,
+      "loss": 0.8439,
+      "step": 31895
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6708446817805604,
+      "learning_rate": 3.068947613384694e-06,
+      "loss": 0.9343,
+      "step": 31896
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.729178734212507,
+      "learning_rate": 3.068397605784981e-06,
+      "loss": 0.8221,
+      "step": 31897
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.98439031258536,
+      "learning_rate": 3.067847638543251e-06,
+      "loss": 0.72,
+      "step": 31898
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5927183157165472,
+      "learning_rate": 3.0672977116627102e-06,
+      "loss": 0.8639,
+      "step": 31899
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6983739034300251,
+      "learning_rate": 3.066747825146558e-06,
+      "loss": 0.6815,
+      "step": 31900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8028356426322747,
+      "learning_rate": 3.06619797899799e-06,
+      "loss": 0.7401,
+      "step": 31901
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5800036729923845,
+      "learning_rate": 3.065648173220217e-06,
+      "loss": 0.7517,
+      "step": 31902
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6085042287187037,
+      "learning_rate": 3.0650984078164325e-06,
+      "loss": 0.792,
+      "step": 31903
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.663414269171894,
+      "learning_rate": 3.064548682789843e-06,
+      "loss": 0.9086,
+      "step": 31904
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4850695741972395,
+      "learning_rate": 3.0639989981436447e-06,
+      "loss": 0.7711,
+      "step": 31905
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8033754654644496,
+      "learning_rate": 3.063449353881043e-06,
+      "loss": 0.8117,
+      "step": 31906
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.557139320299834,
+      "learning_rate": 3.0628997500052314e-06,
+      "loss": 1.0162,
+      "step": 31907
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7174734040081014,
+      "learning_rate": 3.062350186519417e-06,
+      "loss": 0.9225,
+      "step": 31908
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5492179718473404,
+      "learning_rate": 3.061800663426797e-06,
+      "loss": 0.8564,
+      "step": 31909
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4155435766423317,
+      "learning_rate": 3.0612511807305665e-06,
+      "loss": 0.7382,
+      "step": 31910
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7619783375341636,
+      "learning_rate": 3.060701738433929e-06,
+      "loss": 0.8603,
+      "step": 31911
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3743347873100962,
+      "learning_rate": 3.060152336540082e-06,
+      "loss": 0.7658,
+      "step": 31912
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5727539517168148,
+      "learning_rate": 3.0596029750522316e-06,
+      "loss": 0.8264,
+      "step": 31913
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5780482639039943,
+      "learning_rate": 3.059053653973564e-06,
+      "loss": 0.7237,
+      "step": 31914
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8660700945503523,
+      "learning_rate": 3.0585043733072838e-06,
+      "loss": 0.9637,
+      "step": 31915
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7082746706878297,
+      "learning_rate": 3.0579551330565915e-06,
+      "loss": 0.6832,
+      "step": 31916
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.665397897037275,
+      "learning_rate": 3.057405933224682e-06,
+      "loss": 0.7193,
+      "step": 31917
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5666098531460368,
+      "learning_rate": 3.0568567738147505e-06,
+      "loss": 0.8631,
+      "step": 31918
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4656952813621285,
+      "learning_rate": 3.0563076548299965e-06,
+      "loss": 0.8393,
+      "step": 31919
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.514146789719669,
+      "learning_rate": 3.055758576273621e-06,
+      "loss": 0.8707,
+      "step": 31920
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7650222617220688,
+      "learning_rate": 3.0552095381488144e-06,
+      "loss": 0.8456,
+      "step": 31921
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5552353000367747,
+      "learning_rate": 3.0546605404587803e-06,
+      "loss": 0.782,
+      "step": 31922
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.475831165758999,
+      "learning_rate": 3.0541115832067115e-06,
+      "loss": 0.7533,
+      "step": 31923
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6547081010218951,
+      "learning_rate": 3.0535626663957996e-06,
+      "loss": 0.8075,
+      "step": 31924
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8077254702142533,
+      "learning_rate": 3.0530137900292456e-06,
+      "loss": 0.8948,
+      "step": 31925
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5152882417202567,
+      "learning_rate": 3.052464954110249e-06,
+      "loss": 0.8537,
+      "step": 31926
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7802921330486889,
+      "learning_rate": 3.051916158642e-06,
+      "loss": 0.711,
+      "step": 31927
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.555207461048672,
+      "learning_rate": 3.051367403627692e-06,
+      "loss": 0.8103,
+      "step": 31928
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7288408980081038,
+      "learning_rate": 3.0508186890705214e-06,
+      "loss": 0.8296,
+      "step": 31929
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4140184920698524,
+      "learning_rate": 3.050270014973693e-06,
+      "loss": 0.7273,
+      "step": 31930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4245628205528844,
+      "learning_rate": 3.049721381340386e-06,
+      "loss": 0.805,
+      "step": 31931
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7674202198835388,
+      "learning_rate": 3.049172788173801e-06,
+      "loss": 0.6902,
+      "step": 31932
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7627845871340574,
+      "learning_rate": 3.0486242354771322e-06,
+      "loss": 0.7047,
+      "step": 31933
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7397396568137886,
+      "learning_rate": 3.0480757232535773e-06,
+      "loss": 0.7103,
+      "step": 31934
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.001156013956797,
+      "learning_rate": 3.047527251506327e-06,
+      "loss": 0.9031,
+      "step": 31935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.61281505326558,
+      "learning_rate": 3.0469788202385707e-06,
+      "loss": 0.7523,
+      "step": 31936
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.754617528463205,
+      "learning_rate": 3.0464304294535086e-06,
+      "loss": 0.83,
+      "step": 31937
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5210993550526548,
+      "learning_rate": 3.0458820791543263e-06,
+      "loss": 0.7519,
+      "step": 31938
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5648894461867024,
+      "learning_rate": 3.045333769344223e-06,
+      "loss": 0.8876,
+      "step": 31939
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3230989739698304,
+      "learning_rate": 3.044785500026386e-06,
+      "loss": 0.6633,
+      "step": 31940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8989108614851646,
+      "learning_rate": 3.0442372712040123e-06,
+      "loss": 0.8865,
+      "step": 31941
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2239117787292866,
+      "learning_rate": 3.043689082880289e-06,
+      "loss": 0.8843,
+      "step": 31942
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.660627860740885,
+      "learning_rate": 3.043140935058412e-06,
+      "loss": 0.7221,
+      "step": 31943
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5436201951396973,
+      "learning_rate": 3.0425928277415716e-06,
+      "loss": 0.8404,
+      "step": 31944
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.39450632683387,
+      "learning_rate": 3.0420447609329564e-06,
+      "loss": 0.8821,
+      "step": 31945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6367416668652606,
+      "learning_rate": 3.041496734635758e-06,
+      "loss": 0.8726,
+      "step": 31946
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.303766223846228,
+      "learning_rate": 3.040948748853172e-06,
+      "loss": 0.816,
+      "step": 31947
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5861321470988836,
+      "learning_rate": 3.0404008035883857e-06,
+      "loss": 0.8843,
+      "step": 31948
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5600786047757973,
+      "learning_rate": 3.0398528988445875e-06,
+      "loss": 0.8209,
+      "step": 31949
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5624699735922905,
+      "learning_rate": 3.0393050346249677e-06,
+      "loss": 0.8802,
+      "step": 31950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4724557729299197,
+      "learning_rate": 3.0387572109327214e-06,
+      "loss": 0.8117,
+      "step": 31951
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6560269798645573,
+      "learning_rate": 3.0382094277710337e-06,
+      "loss": 0.8852,
+      "step": 31952
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.725936707734405,
+      "learning_rate": 3.037661685143093e-06,
+      "loss": 0.8138,
+      "step": 31953
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6228346008555432,
+      "learning_rate": 3.037113983052089e-06,
+      "loss": 0.9153,
+      "step": 31954
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.690875009210788,
+      "learning_rate": 3.036566321501215e-06,
+      "loss": 0.8114,
+      "step": 31955
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.888019186084607,
+      "learning_rate": 3.036018700493658e-06,
+      "loss": 0.7066,
+      "step": 31956
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4821883858141427,
+      "learning_rate": 3.0354711200326004e-06,
+      "loss": 0.8462,
+      "step": 31957
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9132319628198466,
+      "learning_rate": 3.0349235801212383e-06,
+      "loss": 0.926,
+      "step": 31958
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6714574398103497,
+      "learning_rate": 3.0343760807627533e-06,
+      "loss": 0.7828,
+      "step": 31959
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.472638682967675,
+      "learning_rate": 3.0338286219603385e-06,
+      "loss": 0.8464,
+      "step": 31960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.838769787828842,
+      "learning_rate": 3.033281203717177e-06,
+      "loss": 0.8989,
+      "step": 31961
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6180838299432263,
+      "learning_rate": 3.0327338260364604e-06,
+      "loss": 0.8217,
+      "step": 31962
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.580313515016446,
+      "learning_rate": 3.032186488921369e-06,
+      "loss": 0.8393,
+      "step": 31963
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4574110241137492,
+      "learning_rate": 3.0316391923750986e-06,
+      "loss": 0.8949,
+      "step": 31964
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7958677846017934,
+      "learning_rate": 3.031091936400831e-06,
+      "loss": 0.8314,
+      "step": 31965
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7513084773526881,
+      "learning_rate": 3.030544721001748e-06,
+      "loss": 0.8984,
+      "step": 31966
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6327131694743133,
+      "learning_rate": 3.0299975461810406e-06,
+      "loss": 0.7309,
+      "step": 31967
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5579762244371342,
+      "learning_rate": 3.029450411941898e-06,
+      "loss": 0.8203,
+      "step": 31968
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6027381358053443,
+      "learning_rate": 3.0289033182875015e-06,
+      "loss": 0.8779,
+      "step": 31969
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6044925685089515,
+      "learning_rate": 3.028356265221033e-06,
+      "loss": 0.8477,
+      "step": 31970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7748945318677172,
+      "learning_rate": 3.0278092527456826e-06,
+      "loss": 0.7323,
+      "step": 31971
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.836446044118349,
+      "learning_rate": 3.0272622808646366e-06,
+      "loss": 0.8278,
+      "step": 31972
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.500345272961404,
+      "learning_rate": 3.0267153495810776e-06,
+      "loss": 0.8627,
+      "step": 31973
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9828836918019241,
+      "learning_rate": 3.0261684588981866e-06,
+      "loss": 0.7829,
+      "step": 31974
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.408583016146898,
+      "learning_rate": 3.0256216088191502e-06,
+      "loss": 0.7854,
+      "step": 31975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6977474645133657,
+      "learning_rate": 3.0250747993471563e-06,
+      "loss": 0.8914,
+      "step": 31976
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7463148611649582,
+      "learning_rate": 3.024528030485384e-06,
+      "loss": 1.0102,
+      "step": 31977
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7668146949132206,
+      "learning_rate": 3.023981302237016e-06,
+      "loss": 0.7075,
+      "step": 31978
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5062231665637515,
+      "learning_rate": 3.0234346146052374e-06,
+      "loss": 0.8429,
+      "step": 31979
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4901338733756802,
+      "learning_rate": 3.0228879675932343e-06,
+      "loss": 0.7146,
+      "step": 31980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.499466519988945,
+      "learning_rate": 3.0223413612041865e-06,
+      "loss": 0.8048,
+      "step": 31981
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.931056250428164,
+      "learning_rate": 3.021794795441272e-06,
+      "loss": 0.8866,
+      "step": 31982
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6339597308270586,
+      "learning_rate": 3.0212482703076817e-06,
+      "loss": 0.9124,
+      "step": 31983
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7929215849319613,
+      "learning_rate": 3.0207017858065903e-06,
+      "loss": 0.7443,
+      "step": 31984
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6063439908293982,
+      "learning_rate": 3.0201553419411857e-06,
+      "loss": 0.8551,
+      "step": 31985
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6620393088060623,
+      "learning_rate": 3.0196089387146433e-06,
+      "loss": 0.8897,
+      "step": 31986
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7943957259600598,
+      "learning_rate": 3.0190625761301506e-06,
+      "loss": 0.9313,
+      "step": 31987
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4236447734771223,
+      "learning_rate": 3.018516254190883e-06,
+      "loss": 0.8649,
+      "step": 31988
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.493124287876549,
+      "learning_rate": 3.017969972900027e-06,
+      "loss": 0.7242,
+      "step": 31989
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6838864535457785,
+      "learning_rate": 3.0174237322607593e-06,
+      "loss": 0.7894,
+      "step": 31990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.661278450818535,
+      "learning_rate": 3.016877532276259e-06,
+      "loss": 0.7411,
+      "step": 31991
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6214697646267056,
+      "learning_rate": 3.016331372949709e-06,
+      "loss": 0.7119,
+      "step": 31992
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.465530029547846,
+      "learning_rate": 3.0157852542842914e-06,
+      "loss": 0.8585,
+      "step": 31993
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.363246866567292,
+      "learning_rate": 3.015239176283183e-06,
+      "loss": 0.7521,
+      "step": 31994
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.727976364605568,
+      "learning_rate": 3.0146931389495604e-06,
+      "loss": 0.8289,
+      "step": 31995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6334840198832223,
+      "learning_rate": 3.0141471422866065e-06,
+      "loss": 0.8435,
+      "step": 31996
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.785951725044856,
+      "learning_rate": 3.0136011862975022e-06,
+      "loss": 0.9127,
+      "step": 31997
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3023555599948318,
+      "learning_rate": 3.0130552709854235e-06,
+      "loss": 0.9003,
+      "step": 31998
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7016983699787146,
+      "learning_rate": 3.0125093963535458e-06,
+      "loss": 0.6296,
+      "step": 31999
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.620182029858325,
+      "learning_rate": 3.0119635624050515e-06,
+      "loss": 0.8423,
+      "step": 32000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6882553336415433,
+      "learning_rate": 3.0114177691431203e-06,
+      "loss": 0.7326,
+      "step": 32001
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5555904968052798,
+      "learning_rate": 3.0108720165709273e-06,
+      "loss": 0.8843,
+      "step": 32002
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.570050637686138,
+      "learning_rate": 3.0103263046916464e-06,
+      "loss": 0.8002,
+      "step": 32003
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.734061703198529,
+      "learning_rate": 3.009780633508463e-06,
+      "loss": 0.7375,
+      "step": 32004
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6372745303144605,
+      "learning_rate": 3.0092350030245465e-06,
+      "loss": 0.7939,
+      "step": 32005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7460663024533964,
+      "learning_rate": 3.00868941324308e-06,
+      "loss": 0.7317,
+      "step": 32006
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5637941410604206,
+      "learning_rate": 3.0081438641672334e-06,
+      "loss": 0.7998,
+      "step": 32007
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6077726671995602,
+      "learning_rate": 3.0075983558001898e-06,
+      "loss": 0.7953,
+      "step": 32008
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.631700120364155,
+      "learning_rate": 3.00705288814512e-06,
+      "loss": 0.6749,
+      "step": 32009
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7505768438536753,
+      "learning_rate": 3.0065074612052024e-06,
+      "loss": 0.7173,
+      "step": 32010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7550626241195678,
+      "learning_rate": 3.0059620749836182e-06,
+      "loss": 0.6729,
+      "step": 32011
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7406074442026398,
+      "learning_rate": 3.0054167294835314e-06,
+      "loss": 0.7223,
+      "step": 32012
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5996792361222552,
+      "learning_rate": 3.004871424708121e-06,
+      "loss": 0.7985,
+      "step": 32013
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4497734629088517,
+      "learning_rate": 3.0043261606605656e-06,
+      "loss": 0.8314,
+      "step": 32014
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7436685801049813,
+      "learning_rate": 3.003780937344044e-06,
+      "loss": 0.6636,
+      "step": 32015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9350086906480368,
+      "learning_rate": 3.0032357547617187e-06,
+      "loss": 0.8656,
+      "step": 32016
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5680538322781201,
+      "learning_rate": 3.00269061291677e-06,
+      "loss": 0.8548,
+      "step": 32017
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9747418805523609,
+      "learning_rate": 3.0021455118123745e-06,
+      "loss": 0.8767,
+      "step": 32018
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.663739068562192,
+      "learning_rate": 3.0016004514517005e-06,
+      "loss": 0.8368,
+      "step": 32019
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4933190342724276,
+      "learning_rate": 3.001055431837927e-06,
+      "loss": 0.8116,
+      "step": 32020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.978982648042938,
+      "learning_rate": 3.0005104529742214e-06,
+      "loss": 0.8973,
+      "step": 32021
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5922873841753828,
+      "learning_rate": 2.999965514863764e-06,
+      "loss": 0.8365,
+      "step": 32022
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6970101422531918,
+      "learning_rate": 2.9994206175097196e-06,
+      "loss": 0.8681,
+      "step": 32023
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0745460464798504,
+      "learning_rate": 2.9988757609152686e-06,
+      "loss": 0.8667,
+      "step": 32024
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7732619362332869,
+      "learning_rate": 2.9983309450835785e-06,
+      "loss": 0.7588,
+      "step": 32025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.488800888287776,
+      "learning_rate": 2.9977861700178202e-06,
+      "loss": 0.7948,
+      "step": 32026
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.462750315793651,
+      "learning_rate": 2.9972414357211667e-06,
+      "loss": 0.9518,
+      "step": 32027
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6407705821076901,
+      "learning_rate": 2.996696742196794e-06,
+      "loss": 0.8118,
+      "step": 32028
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.40991044649216,
+      "learning_rate": 2.9961520894478703e-06,
+      "loss": 0.843,
+      "step": 32029
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3736011839383717,
+      "learning_rate": 2.9956074774775636e-06,
+      "loss": 0.7373,
+      "step": 32030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6841611077000351,
+      "learning_rate": 2.995062906289047e-06,
+      "loss": 0.6156,
+      "step": 32031
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4024283270630744,
+      "learning_rate": 2.9945183758854956e-06,
+      "loss": 0.8465,
+      "step": 32032
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.464289014184624,
+      "learning_rate": 2.9939738862700764e-06,
+      "loss": 0.8692,
+      "step": 32033
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5426893447361032,
+      "learning_rate": 2.9934294374459548e-06,
+      "loss": 0.6329,
+      "step": 32034
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7683652220865892,
+      "learning_rate": 2.992885029416307e-06,
+      "loss": 0.6916,
+      "step": 32035
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.617635741165083,
+      "learning_rate": 2.9923406621843034e-06,
+      "loss": 0.9436,
+      "step": 32036
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7143818539664488,
+      "learning_rate": 2.9917963357531112e-06,
+      "loss": 0.7304,
+      "step": 32037
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4577126816902335,
+      "learning_rate": 2.9912520501258958e-06,
+      "loss": 0.7165,
+      "step": 32038
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.813708591009949,
+      "learning_rate": 2.9907078053058325e-06,
+      "loss": 0.8249,
+      "step": 32039
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5856064420963774,
+      "learning_rate": 2.9901636012960856e-06,
+      "loss": 0.8141,
+      "step": 32040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.747350011424818,
+      "learning_rate": 2.9896194380998276e-06,
+      "loss": 0.6942,
+      "step": 32041
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5846494911229128,
+      "learning_rate": 2.989075315720222e-06,
+      "loss": 0.7776,
+      "step": 32042
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7645801638019005,
+      "learning_rate": 2.988531234160443e-06,
+      "loss": 0.7398,
+      "step": 32043
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5608831753903838,
+      "learning_rate": 2.9879871934236525e-06,
+      "loss": 0.8842,
+      "step": 32044
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.771656129943726,
+      "learning_rate": 2.9874431935130223e-06,
+      "loss": 0.9022,
+      "step": 32045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.635130192049915,
+      "learning_rate": 2.9868992344317184e-06,
+      "loss": 0.9679,
+      "step": 32046
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8436667639063196,
+      "learning_rate": 2.986355316182905e-06,
+      "loss": 0.717,
+      "step": 32047
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6285168146009057,
+      "learning_rate": 2.98581143876975e-06,
+      "loss": 0.8147,
+      "step": 32048
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6044594363361653,
+      "learning_rate": 2.985267602195426e-06,
+      "loss": 0.8283,
+      "step": 32049
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5017369528503868,
+      "learning_rate": 2.9847238064630947e-06,
+      "loss": 0.8871,
+      "step": 32050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5461900382232239,
+      "learning_rate": 2.984180051575918e-06,
+      "loss": 0.9029,
+      "step": 32051
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5951835003387087,
+      "learning_rate": 2.983636337537067e-06,
+      "loss": 0.8077,
+      "step": 32052
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8569229097081543,
+      "learning_rate": 2.9830926643497106e-06,
+      "loss": 0.847,
+      "step": 32053
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9023173657506491,
+      "learning_rate": 2.982549032017009e-06,
+      "loss": 0.7769,
+      "step": 32054
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.907779213300474,
+      "learning_rate": 2.9820054405421263e-06,
+      "loss": 0.7393,
+      "step": 32055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.495436197236124,
+      "learning_rate": 2.981461889928229e-06,
+      "loss": 0.902,
+      "step": 32056
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.703582894688267,
+      "learning_rate": 2.980918380178486e-06,
+      "loss": 0.8515,
+      "step": 32057
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6277248228849395,
+      "learning_rate": 2.980374911296058e-06,
+      "loss": 0.8235,
+      "step": 32058
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7775499727250399,
+      "learning_rate": 2.979831483284107e-06,
+      "loss": 0.7215,
+      "step": 32059
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5676495156552943,
+      "learning_rate": 2.9792880961458028e-06,
+      "loss": 0.8178,
+      "step": 32060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.456829192027114,
+      "learning_rate": 2.9787447498843024e-06,
+      "loss": 0.7334,
+      "step": 32061
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5619543427756934,
+      "learning_rate": 2.9782014445027774e-06,
+      "loss": 0.8142,
+      "step": 32062
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5371583891931155,
+      "learning_rate": 2.9776581800043823e-06,
+      "loss": 0.7827,
+      "step": 32063
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7620324899995934,
+      "learning_rate": 2.9771149563922884e-06,
+      "loss": 0.9475,
+      "step": 32064
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.593655926702363,
+      "learning_rate": 2.9765717736696507e-06,
+      "loss": 0.7276,
+      "step": 32065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6287674121083675,
+      "learning_rate": 2.9760286318396394e-06,
+      "loss": 0.8685,
+      "step": 32066
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.733135695239416,
+      "learning_rate": 2.975485530905413e-06,
+      "loss": 0.8762,
+      "step": 32067
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0423312801711684,
+      "learning_rate": 2.9749424708701313e-06,
+      "loss": 0.8287,
+      "step": 32068
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7493638857486442,
+      "learning_rate": 2.9743994517369577e-06,
+      "loss": 0.8853,
+      "step": 32069
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.747490593826454,
+      "learning_rate": 2.973856473509058e-06,
+      "loss": 0.7615,
+      "step": 32070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7860424157482293,
+      "learning_rate": 2.9733135361895905e-06,
+      "loss": 0.7242,
+      "step": 32071
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6011318016727567,
+      "learning_rate": 2.9727706397817136e-06,
+      "loss": 0.8742,
+      "step": 32072
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0452359446735446,
+      "learning_rate": 2.9722277842885893e-06,
+      "loss": 0.8651,
+      "step": 32073
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.714073121945386,
+      "learning_rate": 2.971684969713384e-06,
+      "loss": 0.8689,
+      "step": 32074
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3482965382282273,
+      "learning_rate": 2.9711421960592546e-06,
+      "loss": 0.891,
+      "step": 32075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6772468342876603,
+      "learning_rate": 2.9705994633293567e-06,
+      "loss": 0.8563,
+      "step": 32076
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7143461851745847,
+      "learning_rate": 2.9700567715268546e-06,
+      "loss": 0.7419,
+      "step": 32077
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.868053679537811,
+      "learning_rate": 2.9695141206549107e-06,
+      "loss": 0.8158,
+      "step": 32078
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6705876778489108,
+      "learning_rate": 2.9689715107166806e-06,
+      "loss": 0.7493,
+      "step": 32079
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.446029075566809,
+      "learning_rate": 2.968428941715321e-06,
+      "loss": 0.7154,
+      "step": 32080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4882985927972732,
+      "learning_rate": 2.9678864136539985e-06,
+      "loss": 0.8701,
+      "step": 32081
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7425167842767102,
+      "learning_rate": 2.967343926535864e-06,
+      "loss": 0.9041,
+      "step": 32082
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0176745326758287,
+      "learning_rate": 2.966801480364083e-06,
+      "loss": 0.8607,
+      "step": 32083
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0880689795562324,
+      "learning_rate": 2.9662590751418075e-06,
+      "loss": 0.9072,
+      "step": 32084
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4561686928505202,
+      "learning_rate": 2.9657167108722018e-06,
+      "loss": 0.7561,
+      "step": 32085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.638696839245576,
+      "learning_rate": 2.965174387558416e-06,
+      "loss": 0.7764,
+      "step": 32086
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6484694190728841,
+      "learning_rate": 2.9646321052036164e-06,
+      "loss": 0.9331,
+      "step": 32087
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6477507322128846,
+      "learning_rate": 2.964089863810956e-06,
+      "loss": 0.8686,
+      "step": 32088
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7519324007793432,
+      "learning_rate": 2.9635476633835893e-06,
+      "loss": 0.7115,
+      "step": 32089
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.516948050527072,
+      "learning_rate": 2.963005503924674e-06,
+      "loss": 0.8619,
+      "step": 32090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4974057204003899,
+      "learning_rate": 2.962463385437373e-06,
+      "loss": 0.8131,
+      "step": 32091
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4013184472074747,
+      "learning_rate": 2.961921307924838e-06,
+      "loss": 0.8286,
+      "step": 32092
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6344460658172713,
+      "learning_rate": 2.961379271390222e-06,
+      "loss": 0.7556,
+      "step": 32093
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6956874941103466,
+      "learning_rate": 2.9608372758366842e-06,
+      "loss": 0.7962,
+      "step": 32094
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7209954023021594,
+      "learning_rate": 2.9602953212673813e-06,
+      "loss": 0.8566,
+      "step": 32095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4805929361407366,
+      "learning_rate": 2.959753407685473e-06,
+      "loss": 0.8525,
+      "step": 32096
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7329544506135445,
+      "learning_rate": 2.9592115350941032e-06,
+      "loss": 0.6819,
+      "step": 32097
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5856067252224622,
+      "learning_rate": 2.958669703496434e-06,
+      "loss": 0.8412,
+      "step": 32098
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5846158557332723,
+      "learning_rate": 2.958127912895621e-06,
+      "loss": 0.8486,
+      "step": 32099
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.60497299423398,
+      "learning_rate": 2.9575861632948168e-06,
+      "loss": 0.8234,
+      "step": 32100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6158373504832566,
+      "learning_rate": 2.957044454697172e-06,
+      "loss": 0.8523,
+      "step": 32101
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.588755859781262,
+      "learning_rate": 2.956502787105845e-06,
+      "loss": 0.8226,
+      "step": 32102
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4567901895743671,
+      "learning_rate": 2.9559611605239914e-06,
+      "loss": 0.8668,
+      "step": 32103
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6330062693698781,
+      "learning_rate": 2.955419574954762e-06,
+      "loss": 0.8161,
+      "step": 32104
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7141184449778215,
+      "learning_rate": 2.954878030401308e-06,
+      "loss": 0.82,
+      "step": 32105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6864421458323993,
+      "learning_rate": 2.9543365268667866e-06,
+      "loss": 0.8234,
+      "step": 32106
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.710170047082658,
+      "learning_rate": 2.9537950643543455e-06,
+      "loss": 0.7966,
+      "step": 32107
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5563603402784065,
+      "learning_rate": 2.953253642867141e-06,
+      "loss": 0.8872,
+      "step": 32108
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5901542713706653,
+      "learning_rate": 2.952712262408327e-06,
+      "loss": 0.8627,
+      "step": 32109
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.593330503016378,
+      "learning_rate": 2.9521709229810535e-06,
+      "loss": 0.7483,
+      "step": 32110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6574150996508727,
+      "learning_rate": 2.951629624588469e-06,
+      "loss": 0.8017,
+      "step": 32111
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5969465195732448,
+      "learning_rate": 2.951088367233729e-06,
+      "loss": 0.7799,
+      "step": 32112
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.065593757352318,
+      "learning_rate": 2.95054715091999e-06,
+      "loss": 0.8078,
+      "step": 32113
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.338545252128151,
+      "learning_rate": 2.9500059756503907e-06,
+      "loss": 0.8388,
+      "step": 32114
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4780151869713736,
+      "learning_rate": 2.949464841428089e-06,
+      "loss": 0.8203,
+      "step": 32115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8034893280928254,
+      "learning_rate": 2.948923748256236e-06,
+      "loss": 0.7021,
+      "step": 32116
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.551823669358586,
+      "learning_rate": 2.9483826961379837e-06,
+      "loss": 0.7835,
+      "step": 32117
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6607316222707746,
+      "learning_rate": 2.947841685076479e-06,
+      "loss": 0.7717,
+      "step": 32118
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7166343457663027,
+      "learning_rate": 2.94730071507487e-06,
+      "loss": 0.8707,
+      "step": 32119
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9361243051578014,
+      "learning_rate": 2.946759786136313e-06,
+      "loss": 0.7833,
+      "step": 32120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5821660413393273,
+      "learning_rate": 2.9462188982639494e-06,
+      "loss": 0.8884,
+      "step": 32121
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.541002941944384,
+      "learning_rate": 2.9456780514609364e-06,
+      "loss": 0.7798,
+      "step": 32122
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5473083851507285,
+      "learning_rate": 2.9451372457304163e-06,
+      "loss": 0.7431,
+      "step": 32123
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4932579606036176,
+      "learning_rate": 2.944596481075542e-06,
+      "loss": 0.7879,
+      "step": 32124
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.663837353828135,
+      "learning_rate": 2.94405575749946e-06,
+      "loss": 0.7727,
+      "step": 32125
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7817763194057371,
+      "learning_rate": 2.943515075005321e-06,
+      "loss": 0.6764,
+      "step": 32126
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.913816278997644,
+      "learning_rate": 2.9429744335962717e-06,
+      "loss": 0.895,
+      "step": 32127
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7048892050544158,
+      "learning_rate": 2.9424338332754553e-06,
+      "loss": 0.8264,
+      "step": 32128
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5374751712462322,
+      "learning_rate": 2.941893274046025e-06,
+      "loss": 0.6767,
+      "step": 32129
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.714887495905764,
+      "learning_rate": 2.941352755911129e-06,
+      "loss": 0.6567,
+      "step": 32130
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.575664582062448,
+      "learning_rate": 2.9408122788739124e-06,
+      "loss": 0.7833,
+      "step": 32131
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4969604064627902,
+      "learning_rate": 2.940271842937518e-06,
+      "loss": 0.7403,
+      "step": 32132
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5943329895054312,
+      "learning_rate": 2.9397314481050953e-06,
+      "loss": 0.9471,
+      "step": 32133
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6506101144550316,
+      "learning_rate": 2.9391910943797963e-06,
+      "loss": 0.859,
+      "step": 32134
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5649291331870736,
+      "learning_rate": 2.93865078176476e-06,
+      "loss": 0.8586,
+      "step": 32135
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4736074225393212,
+      "learning_rate": 2.9381105102631334e-06,
+      "loss": 0.8289,
+      "step": 32136
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6066729306433818,
+      "learning_rate": 2.9375702798780626e-06,
+      "loss": 0.8379,
+      "step": 32137
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8648745040164028,
+      "learning_rate": 2.937030090612696e-06,
+      "loss": 0.8832,
+      "step": 32138
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2545308321208806,
+      "learning_rate": 2.9364899424701765e-06,
+      "loss": 0.6592,
+      "step": 32139
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9430153571000564,
+      "learning_rate": 2.9359498354536465e-06,
+      "loss": 0.782,
+      "step": 32140
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4314138992762275,
+      "learning_rate": 2.935409769566255e-06,
+      "loss": 0.7513,
+      "step": 32141
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6529687721368236,
+      "learning_rate": 2.934869744811142e-06,
+      "loss": 0.8204,
+      "step": 32142
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7707508265265338,
+      "learning_rate": 2.934329761191457e-06,
+      "loss": 0.6219,
+      "step": 32143
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7317514905582814,
+      "learning_rate": 2.9337898187103374e-06,
+      "loss": 0.7623,
+      "step": 32144
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0746367368945737,
+      "learning_rate": 2.933249917370935e-06,
+      "loss": 0.8596,
+      "step": 32145
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7103785708474852,
+      "learning_rate": 2.9327100571763843e-06,
+      "loss": 0.8603,
+      "step": 32146
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5507813793435752,
+      "learning_rate": 2.9321702381298378e-06,
+      "loss": 0.8054,
+      "step": 32147
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3265980261119592,
+      "learning_rate": 2.9316304602344324e-06,
+      "loss": 0.805,
+      "step": 32148
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.964192579791654,
+      "learning_rate": 2.9310907234933104e-06,
+      "loss": 0.8389,
+      "step": 32149
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.805290863574419,
+      "learning_rate": 2.930551027909615e-06,
+      "loss": 0.8158,
+      "step": 32150
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.501828898822335,
+      "learning_rate": 2.930011373486494e-06,
+      "loss": 0.8766,
+      "step": 32151
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7802958792958163,
+      "learning_rate": 2.9294717602270837e-06,
+      "loss": 0.7493,
+      "step": 32152
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3896563513658462,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.8164,
+      "step": 32153
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.011088224448752,
+      "learning_rate": 2.9283926572119625e-06,
+      "loss": 0.6983,
+      "step": 32154
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.498674377435085,
+      "learning_rate": 2.927853167462539e-06,
+      "loss": 0.8426,
+      "step": 32155
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.542318205598442,
+      "learning_rate": 2.927313718889393e-06,
+      "loss": 0.8754,
+      "step": 32156
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6608366514767858,
+      "learning_rate": 2.9267743114956635e-06,
+      "loss": 0.8049,
+      "step": 32157
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7251013971241871,
+      "learning_rate": 2.926234945284493e-06,
+      "loss": 0.6877,
+      "step": 32158
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7788027050369533,
+      "learning_rate": 2.9256956202590248e-06,
+      "loss": 0.7962,
+      "step": 32159
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0708458517933703,
+      "learning_rate": 2.9251563364223967e-06,
+      "loss": 0.7993,
+      "step": 32160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0927407221655687,
+      "learning_rate": 2.924617093777745e-06,
+      "loss": 0.8038,
+      "step": 32161
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.669856324289624,
+      "learning_rate": 2.9240778923282163e-06,
+      "loss": 0.8583,
+      "step": 32162
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6198350205877432,
+      "learning_rate": 2.9235387320769427e-06,
+      "loss": 0.935,
+      "step": 32163
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3133309936784316,
+      "learning_rate": 2.92299961302707e-06,
+      "loss": 0.8383,
+      "step": 32164
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7616863402592225,
+      "learning_rate": 2.9224605351817314e-06,
+      "loss": 0.7003,
+      "step": 32165
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4877835253059635,
+      "learning_rate": 2.9219214985440715e-06,
+      "loss": 0.8266,
+      "step": 32166
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.682412670479461,
+      "learning_rate": 2.9213825031172216e-06,
+      "loss": 0.805,
+      "step": 32167
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5997370304285248,
+      "learning_rate": 2.9208435489043276e-06,
+      "loss": 0.8354,
+      "step": 32168
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.810413863667481,
+      "learning_rate": 2.920304635908524e-06,
+      "loss": 0.9058,
+      "step": 32169
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5082451127261598,
+      "learning_rate": 2.9197657641329447e-06,
+      "loss": 0.8795,
+      "step": 32170
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6253648091466677,
+      "learning_rate": 2.91922693358073e-06,
+      "loss": 1.0205,
+      "step": 32171
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8650518014534248,
+      "learning_rate": 2.918688144255022e-06,
+      "loss": 0.7372,
+      "step": 32172
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7190776345546481,
+      "learning_rate": 2.9181493961589526e-06,
+      "loss": 0.8715,
+      "step": 32173
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7545454000177654,
+      "learning_rate": 2.917610689295657e-06,
+      "loss": 0.6495,
+      "step": 32174
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.804621169471203,
+      "learning_rate": 2.9170720236682737e-06,
+      "loss": 0.8066,
+      "step": 32175
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.693331744343695,
+      "learning_rate": 2.9165333992799428e-06,
+      "loss": 0.7501,
+      "step": 32176
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7002738917565405,
+      "learning_rate": 2.915994816133797e-06,
+      "loss": 0.8927,
+      "step": 32177
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5933956460366205,
+      "learning_rate": 2.9154562742329695e-06,
+      "loss": 0.919,
+      "step": 32178
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4999242209857584,
+      "learning_rate": 2.9149177735805977e-06,
+      "loss": 0.92,
+      "step": 32179
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7653426000510115,
+      "learning_rate": 2.9143793141798214e-06,
+      "loss": 0.7185,
+      "step": 32180
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.403009206309841,
+      "learning_rate": 2.9138408960337725e-06,
+      "loss": 0.7242,
+      "step": 32181
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4862365547591088,
+      "learning_rate": 2.913302519145582e-06,
+      "loss": 0.8297,
+      "step": 32182
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6038615080072696,
+      "learning_rate": 2.9127641835183906e-06,
+      "loss": 0.7699,
+      "step": 32183
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5915935427081207,
+      "learning_rate": 2.912225889155327e-06,
+      "loss": 0.8024,
+      "step": 32184
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.029800505728631,
+      "learning_rate": 2.9116876360595304e-06,
+      "loss": 0.8659,
+      "step": 32185
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.709088329491362,
+      "learning_rate": 2.9111494242341297e-06,
+      "loss": 0.8071,
+      "step": 32186
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5225154392945632,
+      "learning_rate": 2.910611253682265e-06,
+      "loss": 0.8223,
+      "step": 32187
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.649891904701666,
+      "learning_rate": 2.9100731244070635e-06,
+      "loss": 0.88,
+      "step": 32188
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7290563181274947,
+      "learning_rate": 2.909535036411664e-06,
+      "loss": 0.6661,
+      "step": 32189
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.550648710724822,
+      "learning_rate": 2.9089969896991956e-06,
+      "loss": 0.7504,
+      "step": 32190
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.450354607138661,
+      "learning_rate": 2.9084589842727883e-06,
+      "loss": 0.8796,
+      "step": 32191
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4533763081434576,
+      "learning_rate": 2.907921020135578e-06,
+      "loss": 0.8802,
+      "step": 32192
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4652551761058727,
+      "learning_rate": 2.907383097290697e-06,
+      "loss": 0.724,
+      "step": 32193
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6486762422518173,
+      "learning_rate": 2.906845215741283e-06,
+      "loss": 0.8707,
+      "step": 32194
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.587211409477784,
+      "learning_rate": 2.906307375490457e-06,
+      "loss": 0.838,
+      "step": 32195
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.050103631041777,
+      "learning_rate": 2.905769576541354e-06,
+      "loss": 0.8961,
+      "step": 32196
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6678376402083317,
+      "learning_rate": 2.9052318188971097e-06,
+      "loss": 0.9003,
+      "step": 32197
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5933715057001872,
+      "learning_rate": 2.904694102560852e-06,
+      "loss": 0.7975,
+      "step": 32198
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6054488051975646,
+      "learning_rate": 2.904156427535709e-06,
+      "loss": 0.8908,
+      "step": 32199
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5040117335560137,
+      "learning_rate": 2.9036187938248126e-06,
+      "loss": 0.8617,
+      "step": 32200
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7228418236613885,
+      "learning_rate": 2.903081201431299e-06,
+      "loss": 0.8158,
+      "step": 32201
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5553960379953928,
+      "learning_rate": 2.9025436503582893e-06,
+      "loss": 0.8413,
+      "step": 32202
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6115566177450769,
+      "learning_rate": 2.9020061406089207e-06,
+      "loss": 0.8129,
+      "step": 32203
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5944987544314204,
+      "learning_rate": 2.9014686721863192e-06,
+      "loss": 0.8857,
+      "step": 32204
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5786043823417908,
+      "learning_rate": 2.900931245093612e-06,
+      "loss": 0.8495,
+      "step": 32205
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5686207902454368,
+      "learning_rate": 2.900393859333931e-06,
+      "loss": 0.9306,
+      "step": 32206
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5834486196940278,
+      "learning_rate": 2.8998565149104075e-06,
+      "loss": 0.8068,
+      "step": 32207
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8591627426005939,
+      "learning_rate": 2.899319211826167e-06,
+      "loss": 0.7982,
+      "step": 32208
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5853765478983013,
+      "learning_rate": 2.898781950084335e-06,
+      "loss": 0.7893,
+      "step": 32209
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.614377747734403,
+      "learning_rate": 2.8982447296880423e-06,
+      "loss": 0.8172,
+      "step": 32210
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6484322834906933,
+      "learning_rate": 2.897707550640424e-06,
+      "loss": 0.8717,
+      "step": 32211
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7052590106716323,
+      "learning_rate": 2.8971704129445944e-06,
+      "loss": 0.84,
+      "step": 32212
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7023682546060979,
+      "learning_rate": 2.8966333166036864e-06,
+      "loss": 0.8999,
+      "step": 32213
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.537780971057724,
+      "learning_rate": 2.89609626162083e-06,
+      "loss": 0.7576,
+      "step": 32214
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7395124928841097,
+      "learning_rate": 2.8955592479991524e-06,
+      "loss": 0.8656,
+      "step": 32215
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7095310386070015,
+      "learning_rate": 2.895022275741778e-06,
+      "loss": 0.8565,
+      "step": 32216
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4831219440039822,
+      "learning_rate": 2.8944853448518295e-06,
+      "loss": 0.7982,
+      "step": 32217
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.581242851774605,
+      "learning_rate": 2.893948455332438e-06,
+      "loss": 0.8062,
+      "step": 32218
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4410704543075967,
+      "learning_rate": 2.8934116071867314e-06,
+      "loss": 0.9005,
+      "step": 32219
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9097034909438366,
+      "learning_rate": 2.892874800417832e-06,
+      "loss": 0.8104,
+      "step": 32220
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.65286492537908,
+      "learning_rate": 2.892338035028862e-06,
+      "loss": 0.7737,
+      "step": 32221
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7072114468635973,
+      "learning_rate": 2.8918013110229535e-06,
+      "loss": 0.9472,
+      "step": 32222
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4706317922228074,
+      "learning_rate": 2.8912646284032264e-06,
+      "loss": 0.7721,
+      "step": 32223
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.463195513004139,
+      "learning_rate": 2.890727987172809e-06,
+      "loss": 0.8914,
+      "step": 32224
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7079974006205436,
+      "learning_rate": 2.8901913873348208e-06,
+      "loss": 0.7655,
+      "step": 32225
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5942338672871845,
+      "learning_rate": 2.889654828892393e-06,
+      "loss": 0.7947,
+      "step": 32226
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.726357585855881,
+      "learning_rate": 2.889118311848643e-06,
+      "loss": 0.6484,
+      "step": 32227
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.618414551937836,
+      "learning_rate": 2.8885818362066997e-06,
+      "loss": 0.7541,
+      "step": 32228
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6182932947418576,
+      "learning_rate": 2.8880454019696846e-06,
+      "loss": 0.7966,
+      "step": 32229
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.518445200112405,
+      "learning_rate": 2.8875090091407187e-06,
+      "loss": 0.8074,
+      "step": 32230
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.609715722077438,
+      "learning_rate": 2.8869726577229264e-06,
+      "loss": 0.7984,
+      "step": 32231
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.286323842166511,
+      "learning_rate": 2.8864363477194336e-06,
+      "loss": 0.7989,
+      "step": 32232
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6042771877756112,
+      "learning_rate": 2.8859000791333613e-06,
+      "loss": 0.8619,
+      "step": 32233
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4535769328340218,
+      "learning_rate": 2.8853638519678285e-06,
+      "loss": 0.9301,
+      "step": 32234
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7086885044589204,
+      "learning_rate": 2.884827666225959e-06,
+      "loss": 0.9351,
+      "step": 32235
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.456085942245219,
+      "learning_rate": 2.884291521910878e-06,
+      "loss": 0.8053,
+      "step": 32236
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7783692849624668,
+      "learning_rate": 2.8837554190257057e-06,
+      "loss": 0.7605,
+      "step": 32237
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3864008203877511,
+      "learning_rate": 2.883219357573559e-06,
+      "loss": 0.7836,
+      "step": 32238
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5353114040266642,
+      "learning_rate": 2.8826833375575615e-06,
+      "loss": 0.7124,
+      "step": 32239
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8668055469468068,
+      "learning_rate": 2.882147358980839e-06,
+      "loss": 0.7275,
+      "step": 32240
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5678892476307373,
+      "learning_rate": 2.8816114218465073e-06,
+      "loss": 0.9085,
+      "step": 32241
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6770922959047148,
+      "learning_rate": 2.881075526157685e-06,
+      "loss": 0.9183,
+      "step": 32242
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7335043144967308,
+      "learning_rate": 2.880539671917497e-06,
+      "loss": 0.8945,
+      "step": 32243
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7127978569655153,
+      "learning_rate": 2.880003859129059e-06,
+      "loss": 0.7899,
+      "step": 32244
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7325989099265127,
+      "learning_rate": 2.879468087795495e-06,
+      "loss": 0.8709,
+      "step": 32245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7010502995418375,
+      "learning_rate": 2.87893235791992e-06,
+      "loss": 0.6269,
+      "step": 32246
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6393770893080337,
+      "learning_rate": 2.8783966695054575e-06,
+      "loss": 1.0419,
+      "step": 32247
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.851457542189568,
+      "learning_rate": 2.877861022555222e-06,
+      "loss": 0.7841,
+      "step": 32248
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5123154658981366,
+      "learning_rate": 2.8773254170723363e-06,
+      "loss": 0.8439,
+      "step": 32249
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.709346768000983,
+      "learning_rate": 2.876789853059918e-06,
+      "loss": 0.8752,
+      "step": 32250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5639828496084653,
+      "learning_rate": 2.876254330521081e-06,
+      "loss": 0.8219,
+      "step": 32251
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6399981543496258,
+      "learning_rate": 2.8757188494589474e-06,
+      "loss": 0.9333,
+      "step": 32252
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6681601913835387,
+      "learning_rate": 2.875183409876636e-06,
+      "loss": 0.8713,
+      "step": 32253
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5054820936322266,
+      "learning_rate": 2.8746480117772626e-06,
+      "loss": 0.8704,
+      "step": 32254
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7702945597428478,
+      "learning_rate": 2.874112655163942e-06,
+      "loss": 0.6655,
+      "step": 32255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.566093854186186,
+      "learning_rate": 2.8735773400397926e-06,
+      "loss": 0.8879,
+      "step": 32256
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0373847959881584,
+      "learning_rate": 2.8730420664079363e-06,
+      "loss": 0.871,
+      "step": 32257
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6968255171695381,
+      "learning_rate": 2.8725068342714847e-06,
+      "loss": 0.7434,
+      "step": 32258
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.697798516252022,
+      "learning_rate": 2.871971643633552e-06,
+      "loss": 0.8343,
+      "step": 32259
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.247213910758304,
+      "learning_rate": 2.871436494497256e-06,
+      "loss": 0.8382,
+      "step": 32260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5337461479739798,
+      "learning_rate": 2.8709013868657186e-06,
+      "loss": 0.7507,
+      "step": 32261
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7983578706256516,
+      "learning_rate": 2.8703663207420494e-06,
+      "loss": 0.8061,
+      "step": 32262
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2127393118032126,
+      "learning_rate": 2.869831296129362e-06,
+      "loss": 1.0721,
+      "step": 32263
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5997884314683268,
+      "learning_rate": 2.869296313030776e-06,
+      "loss": 0.8008,
+      "step": 32264
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5179701388205653,
+      "learning_rate": 2.8687613714494023e-06,
+      "loss": 0.8026,
+      "step": 32265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6839507854525553,
+      "learning_rate": 2.868226471388361e-06,
+      "loss": 0.7422,
+      "step": 32266
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6390842386641489,
+      "learning_rate": 2.867691612850759e-06,
+      "loss": 0.7469,
+      "step": 32267
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7444389477353612,
+      "learning_rate": 2.8671567958397183e-06,
+      "loss": 0.6962,
+      "step": 32268
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5542379744502266,
+      "learning_rate": 2.8666220203583462e-06,
+      "loss": 0.7676,
+      "step": 32269
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.772224775257747,
+      "learning_rate": 2.866087286409761e-06,
+      "loss": 0.8206,
+      "step": 32270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5751079992342543,
+      "learning_rate": 2.865552593997075e-06,
+      "loss": 0.8523,
+      "step": 32271
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5366742371997677,
+      "learning_rate": 2.8650179431233984e-06,
+      "loss": 0.7532,
+      "step": 32272
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5878216102659137,
+      "learning_rate": 2.8644833337918444e-06,
+      "loss": 0.8393,
+      "step": 32273
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8086050054267397,
+      "learning_rate": 2.863948766005531e-06,
+      "loss": 0.74,
+      "step": 32274
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1681010277072197,
+      "learning_rate": 2.8634142397675678e-06,
+      "loss": 0.7911,
+      "step": 32275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6627945814499014,
+      "learning_rate": 2.862879755081063e-06,
+      "loss": 0.6852,
+      "step": 32276
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6623221266595893,
+      "learning_rate": 2.8623453119491316e-06,
+      "loss": 0.8236,
+      "step": 32277
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7494882942633556,
+      "learning_rate": 2.861810910374888e-06,
+      "loss": 0.68,
+      "step": 32278
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7008577107966008,
+      "learning_rate": 2.8612765503614424e-06,
+      "loss": 0.7373,
+      "step": 32279
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.556415490352297,
+      "learning_rate": 2.8607422319119017e-06,
+      "loss": 0.8682,
+      "step": 32280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7718151891223284,
+      "learning_rate": 2.8602079550293792e-06,
+      "loss": 0.8669,
+      "step": 32281
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7628371782797279,
+      "learning_rate": 2.85967371971699e-06,
+      "loss": 0.7531,
+      "step": 32282
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.649905573510686,
+      "learning_rate": 2.859139525977841e-06,
+      "loss": 0.7572,
+      "step": 32283
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1656317110511565,
+      "learning_rate": 2.8586053738150388e-06,
+      "loss": 0.7842,
+      "step": 32284
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6664265800297042,
+      "learning_rate": 2.858071263231701e-06,
+      "loss": 0.7331,
+      "step": 32285
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4610297291407999,
+      "learning_rate": 2.8575371942309293e-06,
+      "loss": 0.7112,
+      "step": 32286
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6531741556443658,
+      "learning_rate": 2.8570031668158403e-06,
+      "loss": 0.9023,
+      "step": 32287
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8047000326232607,
+      "learning_rate": 2.8564691809895383e-06,
+      "loss": 0.8162,
+      "step": 32288
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5034051131791897,
+      "learning_rate": 2.8559352367551375e-06,
+      "loss": 0.8618,
+      "step": 32289
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6548614032031994,
+      "learning_rate": 2.855401334115739e-06,
+      "loss": 0.7652,
+      "step": 32290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.639596163890266,
+      "learning_rate": 2.8548674730744576e-06,
+      "loss": 0.8587,
+      "step": 32291
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8359004225246158,
+      "learning_rate": 2.8543336536344047e-06,
+      "loss": 0.895,
+      "step": 32292
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6345324827348386,
+      "learning_rate": 2.8537998757986773e-06,
+      "loss": 0.8208,
+      "step": 32293
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4956639016048423,
+      "learning_rate": 2.853266139570391e-06,
+      "loss": 0.8856,
+      "step": 32294
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.103301836818949,
+      "learning_rate": 2.8527324449526505e-06,
+      "loss": 0.8824,
+      "step": 32295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.702741639947702,
+      "learning_rate": 2.8521987919485705e-06,
+      "loss": 0.8541,
+      "step": 32296
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5409649982192286,
+      "learning_rate": 2.851665180561247e-06,
+      "loss": 0.8772,
+      "step": 32297
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7741195047037281,
+      "learning_rate": 2.8511316107937912e-06,
+      "loss": 0.7125,
+      "step": 32298
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.776544665760189,
+      "learning_rate": 2.850598082649314e-06,
+      "loss": 0.6905,
+      "step": 32299
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7602381403134466,
+      "learning_rate": 2.850064596130915e-06,
+      "loss": 0.707,
+      "step": 32300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7861220641726762,
+      "learning_rate": 2.849531151241707e-06,
+      "loss": 0.7515,
+      "step": 32301
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.543389992406762,
+      "learning_rate": 2.8489977479847886e-06,
+      "loss": 0.8286,
+      "step": 32302
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5075106968257583,
+      "learning_rate": 2.8484643863632722e-06,
+      "loss": 0.8412,
+      "step": 32303
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8095190928385474,
+      "learning_rate": 2.8479310663802574e-06,
+      "loss": 0.6958,
+      "step": 32304
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7850425539431378,
+      "learning_rate": 2.847397788038856e-06,
+      "loss": 0.9115,
+      "step": 32305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7018827692443883,
+      "learning_rate": 2.8468645513421688e-06,
+      "loss": 0.863,
+      "step": 32306
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6629788001075434,
+      "learning_rate": 2.8463313562932983e-06,
+      "loss": 0.671,
+      "step": 32307
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.487626798974178,
+      "learning_rate": 2.8457982028953502e-06,
+      "loss": 0.9013,
+      "step": 32308
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.734500764898843,
+      "learning_rate": 2.8452650911514336e-06,
+      "loss": 0.8737,
+      "step": 32309
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5241864520447055,
+      "learning_rate": 2.844732021064649e-06,
+      "loss": 0.7839,
+      "step": 32310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6963809900851106,
+      "learning_rate": 2.8441989926380966e-06,
+      "loss": 0.7561,
+      "step": 32311
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6445858819171144,
+      "learning_rate": 2.843666005874883e-06,
+      "loss": 0.8386,
+      "step": 32312
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5583983156096546,
+      "learning_rate": 2.843133060778115e-06,
+      "loss": 0.8275,
+      "step": 32313
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5587268679282074,
+      "learning_rate": 2.842600157350892e-06,
+      "loss": 0.8917,
+      "step": 32314
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.548207142879946,
+      "learning_rate": 2.8420672955963136e-06,
+      "loss": 0.9897,
+      "step": 32315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.737682751834753,
+      "learning_rate": 2.8415344755174855e-06,
+      "loss": 0.8423,
+      "step": 32316
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.821704973998489,
+      "learning_rate": 2.8410016971175136e-06,
+      "loss": 0.7609,
+      "step": 32317
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4890730464166293,
+      "learning_rate": 2.840468960399496e-06,
+      "loss": 0.7985,
+      "step": 32318
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5470243038785052,
+      "learning_rate": 2.8399362653665318e-06,
+      "loss": 0.8089,
+      "step": 32319
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5028902550826286,
+      "learning_rate": 2.8394036120217296e-06,
+      "loss": 0.7967,
+      "step": 32320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5216515089317413,
+      "learning_rate": 2.838871000368182e-06,
+      "loss": 0.8974,
+      "step": 32321
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.433457230998061,
+      "learning_rate": 2.8383384304089978e-06,
+      "loss": 0.805,
+      "step": 32322
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7046733281610071,
+      "learning_rate": 2.8378059021472725e-06,
+      "loss": 0.8948,
+      "step": 32323
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.538971078849944,
+      "learning_rate": 2.8372734155861105e-06,
+      "loss": 0.7457,
+      "step": 32324
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6345763908954483,
+      "learning_rate": 2.8367409707286085e-06,
+      "loss": 0.7399,
+      "step": 32325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.514521706036574,
+      "learning_rate": 2.8362085675778708e-06,
+      "loss": 0.8416,
+      "step": 32326
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.52446657527312,
+      "learning_rate": 2.8356762061369945e-06,
+      "loss": 0.8532,
+      "step": 32327
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6939916316539885,
+      "learning_rate": 2.8351438864090774e-06,
+      "loss": 0.8759,
+      "step": 32328
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5669304129467567,
+      "learning_rate": 2.834611608397221e-06,
+      "loss": 0.8288,
+      "step": 32329
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7325596330596729,
+      "learning_rate": 2.8340793721045266e-06,
+      "loss": 0.8453,
+      "step": 32330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8581597845038835,
+      "learning_rate": 2.8335471775340907e-06,
+      "loss": 0.8473,
+      "step": 32331
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6972325997444249,
+      "learning_rate": 2.8330150246890097e-06,
+      "loss": 0.8678,
+      "step": 32332
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.464648976291235,
+      "learning_rate": 2.832482913572384e-06,
+      "loss": 0.9033,
+      "step": 32333
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7034744588533257,
+      "learning_rate": 2.8319508441873143e-06,
+      "loss": 0.9038,
+      "step": 32334
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.665765229326212,
+      "learning_rate": 2.831418816536897e-06,
+      "loss": 0.8328,
+      "step": 32335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6836281607034054,
+      "learning_rate": 2.8308868306242253e-06,
+      "loss": 0.8352,
+      "step": 32336
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5851207877927476,
+      "learning_rate": 2.8303548864524012e-06,
+      "loss": 0.8543,
+      "step": 32337
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6382474768514974,
+      "learning_rate": 2.829822984024523e-06,
+      "loss": 0.9164,
+      "step": 32338
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7727595545863488,
+      "learning_rate": 2.829291123343686e-06,
+      "loss": 0.6751,
+      "step": 32339
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5751075012694447,
+      "learning_rate": 2.8287593044129825e-06,
+      "loss": 0.9974,
+      "step": 32340
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.692778536123289,
+      "learning_rate": 2.828227527235513e-06,
+      "loss": 0.851,
+      "step": 32341
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6071058662446112,
+      "learning_rate": 2.8276957918143766e-06,
+      "loss": 0.8232,
+      "step": 32342
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4551926608178976,
+      "learning_rate": 2.827164098152666e-06,
+      "loss": 0.8013,
+      "step": 32343
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5461313301332646,
+      "learning_rate": 2.8266324462534733e-06,
+      "loss": 0.8581,
+      "step": 32344
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.611665003847029,
+      "learning_rate": 2.8261008361199017e-06,
+      "loss": 0.9086,
+      "step": 32345
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3798164327580977,
+      "learning_rate": 2.8255692677550385e-06,
+      "loss": 0.6855,
+      "step": 32346
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7257053683796808,
+      "learning_rate": 2.8250377411619845e-06,
+      "loss": 0.7153,
+      "step": 32347
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7891179970500142,
+      "learning_rate": 2.8245062563438306e-06,
+      "loss": 0.8479,
+      "step": 32348
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7480802423606416,
+      "learning_rate": 2.823974813303676e-06,
+      "loss": 0.8027,
+      "step": 32349
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8486482378847087,
+      "learning_rate": 2.8234434120446084e-06,
+      "loss": 0.7935,
+      "step": 32350
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.498669925885865,
+      "learning_rate": 2.8229120525697285e-06,
+      "loss": 0.8359,
+      "step": 32351
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6158000879972765,
+      "learning_rate": 2.8223807348821273e-06,
+      "loss": 0.7849,
+      "step": 32352
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7223335664679849,
+      "learning_rate": 2.821849458984894e-06,
+      "loss": 0.6874,
+      "step": 32353
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0042717536681303,
+      "learning_rate": 2.8213182248811265e-06,
+      "loss": 0.761,
+      "step": 32354
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8096738175570444,
+      "learning_rate": 2.8207870325739196e-06,
+      "loss": 0.9144,
+      "step": 32355
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6338755542100538,
+      "learning_rate": 2.820255882066364e-06,
+      "loss": 0.8375,
+      "step": 32356
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8013657072703764,
+      "learning_rate": 2.8197247733615474e-06,
+      "loss": 0.7274,
+      "step": 32357
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.790881386425575,
+      "learning_rate": 2.8191937064625675e-06,
+      "loss": 0.9439,
+      "step": 32358
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3903640541337083,
+      "learning_rate": 2.818662681372518e-06,
+      "loss": 0.8418,
+      "step": 32359
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8080153236766792,
+      "learning_rate": 2.8181316980944884e-06,
+      "loss": 0.8622,
+      "step": 32360
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8001098310531829,
+      "learning_rate": 2.8176007566315665e-06,
+      "loss": 0.7127,
+      "step": 32361
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5330279517861776,
+      "learning_rate": 2.817069856986846e-06,
+      "loss": 0.9061,
+      "step": 32362
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.579976186422907,
+      "learning_rate": 2.816538999163423e-06,
+      "loss": 0.7762,
+      "step": 32363
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5649314567618609,
+      "learning_rate": 2.8160081831643836e-06,
+      "loss": 0.7656,
+      "step": 32364
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6972236887732823,
+      "learning_rate": 2.8154774089928172e-06,
+      "loss": 0.9038,
+      "step": 32365
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.782397826176686,
+      "learning_rate": 2.8149466766518175e-06,
+      "loss": 0.8124,
+      "step": 32366
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5217581608294615,
+      "learning_rate": 2.814415986144471e-06,
+      "loss": 0.6746,
+      "step": 32367
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.895096999225569,
+      "learning_rate": 2.8138853374738717e-06,
+      "loss": 0.758,
+      "step": 32368
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7601623303906785,
+      "learning_rate": 2.813354730643104e-06,
+      "loss": 0.7185,
+      "step": 32369
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7352013764970025,
+      "learning_rate": 2.8128241656552637e-06,
+      "loss": 0.8694,
+      "step": 32370
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5423766413418536,
+      "learning_rate": 2.812293642513434e-06,
+      "loss": 0.7239,
+      "step": 32371
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.556414850687604,
+      "learning_rate": 2.8117631612207084e-06,
+      "loss": 0.8552,
+      "step": 32372
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.553035411465646,
+      "learning_rate": 2.811232721780174e-06,
+      "loss": 0.7685,
+      "step": 32373
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.842072746478816,
+      "learning_rate": 2.810702324194915e-06,
+      "loss": 0.7182,
+      "step": 32374
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.829999617307067,
+      "learning_rate": 2.810171968468024e-06,
+      "loss": 0.7022,
+      "step": 32375
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6027385325573145,
+      "learning_rate": 2.809641654602587e-06,
+      "loss": 0.8474,
+      "step": 32376
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.574960645686489,
+      "learning_rate": 2.8091113826017004e-06,
+      "loss": 0.7607,
+      "step": 32377
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6213336433330026,
+      "learning_rate": 2.808581152468436e-06,
+      "loss": 0.7826,
+      "step": 32378
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.705115741038839,
+      "learning_rate": 2.8080509642058894e-06,
+      "loss": 0.7748,
+      "step": 32379
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4513655576966609,
+      "learning_rate": 2.8075208178171507e-06,
+      "loss": 0.8693,
+      "step": 32380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.668236777287976,
+      "learning_rate": 2.8069907133053033e-06,
+      "loss": 0.7403,
+      "step": 32381
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6823040264117135,
+      "learning_rate": 2.8064606506734292e-06,
+      "loss": 0.737,
+      "step": 32382
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7068682839746727,
+      "learning_rate": 2.805930629924619e-06,
+      "loss": 0.7508,
+      "step": 32383
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7383654904658759,
+      "learning_rate": 2.8054006510619615e-06,
+      "loss": 0.672,
+      "step": 32384
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5841257299272504,
+      "learning_rate": 2.804870714088539e-06,
+      "loss": 0.8895,
+      "step": 32385
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8579071559487026,
+      "learning_rate": 2.804340819007434e-06,
+      "loss": 0.8745,
+      "step": 32386
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.653977059741377,
+      "learning_rate": 2.803810965821737e-06,
+      "loss": 0.812,
+      "step": 32387
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6124875548660111,
+      "learning_rate": 2.8032811545345294e-06,
+      "loss": 0.8629,
+      "step": 32388
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7203578046370768,
+      "learning_rate": 2.8027513851488963e-06,
+      "loss": 0.6479,
+      "step": 32389
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.569438411155337,
+      "learning_rate": 2.802221657667927e-06,
+      "loss": 0.742,
+      "step": 32390
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6717950662237808,
+      "learning_rate": 2.8016919720947024e-06,
+      "loss": 1.0015,
+      "step": 32391
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.618202340531434,
+      "learning_rate": 2.801162328432302e-06,
+      "loss": 0.8005,
+      "step": 32392
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.658847294618581,
+      "learning_rate": 2.800632726683814e-06,
+      "loss": 0.9561,
+      "step": 32393
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9283935835354629,
+      "learning_rate": 2.8001031668523273e-06,
+      "loss": 0.9272,
+      "step": 32394
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7727553075205592,
+      "learning_rate": 2.799573648940913e-06,
+      "loss": 0.7058,
+      "step": 32395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.537626695873553,
+      "learning_rate": 2.7990441729526617e-06,
+      "loss": 0.9375,
+      "step": 32396
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.624237709822225,
+      "learning_rate": 2.798514738890654e-06,
+      "loss": 0.8579,
+      "step": 32397
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.513667541236211,
+      "learning_rate": 2.7979853467579777e-06,
+      "loss": 0.8394,
+      "step": 32398
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6981193075307692,
+      "learning_rate": 2.7974559965577097e-06,
+      "loss": 0.7854,
+      "step": 32399
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4710019389355615,
+      "learning_rate": 2.7969266882929304e-06,
+      "loss": 0.7622,
+      "step": 32400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7316195416923024,
+      "learning_rate": 2.796397421966728e-06,
+      "loss": 0.8019,
+      "step": 32401
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.524588154118203,
+      "learning_rate": 2.7958681975821767e-06,
+      "loss": 0.8451,
+      "step": 32402
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6844105653068997,
+      "learning_rate": 2.795339015142364e-06,
+      "loss": 0.7216,
+      "step": 32403
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8542572467172662,
+      "learning_rate": 2.794809874650367e-06,
+      "loss": 0.8525,
+      "step": 32404
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6435044418580336,
+      "learning_rate": 2.7942807761092704e-06,
+      "loss": 0.901,
+      "step": 32405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5441457876620768,
+      "learning_rate": 2.7937517195221498e-06,
+      "loss": 0.7999,
+      "step": 32406
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.639246303684726,
+      "learning_rate": 2.79322270489209e-06,
+      "loss": 0.7916,
+      "step": 32407
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7768063113709363,
+      "learning_rate": 2.7926937322221705e-06,
+      "loss": 0.8301,
+      "step": 32408
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.118215955690618,
+      "learning_rate": 2.7921648015154655e-06,
+      "loss": 0.8063,
+      "step": 32409
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4536503265162537,
+      "learning_rate": 2.7916359127750605e-06,
+      "loss": 0.7493,
+      "step": 32410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3905291547722618,
+      "learning_rate": 2.791107066004035e-06,
+      "loss": 0.8801,
+      "step": 32411
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.704997122360729,
+      "learning_rate": 2.7905782612054676e-06,
+      "loss": 0.7445,
+      "step": 32412
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4950280218793106,
+      "learning_rate": 2.790049498382432e-06,
+      "loss": 0.7966,
+      "step": 32413
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8870482377802928,
+      "learning_rate": 2.7895207775380116e-06,
+      "loss": 0.8394,
+      "step": 32414
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.691524602341267,
+      "learning_rate": 2.788992098675287e-06,
+      "loss": 0.6746,
+      "step": 32415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8902677902392901,
+      "learning_rate": 2.788463461797333e-06,
+      "loss": 0.9213,
+      "step": 32416
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.591368692733793,
+      "learning_rate": 2.787934866907225e-06,
+      "loss": 0.8132,
+      "step": 32417
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6973653445509205,
+      "learning_rate": 2.787406314008043e-06,
+      "loss": 0.8099,
+      "step": 32418
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.900214870051475,
+      "learning_rate": 2.7868778031028688e-06,
+      "loss": 0.7619,
+      "step": 32419
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7039830193535414,
+      "learning_rate": 2.7863493341947756e-06,
+      "loss": 0.671,
+      "step": 32420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6766263514566977,
+      "learning_rate": 2.785820907286837e-06,
+      "loss": 0.7449,
+      "step": 32421
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9254412660461317,
+      "learning_rate": 2.785292522382137e-06,
+      "loss": 0.8281,
+      "step": 32422
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4995684048981899,
+      "learning_rate": 2.7847641794837432e-06,
+      "loss": 0.8128,
+      "step": 32423
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7332340162833582,
+      "learning_rate": 2.784235878594741e-06,
+      "loss": 0.8938,
+      "step": 32424
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.516477652008726,
+      "learning_rate": 2.7837076197181987e-06,
+      "loss": 0.8611,
+      "step": 32425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4837227502381432,
+      "learning_rate": 2.783179402857198e-06,
+      "loss": 1.0002,
+      "step": 32426
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.016191552162309,
+      "learning_rate": 2.782651228014809e-06,
+      "loss": 0.7779,
+      "step": 32427
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5315504207042436,
+      "learning_rate": 2.7821230951941125e-06,
+      "loss": 0.9038,
+      "step": 32428
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6850502529672027,
+      "learning_rate": 2.781595004398181e-06,
+      "loss": 0.8617,
+      "step": 32429
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5358913371506102,
+      "learning_rate": 2.781066955630085e-06,
+      "loss": 0.8923,
+      "step": 32430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0272080771921948,
+      "learning_rate": 2.7805389488929025e-06,
+      "loss": 0.7464,
+      "step": 32431
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5760697736820135,
+      "learning_rate": 2.7800109841897105e-06,
+      "loss": 0.7038,
+      "step": 32432
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4538280693689651,
+      "learning_rate": 2.7794830615235812e-06,
+      "loss": 0.9062,
+      "step": 32433
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7603197476656143,
+      "learning_rate": 2.7789551808975843e-06,
+      "loss": 0.8886,
+      "step": 32434
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6203350219039585,
+      "learning_rate": 2.7784273423147966e-06,
+      "loss": 0.7889,
+      "step": 32435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.490007201473502,
+      "learning_rate": 2.777899545778294e-06,
+      "loss": 0.7252,
+      "step": 32436
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.770889121529183,
+      "learning_rate": 2.777371791291147e-06,
+      "loss": 0.7952,
+      "step": 32437
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4939351997542532,
+      "learning_rate": 2.7768440788564243e-06,
+      "loss": 0.7676,
+      "step": 32438
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6525814203075646,
+      "learning_rate": 2.7763164084772033e-06,
+      "loss": 0.802,
+      "step": 32439
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5375470205606994,
+      "learning_rate": 2.7757887801565577e-06,
+      "loss": 0.753,
+      "step": 32440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.600279095834102,
+      "learning_rate": 2.7752611938975573e-06,
+      "loss": 0.7989,
+      "step": 32441
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5915406889978254,
+      "learning_rate": 2.7747336497032697e-06,
+      "loss": 0.6574,
+      "step": 32442
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.710117627234131,
+      "learning_rate": 2.7742061475767735e-06,
+      "loss": 0.8689,
+      "step": 32443
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0434652970088742,
+      "learning_rate": 2.773678687521134e-06,
+      "loss": 0.8057,
+      "step": 32444
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6724589150121798,
+      "learning_rate": 2.773151269539428e-06,
+      "loss": 0.8188,
+      "step": 32445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.770601078244265,
+      "learning_rate": 2.7726238936347204e-06,
+      "loss": 0.771,
+      "step": 32446
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6101582518233533,
+      "learning_rate": 2.7720965598100867e-06,
+      "loss": 0.8984,
+      "step": 32447
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9095124262972092,
+      "learning_rate": 2.7715692680685933e-06,
+      "loss": 0.8592,
+      "step": 32448
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.318924021532783,
+      "learning_rate": 2.7710420184133148e-06,
+      "loss": 0.7777,
+      "step": 32449
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3634604691731338,
+      "learning_rate": 2.7705148108473177e-06,
+      "loss": 0.8553,
+      "step": 32450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4035956747046527,
+      "learning_rate": 2.769987645373671e-06,
+      "loss": 0.8603,
+      "step": 32451
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4885043044893747,
+      "learning_rate": 2.769460521995443e-06,
+      "loss": 0.8442,
+      "step": 32452
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5049730998048139,
+      "learning_rate": 2.7689334407157086e-06,
+      "loss": 0.9156,
+      "step": 32453
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.922389678568545,
+      "learning_rate": 2.7684064015375333e-06,
+      "loss": 0.8053,
+      "step": 32454
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6545353688206594,
+      "learning_rate": 2.7678794044639835e-06,
+      "loss": 0.9189,
+      "step": 32455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5654315749544911,
+      "learning_rate": 2.7673524494981275e-06,
+      "loss": 1.0079,
+      "step": 32456
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5518626226591912,
+      "learning_rate": 2.7668255366430397e-06,
+      "loss": 0.9436,
+      "step": 32457
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6860242847282148,
+      "learning_rate": 2.7662986659017833e-06,
+      "loss": 0.9201,
+      "step": 32458
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.795949062359374,
+      "learning_rate": 2.7657718372774233e-06,
+      "loss": 0.8386,
+      "step": 32459
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.500378820179118,
+      "learning_rate": 2.7652450507730297e-06,
+      "loss": 0.8544,
+      "step": 32460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5389945615726695,
+      "learning_rate": 2.764718306391673e-06,
+      "loss": 0.8977,
+      "step": 32461
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6331326458011008,
+      "learning_rate": 2.764191604136417e-06,
+      "loss": 1.0304,
+      "step": 32462
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.565837691260766,
+      "learning_rate": 2.7636649440103247e-06,
+      "loss": 0.7417,
+      "step": 32463
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4755205514324938,
+      "learning_rate": 2.763138326016467e-06,
+      "loss": 0.831,
+      "step": 32464
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.502915057877333,
+      "learning_rate": 2.762611750157912e-06,
+      "loss": 0.8992,
+      "step": 32465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5670570501072896,
+      "learning_rate": 2.762085216437722e-06,
+      "loss": 0.786,
+      "step": 32466
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7320379545746254,
+      "learning_rate": 2.7615587248589614e-06,
+      "loss": 0.8898,
+      "step": 32467
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.450448961584737,
+      "learning_rate": 2.7610322754247e-06,
+      "loss": 0.8206,
+      "step": 32468
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.096187769197245,
+      "learning_rate": 2.760505868137997e-06,
+      "loss": 0.8785,
+      "step": 32469
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5428593350046942,
+      "learning_rate": 2.7599795030019237e-06,
+      "loss": 0.7663,
+      "step": 32470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.651013290778896,
+      "learning_rate": 2.75945318001954e-06,
+      "loss": 0.8272,
+      "step": 32471
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4397728559350254,
+      "learning_rate": 2.7589268991939145e-06,
+      "loss": 0.78,
+      "step": 32472
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8343927011067196,
+      "learning_rate": 2.758400660528107e-06,
+      "loss": 0.7235,
+      "step": 32473
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4867277739091116,
+      "learning_rate": 2.757874464025182e-06,
+      "loss": 0.7787,
+      "step": 32474
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7251208362107144,
+      "learning_rate": 2.7573483096882114e-06,
+      "loss": 0.8177,
+      "step": 32475
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5063484149773294,
+      "learning_rate": 2.7568221975202468e-06,
+      "loss": 0.9526,
+      "step": 32476
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5702325778677462,
+      "learning_rate": 2.756296127524355e-06,
+      "loss": 0.8106,
+      "step": 32477
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4896736471044079,
+      "learning_rate": 2.755770099703602e-06,
+      "loss": 0.8287,
+      "step": 32478
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3917774514254324,
+      "learning_rate": 2.755244114061054e-06,
+      "loss": 0.8462,
+      "step": 32479
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8188208463792295,
+      "learning_rate": 2.7547181705997626e-06,
+      "loss": 0.81,
+      "step": 32480
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5228275155351765,
+      "learning_rate": 2.754192269322795e-06,
+      "loss": 0.6464,
+      "step": 32481
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4490073630576663,
+      "learning_rate": 2.7536664102332177e-06,
+      "loss": 0.8323,
+      "step": 32482
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7286842305716392,
+      "learning_rate": 2.7531405933340894e-06,
+      "loss": 0.8108,
+      "step": 32483
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.567184538217592,
+      "learning_rate": 2.7526148186284674e-06,
+      "loss": 0.7205,
+      "step": 32484
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.614810309511677,
+      "learning_rate": 2.752089086119416e-06,
+      "loss": 0.7727,
+      "step": 32485
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4573204504206994,
+      "learning_rate": 2.7515633958099998e-06,
+      "loss": 0.8049,
+      "step": 32486
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5580356624701732,
+      "learning_rate": 2.7510377477032735e-06,
+      "loss": 0.8133,
+      "step": 32487
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8286167225595238,
+      "learning_rate": 2.750512141802304e-06,
+      "loss": 0.8425,
+      "step": 32488
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5909619812551088,
+      "learning_rate": 2.7499865781101466e-06,
+      "loss": 0.8367,
+      "step": 32489
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7539332594907873,
+      "learning_rate": 2.7494610566298608e-06,
+      "loss": 0.7958,
+      "step": 32490
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5642831313353125,
+      "learning_rate": 2.7489355773645078e-06,
+      "loss": 0.7864,
+      "step": 32491
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7957621357087878,
+      "learning_rate": 2.7484101403171505e-06,
+      "loss": 0.7746,
+      "step": 32492
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5436607488953122,
+      "learning_rate": 2.7478847454908454e-06,
+      "loss": 0.823,
+      "step": 32493
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6880821483002768,
+      "learning_rate": 2.747359392888648e-06,
+      "loss": 0.8427,
+      "step": 32494
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7697117876863501,
+      "learning_rate": 2.74683408251362e-06,
+      "loss": 0.7333,
+      "step": 32495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.401159453135072,
+      "learning_rate": 2.7463088143688234e-06,
+      "loss": 0.7578,
+      "step": 32496
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6875157071477815,
+      "learning_rate": 2.7457835884573137e-06,
+      "loss": 0.7968,
+      "step": 32497
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9400334475453156,
+      "learning_rate": 2.7452584047821452e-06,
+      "loss": 0.8647,
+      "step": 32498
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.541097421988596,
+      "learning_rate": 2.7447332633463795e-06,
+      "loss": 0.7889,
+      "step": 32499
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8639055002723508,
+      "learning_rate": 2.744208164153077e-06,
+      "loss": 0.8275,
+      "step": 32500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6455825217419466,
+      "learning_rate": 2.7436831072052915e-06,
+      "loss": 0.7104,
+      "step": 32501
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7138480336356574,
+      "learning_rate": 2.7431580925060762e-06,
+      "loss": 0.9116,
+      "step": 32502
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8020995495838503,
+      "learning_rate": 2.7426331200584966e-06,
+      "loss": 0.7566,
+      "step": 32503
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7497508155357461,
+      "learning_rate": 2.742108189865601e-06,
+      "loss": 0.7142,
+      "step": 32504
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5145913804843856,
+      "learning_rate": 2.7415833019304517e-06,
+      "loss": 0.8777,
+      "step": 32505
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6560018895058202,
+      "learning_rate": 2.7410584562561004e-06,
+      "loss": 0.7595,
+      "step": 32506
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5275290508576862,
+      "learning_rate": 2.7405336528456073e-06,
+      "loss": 0.7886,
+      "step": 32507
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5587226617661325,
+      "learning_rate": 2.740008891702023e-06,
+      "loss": 0.8971,
+      "step": 32508
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6247315710575256,
+      "learning_rate": 2.7394841728284083e-06,
+      "loss": 0.8013,
+      "step": 32509
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0334247187710854,
+      "learning_rate": 2.738959496227814e-06,
+      "loss": 0.7904,
+      "step": 32510
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.535736219541319,
+      "learning_rate": 2.738434861903295e-06,
+      "loss": 0.8322,
+      "step": 32511
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7195979164967528,
+      "learning_rate": 2.7379102698579075e-06,
+      "loss": 0.9537,
+      "step": 32512
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.489524975443844,
+      "learning_rate": 2.7373857200947073e-06,
+      "loss": 0.8944,
+      "step": 32513
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5375516590797222,
+      "learning_rate": 2.736861212616747e-06,
+      "loss": 0.858,
+      "step": 32514
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5966343645420573,
+      "learning_rate": 2.7363367474270774e-06,
+      "loss": 0.8762,
+      "step": 32515
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8111805952830679,
+      "learning_rate": 2.7358123245287548e-06,
+      "loss": 0.7175,
+      "step": 32516
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6793606409596358,
+      "learning_rate": 2.7352879439248358e-06,
+      "loss": 0.9939,
+      "step": 32517
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.063737999173893,
+      "learning_rate": 2.73476360561837e-06,
+      "loss": 0.744,
+      "step": 32518
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7509814984304479,
+      "learning_rate": 2.734239309612409e-06,
+      "loss": 0.6748,
+      "step": 32519
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.428436295596562,
+      "learning_rate": 2.7337150559100055e-06,
+      "loss": 0.7957,
+      "step": 32520
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5151473441674945,
+      "learning_rate": 2.7331908445142165e-06,
+      "loss": 0.823,
+      "step": 32521
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6240225281077734,
+      "learning_rate": 2.7326666754280918e-06,
+      "loss": 0.8257,
+      "step": 32522
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5281542356431876,
+      "learning_rate": 2.7321425486546792e-06,
+      "loss": 0.7678,
+      "step": 32523
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.605450130691713,
+      "learning_rate": 2.7316184641970367e-06,
+      "loss": 0.8035,
+      "step": 32524
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6074179521875789,
+      "learning_rate": 2.7310944220582092e-06,
+      "loss": 0.7988,
+      "step": 32525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6971889844198849,
+      "learning_rate": 2.7305704222412543e-06,
+      "loss": 0.88,
+      "step": 32526
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.268841558730516,
+      "learning_rate": 2.7300464647492177e-06,
+      "loss": 0.825,
+      "step": 32527
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.537411093902003,
+      "learning_rate": 2.729522549585154e-06,
+      "loss": 0.7676,
+      "step": 32528
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7366440071526994,
+      "learning_rate": 2.7289986767521105e-06,
+      "loss": 0.7244,
+      "step": 32529
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.588743307110418,
+      "learning_rate": 2.7284748462531394e-06,
+      "loss": 0.8267,
+      "step": 32530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7000649837977533,
+      "learning_rate": 2.7279510580912916e-06,
+      "loss": 0.8057,
+      "step": 32531
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6507455054536537,
+      "learning_rate": 2.7274273122696118e-06,
+      "loss": 0.7825,
+      "step": 32532
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6011825562243225,
+      "learning_rate": 2.7269036087911525e-06,
+      "loss": 0.7213,
+      "step": 32533
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5017429097423816,
+      "learning_rate": 2.726379947658966e-06,
+      "loss": 0.8385,
+      "step": 32534
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9076527081835608,
+      "learning_rate": 2.7258563288760976e-06,
+      "loss": 0.8439,
+      "step": 32535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7356620086007446,
+      "learning_rate": 2.725332752445594e-06,
+      "loss": 0.9001,
+      "step": 32536
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0285734363407046,
+      "learning_rate": 2.7248092183705065e-06,
+      "loss": 0.8583,
+      "step": 32537
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6173629581875675,
+      "learning_rate": 2.7242857266538857e-06,
+      "loss": 0.9128,
+      "step": 32538
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5944943884641585,
+      "learning_rate": 2.7237622772987772e-06,
+      "loss": 0.812,
+      "step": 32539
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7610442578485423,
+      "learning_rate": 2.7232388703082258e-06,
+      "loss": 0.7004,
+      "step": 32540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7843848883094469,
+      "learning_rate": 2.7227155056852806e-06,
+      "loss": 0.9016,
+      "step": 32541
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5810985703597293,
+      "learning_rate": 2.7221921834329923e-06,
+      "loss": 0.9212,
+      "step": 32542
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4965605297457625,
+      "learning_rate": 2.7216689035544064e-06,
+      "loss": 0.8855,
+      "step": 32543
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.341282669054127,
+      "learning_rate": 2.7211456660525647e-06,
+      "loss": 0.7419,
+      "step": 32544
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6301188643533933,
+      "learning_rate": 2.7206224709305206e-06,
+      "loss": 0.7122,
+      "step": 32545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5445651161723466,
+      "learning_rate": 2.7200993181913137e-06,
+      "loss": 0.916,
+      "step": 32546
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6872527971199893,
+      "learning_rate": 2.719576207837997e-06,
+      "loss": 0.8067,
+      "step": 32547
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6163069139028994,
+      "learning_rate": 2.7190531398736098e-06,
+      "loss": 0.7838,
+      "step": 32548
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.738930530922402,
+      "learning_rate": 2.718530114301202e-06,
+      "loss": 0.7359,
+      "step": 32549
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0563659312727434,
+      "learning_rate": 2.7180071311238152e-06,
+      "loss": 0.9013,
+      "step": 32550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6070778627887143,
+      "learning_rate": 2.717484190344499e-06,
+      "loss": 0.9037,
+      "step": 32551
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6386595083393523,
+      "learning_rate": 2.716961291966296e-06,
+      "loss": 0.9397,
+      "step": 32552
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7810391964514827,
+      "learning_rate": 2.7164384359922467e-06,
+      "loss": 0.7669,
+      "step": 32553
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.572817406860969,
+      "learning_rate": 2.715915622425398e-06,
+      "loss": 0.8132,
+      "step": 32554
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1275002179707516,
+      "learning_rate": 2.715392851268799e-06,
+      "loss": 0.9395,
+      "step": 32555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6042292933535516,
+      "learning_rate": 2.714870122525488e-06,
+      "loss": 0.8887,
+      "step": 32556
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7962061910420553,
+      "learning_rate": 2.7143474361985065e-06,
+      "loss": 0.7767,
+      "step": 32557
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5803535132882052,
+      "learning_rate": 2.7138247922909024e-06,
+      "loss": 0.8313,
+      "step": 32558
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5411633258614514,
+      "learning_rate": 2.713302190805719e-06,
+      "loss": 0.8018,
+      "step": 32559
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.622012616725581,
+      "learning_rate": 2.7127796317459976e-06,
+      "loss": 0.8054,
+      "step": 32560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5901812584820332,
+      "learning_rate": 2.7122571151147757e-06,
+      "loss": 0.888,
+      "step": 32561
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7631889725038954,
+      "learning_rate": 2.7117346409151023e-06,
+      "loss": 0.8974,
+      "step": 32562
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6294490467430782,
+      "learning_rate": 2.7112122091500194e-06,
+      "loss": 0.7681,
+      "step": 32563
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7289255358825778,
+      "learning_rate": 2.710689819822566e-06,
+      "loss": 0.6976,
+      "step": 32564
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.775616754644538,
+      "learning_rate": 2.7101674729357817e-06,
+      "loss": 0.7156,
+      "step": 32565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5985387374724904,
+      "learning_rate": 2.7096451684927138e-06,
+      "loss": 0.9413,
+      "step": 32566
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5519147417190307,
+      "learning_rate": 2.7091229064963964e-06,
+      "loss": 0.831,
+      "step": 32567
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.709082716837125,
+      "learning_rate": 2.708600686949876e-06,
+      "loss": 0.845,
+      "step": 32568
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5889289128271953,
+      "learning_rate": 2.708078509856189e-06,
+      "loss": 0.8786,
+      "step": 32569
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5726908211756334,
+      "learning_rate": 2.70755637521838e-06,
+      "loss": 0.8166,
+      "step": 32570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.641817037339243,
+      "learning_rate": 2.7070342830394836e-06,
+      "loss": 0.8564,
+      "step": 32571
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.575245346478918,
+      "learning_rate": 2.7065122333225426e-06,
+      "loss": 0.7878,
+      "step": 32572
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.756562186032002,
+      "learning_rate": 2.705990226070602e-06,
+      "loss": 0.7816,
+      "step": 32573
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5322113243011473,
+      "learning_rate": 2.70546826128669e-06,
+      "loss": 0.7226,
+      "step": 32574
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5225702578606324,
+      "learning_rate": 2.7049463389738506e-06,
+      "loss": 0.8592,
+      "step": 32575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5135544725618852,
+      "learning_rate": 2.704424459135123e-06,
+      "loss": 0.8789,
+      "step": 32576
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5982093223890392,
+      "learning_rate": 2.703902621773553e-06,
+      "loss": 0.9065,
+      "step": 32577
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.481262969373208,
+      "learning_rate": 2.7033808268921658e-06,
+      "loss": 0.8674,
+      "step": 32578
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7721950335164138,
+      "learning_rate": 2.702859074494004e-06,
+      "loss": 0.818,
+      "step": 32579
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.381207564117187,
+      "learning_rate": 2.7023373645821115e-06,
+      "loss": 0.9389,
+      "step": 32580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.156375133351349,
+      "learning_rate": 2.7018156971595177e-06,
+      "loss": 0.7961,
+      "step": 32581
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6812310748121957,
+      "learning_rate": 2.701294072229267e-06,
+      "loss": 0.8787,
+      "step": 32582
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.459454443557736,
+      "learning_rate": 2.7007724897943887e-06,
+      "loss": 0.7906,
+      "step": 32583
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.594703863991828,
+      "learning_rate": 2.7002509498579275e-06,
+      "loss": 0.8228,
+      "step": 32584
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6364239319441556,
+      "learning_rate": 2.6997294524229145e-06,
+      "loss": 0.9207,
+      "step": 32585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6442939520270823,
+      "learning_rate": 2.699207997492389e-06,
+      "loss": 0.8027,
+      "step": 32586
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4157346872227161,
+      "learning_rate": 2.6986865850693877e-06,
+      "loss": 0.9687,
+      "step": 32587
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7308941147390258,
+      "learning_rate": 2.6981652151569404e-06,
+      "loss": 0.9187,
+      "step": 32588
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6636787563091198,
+      "learning_rate": 2.6976438877580867e-06,
+      "loss": 0.9388,
+      "step": 32589
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6841393864283534,
+      "learning_rate": 2.6971226028758655e-06,
+      "loss": 0.7951,
+      "step": 32590
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.626577728667814,
+      "learning_rate": 2.696601360513309e-06,
+      "loss": 0.8509,
+      "step": 32591
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8934752510249397,
+      "learning_rate": 2.6960801606734487e-06,
+      "loss": 0.8115,
+      "step": 32592
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6294604957343972,
+      "learning_rate": 2.6955590033593217e-06,
+      "loss": 0.8475,
+      "step": 32593
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6216278235184898,
+      "learning_rate": 2.695037888573966e-06,
+      "loss": 0.9048,
+      "step": 32594
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7023750808193556,
+      "learning_rate": 2.6945168163204116e-06,
+      "loss": 0.7917,
+      "step": 32595
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8825874817867296,
+      "learning_rate": 2.693995786601691e-06,
+      "loss": 0.7299,
+      "step": 32596
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6208791154435942,
+      "learning_rate": 2.6934747994208397e-06,
+      "loss": 0.8193,
+      "step": 32597
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8407083876219086,
+      "learning_rate": 2.692953854780894e-06,
+      "loss": 0.8609,
+      "step": 32598
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6272652130363603,
+      "learning_rate": 2.692432952684886e-06,
+      "loss": 0.7783,
+      "step": 32599
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7916757239543493,
+      "learning_rate": 2.691912093135842e-06,
+      "loss": 0.8179,
+      "step": 32600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5502918634072815,
+      "learning_rate": 2.6913912761368e-06,
+      "loss": 0.8221,
+      "step": 32601
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7353389344033652,
+      "learning_rate": 2.6908705016907956e-06,
+      "loss": 0.8965,
+      "step": 32602
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5597430385206181,
+      "learning_rate": 2.6903497698008564e-06,
+      "loss": 0.804,
+      "step": 32603
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5065905316255888,
+      "learning_rate": 2.689829080470011e-06,
+      "loss": 0.844,
+      "step": 32604
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5291179629060412,
+      "learning_rate": 2.6893084337013e-06,
+      "loss": 0.7957,
+      "step": 32605
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8138226109662643,
+      "learning_rate": 2.6887878294977456e-06,
+      "loss": 0.8452,
+      "step": 32606
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.641038501094703,
+      "learning_rate": 2.688267267862387e-06,
+      "loss": 0.9413,
+      "step": 32607
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5463381918654788,
+      "learning_rate": 2.687746748798248e-06,
+      "loss": 0.9472,
+      "step": 32608
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9837604855634636,
+      "learning_rate": 2.687226272308365e-06,
+      "loss": 0.8211,
+      "step": 32609
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7601967344972813,
+      "learning_rate": 2.6867058383957633e-06,
+      "loss": 0.8536,
+      "step": 32610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7566022936067969,
+      "learning_rate": 2.6861854470634784e-06,
+      "loss": 0.9955,
+      "step": 32611
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5697798096024664,
+      "learning_rate": 2.685665098314537e-06,
+      "loss": 0.7848,
+      "step": 32612
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6216372453535637,
+      "learning_rate": 2.685144792151966e-06,
+      "loss": 0.793,
+      "step": 32613
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7924866291741461,
+      "learning_rate": 2.6846245285787985e-06,
+      "loss": 0.8455,
+      "step": 32614
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7031794978865424,
+      "learning_rate": 2.6841043075980653e-06,
+      "loss": 0.8243,
+      "step": 32615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.628811031298332,
+      "learning_rate": 2.683584129212794e-06,
+      "loss": 0.7623,
+      "step": 32616
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7679101627823903,
+      "learning_rate": 2.6830639934260084e-06,
+      "loss": 0.8171,
+      "step": 32617
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2012351989734196,
+      "learning_rate": 2.6825439002407405e-06,
+      "loss": 0.7456,
+      "step": 32618
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.610518807213718,
+      "learning_rate": 2.682023849660023e-06,
+      "loss": 0.8802,
+      "step": 32619
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6785788939410473,
+      "learning_rate": 2.681503841686879e-06,
+      "loss": 0.8509,
+      "step": 32620
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6035037135737111,
+      "learning_rate": 2.6809838763243325e-06,
+      "loss": 0.8103,
+      "step": 32621
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.450650088490838,
+      "learning_rate": 2.680463953575416e-06,
+      "loss": 0.846,
+      "step": 32622
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.723598636461957,
+      "learning_rate": 2.679944073443158e-06,
+      "loss": 0.8031,
+      "step": 32623
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7045291275779213,
+      "learning_rate": 2.6794242359305843e-06,
+      "loss": 0.8497,
+      "step": 32624
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4923373299144533,
+      "learning_rate": 2.6789044410407162e-06,
+      "loss": 0.9232,
+      "step": 32625
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5910591532507874,
+      "learning_rate": 2.6783846887765886e-06,
+      "loss": 0.765,
+      "step": 32626
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5689879313375656,
+      "learning_rate": 2.6778649791412194e-06,
+      "loss": 0.7665,
+      "step": 32627
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.877241561159656,
+      "learning_rate": 2.6773453121376414e-06,
+      "loss": 0.865,
+      "step": 32628
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7070743656926226,
+      "learning_rate": 2.6768256877688735e-06,
+      "loss": 0.7736,
+      "step": 32629
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8711175896839607,
+      "learning_rate": 2.6763061060379492e-06,
+      "loss": 0.921,
+      "step": 32630
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8066301658688468,
+      "learning_rate": 2.6757865669478854e-06,
+      "loss": 0.8593,
+      "step": 32631
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.542589821943814,
+      "learning_rate": 2.6752670705017148e-06,
+      "loss": 0.6499,
+      "step": 32632
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5868036653192172,
+      "learning_rate": 2.6747476167024587e-06,
+      "loss": 0.9004,
+      "step": 32633
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4634457185483805,
+      "learning_rate": 2.674228205553138e-06,
+      "loss": 0.8564,
+      "step": 32634
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4280752200147546,
+      "learning_rate": 2.673708837056779e-06,
+      "loss": 0.79,
+      "step": 32635
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.718832861023259,
+      "learning_rate": 2.6731895112164096e-06,
+      "loss": 0.7495,
+      "step": 32636
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5308257738656024,
+      "learning_rate": 2.6726702280350513e-06,
+      "loss": 0.941,
+      "step": 32637
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5717884315450417,
+      "learning_rate": 2.672150987515724e-06,
+      "loss": 0.8208,
+      "step": 32638
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7855515449685986,
+      "learning_rate": 2.671631789661453e-06,
+      "loss": 0.8572,
+      "step": 32639
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6406864595396695,
+      "learning_rate": 2.6711126344752647e-06,
+      "loss": 0.9276,
+      "step": 32640
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.754123606515701,
+      "learning_rate": 2.6705935219601788e-06,
+      "loss": 0.7311,
+      "step": 32641
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.632615922065591,
+      "learning_rate": 2.670074452119216e-06,
+      "loss": 0.7712,
+      "step": 32642
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6001054555451424,
+      "learning_rate": 2.669555424955399e-06,
+      "loss": 0.8748,
+      "step": 32643
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.528087323131028,
+      "learning_rate": 2.669036440471754e-06,
+      "loss": 0.9732,
+      "step": 32644
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7759462200430781,
+      "learning_rate": 2.6685174986713012e-06,
+      "loss": 0.7425,
+      "step": 32645
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7597493666830324,
+      "learning_rate": 2.667998599557057e-06,
+      "loss": 0.7173,
+      "step": 32646
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.64149027581005,
+      "learning_rate": 2.667479743132049e-06,
+      "loss": 0.9278,
+      "step": 32647
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.474129788559318,
+      "learning_rate": 2.6669609293992914e-06,
+      "loss": 0.8318,
+      "step": 32648
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5081373870958954,
+      "learning_rate": 2.6664421583618138e-06,
+      "loss": 0.7454,
+      "step": 32649
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4810199741073764,
+      "learning_rate": 2.665923430022628e-06,
+      "loss": 0.7738,
+      "step": 32650
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6878803245100304,
+      "learning_rate": 2.6654047443847607e-06,
+      "loss": 0.769,
+      "step": 32651
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5482680933582258,
+      "learning_rate": 2.6648861014512263e-06,
+      "loss": 0.851,
+      "step": 32652
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8590728195126474,
+      "learning_rate": 2.664367501225049e-06,
+      "loss": 0.9068,
+      "step": 32653
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5740599485055893,
+      "learning_rate": 2.6638489437092486e-06,
+      "loss": 1.0539,
+      "step": 32654
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7129414094774187,
+      "learning_rate": 2.6633304289068384e-06,
+      "loss": 0.8407,
+      "step": 32655
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5133411564164694,
+      "learning_rate": 2.662811956820841e-06,
+      "loss": 0.7823,
+      "step": 32656
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5162598787642203,
+      "learning_rate": 2.662293527454275e-06,
+      "loss": 0.8344,
+      "step": 32657
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7365726074320043,
+      "learning_rate": 2.6617751408101655e-06,
+      "loss": 0.6762,
+      "step": 32658
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6525033947379146,
+      "learning_rate": 2.661256796891518e-06,
+      "loss": 0.8716,
+      "step": 32659
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7766876753551136,
+      "learning_rate": 2.6607384957013583e-06,
+      "loss": 0.6896,
+      "step": 32660
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7450348234779463,
+      "learning_rate": 2.6602202372427046e-06,
+      "loss": 0.8804,
+      "step": 32661
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7954540243760062,
+      "learning_rate": 2.659702021518572e-06,
+      "loss": 0.8745,
+      "step": 32662
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7175647960686915,
+      "learning_rate": 2.6591838485319754e-06,
+      "loss": 0.8531,
+      "step": 32663
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6242884507616338,
+      "learning_rate": 2.658665718285934e-06,
+      "loss": 0.8773,
+      "step": 32664
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.802795320564319,
+      "learning_rate": 2.6581476307834697e-06,
+      "loss": 0.7751,
+      "step": 32665
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4353739666364684,
+      "learning_rate": 2.6576295860275927e-06,
+      "loss": 0.8828,
+      "step": 32666
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.05815336580643,
+      "learning_rate": 2.657111584021318e-06,
+      "loss": 0.9069,
+      "step": 32667
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7527083342576893,
+      "learning_rate": 2.6565936247676684e-06,
+      "loss": 0.7059,
+      "step": 32668
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5808447976065616,
+      "learning_rate": 2.656075708269652e-06,
+      "loss": 0.7535,
+      "step": 32669
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4813811531346721,
+      "learning_rate": 2.655557834530288e-06,
+      "loss": 0.836,
+      "step": 32670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8577714718894667,
+      "learning_rate": 2.6550400035525936e-06,
+      "loss": 0.8795,
+      "step": 32671
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9368165431762867,
+      "learning_rate": 2.6545222153395822e-06,
+      "loss": 0.9148,
+      "step": 32672
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8888227188089726,
+      "learning_rate": 2.6540044698942655e-06,
+      "loss": 0.6247,
+      "step": 32673
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6556586412169993,
+      "learning_rate": 2.6534867672196594e-06,
+      "loss": 0.948,
+      "step": 32674
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1370088494998307,
+      "learning_rate": 2.652969107318787e-06,
+      "loss": 0.9298,
+      "step": 32675
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5325241008156365,
+      "learning_rate": 2.6524514901946473e-06,
+      "loss": 0.7664,
+      "step": 32676
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7336707874870894,
+      "learning_rate": 2.651933915850261e-06,
+      "loss": 0.8335,
+      "step": 32677
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6556691709968998,
+      "learning_rate": 2.651416384288642e-06,
+      "loss": 0.7443,
+      "step": 32678
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4782822581582944,
+      "learning_rate": 2.6508988955128067e-06,
+      "loss": 0.7439,
+      "step": 32679
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6658031653008554,
+      "learning_rate": 2.6503814495257652e-06,
+      "loss": 0.8431,
+      "step": 32680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5482899041958185,
+      "learning_rate": 2.649864046330526e-06,
+      "loss": 0.8427,
+      "step": 32681
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.809490689879481,
+      "learning_rate": 2.6493466859301074e-06,
+      "loss": 1.0144,
+      "step": 32682
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8004297508298817,
+      "learning_rate": 2.6488293683275177e-06,
+      "loss": 0.6991,
+      "step": 32683
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8150233132265206,
+      "learning_rate": 2.6483120935257744e-06,
+      "loss": 0.7979,
+      "step": 32684
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7633614237581625,
+      "learning_rate": 2.6477948615278813e-06,
+      "loss": 0.7038,
+      "step": 32685
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7071314670171909,
+      "learning_rate": 2.6472776723368577e-06,
+      "loss": 0.7992,
+      "step": 32686
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5646277135591526,
+      "learning_rate": 2.646760525955708e-06,
+      "loss": 0.7985,
+      "step": 32687
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.538031208298346,
+      "learning_rate": 2.64624342238745e-06,
+      "loss": 0.6705,
+      "step": 32688
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7913839136664498,
+      "learning_rate": 2.6457263616350915e-06,
+      "loss": 0.727,
+      "step": 32689
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7948505050826826,
+      "learning_rate": 2.645209343701638e-06,
+      "loss": 0.7726,
+      "step": 32690
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5235043157236547,
+      "learning_rate": 2.644692368590105e-06,
+      "loss": 0.9581,
+      "step": 32691
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.823985610583316,
+      "learning_rate": 2.6441754363035045e-06,
+      "loss": 0.9174,
+      "step": 32692
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9390273045090904,
+      "learning_rate": 2.6436585468448438e-06,
+      "loss": 0.8134,
+      "step": 32693
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4639555523571608,
+      "learning_rate": 2.643141700217129e-06,
+      "loss": 0.7975,
+      "step": 32694
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.248673521023939,
+      "learning_rate": 2.6426248964233725e-06,
+      "loss": 0.7945,
+      "step": 32695
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.762426738623155,
+      "learning_rate": 2.6421081354665856e-06,
+      "loss": 0.8905,
+      "step": 32696
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7199270689934278,
+      "learning_rate": 2.6415914173497746e-06,
+      "loss": 0.852,
+      "step": 32697
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.688868936274695,
+      "learning_rate": 2.6410747420759455e-06,
+      "loss": 0.8786,
+      "step": 32698
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.214627445811065,
+      "learning_rate": 2.6405581096481083e-06,
+      "loss": 0.7953,
+      "step": 32699
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.536159403652508,
+      "learning_rate": 2.6400415200692754e-06,
+      "loss": 0.8272,
+      "step": 32700
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.847957422310279,
+      "learning_rate": 2.63952497334245e-06,
+      "loss": 0.8006,
+      "step": 32701
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6046582900818689,
+      "learning_rate": 2.639008469470639e-06,
+      "loss": 0.8492,
+      "step": 32702
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4642603003927381,
+      "learning_rate": 2.638492008456853e-06,
+      "loss": 0.8471,
+      "step": 32703
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.718982498725891,
+      "learning_rate": 2.6379755903040947e-06,
+      "loss": 0.8839,
+      "step": 32704
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7050346735137547,
+      "learning_rate": 2.6374592150153755e-06,
+      "loss": 0.8702,
+      "step": 32705
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.65754110201127,
+      "learning_rate": 2.6369428825936973e-06,
+      "loss": 0.8257,
+      "step": 32706
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8302629571282492,
+      "learning_rate": 2.636426593042072e-06,
+      "loss": 0.7747,
+      "step": 32707
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6410409613995374,
+      "learning_rate": 2.635910346363499e-06,
+      "loss": 0.8772,
+      "step": 32708
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.575530776930972,
+      "learning_rate": 2.6353941425609907e-06,
+      "loss": 0.921,
+      "step": 32709
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5586995698422452,
+      "learning_rate": 2.634877981637548e-06,
+      "loss": 0.7851,
+      "step": 32710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9977298357122015,
+      "learning_rate": 2.6343618635961765e-06,
+      "loss": 0.9021,
+      "step": 32711
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6281843786402583,
+      "learning_rate": 2.6338457884398805e-06,
+      "loss": 0.806,
+      "step": 32712
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9889009544176293,
+      "learning_rate": 2.6333297561716696e-06,
+      "loss": 0.7807,
+      "step": 32713
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6837686266846512,
+      "learning_rate": 2.6328137667945463e-06,
+      "loss": 0.8338,
+      "step": 32714
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7795051575009455,
+      "learning_rate": 2.6322978203115102e-06,
+      "loss": 0.6544,
+      "step": 32715
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.922462206654627,
+      "learning_rate": 2.6317819167255676e-06,
+      "loss": 0.8118,
+      "step": 32716
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5189551586840546,
+      "learning_rate": 2.6312660560397273e-06,
+      "loss": 0.8825,
+      "step": 32717
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.768923648450779,
+      "learning_rate": 2.6307502382569894e-06,
+      "loss": 0.7862,
+      "step": 32718
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8026385389681807,
+      "learning_rate": 2.6302344633803534e-06,
+      "loss": 0.8531,
+      "step": 32719
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6875282148024056,
+      "learning_rate": 2.6297187314128247e-06,
+      "loss": 0.7568,
+      "step": 32720
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5064891177059907,
+      "learning_rate": 2.629203042357411e-06,
+      "loss": 0.783,
+      "step": 32721
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5743041579222194,
+      "learning_rate": 2.62868739621711e-06,
+      "loss": 0.7779,
+      "step": 32722
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4852438040525284,
+      "learning_rate": 2.6281717929949226e-06,
+      "loss": 0.7777,
+      "step": 32723
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7949243995346114,
+      "learning_rate": 2.6276562326938516e-06,
+      "loss": 0.6605,
+      "step": 32724
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.654373024971895,
+      "learning_rate": 2.6271407153169036e-06,
+      "loss": 0.8204,
+      "step": 32725
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7503747596833914,
+      "learning_rate": 2.6266252408670776e-06,
+      "loss": 1.0364,
+      "step": 32726
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6387048190937719,
+      "learning_rate": 2.626109809347369e-06,
+      "loss": 0.8387,
+      "step": 32727
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4054948143708408,
+      "learning_rate": 2.6255944207607885e-06,
+      "loss": 0.7291,
+      "step": 32728
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5281214818561704,
+      "learning_rate": 2.6250790751103282e-06,
+      "loss": 0.8108,
+      "step": 32729
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4355897482443307,
+      "learning_rate": 2.6245637723989946e-06,
+      "loss": 0.757,
+      "step": 32730
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7898697121951392,
+      "learning_rate": 2.6240485126297834e-06,
+      "loss": 0.7104,
+      "step": 32731
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5836281342024583,
+      "learning_rate": 2.6235332958057004e-06,
+      "loss": 0.7651,
+      "step": 32732
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6811534637330903,
+      "learning_rate": 2.6230181219297378e-06,
+      "loss": 0.8914,
+      "step": 32733
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7052033908080007,
+      "learning_rate": 2.6225029910049037e-06,
+      "loss": 0.9318,
+      "step": 32734
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5066972295622425,
+      "learning_rate": 2.621987903034192e-06,
+      "loss": 0.8516,
+      "step": 32735
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.57029951795833,
+      "learning_rate": 2.6214728580205997e-06,
+      "loss": 0.8801,
+      "step": 32736
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6824648998737872,
+      "learning_rate": 2.620957855967128e-06,
+      "loss": 0.7297,
+      "step": 32737
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6416423100838544,
+      "learning_rate": 2.620442896876779e-06,
+      "loss": 0.8533,
+      "step": 32738
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6818907658613491,
+      "learning_rate": 2.619927980752548e-06,
+      "loss": 0.7835,
+      "step": 32739
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8403428748669868,
+      "learning_rate": 2.6194131075974294e-06,
+      "loss": 0.8196,
+      "step": 32740
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7902568471025226,
+      "learning_rate": 2.6188982774144245e-06,
+      "loss": 0.7126,
+      "step": 32741
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5748675694434517,
+      "learning_rate": 2.618383490206534e-06,
+      "loss": 0.7876,
+      "step": 32742
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7818148821076787,
+      "learning_rate": 2.6178687459767514e-06,
+      "loss": 0.8613,
+      "step": 32743
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.977637455282083,
+      "learning_rate": 2.617354044728072e-06,
+      "loss": 0.956,
+      "step": 32744
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7482798339045645,
+      "learning_rate": 2.6168393864634935e-06,
+      "loss": 0.6787,
+      "step": 32745
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.705515242104592,
+      "learning_rate": 2.6163247711860173e-06,
+      "loss": 0.8411,
+      "step": 32746
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7077887523369137,
+      "learning_rate": 2.615810198898635e-06,
+      "loss": 0.794,
+      "step": 32747
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8026477822887457,
+      "learning_rate": 2.615295669604342e-06,
+      "loss": 0.8969,
+      "step": 32748
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8290324731904617,
+      "learning_rate": 2.6147811833061376e-06,
+      "loss": 0.7334,
+      "step": 32749
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6414329731504707,
+      "learning_rate": 2.614266740007012e-06,
+      "loss": 0.8175,
+      "step": 32750
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4669929808153084,
+      "learning_rate": 2.6137523397099672e-06,
+      "loss": 0.8206,
+      "step": 32751
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5072790143884125,
+      "learning_rate": 2.6132379824179923e-06,
+      "loss": 0.7493,
+      "step": 32752
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4316944590981555,
+      "learning_rate": 2.612723668134086e-06,
+      "loss": 0.8008,
+      "step": 32753
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6135417457536079,
+      "learning_rate": 2.6122093968612393e-06,
+      "loss": 0.7894,
+      "step": 32754
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5853196709582027,
+      "learning_rate": 2.611695168602448e-06,
+      "loss": 0.9,
+      "step": 32755
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4990241474746364,
+      "learning_rate": 2.6111809833607126e-06,
+      "loss": 0.9582,
+      "step": 32756
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.530851436815641,
+      "learning_rate": 2.6106668411390156e-06,
+      "loss": 0.8343,
+      "step": 32757
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5679929048739196,
+      "learning_rate": 2.610152741940355e-06,
+      "loss": 0.8154,
+      "step": 32758
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.829247563015638,
+      "learning_rate": 2.609638685767725e-06,
+      "loss": 0.8411,
+      "step": 32759
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.007233043499222,
+      "learning_rate": 2.609124672624124e-06,
+      "loss": 0.7468,
+      "step": 32760
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.712031317464411,
+      "learning_rate": 2.608610702512533e-06,
+      "loss": 0.8172,
+      "step": 32761
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7975402681852121,
+      "learning_rate": 2.608096775435951e-06,
+      "loss": 0.6897,
+      "step": 32762
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8502741926952126,
+      "learning_rate": 2.607582891397372e-06,
+      "loss": 0.7175,
+      "step": 32763
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.642248224794716,
+      "learning_rate": 2.6070690503997866e-06,
+      "loss": 0.6992,
+      "step": 32764
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8371722061155848,
+      "learning_rate": 2.6065552524461823e-06,
+      "loss": 0.9266,
+      "step": 32765
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.050781422512865,
+      "learning_rate": 2.6060414975395543e-06,
+      "loss": 0.6419,
+      "step": 32766
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6601544605624148,
+      "learning_rate": 2.6055277856828964e-06,
+      "loss": 0.7435,
+      "step": 32767
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6696468052828095,
+      "learning_rate": 2.605014116879193e-06,
+      "loss": 0.9455,
+      "step": 32768
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8507768514791939,
+      "learning_rate": 2.604500491131443e-06,
+      "loss": 0.7508,
+      "step": 32769
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6997811442515232,
+      "learning_rate": 2.603986908442632e-06,
+      "loss": 0.8776,
+      "step": 32770
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5349890368973451,
+      "learning_rate": 2.6034733688157475e-06,
+      "loss": 0.7975,
+      "step": 32771
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8092256724195996,
+      "learning_rate": 2.602959872253783e-06,
+      "loss": 0.7817,
+      "step": 32772
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8671814327859666,
+      "learning_rate": 2.602446418759731e-06,
+      "loss": 0.7831,
+      "step": 32773
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8409315945180384,
+      "learning_rate": 2.6019330083365778e-06,
+      "loss": 0.8513,
+      "step": 32774
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.435367893846145,
+      "learning_rate": 2.6014196409873104e-06,
+      "loss": 0.8257,
+      "step": 32775
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7206730303449134,
+      "learning_rate": 2.6009063167149195e-06,
+      "loss": 0.853,
+      "step": 32776
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7718523927765457,
+      "learning_rate": 2.600393035522398e-06,
+      "loss": 0.926,
+      "step": 32777
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5281464990140678,
+      "learning_rate": 2.599879797412731e-06,
+      "loss": 0.7181,
+      "step": 32778
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6793544064058545,
+      "learning_rate": 2.599366602388903e-06,
+      "loss": 0.7798,
+      "step": 32779
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9399453409026863,
+      "learning_rate": 2.5988534504539065e-06,
+      "loss": 0.8693,
+      "step": 32780
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7831297121755993,
+      "learning_rate": 2.598340341610731e-06,
+      "loss": 0.8141,
+      "step": 32781
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.257253201529947,
+      "learning_rate": 2.597827275862361e-06,
+      "loss": 0.7828,
+      "step": 32782
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5984956439478835,
+      "learning_rate": 2.597314253211781e-06,
+      "loss": 0.8267,
+      "step": 32783
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5971242442176767,
+      "learning_rate": 2.596801273661985e-06,
+      "loss": 0.862,
+      "step": 32784
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0209219744254927,
+      "learning_rate": 2.596288337215952e-06,
+      "loss": 0.731,
+      "step": 32785
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8170753036892282,
+      "learning_rate": 2.595775443876676e-06,
+      "loss": 0.6851,
+      "step": 32786
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2639622133620487,
+      "learning_rate": 2.5952625936471354e-06,
+      "loss": 0.7574,
+      "step": 32787
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7348125035398387,
+      "learning_rate": 2.594749786530324e-06,
+      "loss": 0.8166,
+      "step": 32788
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5308455261951042,
+      "learning_rate": 2.5942370225292214e-06,
+      "loss": 0.7654,
+      "step": 32789
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.888964985186458,
+      "learning_rate": 2.593724301646817e-06,
+      "loss": 0.8172,
+      "step": 32790
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5191665600170952,
+      "learning_rate": 2.593211623886096e-06,
+      "loss": 0.801,
+      "step": 32791
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5531671781564553,
+      "learning_rate": 2.5926989892500376e-06,
+      "loss": 0.8648,
+      "step": 32792
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6590609187666663,
+      "learning_rate": 2.5921863977416316e-06,
+      "loss": 0.8829,
+      "step": 32793
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6838099241734317,
+      "learning_rate": 2.5916738493638638e-06,
+      "loss": 0.757,
+      "step": 32794
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9223463814648782,
+      "learning_rate": 2.591161344119717e-06,
+      "loss": 0.8284,
+      "step": 32795
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6410696877739381,
+      "learning_rate": 2.590648882012172e-06,
+      "loss": 0.7488,
+      "step": 32796
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.918600805373053,
+      "learning_rate": 2.5901364630442136e-06,
+      "loss": 0.8634,
+      "step": 32797
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7497708925715414,
+      "learning_rate": 2.5896240872188305e-06,
+      "loss": 0.8013,
+      "step": 32798
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5124145494142067,
+      "learning_rate": 2.589111754539002e-06,
+      "loss": 0.8376,
+      "step": 32799
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7426358158285767,
+      "learning_rate": 2.5885994650077086e-06,
+      "loss": 0.6879,
+      "step": 32800
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5992217280849217,
+      "learning_rate": 2.588087218627935e-06,
+      "loss": 0.9156,
+      "step": 32801
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6911188501161813,
+      "learning_rate": 2.587575015402667e-06,
+      "loss": 0.8373,
+      "step": 32802
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5211911359999069,
+      "learning_rate": 2.587062855334884e-06,
+      "loss": 0.9656,
+      "step": 32803
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.875122346214352,
+      "learning_rate": 2.5865507384275646e-06,
+      "loss": 0.8893,
+      "step": 32804
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7645234211534959,
+      "learning_rate": 2.586038664683698e-06,
+      "loss": 0.6655,
+      "step": 32805
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7665036271629018,
+      "learning_rate": 2.585526634106258e-06,
+      "loss": 0.8511,
+      "step": 32806
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8307199568762555,
+      "learning_rate": 2.585014646698232e-06,
+      "loss": 0.9501,
+      "step": 32807
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7184728330207157,
+      "learning_rate": 2.584502702462596e-06,
+      "loss": 0.844,
+      "step": 32808
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5647184312409388,
+      "learning_rate": 2.5839908014023353e-06,
+      "loss": 0.7955,
+      "step": 32809
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6245419128889458,
+      "learning_rate": 2.5834789435204245e-06,
+      "loss": 0.8996,
+      "step": 32810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6073189841318534,
+      "learning_rate": 2.582967128819851e-06,
+      "loss": 0.8245,
+      "step": 32811
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8755047938519491,
+      "learning_rate": 2.5824553573035914e-06,
+      "loss": 0.799,
+      "step": 32812
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6576753920810152,
+      "learning_rate": 2.581943628974621e-06,
+      "loss": 0.8117,
+      "step": 32813
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7228241505686586,
+      "learning_rate": 2.581431943835925e-06,
+      "loss": 0.8672,
+      "step": 32814
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5468249501032918,
+      "learning_rate": 2.5809203018904828e-06,
+      "loss": 0.7745,
+      "step": 32815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.985118662937876,
+      "learning_rate": 2.5804087031412705e-06,
+      "loss": 0.8398,
+      "step": 32816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7939517741157377,
+      "learning_rate": 2.579897147591266e-06,
+      "loss": 0.7228,
+      "step": 32817
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5905645902889765,
+      "learning_rate": 2.57938563524345e-06,
+      "loss": 0.9127,
+      "step": 32818
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4612225747688738,
+      "learning_rate": 2.578874166100802e-06,
+      "loss": 0.9805,
+      "step": 32819
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5719890508750958,
+      "learning_rate": 2.578362740166298e-06,
+      "loss": 0.7627,
+      "step": 32820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5206332899531492,
+      "learning_rate": 2.5778513574429143e-06,
+      "loss": 0.8758,
+      "step": 32821
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6752051319240033,
+      "learning_rate": 2.577340017933628e-06,
+      "loss": 0.859,
+      "step": 32822
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5615273882876761,
+      "learning_rate": 2.5768287216414224e-06,
+      "loss": 0.92,
+      "step": 32823
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4807394625043404,
+      "learning_rate": 2.5763174685692693e-06,
+      "loss": 0.9317,
+      "step": 32824
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5744845254374042,
+      "learning_rate": 2.5758062587201436e-06,
+      "loss": 0.6843,
+      "step": 32825
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5498472481812064,
+      "learning_rate": 2.5752950920970278e-06,
+      "loss": 0.9042,
+      "step": 32826
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.755079854990465,
+      "learning_rate": 2.574783968702891e-06,
+      "loss": 0.7743,
+      "step": 32827
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.029861000202127,
+      "learning_rate": 2.5742728885407153e-06,
+      "loss": 0.9769,
+      "step": 32828
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6049776625964123,
+      "learning_rate": 2.573761851613471e-06,
+      "loss": 0.9077,
+      "step": 32829
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7894419890484946,
+      "learning_rate": 2.57325085792414e-06,
+      "loss": 0.7055,
+      "step": 32830
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.678512993887734,
+      "learning_rate": 2.57273990747569e-06,
+      "loss": 0.8612,
+      "step": 32831
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5666456541617262,
+      "learning_rate": 2.5722290002711025e-06,
+      "loss": 0.8027,
+      "step": 32832
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5688321435311587,
+      "learning_rate": 2.57171813631335e-06,
+      "loss": 0.8858,
+      "step": 32833
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9831649485772456,
+      "learning_rate": 2.571207315605404e-06,
+      "loss": 0.9431,
+      "step": 32834
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7112067659022633,
+      "learning_rate": 2.5706965381502392e-06,
+      "loss": 0.7089,
+      "step": 32835
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.692493244953774,
+      "learning_rate": 2.5701858039508354e-06,
+      "loss": 0.8914,
+      "step": 32836
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5880686158464001,
+      "learning_rate": 2.5696751130101614e-06,
+      "loss": 0.8803,
+      "step": 32837
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4366481039978511,
+      "learning_rate": 2.5691644653311887e-06,
+      "loss": 0.8152,
+      "step": 32838
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7538061332928323,
+      "learning_rate": 2.5686538609168932e-06,
+      "loss": 0.715,
+      "step": 32839
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1320637060524223,
+      "learning_rate": 2.568143299770247e-06,
+      "loss": 0.8676,
+      "step": 32840
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5487457537567297,
+      "learning_rate": 2.5676327818942305e-06,
+      "loss": 0.7115,
+      "step": 32841
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2306682160886773,
+      "learning_rate": 2.567122307291803e-06,
+      "loss": 0.7049,
+      "step": 32842
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6762969629651383,
+      "learning_rate": 2.566611875965941e-06,
+      "loss": 0.814,
+      "step": 32843
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7666162560052487,
+      "learning_rate": 2.5661014879196223e-06,
+      "loss": 0.8941,
+      "step": 32844
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5174632645931998,
+      "learning_rate": 2.5655911431558145e-06,
+      "loss": 0.809,
+      "step": 32845
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7068359859844457,
+      "learning_rate": 2.565080841677484e-06,
+      "loss": 0.709,
+      "step": 32846
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6608791582350109,
+      "learning_rate": 2.564570583487609e-06,
+      "loss": 0.8432,
+      "step": 32847
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7241365416989893,
+      "learning_rate": 2.5640603685891597e-06,
+      "loss": 0.9294,
+      "step": 32848
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5894113391660323,
+      "learning_rate": 2.563550196985105e-06,
+      "loss": 0.8914,
+      "step": 32849
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6589012592974386,
+      "learning_rate": 2.5630400686784128e-06,
+      "loss": 0.8515,
+      "step": 32850
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.833951224457186,
+      "learning_rate": 2.5625299836720597e-06,
+      "loss": 0.7763,
+      "step": 32851
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7786020676441883,
+      "learning_rate": 2.5620199419690074e-06,
+      "loss": 0.7007,
+      "step": 32852
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4446158477420508,
+      "learning_rate": 2.56150994357223e-06,
+      "loss": 0.7474,
+      "step": 32853
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5304739542814096,
+      "learning_rate": 2.5609999884847003e-06,
+      "loss": 0.7096,
+      "step": 32854
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.612493845517012,
+      "learning_rate": 2.560490076709383e-06,
+      "loss": 0.8874,
+      "step": 32855
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5583659507108916,
+      "learning_rate": 2.559980208249245e-06,
+      "loss": 0.9204,
+      "step": 32856
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.641697599881944,
+      "learning_rate": 2.5594703831072586e-06,
+      "loss": 0.7905,
+      "step": 32857
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7358758991557048,
+      "learning_rate": 2.5589606012863968e-06,
+      "loss": 0.8283,
+      "step": 32858
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7766465819281427,
+      "learning_rate": 2.5584508627896155e-06,
+      "loss": 0.719,
+      "step": 32859
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6782297873747836,
+      "learning_rate": 2.5579411676198897e-06,
+      "loss": 0.8226,
+      "step": 32860
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9795786424139228,
+      "learning_rate": 2.5574315157801867e-06,
+      "loss": 0.8563,
+      "step": 32861
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7305232416733655,
+      "learning_rate": 2.55692190727348e-06,
+      "loss": 0.6827,
+      "step": 32862
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.64516669410839,
+      "learning_rate": 2.556412342102723e-06,
+      "loss": 0.8215,
+      "step": 32863
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6338905692276031,
+      "learning_rate": 2.5559028202708913e-06,
+      "loss": 0.9865,
+      "step": 32864
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6189252149251367,
+      "learning_rate": 2.555393341780953e-06,
+      "loss": 0.8873,
+      "step": 32865
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.506382167117885,
+      "learning_rate": 2.554883906635868e-06,
+      "loss": 0.8619,
+      "step": 32866
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.051528965704495,
+      "learning_rate": 2.554374514838609e-06,
+      "loss": 0.7872,
+      "step": 32867
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4414462356009294,
+      "learning_rate": 2.5538651663921367e-06,
+      "loss": 0.7829,
+      "step": 32868
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7566555756030933,
+      "learning_rate": 2.5533558612994213e-06,
+      "loss": 0.9399,
+      "step": 32869
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5550144727221131,
+      "learning_rate": 2.5528465995634223e-06,
+      "loss": 0.8187,
+      "step": 32870
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.564440706255999,
+      "learning_rate": 2.552337381187111e-06,
+      "loss": 0.7001,
+      "step": 32871
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.690473522411835,
+      "learning_rate": 2.5518282061734502e-06,
+      "loss": 0.7805,
+      "step": 32872
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1454120548945124,
+      "learning_rate": 2.5513190745254e-06,
+      "loss": 0.7476,
+      "step": 32873
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.256219256891183,
+      "learning_rate": 2.550809986245929e-06,
+      "loss": 0.9052,
+      "step": 32874
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5961287120365415,
+      "learning_rate": 2.5503009413380033e-06,
+      "loss": 0.8843,
+      "step": 32875
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4594214469652598,
+      "learning_rate": 2.5497919398045833e-06,
+      "loss": 0.8071,
+      "step": 32876
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8153781575432906,
+      "learning_rate": 2.549282981648631e-06,
+      "loss": 0.8557,
+      "step": 32877
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5147995778845846,
+      "learning_rate": 2.548774066873112e-06,
+      "loss": 0.8417,
+      "step": 32878
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5460658523846456,
+      "learning_rate": 2.548265195480992e-06,
+      "loss": 0.8192,
+      "step": 32879
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7435332325595967,
+      "learning_rate": 2.5477563674752327e-06,
+      "loss": 0.8303,
+      "step": 32880
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6275461929029877,
+      "learning_rate": 2.5472475828587906e-06,
+      "loss": 1.0199,
+      "step": 32881
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6301997845889302,
+      "learning_rate": 2.546738841634633e-06,
+      "loss": 0.8936,
+      "step": 32882
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7652847061759771,
+      "learning_rate": 2.5462301438057245e-06,
+      "loss": 0.7765,
+      "step": 32883
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6351931793621537,
+      "learning_rate": 2.545721489375024e-06,
+      "loss": 0.7665,
+      "step": 32884
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5718881855571247,
+      "learning_rate": 2.54521287834549e-06,
+      "loss": 0.7464,
+      "step": 32885
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7061064895129214,
+      "learning_rate": 2.5447043107200895e-06,
+      "loss": 0.8731,
+      "step": 32886
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6529294933281566,
+      "learning_rate": 2.5441957865017784e-06,
+      "loss": 0.7606,
+      "step": 32887
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4772743424966606,
+      "learning_rate": 2.543687305693523e-06,
+      "loss": 0.8388,
+      "step": 32888
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5122112297245247,
+      "learning_rate": 2.543178868298277e-06,
+      "loss": 0.8757,
+      "step": 32889
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.123120783993959,
+      "learning_rate": 2.542670474319008e-06,
+      "loss": 0.8273,
+      "step": 32890
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5478076245663208,
+      "learning_rate": 2.542162123758669e-06,
+      "loss": 0.7603,
+      "step": 32891
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0232627874030746,
+      "learning_rate": 2.5416538166202267e-06,
+      "loss": 0.7958,
+      "step": 32892
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6678304823618215,
+      "learning_rate": 2.5411455529066363e-06,
+      "loss": 0.8216,
+      "step": 32893
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4491216377263318,
+      "learning_rate": 2.5406373326208554e-06,
+      "loss": 0.8189,
+      "step": 32894
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6415677845442356,
+      "learning_rate": 2.5401291557658457e-06,
+      "loss": 0.9447,
+      "step": 32895
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6850466526053862,
+      "learning_rate": 2.5396210223445684e-06,
+      "loss": 0.8373,
+      "step": 32896
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7614405773963633,
+      "learning_rate": 2.5391129323599796e-06,
+      "loss": 0.8711,
+      "step": 32897
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8788359648183102,
+      "learning_rate": 2.5386048858150346e-06,
+      "loss": 0.9083,
+      "step": 32898
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5472759103447635,
+      "learning_rate": 2.5380968827126937e-06,
+      "loss": 0.8384,
+      "step": 32899
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.775741126311582,
+      "learning_rate": 2.5375889230559192e-06,
+      "loss": 0.8461,
+      "step": 32900
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6796406140721412,
+      "learning_rate": 2.5370810068476637e-06,
+      "loss": 0.8231,
+      "step": 32901
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4653245055931128,
+      "learning_rate": 2.536573134090883e-06,
+      "loss": 0.8016,
+      "step": 32902
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5013699962346343,
+      "learning_rate": 2.5360653047885352e-06,
+      "loss": 0.809,
+      "step": 32903
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.859223178674082,
+      "learning_rate": 2.535557518943582e-06,
+      "loss": 0.8405,
+      "step": 32904
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5392811703900295,
+      "learning_rate": 2.5350497765589765e-06,
+      "loss": 0.788,
+      "step": 32905
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5625054667980702,
+      "learning_rate": 2.5345420776376706e-06,
+      "loss": 0.9114,
+      "step": 32906
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.694105841591771,
+      "learning_rate": 2.5340344221826287e-06,
+      "loss": 0.8123,
+      "step": 32907
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8101355254298463,
+      "learning_rate": 2.533526810196798e-06,
+      "loss": 0.6765,
+      "step": 32908
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7583564893969102,
+      "learning_rate": 2.5330192416831413e-06,
+      "loss": 0.7068,
+      "step": 32909
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8342218761455045,
+      "learning_rate": 2.5325117166446065e-06,
+      "loss": 0.7461,
+      "step": 32910
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5580418685674988,
+      "learning_rate": 2.5320042350841577e-06,
+      "loss": 0.9089,
+      "step": 32911
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7161402384394127,
+      "learning_rate": 2.53149679700474e-06,
+      "loss": 0.853,
+      "step": 32912
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4417767580772691,
+      "learning_rate": 2.5309894024093163e-06,
+      "loss": 0.8462,
+      "step": 32913
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.710881031781218,
+      "learning_rate": 2.5304820513008364e-06,
+      "loss": 0.9003,
+      "step": 32914
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.557244530332757,
+      "learning_rate": 2.529974743682252e-06,
+      "loss": 0.8815,
+      "step": 32915
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.74029722228033,
+      "learning_rate": 2.52946747955652e-06,
+      "loss": 0.6569,
+      "step": 32916
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5705577272025664,
+      "learning_rate": 2.528960258926596e-06,
+      "loss": 0.8122,
+      "step": 32917
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4140403284831187,
+      "learning_rate": 2.52845308179543e-06,
+      "loss": 0.8312,
+      "step": 32918
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8692989961621467,
+      "learning_rate": 2.527945948165974e-06,
+      "loss": 0.7821,
+      "step": 32919
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6996923122635796,
+      "learning_rate": 2.5274388580411813e-06,
+      "loss": 0.9057,
+      "step": 32920
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.885341263331393,
+      "learning_rate": 2.5269318114240092e-06,
+      "loss": 0.845,
+      "step": 32921
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7696553182802881,
+      "learning_rate": 2.5264248083174047e-06,
+      "loss": 0.6639,
+      "step": 32922
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5481896050120993,
+      "learning_rate": 2.5259178487243197e-06,
+      "loss": 0.8383,
+      "step": 32923
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7611467900701938,
+      "learning_rate": 2.525410932647706e-06,
+      "loss": 0.8556,
+      "step": 32924
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3176425475513405,
+      "learning_rate": 2.5249040600905195e-06,
+      "loss": 0.7569,
+      "step": 32925
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4843896966876418,
+      "learning_rate": 2.5243972310557087e-06,
+      "loss": 0.8485,
+      "step": 32926
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.682046825952492,
+      "learning_rate": 2.523890445546221e-06,
+      "loss": 0.7748,
+      "step": 32927
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5548905928003753,
+      "learning_rate": 2.523383703565012e-06,
+      "loss": 0.9043,
+      "step": 32928
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6432819487762957,
+      "learning_rate": 2.5228770051150274e-06,
+      "loss": 0.8807,
+      "step": 32929
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7261292478254173,
+      "learning_rate": 2.5223703501992234e-06,
+      "loss": 0.6706,
+      "step": 32930
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5668984629970137,
+      "learning_rate": 2.521863738820545e-06,
+      "loss": 0.8163,
+      "step": 32931
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.04463042738263,
+      "learning_rate": 2.521357170981944e-06,
+      "loss": 0.7827,
+      "step": 32932
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6053746351329126,
+      "learning_rate": 2.520850646686368e-06,
+      "loss": 0.836,
+      "step": 32933
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7598461383990174,
+      "learning_rate": 2.52034416593677e-06,
+      "loss": 0.7302,
+      "step": 32934
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1162966609132314,
+      "learning_rate": 2.5198377287360954e-06,
+      "loss": 0.9355,
+      "step": 32935
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.542709780735885,
+      "learning_rate": 2.519331335087292e-06,
+      "loss": 0.8103,
+      "step": 32936
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8570619342172805,
+      "learning_rate": 2.5188249849933087e-06,
+      "loss": 0.7386,
+      "step": 32937
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7016474160643782,
+      "learning_rate": 2.5183186784570957e-06,
+      "loss": 0.7469,
+      "step": 32938
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7523323298264502,
+      "learning_rate": 2.5178124154816052e-06,
+      "loss": 0.7631,
+      "step": 32939
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.794362034619696,
+      "learning_rate": 2.5173061960697744e-06,
+      "loss": 0.8598,
+      "step": 32940
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7692705654104055,
+      "learning_rate": 2.516800020224556e-06,
+      "loss": 0.7761,
+      "step": 32941
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5974520247668784,
+      "learning_rate": 2.5162938879488996e-06,
+      "loss": 0.7996,
+      "step": 32942
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.786794145847017,
+      "learning_rate": 2.5157877992457493e-06,
+      "loss": 0.7052,
+      "step": 32943
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5964943978796777,
+      "learning_rate": 2.5152817541180484e-06,
+      "loss": 0.6847,
+      "step": 32944
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4643457830526287,
+      "learning_rate": 2.5147757525687477e-06,
+      "loss": 0.9359,
+      "step": 32945
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5146029214065857,
+      "learning_rate": 2.5142697946007955e-06,
+      "loss": 0.829,
+      "step": 32946
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.153951686871119,
+      "learning_rate": 2.5137638802171337e-06,
+      "loss": 0.8338,
+      "step": 32947
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4882709063679902,
+      "learning_rate": 2.513258009420707e-06,
+      "loss": 0.7947,
+      "step": 32948
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7200013201806208,
+      "learning_rate": 2.512752182214464e-06,
+      "loss": 0.9424,
+      "step": 32949
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9965458028712912,
+      "learning_rate": 2.5122463986013458e-06,
+      "loss": 0.8275,
+      "step": 32950
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7806750468474356,
+      "learning_rate": 2.5117406585842998e-06,
+      "loss": 0.7306,
+      "step": 32951
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7137986063820807,
+      "learning_rate": 2.5112349621662736e-06,
+      "loss": 0.864,
+      "step": 32952
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.708561924508719,
+      "learning_rate": 2.510729309350208e-06,
+      "loss": 0.8483,
+      "step": 32953
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6852511185353152,
+      "learning_rate": 2.510223700139044e-06,
+      "loss": 0.8578,
+      "step": 32954
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5927856705846022,
+      "learning_rate": 2.509718134535729e-06,
+      "loss": 0.7902,
+      "step": 32955
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8908371056699131,
+      "learning_rate": 2.5092126125432137e-06,
+      "loss": 0.8567,
+      "step": 32956
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.502179252534182,
+      "learning_rate": 2.5087071341644265e-06,
+      "loss": 0.8792,
+      "step": 32957
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6320887834292535,
+      "learning_rate": 2.5082016994023185e-06,
+      "loss": 0.8394,
+      "step": 32958
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5507261083799233,
+      "learning_rate": 2.5076963082598327e-06,
+      "loss": 0.9315,
+      "step": 32959
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.519798156824374,
+      "learning_rate": 2.507190960739917e-06,
+      "loss": 0.9018,
+      "step": 32960
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3554151566844976,
+      "learning_rate": 2.5066856568455012e-06,
+      "loss": 0.8608,
+      "step": 32961
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6124239387346053,
+      "learning_rate": 2.506180396579534e-06,
+      "loss": 0.824,
+      "step": 32962
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.530883885562992,
+      "learning_rate": 2.505675179944956e-06,
+      "loss": 0.79,
+      "step": 32963
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6073009462244594,
+      "learning_rate": 2.5051700069447136e-06,
+      "loss": 0.8225,
+      "step": 32964
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5265390222324098,
+      "learning_rate": 2.504664877581745e-06,
+      "loss": 0.778,
+      "step": 32965
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8328552155424902,
+      "learning_rate": 2.5041597918589867e-06,
+      "loss": 0.8611,
+      "step": 32966
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7628975352660707,
+      "learning_rate": 2.5036547497793863e-06,
+      "loss": 0.9432,
+      "step": 32967
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6500732804342324,
+      "learning_rate": 2.503149751345878e-06,
+      "loss": 0.7767,
+      "step": 32968
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.719678724283998,
+      "learning_rate": 2.5026447965614086e-06,
+      "loss": 0.7472,
+      "step": 32969
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9604550080671137,
+      "learning_rate": 2.5021398854289124e-06,
+      "loss": 0.7323,
+      "step": 32970
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.090529342745213,
+      "learning_rate": 2.501635017951334e-06,
+      "loss": 0.7468,
+      "step": 32971
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6991171925542048,
+      "learning_rate": 2.501130194131608e-06,
+      "loss": 0.8391,
+      "step": 32972
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5496879919912214,
+      "learning_rate": 2.500625413972678e-06,
+      "loss": 0.704,
+      "step": 32973
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5629520209088785,
+      "learning_rate": 2.5001206774774813e-06,
+      "loss": 0.8038,
+      "step": 32974
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.169751460238314,
+      "learning_rate": 2.4996159846489544e-06,
+      "loss": 0.8912,
+      "step": 32975
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6195502144552083,
+      "learning_rate": 2.4991113354900374e-06,
+      "loss": 0.7447,
+      "step": 32976
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7280729938206612,
+      "learning_rate": 2.4986067300036708e-06,
+      "loss": 0.8195,
+      "step": 32977
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6628661563474119,
+      "learning_rate": 2.498102168192792e-06,
+      "loss": 0.8846,
+      "step": 32978
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.48113145290053,
+      "learning_rate": 2.4975976500603348e-06,
+      "loss": 0.8922,
+      "step": 32979
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4987412403676754,
+      "learning_rate": 2.4970931756092387e-06,
+      "loss": 0.8459,
+      "step": 32980
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.693419558465922,
+      "learning_rate": 2.4965887448424443e-06,
+      "loss": 0.7637,
+      "step": 32981
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6919333232282976,
+      "learning_rate": 2.4960843577628858e-06,
+      "loss": 0.9294,
+      "step": 32982
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7809853584433059,
+      "learning_rate": 2.4955800143734966e-06,
+      "loss": 0.6962,
+      "step": 32983
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0759118841194915,
+      "learning_rate": 2.4950757146772175e-06,
+      "loss": 0.9765,
+      "step": 32984
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7649878227384574,
+      "learning_rate": 2.4945714586769854e-06,
+      "loss": 0.8514,
+      "step": 32985
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5274882800044582,
+      "learning_rate": 2.4940672463757353e-06,
+      "loss": 0.7621,
+      "step": 32986
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5632187173079712,
+      "learning_rate": 2.493563077776399e-06,
+      "loss": 0.8611,
+      "step": 32987
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5355194451735912,
+      "learning_rate": 2.4930589528819172e-06,
+      "loss": 0.9199,
+      "step": 32988
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7222778779652994,
+      "learning_rate": 2.492554871695221e-06,
+      "loss": 0.9112,
+      "step": 32989
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7873867367193488,
+      "learning_rate": 2.4920508342192485e-06,
+      "loss": 0.7426,
+      "step": 32990
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4860035545089993,
+      "learning_rate": 2.491546840456931e-06,
+      "loss": 0.8547,
+      "step": 32991
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6266855369571527,
+      "learning_rate": 2.491042890411208e-06,
+      "loss": 0.8794,
+      "step": 32992
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7659128252212775,
+      "learning_rate": 2.490538984085008e-06,
+      "loss": 0.7913,
+      "step": 32993
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6785428585157922,
+      "learning_rate": 2.4900351214812703e-06,
+      "loss": 0.9073,
+      "step": 32994
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6479658673469784,
+      "learning_rate": 2.4895313026029243e-06,
+      "loss": 0.8756,
+      "step": 32995
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5703389075083278,
+      "learning_rate": 2.489027527452903e-06,
+      "loss": 0.8517,
+      "step": 32996
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.158504440722688,
+      "learning_rate": 2.4885237960341413e-06,
+      "loss": 0.7181,
+      "step": 32997
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.503530919432247,
+      "learning_rate": 2.488020108349576e-06,
+      "loss": 0.8498,
+      "step": 32998
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5447826387558679,
+      "learning_rate": 2.4875164644021343e-06,
+      "loss": 0.7859,
+      "step": 32999
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5842212265361055,
+      "learning_rate": 2.487012864194748e-06,
+      "loss": 0.8623,
+      "step": 33000
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5555462933140887,
+      "learning_rate": 2.4865093077303514e-06,
+      "loss": 0.8339,
+      "step": 33001
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9038062739523218,
+      "learning_rate": 2.48600579501188e-06,
+      "loss": 0.8718,
+      "step": 33002
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7156059368209318,
+      "learning_rate": 2.4855023260422605e-06,
+      "loss": 0.9139,
+      "step": 33003
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.661771249018875,
+      "learning_rate": 2.4849989008244224e-06,
+      "loss": 0.8617,
+      "step": 33004
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6003535122955426,
+      "learning_rate": 2.4844955193613007e-06,
+      "loss": 0.7999,
+      "step": 33005
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.634627431095332,
+      "learning_rate": 2.483992181655829e-06,
+      "loss": 0.7485,
+      "step": 33006
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.758645530098284,
+      "learning_rate": 2.483488887710933e-06,
+      "loss": 0.8452,
+      "step": 33007
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4754938282894852,
+      "learning_rate": 2.4829856375295426e-06,
+      "loss": 0.8517,
+      "step": 33008
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.797219311920756,
+      "learning_rate": 2.482482431114591e-06,
+      "loss": 0.7623,
+      "step": 33009
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6931809895954557,
+      "learning_rate": 2.4819792684690046e-06,
+      "loss": 0.7654,
+      "step": 33010
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.005907809979509,
+      "learning_rate": 2.4814761495957183e-06,
+      "loss": 0.7926,
+      "step": 33011
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9018182456446897,
+      "learning_rate": 2.4809730744976556e-06,
+      "loss": 0.8832,
+      "step": 33012
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4687750966303885,
+      "learning_rate": 2.480470043177751e-06,
+      "loss": 0.8634,
+      "step": 33013
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7512711098780943,
+      "learning_rate": 2.4799670556389267e-06,
+      "loss": 0.66,
+      "step": 33014
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7237137161900362,
+      "learning_rate": 2.479464111884118e-06,
+      "loss": 0.8109,
+      "step": 33015
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5821818865406159,
+      "learning_rate": 2.478961211916251e-06,
+      "loss": 0.8103,
+      "step": 33016
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6338471932203507,
+      "learning_rate": 2.4784583557382492e-06,
+      "loss": 0.8028,
+      "step": 33017
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8058418676178845,
+      "learning_rate": 2.4779555433530446e-06,
+      "loss": 0.9972,
+      "step": 33018
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5916765263323804,
+      "learning_rate": 2.4774527747635667e-06,
+      "loss": 0.8906,
+      "step": 33019
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.626732709355383,
+      "learning_rate": 2.4769500499727417e-06,
+      "loss": 1.0409,
+      "step": 33020
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6349386278518843,
+      "learning_rate": 2.4764473689834912e-06,
+      "loss": 0.7591,
+      "step": 33021
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.526715549270257,
+      "learning_rate": 2.475944731798746e-06,
+      "loss": 0.7462,
+      "step": 33022
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6911000965190066,
+      "learning_rate": 2.4754421384214355e-06,
+      "loss": 0.9534,
+      "step": 33023
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5199814039035222,
+      "learning_rate": 2.4749395888544837e-06,
+      "loss": 0.8796,
+      "step": 33024
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.485151607435248,
+      "learning_rate": 2.4744370831008125e-06,
+      "loss": 0.8302,
+      "step": 33025
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6836500809786836,
+      "learning_rate": 2.473934621163352e-06,
+      "loss": 0.8641,
+      "step": 33026
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.590717610985637,
+      "learning_rate": 2.4734322030450286e-06,
+      "loss": 0.9058,
+      "step": 33027
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.52931615397419,
+      "learning_rate": 2.472929828748767e-06,
+      "loss": 0.84,
+      "step": 33028
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5623008978732214,
+      "learning_rate": 2.4724274982774866e-06,
+      "loss": 0.8897,
+      "step": 33029
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6310332588906098,
+      "learning_rate": 2.4719252116341207e-06,
+      "loss": 0.7539,
+      "step": 33030
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4724856420194146,
+      "learning_rate": 2.4714229688215864e-06,
+      "loss": 0.9327,
+      "step": 33031
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4782769108782143,
+      "learning_rate": 2.470920769842814e-06,
+      "loss": 0.9003,
+      "step": 33032
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7125688118563067,
+      "learning_rate": 2.4704186147007216e-06,
+      "loss": 0.913,
+      "step": 33033
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5360055408148703,
+      "learning_rate": 2.4699165033982377e-06,
+      "loss": 0.7976,
+      "step": 33034
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.244447045292176,
+      "learning_rate": 2.469414435938282e-06,
+      "loss": 0.8574,
+      "step": 33035
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7029686187020479,
+      "learning_rate": 2.4689124123237795e-06,
+      "loss": 0.7895,
+      "step": 33036
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4836312142406622,
+      "learning_rate": 2.4684104325576595e-06,
+      "loss": 0.8322,
+      "step": 33037
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5831519549171922,
+      "learning_rate": 2.467908496642831e-06,
+      "loss": 0.929,
+      "step": 33038
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6172660317478496,
+      "learning_rate": 2.4674066045822256e-06,
+      "loss": 0.9861,
+      "step": 33039
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8376055943735536,
+      "learning_rate": 2.466904756378763e-06,
+      "loss": 1.005,
+      "step": 33040
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.566828391507528,
+      "learning_rate": 2.4664029520353716e-06,
+      "loss": 0.7239,
+      "step": 33041
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7859655752852763,
+      "learning_rate": 2.4659011915549626e-06,
+      "loss": 0.7573,
+      "step": 33042
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.137567110065376,
+      "learning_rate": 2.465399474940461e-06,
+      "loss": 0.7593,
+      "step": 33043
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8036006172397712,
+      "learning_rate": 2.4648978021947923e-06,
+      "loss": 0.8359,
+      "step": 33044
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5217939368238216,
+      "learning_rate": 2.464396173320873e-06,
+      "loss": 0.8445,
+      "step": 33045
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.669479151212522,
+      "learning_rate": 2.4638945883216236e-06,
+      "loss": 0.8516,
+      "step": 33046
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5320487980312376,
+      "learning_rate": 2.4633930471999657e-06,
+      "loss": 0.9184,
+      "step": 33047
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7374654586149287,
+      "learning_rate": 2.4628915499588213e-06,
+      "loss": 0.9995,
+      "step": 33048
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9129081720644185,
+      "learning_rate": 2.4623900966011062e-06,
+      "loss": 0.8263,
+      "step": 33049
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.555402477309834,
+      "learning_rate": 2.4618886871297454e-06,
+      "loss": 0.9488,
+      "step": 33050
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8135443844277103,
+      "learning_rate": 2.461387321547656e-06,
+      "loss": 0.7608,
+      "step": 33051
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.755556695288661,
+      "learning_rate": 2.4608859998577515e-06,
+      "loss": 0.9745,
+      "step": 33052
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3860365042483032,
+      "learning_rate": 2.4603847220629575e-06,
+      "loss": 0.8137,
+      "step": 33053
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.67111567384339,
+      "learning_rate": 2.4598834881661914e-06,
+      "loss": 0.9335,
+      "step": 33054
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.51966722733539,
+      "learning_rate": 2.459382298170373e-06,
+      "loss": 0.764,
+      "step": 33055
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8251106396874445,
+      "learning_rate": 2.4588811520784138e-06,
+      "loss": 0.695,
+      "step": 33056
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7950536854736099,
+      "learning_rate": 2.4583800498932365e-06,
+      "loss": 0.7345,
+      "step": 33057
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7638977938925028,
+      "learning_rate": 2.4578789916177614e-06,
+      "loss": 0.6815,
+      "step": 33058
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.839628711494292,
+      "learning_rate": 2.4573779772549023e-06,
+      "loss": 0.8219,
+      "step": 33059
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7333196753922517,
+      "learning_rate": 2.456877006807573e-06,
+      "loss": 0.8262,
+      "step": 33060
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.836697045503976,
+      "learning_rate": 2.456376080278695e-06,
+      "loss": 0.8462,
+      "step": 33061
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7323952284613489,
+      "learning_rate": 2.4558751976711857e-06,
+      "loss": 0.8556,
+      "step": 33062
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6806389930704535,
+      "learning_rate": 2.45537435898796e-06,
+      "loss": 0.7767,
+      "step": 33063
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9625747305634154,
+      "learning_rate": 2.4548735642319297e-06,
+      "loss": 0.8479,
+      "step": 33064
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5788485023147192,
+      "learning_rate": 2.4543728134060186e-06,
+      "loss": 0.8513,
+      "step": 33065
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.762454730546949,
+      "learning_rate": 2.4538721065131344e-06,
+      "loss": 0.7496,
+      "step": 33066
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.591715118438324,
+      "learning_rate": 2.453371443556198e-06,
+      "loss": 0.8255,
+      "step": 33067
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5713178160509953,
+      "learning_rate": 2.4528708245381203e-06,
+      "loss": 0.8731,
+      "step": 33068
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6211690202654399,
+      "learning_rate": 2.4523702494618206e-06,
+      "loss": 0.7532,
+      "step": 33069
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4905652675094847,
+      "learning_rate": 2.4518697183302076e-06,
+      "loss": 0.711,
+      "step": 33070
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7850920715415906,
+      "learning_rate": 2.4513692311462024e-06,
+      "loss": 0.677,
+      "step": 33071
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5580007475656856,
+      "learning_rate": 2.4508687879127145e-06,
+      "loss": 0.7487,
+      "step": 33072
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.561602405674738,
+      "learning_rate": 2.4503683886326557e-06,
+      "loss": 0.9533,
+      "step": 33073
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7238992606330852,
+      "learning_rate": 2.449868033308943e-06,
+      "loss": 0.7794,
+      "step": 33074
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5601506744814269,
+      "learning_rate": 2.449367721944492e-06,
+      "loss": 0.9073,
+      "step": 33075
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6282233760357172,
+      "learning_rate": 2.448867454542212e-06,
+      "loss": 0.9266,
+      "step": 33076
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4818755590373194,
+      "learning_rate": 2.4483672311050134e-06,
+      "loss": 0.8199,
+      "step": 33077
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7963395534727924,
+      "learning_rate": 2.4478670516358116e-06,
+      "loss": 0.7696,
+      "step": 33078
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5773554764218916,
+      "learning_rate": 2.4473669161375225e-06,
+      "loss": 0.8855,
+      "step": 33079
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.826442538407803,
+      "learning_rate": 2.4468668246130534e-06,
+      "loss": 0.882,
+      "step": 33080
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7027749975688138,
+      "learning_rate": 2.4463667770653144e-06,
+      "loss": 0.794,
+      "step": 33081
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5909744780235948,
+      "learning_rate": 2.4458667734972196e-06,
+      "loss": 0.7564,
+      "step": 33082
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5863602918586048,
+      "learning_rate": 2.4453668139116824e-06,
+      "loss": 0.8078,
+      "step": 33083
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5975541331229333,
+      "learning_rate": 2.444866898311612e-06,
+      "loss": 0.811,
+      "step": 33084
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5313782689169404,
+      "learning_rate": 2.4443670266999155e-06,
+      "loss": 0.7444,
+      "step": 33085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.541122822227094,
+      "learning_rate": 2.443867199079506e-06,
+      "loss": 0.8771,
+      "step": 33086
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6247767060889216,
+      "learning_rate": 2.443367415453297e-06,
+      "loss": 0.8528,
+      "step": 33087
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7324630344269916,
+      "learning_rate": 2.4428676758241953e-06,
+      "loss": 0.6836,
+      "step": 33088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9232694274493418,
+      "learning_rate": 2.4423679801951073e-06,
+      "loss": 0.9264,
+      "step": 33089
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7912251989760019,
+      "learning_rate": 2.441868328568948e-06,
+      "loss": 0.7372,
+      "step": 33090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5577172680507338,
+      "learning_rate": 2.4413687209486223e-06,
+      "loss": 0.8082,
+      "step": 33091
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7527259623790993,
+      "learning_rate": 2.4408691573370438e-06,
+      "loss": 0.6488,
+      "step": 33092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.48401461594073,
+      "learning_rate": 2.4403696377371146e-06,
+      "loss": 0.8521,
+      "step": 33093
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6194328964197162,
+      "learning_rate": 2.4398701621517507e-06,
+      "loss": 0.9009,
+      "step": 33094
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.529733731454599,
+      "learning_rate": 2.439370730583852e-06,
+      "loss": 0.8498,
+      "step": 33095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7283255786863926,
+      "learning_rate": 2.4388713430363343e-06,
+      "loss": 0.6769,
+      "step": 33096
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6514492022476734,
+      "learning_rate": 2.4383719995121013e-06,
+      "loss": 0.9683,
+      "step": 33097
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.034260264430451,
+      "learning_rate": 2.437872700014058e-06,
+      "loss": 0.7774,
+      "step": 33098
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4803345197985203,
+      "learning_rate": 2.4373734445451127e-06,
+      "loss": 0.7718,
+      "step": 33099
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4754676260576414,
+      "learning_rate": 2.4368742331081775e-06,
+      "loss": 0.7802,
+      "step": 33100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6011606832865446,
+      "learning_rate": 2.4363750657061546e-06,
+      "loss": 0.7566,
+      "step": 33101
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5506496353511838,
+      "learning_rate": 2.4358759423419476e-06,
+      "loss": 0.7195,
+      "step": 33102
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7142554405289057,
+      "learning_rate": 2.4353768630184658e-06,
+      "loss": 0.8455,
+      "step": 33103
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5158421195030727,
+      "learning_rate": 2.4348778277386177e-06,
+      "loss": 0.8706,
+      "step": 33104
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5163872706583703,
+      "learning_rate": 2.4343788365053066e-06,
+      "loss": 0.9333,
+      "step": 33105
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6270050762730175,
+      "learning_rate": 2.4338798893214332e-06,
+      "loss": 0.8669,
+      "step": 33106
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6011881779753698,
+      "learning_rate": 2.433380986189907e-06,
+      "loss": 0.8042,
+      "step": 33107
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6460482435755042,
+      "learning_rate": 2.432882127113635e-06,
+      "loss": 0.7787,
+      "step": 33108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.492325496959144,
+      "learning_rate": 2.4323833120955197e-06,
+      "loss": 0.8137,
+      "step": 33109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.555276912166214,
+      "learning_rate": 2.4318845411384608e-06,
+      "loss": 0.8375,
+      "step": 33110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9643874792503335,
+      "learning_rate": 2.4313858142453695e-06,
+      "loss": 0.7619,
+      "step": 33111
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0779410635640767,
+      "learning_rate": 2.4308871314191427e-06,
+      "loss": 0.8645,
+      "step": 33112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.646754735928076,
+      "learning_rate": 2.430388492662691e-06,
+      "loss": 0.7948,
+      "step": 33113
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6537181602789193,
+      "learning_rate": 2.429889897978911e-06,
+      "loss": 0.8001,
+      "step": 33114
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7342546226556466,
+      "learning_rate": 2.4293913473707108e-06,
+      "loss": 0.7409,
+      "step": 33115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5270332741490058,
+      "learning_rate": 2.4288928408409885e-06,
+      "loss": 0.7911,
+      "step": 33116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7668506899168737,
+      "learning_rate": 2.4283943783926523e-06,
+      "loss": 0.6786,
+      "step": 33117
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.64141218723034,
+      "learning_rate": 2.427895960028601e-06,
+      "loss": 0.8679,
+      "step": 33118
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6329858676603664,
+      "learning_rate": 2.427397585751734e-06,
+      "loss": 0.7818,
+      "step": 33119
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.570726751518713,
+      "learning_rate": 2.4268992555649553e-06,
+      "loss": 0.8552,
+      "step": 33120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9412874761217733,
+      "learning_rate": 2.4264009694711665e-06,
+      "loss": 0.861,
+      "step": 33121
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8056711866143281,
+      "learning_rate": 2.4259027274732748e-06,
+      "loss": 0.715,
+      "step": 33122
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5198671889444224,
+      "learning_rate": 2.42540452957417e-06,
+      "loss": 0.7957,
+      "step": 33123
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.77986996648781,
+      "learning_rate": 2.4249063757767587e-06,
+      "loss": 0.911,
+      "step": 33124
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.648898521792617,
+      "learning_rate": 2.424408266083943e-06,
+      "loss": 0.8351,
+      "step": 33125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8831554134793722,
+      "learning_rate": 2.4239102004986203e-06,
+      "loss": 0.7431,
+      "step": 33126
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8684837371400207,
+      "learning_rate": 2.4234121790236887e-06,
+      "loss": 0.874,
+      "step": 33127
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7770383528233495,
+      "learning_rate": 2.42291420166205e-06,
+      "loss": 0.6742,
+      "step": 33128
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6329970390806854,
+      "learning_rate": 2.4224162684166064e-06,
+      "loss": 0.8986,
+      "step": 33129
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8541774259470352,
+      "learning_rate": 2.4219183792902544e-06,
+      "loss": 0.8476,
+      "step": 33130
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5892636745433726,
+      "learning_rate": 2.4214205342858897e-06,
+      "loss": 0.8277,
+      "step": 33131
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4476111118889683,
+      "learning_rate": 2.4209227334064167e-06,
+      "loss": 0.8408,
+      "step": 33132
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.608905172477354,
+      "learning_rate": 2.4204249766547295e-06,
+      "loss": 0.7991,
+      "step": 33133
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3666318857765898,
+      "learning_rate": 2.4199272640337256e-06,
+      "loss": 0.9215,
+      "step": 33134
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.605540879100124,
+      "learning_rate": 2.41942959554631e-06,
+      "loss": 0.7056,
+      "step": 33135
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5136227603219754,
+      "learning_rate": 2.418931971195374e-06,
+      "loss": 0.8203,
+      "step": 33136
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5912688373393824,
+      "learning_rate": 2.4184343909838135e-06,
+      "loss": 0.8194,
+      "step": 33137
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.655928532673977,
+      "learning_rate": 2.4179368549145286e-06,
+      "loss": 0.7445,
+      "step": 33138
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5115905779848107,
+      "learning_rate": 2.4174393629904214e-06,
+      "loss": 0.9311,
+      "step": 33139
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.854057571756143,
+      "learning_rate": 2.416941915214377e-06,
+      "loss": 0.8644,
+      "step": 33140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7019555100979573,
+      "learning_rate": 2.4164445115892966e-06,
+      "loss": 0.7753,
+      "step": 33141
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.679163631037557,
+      "learning_rate": 2.415947152118079e-06,
+      "loss": 0.7673,
+      "step": 33142
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.627489828983251,
+      "learning_rate": 2.4154498368036226e-06,
+      "loss": 0.8642,
+      "step": 33143
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.566435606740398,
+      "learning_rate": 2.4149525656488128e-06,
+      "loss": 0.8582,
+      "step": 33144
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5196348473854586,
+      "learning_rate": 2.4144553386565504e-06,
+      "loss": 0.8803,
+      "step": 33145
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4706081914863862,
+      "learning_rate": 2.4139581558297344e-06,
+      "loss": 0.8475,
+      "step": 33146
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6432164575899213,
+      "learning_rate": 2.4134610171712525e-06,
+      "loss": 0.8686,
+      "step": 33147
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6096262723912858,
+      "learning_rate": 2.412963922684004e-06,
+      "loss": 0.8483,
+      "step": 33148
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7265503234495938,
+      "learning_rate": 2.41246687237088e-06,
+      "loss": 0.7611,
+      "step": 33149
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5438709458655702,
+      "learning_rate": 2.411969866234779e-06,
+      "loss": 0.7048,
+      "step": 33150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6698333575605289,
+      "learning_rate": 2.411472904278589e-06,
+      "loss": 0.7677,
+      "step": 33151
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6310202009064558,
+      "learning_rate": 2.4109759865052084e-06,
+      "loss": 0.8142,
+      "step": 33152
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.495702781618676,
+      "learning_rate": 2.4104791129175286e-06,
+      "loss": 0.9265,
+      "step": 33153
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7303903597409842,
+      "learning_rate": 2.4099822835184392e-06,
+      "loss": 0.7439,
+      "step": 33154
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.736560231908022,
+      "learning_rate": 2.409485498310835e-06,
+      "loss": 0.7107,
+      "step": 33155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7592159404482277,
+      "learning_rate": 2.4089887572976134e-06,
+      "loss": 0.8185,
+      "step": 33156
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6728159037194987,
+      "learning_rate": 2.408492060481663e-06,
+      "loss": 0.8319,
+      "step": 33157
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8091079475339298,
+      "learning_rate": 2.407995407865872e-06,
+      "loss": 0.7726,
+      "step": 33158
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6970277937262939,
+      "learning_rate": 2.407498799453134e-06,
+      "loss": 0.8136,
+      "step": 33159
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.627900952640927,
+      "learning_rate": 2.407002235246346e-06,
+      "loss": 0.7763,
+      "step": 33160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.579167698458378,
+      "learning_rate": 2.406505715248395e-06,
+      "loss": 0.8648,
+      "step": 33161
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.858573937830361,
+      "learning_rate": 2.406009239462168e-06,
+      "loss": 0.8378,
+      "step": 33162
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5970278994372529,
+      "learning_rate": 2.40551280789056e-06,
+      "loss": 1.005,
+      "step": 33163
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7753238920260535,
+      "learning_rate": 2.4050164205364633e-06,
+      "loss": 0.6751,
+      "step": 33164
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4472664207338548,
+      "learning_rate": 2.4045200774027653e-06,
+      "loss": 0.6906,
+      "step": 33165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5830784651514187,
+      "learning_rate": 2.404023778492354e-06,
+      "loss": 0.8151,
+      "step": 33166
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8456129686914309,
+      "learning_rate": 2.4035275238081236e-06,
+      "loss": 0.7301,
+      "step": 33167
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.096391724418559,
+      "learning_rate": 2.4030313133529583e-06,
+      "loss": 0.816,
+      "step": 33168
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8375926533149611,
+      "learning_rate": 2.402535147129753e-06,
+      "loss": 0.8727,
+      "step": 33169
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.506358099449905,
+      "learning_rate": 2.4020390251413893e-06,
+      "loss": 0.9116,
+      "step": 33170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6032284095971654,
+      "learning_rate": 2.4015429473907637e-06,
+      "loss": 0.8085,
+      "step": 33171
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6419386678561667,
+      "learning_rate": 2.4010469138807578e-06,
+      "loss": 0.8136,
+      "step": 33172
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.560085181709657,
+      "learning_rate": 2.4005509246142655e-06,
+      "loss": 0.8608,
+      "step": 33173
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6246154576175325,
+      "learning_rate": 2.400054979594172e-06,
+      "loss": 0.9166,
+      "step": 33174
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7844400380030234,
+      "learning_rate": 2.3995590788233604e-06,
+      "loss": 0.7537,
+      "step": 33175
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5141686666583776,
+      "learning_rate": 2.399063222304723e-06,
+      "loss": 0.7911,
+      "step": 33176
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6208753180127713,
+      "learning_rate": 2.3985674100411484e-06,
+      "loss": 0.8674,
+      "step": 33177
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7386185208625972,
+      "learning_rate": 2.398071642035521e-06,
+      "loss": 0.6937,
+      "step": 33178
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.771121508558164,
+      "learning_rate": 2.3975759182907244e-06,
+      "loss": 0.8279,
+      "step": 33179
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8573701633649204,
+      "learning_rate": 2.3970802388096483e-06,
+      "loss": 0.7163,
+      "step": 33180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6388604930244968,
+      "learning_rate": 2.39658460359518e-06,
+      "loss": 0.8032,
+      "step": 33181
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8188293074635997,
+      "learning_rate": 2.3960890126502034e-06,
+      "loss": 0.7934,
+      "step": 33182
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5572034872008123,
+      "learning_rate": 2.3955934659776013e-06,
+      "loss": 0.8329,
+      "step": 33183
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.514559562099799,
+      "learning_rate": 2.3950979635802617e-06,
+      "loss": 0.8227,
+      "step": 33184
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.469568199635137,
+      "learning_rate": 2.3946025054610722e-06,
+      "loss": 0.8899,
+      "step": 33185
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7244266183533006,
+      "learning_rate": 2.394107091622915e-06,
+      "loss": 0.9286,
+      "step": 33186
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7738628064182492,
+      "learning_rate": 2.3936117220686717e-06,
+      "loss": 0.6863,
+      "step": 33187
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6458149006862695,
+      "learning_rate": 2.3931163968012315e-06,
+      "loss": 0.919,
+      "step": 33188
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8302451961428263,
+      "learning_rate": 2.3926211158234725e-06,
+      "loss": 0.7784,
+      "step": 33189
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.816988956409614,
+      "learning_rate": 2.392125879138286e-06,
+      "loss": 0.9387,
+      "step": 33190
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6947069557000582,
+      "learning_rate": 2.3916306867485474e-06,
+      "loss": 0.9183,
+      "step": 33191
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6976593787431193,
+      "learning_rate": 2.3911355386571466e-06,
+      "loss": 0.7267,
+      "step": 33192
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.549876998716431,
+      "learning_rate": 2.390640434866961e-06,
+      "loss": 0.7729,
+      "step": 33193
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5528331889906064,
+      "learning_rate": 2.390145375380879e-06,
+      "loss": 0.969,
+      "step": 33194
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7471273963754557,
+      "learning_rate": 2.3896503602017794e-06,
+      "loss": 0.7449,
+      "step": 33195
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7782019253998901,
+      "learning_rate": 2.3891553893325424e-06,
+      "loss": 0.6715,
+      "step": 33196
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.784096228181408,
+      "learning_rate": 2.388660462776051e-06,
+      "loss": 0.9108,
+      "step": 33197
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4938439087827902,
+      "learning_rate": 2.388165580535192e-06,
+      "loss": 0.9154,
+      "step": 33198
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5062088658602562,
+      "learning_rate": 2.387670742612843e-06,
+      "loss": 0.8426,
+      "step": 33199
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.523674553395637,
+      "learning_rate": 2.387175949011882e-06,
+      "loss": 0.8008,
+      "step": 33200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.535491955906973,
+      "learning_rate": 2.3866811997351925e-06,
+      "loss": 0.76,
+      "step": 33201
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.459690343118976,
+      "learning_rate": 2.3861864947856583e-06,
+      "loss": 0.8599,
+      "step": 33202
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4875575880352758,
+      "learning_rate": 2.385691834166157e-06,
+      "loss": 0.826,
+      "step": 33203
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4128819891150175,
+      "learning_rate": 2.385197217879566e-06,
+      "loss": 0.7513,
+      "step": 33204
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.794166171459716,
+      "learning_rate": 2.384702645928767e-06,
+      "loss": 0.7367,
+      "step": 33205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.650264366229098,
+      "learning_rate": 2.3842081183166434e-06,
+      "loss": 0.9483,
+      "step": 33206
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.690642769972394,
+      "learning_rate": 2.383713635046071e-06,
+      "loss": 0.8931,
+      "step": 33207
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.521328632372662,
+      "learning_rate": 2.383219196119927e-06,
+      "loss": 0.8575,
+      "step": 33208
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7829187729561953,
+      "learning_rate": 2.3827248015410907e-06,
+      "loss": 0.712,
+      "step": 33209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3957972674539765,
+      "learning_rate": 2.382230451312446e-06,
+      "loss": 0.7879,
+      "step": 33210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.576305836322736,
+      "learning_rate": 2.381736145436867e-06,
+      "loss": 0.9373,
+      "step": 33211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.70791965679566,
+      "learning_rate": 2.38124188391723e-06,
+      "loss": 0.863,
+      "step": 33212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5109452256321207,
+      "learning_rate": 2.3807476667564168e-06,
+      "loss": 0.8937,
+      "step": 33213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.609938413962707,
+      "learning_rate": 2.3802534939573006e-06,
+      "loss": 0.8815,
+      "step": 33214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.438821996424103,
+      "learning_rate": 2.3797593655227637e-06,
+      "loss": 0.9467,
+      "step": 33215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.388034379580649,
+      "learning_rate": 2.379265281455677e-06,
+      "loss": 0.7471,
+      "step": 33216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6505847691688504,
+      "learning_rate": 2.378771241758924e-06,
+      "loss": 0.8117,
+      "step": 33217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.525952507837931,
+      "learning_rate": 2.378277246435374e-06,
+      "loss": 0.7645,
+      "step": 33218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7149629864230418,
+      "learning_rate": 2.3777832954879075e-06,
+      "loss": 0.8321,
+      "step": 33219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6874946742751933,
+      "learning_rate": 2.377289388919406e-06,
+      "loss": 0.8605,
+      "step": 33220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1788717726802984,
+      "learning_rate": 2.3767955267327325e-06,
+      "loss": 0.7505,
+      "step": 33221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7704021993915364,
+      "learning_rate": 2.3763017089307692e-06,
+      "loss": 0.6689,
+      "step": 33222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6189096564175702,
+      "learning_rate": 2.375807935516391e-06,
+      "loss": 0.7819,
+      "step": 33223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6918924245298297,
+      "learning_rate": 2.3753142064924785e-06,
+      "loss": 0.7599,
+      "step": 33224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9547660651949144,
+      "learning_rate": 2.374820521861895e-06,
+      "loss": 0.9649,
+      "step": 33225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5355167111705288,
+      "learning_rate": 2.37432688162752e-06,
+      "loss": 0.9619,
+      "step": 33226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8075549547539597,
+      "learning_rate": 2.3738332857922307e-06,
+      "loss": 0.743,
+      "step": 33227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9118676133988013,
+      "learning_rate": 2.3733397343588994e-06,
+      "loss": 0.8032,
+      "step": 33228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6180779479718137,
+      "learning_rate": 2.3728462273303954e-06,
+      "loss": 0.8526,
+      "step": 33229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6281984765191522,
+      "learning_rate": 2.3723527647095956e-06,
+      "loss": 0.8683,
+      "step": 33230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.767924619402887,
+      "learning_rate": 2.3718593464993757e-06,
+      "loss": 0.8384,
+      "step": 33231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6993499045139135,
+      "learning_rate": 2.3713659727026016e-06,
+      "loss": 0.9035,
+      "step": 33232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7869061930545225,
+      "learning_rate": 2.370872643322154e-06,
+      "loss": 0.8508,
+      "step": 33233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8356871892890325,
+      "learning_rate": 2.3703793583609013e-06,
+      "loss": 0.8087,
+      "step": 33234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.631295166798942,
+      "learning_rate": 2.3698861178217125e-06,
+      "loss": 0.9488,
+      "step": 33235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5290481394904745,
+      "learning_rate": 2.369392921707463e-06,
+      "loss": 0.8692,
+      "step": 33236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6581621770512285,
+      "learning_rate": 2.368899770021026e-06,
+      "loss": 0.8051,
+      "step": 33237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6441584076175215,
+      "learning_rate": 2.36840666276527e-06,
+      "loss": 0.9382,
+      "step": 33238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7147319425258146,
+      "learning_rate": 2.3679135999430646e-06,
+      "loss": 0.689,
+      "step": 33239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5300354283326991,
+      "learning_rate": 2.3674205815572816e-06,
+      "loss": 0.7349,
+      "step": 33240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.720536523280156,
+      "learning_rate": 2.3669276076107985e-06,
+      "loss": 0.6601,
+      "step": 33241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5160731458112189,
+      "learning_rate": 2.366434678106474e-06,
+      "loss": 0.8445,
+      "step": 33242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.420936945019825,
+      "learning_rate": 2.365941793047184e-06,
+      "loss": 0.665,
+      "step": 33243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5355367893932126,
+      "learning_rate": 2.365448952435797e-06,
+      "loss": 0.7956,
+      "step": 33244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6537107233709862,
+      "learning_rate": 2.3649561562751853e-06,
+      "loss": 0.836,
+      "step": 33245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5516987896128835,
+      "learning_rate": 2.364463404568218e-06,
+      "loss": 0.6459,
+      "step": 33246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6883717121577988,
+      "learning_rate": 2.3639706973177577e-06,
+      "loss": 0.8389,
+      "step": 33247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5545339971516563,
+      "learning_rate": 2.3634780345266805e-06,
+      "loss": 0.8513,
+      "step": 33248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6616357438695446,
+      "learning_rate": 2.3629854161978505e-06,
+      "loss": 0.7544,
+      "step": 33249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5029628612240238,
+      "learning_rate": 2.3624928423341387e-06,
+      "loss": 0.8905,
+      "step": 33250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.57286391277668,
+      "learning_rate": 2.3620003129384095e-06,
+      "loss": 0.7543,
+      "step": 33251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.680062247333405,
+      "learning_rate": 2.361507828013535e-06,
+      "loss": 0.8209,
+      "step": 33252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7532199801717172,
+      "learning_rate": 2.3610153875623777e-06,
+      "loss": 0.867,
+      "step": 33253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6094309797051711,
+      "learning_rate": 2.3605229915878093e-06,
+      "loss": 0.8183,
+      "step": 33254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7295279887047699,
+      "learning_rate": 2.3600306400926955e-06,
+      "loss": 0.7895,
+      "step": 33255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5641465956831335,
+      "learning_rate": 2.359538333079898e-06,
+      "loss": 0.828,
+      "step": 33256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.007200676079349,
+      "learning_rate": 2.359046070552289e-06,
+      "loss": 0.7586,
+      "step": 33257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.64587710888896,
+      "learning_rate": 2.358553852512735e-06,
+      "loss": 0.7804,
+      "step": 33258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4898588027488098,
+      "learning_rate": 2.3580616789640997e-06,
+      "loss": 0.8675,
+      "step": 33259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7819476045623477,
+      "learning_rate": 2.3575695499092455e-06,
+      "loss": 0.6763,
+      "step": 33260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.96907673907576,
+      "learning_rate": 2.3570774653510407e-06,
+      "loss": 0.9209,
+      "step": 33261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7401223494816158,
+      "learning_rate": 2.3565854252923547e-06,
+      "loss": 0.8137,
+      "step": 33262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4376010230027365,
+      "learning_rate": 2.356093429736047e-06,
+      "loss": 0.8175,
+      "step": 33263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.640145995186809,
+      "learning_rate": 2.355601478684981e-06,
+      "loss": 0.8792,
+      "step": 33264
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4473740585379993,
+      "learning_rate": 2.355109572142024e-06,
+      "loss": 0.7932,
+      "step": 33265
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6360724389599317,
+      "learning_rate": 2.3546177101100422e-06,
+      "loss": 0.814,
+      "step": 33266
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.200059986429094,
+      "learning_rate": 2.354125892591895e-06,
+      "loss": 0.8502,
+      "step": 33267
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6267955028633443,
+      "learning_rate": 2.3536341195904455e-06,
+      "loss": 0.8072,
+      "step": 33268
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6176685867212912,
+      "learning_rate": 2.3531423911085626e-06,
+      "loss": 0.9241,
+      "step": 33269
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.634880686631493,
+      "learning_rate": 2.3526507071491023e-06,
+      "loss": 0.9321,
+      "step": 33270
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7596039372064947,
+      "learning_rate": 2.352159067714933e-06,
+      "loss": 0.6665,
+      "step": 33271
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4177341304023816,
+      "learning_rate": 2.3516674728089127e-06,
+      "loss": 0.8196,
+      "step": 33272
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.670768198713941,
+      "learning_rate": 2.351175922433908e-06,
+      "loss": 0.8516,
+      "step": 33273
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.992024739494049,
+      "learning_rate": 2.3506844165927765e-06,
+      "loss": 0.8087,
+      "step": 33274
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4720863127085626,
+      "learning_rate": 2.350192955288385e-06,
+      "loss": 0.733,
+      "step": 33275
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6149496321717505,
+      "learning_rate": 2.349701538523591e-06,
+      "loss": 0.7271,
+      "step": 33276
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5331042664465044,
+      "learning_rate": 2.3492101663012546e-06,
+      "loss": 0.7666,
+      "step": 33277
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.724267990845045,
+      "learning_rate": 2.3487188386242387e-06,
+      "loss": 0.7829,
+      "step": 33278
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4451325994888549,
+      "learning_rate": 2.348227555495407e-06,
+      "loss": 0.7731,
+      "step": 33279
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4874843995967084,
+      "learning_rate": 2.347736316917617e-06,
+      "loss": 0.8486,
+      "step": 33280
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7189831049174209,
+      "learning_rate": 2.3472451228937254e-06,
+      "loss": 0.6803,
+      "step": 33281
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.745366863786091,
+      "learning_rate": 2.3467539734265966e-06,
+      "loss": 0.869,
+      "step": 33282
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6340926097178303,
+      "learning_rate": 2.3462628685190913e-06,
+      "loss": 0.8133,
+      "step": 33283
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5724982945438826,
+      "learning_rate": 2.3457718081740664e-06,
+      "loss": 0.7796,
+      "step": 33284
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.59346057558437,
+      "learning_rate": 2.345280792394379e-06,
+      "loss": 0.7979,
+      "step": 33285
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6466339724330605,
+      "learning_rate": 2.3447898211828903e-06,
+      "loss": 0.9731,
+      "step": 33286
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5178301252824373,
+      "learning_rate": 2.3442988945424617e-06,
+      "loss": 0.9149,
+      "step": 33287
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5376775907771465,
+      "learning_rate": 2.3438080124759476e-06,
+      "loss": 0.825,
+      "step": 33288
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6856238387073825,
+      "learning_rate": 2.343317174986206e-06,
+      "loss": 0.8118,
+      "step": 33289
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8085668353554613,
+      "learning_rate": 2.342826382076098e-06,
+      "loss": 0.6831,
+      "step": 33290
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7689690331083102,
+      "learning_rate": 2.342335633748476e-06,
+      "loss": 0.7267,
+      "step": 33291
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6555815167731214,
+      "learning_rate": 2.341844930006203e-06,
+      "loss": 0.8417,
+      "step": 33292
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5287447156158926,
+      "learning_rate": 2.341354270852131e-06,
+      "loss": 0.8579,
+      "step": 33293
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.661987920933307,
+      "learning_rate": 2.3408636562891218e-06,
+      "loss": 1.052,
+      "step": 33294
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8604236907317888,
+      "learning_rate": 2.3403730863200256e-06,
+      "loss": 0.7747,
+      "step": 33295
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.55651628721316,
+      "learning_rate": 2.339882560947706e-06,
+      "loss": 0.8038,
+      "step": 33296
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9951043604691903,
+      "learning_rate": 2.3393920801750157e-06,
+      "loss": 0.9511,
+      "step": 33297
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4967170958922886,
+      "learning_rate": 2.3389016440048063e-06,
+      "loss": 0.9021,
+      "step": 33298
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.764018126983876,
+      "learning_rate": 2.338411252439937e-06,
+      "loss": 0.7269,
+      "step": 33299
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5728064017252934,
+      "learning_rate": 2.337920905483266e-06,
+      "loss": 0.8174,
+      "step": 33300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6780023318173634,
+      "learning_rate": 2.337430603137646e-06,
+      "loss": 0.8077,
+      "step": 33301
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5688800551675108,
+      "learning_rate": 2.336940345405928e-06,
+      "loss": 0.9098,
+      "step": 33302
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7214844389991533,
+      "learning_rate": 2.3364501322909685e-06,
+      "loss": 0.6646,
+      "step": 33303
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3270246343725374,
+      "learning_rate": 2.3359599637956264e-06,
+      "loss": 0.802,
+      "step": 33304
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.63858930428898,
+      "learning_rate": 2.3354698399227504e-06,
+      "loss": 0.9264,
+      "step": 33305
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5257911159246889,
+      "learning_rate": 2.3349797606751945e-06,
+      "loss": 0.9169,
+      "step": 33306
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7678309906909558,
+      "learning_rate": 2.334489726055812e-06,
+      "loss": 0.6573,
+      "step": 33307
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5522140166136802,
+      "learning_rate": 2.3339997360674606e-06,
+      "loss": 0.955,
+      "step": 33308
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.522125121713872,
+      "learning_rate": 2.333509790712989e-06,
+      "loss": 0.8264,
+      "step": 33309
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8866174085071281,
+      "learning_rate": 2.3330198899952484e-06,
+      "loss": 0.8748,
+      "step": 33310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5106026612631507,
+      "learning_rate": 2.3325300339170954e-06,
+      "loss": 0.8725,
+      "step": 33311
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5556445951756863,
+      "learning_rate": 2.3320402224813776e-06,
+      "loss": 0.813,
+      "step": 33312
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3280818817906974,
+      "learning_rate": 2.3315504556909517e-06,
+      "loss": 0.8771,
+      "step": 33313
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.612483815013158,
+      "learning_rate": 2.3310607335486645e-06,
+      "loss": 0.7481,
+      "step": 33314
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2306686167139835,
+      "learning_rate": 2.3305710560573723e-06,
+      "loss": 0.8651,
+      "step": 33315
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.497987670642415,
+      "learning_rate": 2.330081423219921e-06,
+      "loss": 0.8471,
+      "step": 33316
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9307733398203488,
+      "learning_rate": 2.3295918350391632e-06,
+      "loss": 0.8152,
+      "step": 33317
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4626077704120672,
+      "learning_rate": 2.3291022915179575e-06,
+      "loss": 0.7604,
+      "step": 33318
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.470132369116615,
+      "learning_rate": 2.3286127926591397e-06,
+      "loss": 0.8232,
+      "step": 33319
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6308769124481515,
+      "learning_rate": 2.3281233384655665e-06,
+      "loss": 0.8416,
+      "step": 33320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8047085263849401,
+      "learning_rate": 2.3276339289400895e-06,
+      "loss": 0.7167,
+      "step": 33321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6779476746859818,
+      "learning_rate": 2.3271445640855618e-06,
+      "loss": 0.7313,
+      "step": 33322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7186129829672057,
+      "learning_rate": 2.326655243904822e-06,
+      "loss": 0.8044,
+      "step": 33323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6202448408895291,
+      "learning_rate": 2.326165968400724e-06,
+      "loss": 0.7854,
+      "step": 33324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7084229089216951,
+      "learning_rate": 2.325676737576121e-06,
+      "loss": 0.92,
+      "step": 33325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.033728155866505,
+      "learning_rate": 2.3251875514338575e-06,
+      "loss": 0.7626,
+      "step": 33326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4261227115967419,
+      "learning_rate": 2.324698409976779e-06,
+      "loss": 0.8124,
+      "step": 33327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7062307128220342,
+      "learning_rate": 2.324209313207736e-06,
+      "loss": 0.8592,
+      "step": 33328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6940230590330452,
+      "learning_rate": 2.3237202611295795e-06,
+      "loss": 0.885,
+      "step": 33329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.767716639910951,
+      "learning_rate": 2.323231253745151e-06,
+      "loss": 0.792,
+      "step": 33330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4435781004968566,
+      "learning_rate": 2.322742291057304e-06,
+      "loss": 0.7491,
+      "step": 33331
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8160874419685334,
+      "learning_rate": 2.322253373068879e-06,
+      "loss": 0.7562,
+      "step": 33332
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5144004515009153,
+      "learning_rate": 2.321764499782727e-06,
+      "loss": 0.938,
+      "step": 33333
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7443922035180175,
+      "learning_rate": 2.3212756712016913e-06,
+      "loss": 0.8109,
+      "step": 33334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4906050145034557,
+      "learning_rate": 2.3207868873286232e-06,
+      "loss": 0.9579,
+      "step": 33335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5457452262781815,
+      "learning_rate": 2.3202981481663646e-06,
+      "loss": 0.9009,
+      "step": 33336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6297168601134757,
+      "learning_rate": 2.319809453717757e-06,
+      "loss": 0.8533,
+      "step": 33337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6204897300503127,
+      "learning_rate": 2.319320803985653e-06,
+      "loss": 0.8845,
+      "step": 33338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.849900485439673,
+      "learning_rate": 2.3188321989728955e-06,
+      "loss": 0.7889,
+      "step": 33339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6442097034833807,
+      "learning_rate": 2.31834363868233e-06,
+      "loss": 0.6914,
+      "step": 33340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5401324072914386,
+      "learning_rate": 2.3178551231167968e-06,
+      "loss": 0.8153,
+      "step": 33341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6182925317408579,
+      "learning_rate": 2.317366652279143e-06,
+      "loss": 0.8477,
+      "step": 33342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5318059399957829,
+      "learning_rate": 2.3168782261722156e-06,
+      "loss": 0.9215,
+      "step": 33343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5284459046665317,
+      "learning_rate": 2.3163898447988562e-06,
+      "loss": 0.7864,
+      "step": 33344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5799754486847741,
+      "learning_rate": 2.315901508161904e-06,
+      "loss": 0.8373,
+      "step": 33345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.71750414419159,
+      "learning_rate": 2.315413216264206e-06,
+      "loss": 0.7574,
+      "step": 33346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8070782582769827,
+      "learning_rate": 2.3149249691086094e-06,
+      "loss": 0.9091,
+      "step": 33347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7336723101206746,
+      "learning_rate": 2.3144367666979517e-06,
+      "loss": 0.8107,
+      "step": 33348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8809525300024763,
+      "learning_rate": 2.3139486090350727e-06,
+      "loss": 0.8365,
+      "step": 33349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4088247612088975,
+      "learning_rate": 2.3134604961228225e-06,
+      "loss": 0.6682,
+      "step": 33350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.633848423319082,
+      "learning_rate": 2.3129724279640355e-06,
+      "loss": 0.776,
+      "step": 33351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.737325259373546,
+      "learning_rate": 2.312484404561559e-06,
+      "loss": 0.9149,
+      "step": 33352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4849523707630286,
+      "learning_rate": 2.311996425918229e-06,
+      "loss": 0.8073,
+      "step": 33353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6596786651802538,
+      "learning_rate": 2.3115084920368935e-06,
+      "loss": 0.8828,
+      "step": 33354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0123559245658207,
+      "learning_rate": 2.3110206029203872e-06,
+      "loss": 0.8829,
+      "step": 33355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4792836975800685,
+      "learning_rate": 2.310532758571555e-06,
+      "loss": 0.9099,
+      "step": 33356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6500376994817387,
+      "learning_rate": 2.310044958993236e-06,
+      "loss": 0.7847,
+      "step": 33357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5637192913736944,
+      "learning_rate": 2.3095572041882675e-06,
+      "loss": 0.8755,
+      "step": 33358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7547441755590842,
+      "learning_rate": 2.3090694941594914e-06,
+      "loss": 0.8952,
+      "step": 33359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.804339018865267,
+      "learning_rate": 2.3085818289097506e-06,
+      "loss": 0.7514,
+      "step": 33360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5970718090402114,
+      "learning_rate": 2.308094208441881e-06,
+      "loss": 0.8468,
+      "step": 33361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.525202403383781,
+      "learning_rate": 2.307606632758719e-06,
+      "loss": 0.8909,
+      "step": 33362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6372679269902577,
+      "learning_rate": 2.307119101863108e-06,
+      "loss": 0.8122,
+      "step": 33363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5548695719437537,
+      "learning_rate": 2.306631615757886e-06,
+      "loss": 0.7882,
+      "step": 33364
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5997449583220043,
+      "learning_rate": 2.306144174445891e-06,
+      "loss": 0.9412,
+      "step": 33365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7543798655871405,
+      "learning_rate": 2.3056567779299588e-06,
+      "loss": 0.671,
+      "step": 33366
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.032899862722545,
+      "learning_rate": 2.305169426212928e-06,
+      "loss": 0.8383,
+      "step": 33367
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5852842094537944,
+      "learning_rate": 2.30468211929764e-06,
+      "loss": 0.9404,
+      "step": 33368
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.485851026189584,
+      "learning_rate": 2.304194857186929e-06,
+      "loss": 0.7997,
+      "step": 33369
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.65646719527625,
+      "learning_rate": 2.303707639883629e-06,
+      "loss": 0.8714,
+      "step": 33370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.481500199915382,
+      "learning_rate": 2.303220467390582e-06,
+      "loss": 0.84,
+      "step": 33371
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5760812544452925,
+      "learning_rate": 2.30273333971062e-06,
+      "loss": 0.8252,
+      "step": 33372
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7461144945074949,
+      "learning_rate": 2.3022462568465843e-06,
+      "loss": 0.704,
+      "step": 33373
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.497320629525737,
+      "learning_rate": 2.301759218801305e-06,
+      "loss": 0.9278,
+      "step": 33374
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5076237157743506,
+      "learning_rate": 2.3012722255776234e-06,
+      "loss": 0.8768,
+      "step": 33375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4791989564856047,
+      "learning_rate": 2.300785277178369e-06,
+      "loss": 0.8576,
+      "step": 33376
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.557806073947668,
+      "learning_rate": 2.3002983736063843e-06,
+      "loss": 0.7336,
+      "step": 33377
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8584521356543238,
+      "learning_rate": 2.2998115148644997e-06,
+      "loss": 0.8016,
+      "step": 33378
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.494069999392311,
+      "learning_rate": 2.2993247009555463e-06,
+      "loss": 0.8388,
+      "step": 33379
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6567371172989418,
+      "learning_rate": 2.298837931882364e-06,
+      "loss": 0.8363,
+      "step": 33380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5908009307187332,
+      "learning_rate": 2.2983512076477875e-06,
+      "loss": 0.9242,
+      "step": 33381
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6485162966576883,
+      "learning_rate": 2.2978645282546487e-06,
+      "loss": 0.8195,
+      "step": 33382
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4852880629547849,
+      "learning_rate": 2.2973778937057787e-06,
+      "loss": 0.935,
+      "step": 33383
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4685454795491717,
+      "learning_rate": 2.296891304004013e-06,
+      "loss": 0.8297,
+      "step": 33384
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6891860387310256,
+      "learning_rate": 2.296404759152188e-06,
+      "loss": 0.8539,
+      "step": 33385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.876814678725465,
+      "learning_rate": 2.2959182591531327e-06,
+      "loss": 0.9652,
+      "step": 33386
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7278664860975295,
+      "learning_rate": 2.295431804009678e-06,
+      "loss": 0.8879,
+      "step": 33387
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5442855196875676,
+      "learning_rate": 2.294945393724658e-06,
+      "loss": 0.8853,
+      "step": 33388
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.294232458808314,
+      "learning_rate": 2.2944590283009083e-06,
+      "loss": 0.8955,
+      "step": 33389
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6553967051265772,
+      "learning_rate": 2.2939727077412578e-06,
+      "loss": 0.7738,
+      "step": 33390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3908562389752355,
+      "learning_rate": 2.293486432048535e-06,
+      "loss": 0.8422,
+      "step": 33391
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8827861210655223,
+      "learning_rate": 2.293000201225577e-06,
+      "loss": 0.81,
+      "step": 33392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6218974655384508,
+      "learning_rate": 2.292514015275209e-06,
+      "loss": 0.7878,
+      "step": 33393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.016784469833759,
+      "learning_rate": 2.2920278742002677e-06,
+      "loss": 0.8123,
+      "step": 33394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.636685717668739,
+      "learning_rate": 2.2915417780035763e-06,
+      "loss": 0.8216,
+      "step": 33395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.411850550426937,
+      "learning_rate": 2.2910557266879742e-06,
+      "loss": 0.6975,
+      "step": 33396
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4629458321636657,
+      "learning_rate": 2.2905697202562817e-06,
+      "loss": 0.8465,
+      "step": 33397
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5134193039336383,
+      "learning_rate": 2.2900837587113366e-06,
+      "loss": 0.7975,
+      "step": 33398
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.436610837495852,
+      "learning_rate": 2.289597842055964e-06,
+      "loss": 0.8042,
+      "step": 33399
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.81217406802448,
+      "learning_rate": 2.289111970292992e-06,
+      "loss": 0.7769,
+      "step": 33400
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5298139113493086,
+      "learning_rate": 2.28862614342525e-06,
+      "loss": 0.7546,
+      "step": 33401
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5681967646167025,
+      "learning_rate": 2.288140361455572e-06,
+      "loss": 0.8112,
+      "step": 33402
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6462828206224467,
+      "learning_rate": 2.2876546243867826e-06,
+      "loss": 0.7835,
+      "step": 33403
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5805396316609803,
+      "learning_rate": 2.287168932221706e-06,
+      "loss": 0.8502,
+      "step": 33404
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.42352994455785,
+      "learning_rate": 2.286683284963174e-06,
+      "loss": 0.8152,
+      "step": 33405
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5770348401317793,
+      "learning_rate": 2.2861976826140164e-06,
+      "loss": 0.7961,
+      "step": 33406
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6004679037866174,
+      "learning_rate": 2.2857121251770585e-06,
+      "loss": 0.9431,
+      "step": 33407
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.724823595403026,
+      "learning_rate": 2.285226612655124e-06,
+      "loss": 0.9839,
+      "step": 33408
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8137452919081847,
+      "learning_rate": 2.2847411450510426e-06,
+      "loss": 0.9047,
+      "step": 33409
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.567626933185617,
+      "learning_rate": 2.284255722367643e-06,
+      "loss": 0.8103,
+      "step": 33410
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.552221730127434,
+      "learning_rate": 2.2837703446077496e-06,
+      "loss": 0.7989,
+      "step": 33411
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5683670930773468,
+      "learning_rate": 2.2832850117741866e-06,
+      "loss": 0.8225,
+      "step": 33412
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5552967774902664,
+      "learning_rate": 2.282799723869783e-06,
+      "loss": 0.8167,
+      "step": 33413
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6525391015878534,
+      "learning_rate": 2.2823144808973596e-06,
+      "loss": 0.9157,
+      "step": 33414
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6194064443892129,
+      "learning_rate": 2.2818292828597457e-06,
+      "loss": 0.6973,
+      "step": 33415
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7773981695994234,
+      "learning_rate": 2.2813441297597673e-06,
+      "loss": 0.7835,
+      "step": 33416
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6499931035518003,
+      "learning_rate": 2.280859021600247e-06,
+      "loss": 0.9045,
+      "step": 33417
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.527728927374933,
+      "learning_rate": 2.2803739583840066e-06,
+      "loss": 0.6756,
+      "step": 33418
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8408125348679387,
+      "learning_rate": 2.2798889401138735e-06,
+      "loss": 0.7228,
+      "step": 33419
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6765119953405447,
+      "learning_rate": 2.279403966792676e-06,
+      "loss": 0.9072,
+      "step": 33420
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4766777925437486,
+      "learning_rate": 2.2789190384232286e-06,
+      "loss": 0.7338,
+      "step": 33421
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7417315737789958,
+      "learning_rate": 2.2784341550083577e-06,
+      "loss": 0.9176,
+      "step": 33422
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.059875520947329,
+      "learning_rate": 2.277949316550888e-06,
+      "loss": 0.9783,
+      "step": 33423
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5689643771327382,
+      "learning_rate": 2.2774645230536485e-06,
+      "loss": 0.6753,
+      "step": 33424
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8972388356468544,
+      "learning_rate": 2.276979774519449e-06,
+      "loss": 0.9535,
+      "step": 33425
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5083968533909309,
+      "learning_rate": 2.276495070951118e-06,
+      "loss": 0.8501,
+      "step": 33426
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.168420601500449,
+      "learning_rate": 2.2760104123514814e-06,
+      "loss": 0.8401,
+      "step": 33427
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5945975829433754,
+      "learning_rate": 2.275525798723355e-06,
+      "loss": 0.8456,
+      "step": 33428
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6814086347643324,
+      "learning_rate": 2.2750412300695656e-06,
+      "loss": 0.7312,
+      "step": 33429
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5454157178900743,
+      "learning_rate": 2.2745567063929285e-06,
+      "loss": 0.9725,
+      "step": 33430
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8979499017114592,
+      "learning_rate": 2.2740722276962714e-06,
+      "loss": 0.946,
+      "step": 33431
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6521775516152404,
+      "learning_rate": 2.273587793982409e-06,
+      "loss": 0.8605,
+      "step": 33432
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5065896297988488,
+      "learning_rate": 2.273103405254167e-06,
+      "loss": 0.8176,
+      "step": 33433
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.518086387880684,
+      "learning_rate": 2.2726190615143638e-06,
+      "loss": 0.6939,
+      "step": 33434
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4011990306312911,
+      "learning_rate": 2.272134762765815e-06,
+      "loss": 0.7426,
+      "step": 33435
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7155467313725605,
+      "learning_rate": 2.271650509011346e-06,
+      "loss": 0.8579,
+      "step": 33436
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5228831773018003,
+      "learning_rate": 2.2711663002537754e-06,
+      "loss": 0.9608,
+      "step": 33437
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5931949063694568,
+      "learning_rate": 2.270682136495923e-06,
+      "loss": 0.8507,
+      "step": 33438
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5399506914967955,
+      "learning_rate": 2.2701980177406026e-06,
+      "loss": 0.7883,
+      "step": 33439
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4847541510137183,
+      "learning_rate": 2.2697139439906367e-06,
+      "loss": 0.8396,
+      "step": 33440
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5986101477186716,
+      "learning_rate": 2.269229915248846e-06,
+      "loss": 0.8086,
+      "step": 33441
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9556069630472492,
+      "learning_rate": 2.2687459315180473e-06,
+      "loss": 0.8793,
+      "step": 33442
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.581660262550572,
+      "learning_rate": 2.268261992801054e-06,
+      "loss": 0.872,
+      "step": 33443
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7419590087928822,
+      "learning_rate": 2.267778099100687e-06,
+      "loss": 0.6624,
+      "step": 33444
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.212060028287821,
+      "learning_rate": 2.267294250419767e-06,
+      "loss": 0.8933,
+      "step": 33445
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7735954437491368,
+      "learning_rate": 2.2668104467611087e-06,
+      "loss": 0.8342,
+      "step": 33446
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.686447873011017,
+      "learning_rate": 2.2663266881275247e-06,
+      "loss": 0.8025,
+      "step": 33447
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6242086868620325,
+      "learning_rate": 2.265842974521838e-06,
+      "loss": 0.9175,
+      "step": 33448
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5277418974035473,
+      "learning_rate": 2.265359305946859e-06,
+      "loss": 0.7941,
+      "step": 33449
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8477240703035924,
+      "learning_rate": 2.2648756824054097e-06,
+      "loss": 0.7698,
+      "step": 33450
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.771431645626686,
+      "learning_rate": 2.2643921039003013e-06,
+      "loss": 0.8167,
+      "step": 33451
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5558872552575194,
+      "learning_rate": 2.263908570434352e-06,
+      "loss": 0.7562,
+      "step": 33452
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5570605414551766,
+      "learning_rate": 2.2634250820103752e-06,
+      "loss": 0.7444,
+      "step": 33453
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6421930179566477,
+      "learning_rate": 2.26294163863119e-06,
+      "loss": 0.8831,
+      "step": 33454
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0894774248605965,
+      "learning_rate": 2.2624582402996066e-06,
+      "loss": 0.8066,
+      "step": 33455
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5536867833447874,
+      "learning_rate": 2.261974887018439e-06,
+      "loss": 0.8107,
+      "step": 33456
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8914630319864503,
+      "learning_rate": 2.2614915787905024e-06,
+      "loss": 0.8343,
+      "step": 33457
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8432213902009054,
+      "learning_rate": 2.261008315618616e-06,
+      "loss": 0.7908,
+      "step": 33458
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7877176830153736,
+      "learning_rate": 2.260525097505588e-06,
+      "loss": 0.6931,
+      "step": 33459
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.458771128806158,
+      "learning_rate": 2.2600419244542316e-06,
+      "loss": 0.8037,
+      "step": 33460
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8952878437806682,
+      "learning_rate": 2.2595587964673614e-06,
+      "loss": 0.8771,
+      "step": 33461
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.59147421008894,
+      "learning_rate": 2.2590757135477927e-06,
+      "loss": 0.7766,
+      "step": 33462
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5509793506498057,
+      "learning_rate": 2.2585926756983355e-06,
+      "loss": 0.8194,
+      "step": 33463
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5189460860411728,
+      "learning_rate": 2.2581096829218008e-06,
+      "loss": 0.8128,
+      "step": 33464
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5144926279981388,
+      "learning_rate": 2.2576267352210022e-06,
+      "loss": 0.8297,
+      "step": 33465
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7717468910034431,
+      "learning_rate": 2.257143832598755e-06,
+      "loss": 0.6896,
+      "step": 33466
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4573193262648916,
+      "learning_rate": 2.256660975057867e-06,
+      "loss": 0.8058,
+      "step": 33467
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6479501120600222,
+      "learning_rate": 2.256178162601148e-06,
+      "loss": 0.7937,
+      "step": 33468
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6951926730216573,
+      "learning_rate": 2.255695395231412e-06,
+      "loss": 0.7796,
+      "step": 33469
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5880742626701703,
+      "learning_rate": 2.255212672951471e-06,
+      "loss": 0.7993,
+      "step": 33470
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8265108246576753,
+      "learning_rate": 2.2547299957641345e-06,
+      "loss": 0.8346,
+      "step": 33471
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4157389400088018,
+      "learning_rate": 2.2542473636722095e-06,
+      "loss": 0.7379,
+      "step": 33472
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4261874977587383,
+      "learning_rate": 2.2537647766785107e-06,
+      "loss": 0.7982,
+      "step": 33473
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6038735999416978,
+      "learning_rate": 2.253282234785843e-06,
+      "loss": 0.851,
+      "step": 33474
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.681533956457174,
+      "learning_rate": 2.2527997379970214e-06,
+      "loss": 0.7667,
+      "step": 33475
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7730294049974036,
+      "learning_rate": 2.25231728631485e-06,
+      "loss": 0.9193,
+      "step": 33476
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.734688685208965,
+      "learning_rate": 2.251834879742143e-06,
+      "loss": 0.7141,
+      "step": 33477
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.64365039091426,
+      "learning_rate": 2.251352518281704e-06,
+      "loss": 0.7889,
+      "step": 33478
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6407236164333268,
+      "learning_rate": 2.250870201936346e-06,
+      "loss": 0.8057,
+      "step": 33479
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8382399466607788,
+      "learning_rate": 2.250387930708875e-06,
+      "loss": 0.8807,
+      "step": 33480
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7962684272880837,
+      "learning_rate": 2.2499057046020966e-06,
+      "loss": 0.7162,
+      "step": 33481
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5594263049111168,
+      "learning_rate": 2.24942352361882e-06,
+      "loss": 0.7702,
+      "step": 33482
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.700380622810718,
+      "learning_rate": 2.248941387761857e-06,
+      "loss": 0.861,
+      "step": 33483
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7924616502747166,
+      "learning_rate": 2.2484592970340115e-06,
+      "loss": 0.7897,
+      "step": 33484
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.656643064230803,
+      "learning_rate": 2.247977251438086e-06,
+      "loss": 0.7995,
+      "step": 33485
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7567049147798126,
+      "learning_rate": 2.247495250976892e-06,
+      "loss": 0.7745,
+      "step": 33486
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6619156525548848,
+      "learning_rate": 2.2470132956532376e-06,
+      "loss": 0.7606,
+      "step": 33487
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7929350009870977,
+      "learning_rate": 2.2465313854699265e-06,
+      "loss": 0.8059,
+      "step": 33488
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7085881425952656,
+      "learning_rate": 2.2460495204297615e-06,
+      "loss": 0.7966,
+      "step": 33489
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9529853778941917,
+      "learning_rate": 2.2455677005355513e-06,
+      "loss": 0.815,
+      "step": 33490
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8217736119139063,
+      "learning_rate": 2.245085925790104e-06,
+      "loss": 0.8564,
+      "step": 33491
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.587398898557458,
+      "learning_rate": 2.2446041961962207e-06,
+      "loss": 0.7593,
+      "step": 33492
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0113976163958927,
+      "learning_rate": 2.2441225117567056e-06,
+      "loss": 0.8307,
+      "step": 33493
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7704695657786433,
+      "learning_rate": 2.2436408724743664e-06,
+      "loss": 0.9302,
+      "step": 33494
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8075465129749543,
+      "learning_rate": 2.243159278352003e-06,
+      "loss": 0.7035,
+      "step": 33495
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8072206513650768,
+      "learning_rate": 2.242677729392425e-06,
+      "loss": 0.7656,
+      "step": 33496
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4385717807832243,
+      "learning_rate": 2.242196225598431e-06,
+      "loss": 0.8409,
+      "step": 33497
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.778600378007498,
+      "learning_rate": 2.241714766972828e-06,
+      "loss": 0.7262,
+      "step": 33498
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0255709480522635,
+      "learning_rate": 2.241233353518416e-06,
+      "loss": 0.7585,
+      "step": 33499
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.423521405829967,
+      "learning_rate": 2.240751985238e-06,
+      "loss": 0.7804,
+      "step": 33500
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3773830349092084,
+      "learning_rate": 2.2402706621343873e-06,
+      "loss": 0.8682,
+      "step": 33501
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7743475036939274,
+      "learning_rate": 2.2397893842103714e-06,
+      "loss": 0.7136,
+      "step": 33502
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4385259417739376,
+      "learning_rate": 2.2393081514687564e-06,
+      "loss": 0.8414,
+      "step": 33503
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.579645365081228,
+      "learning_rate": 2.2388269639123482e-06,
+      "loss": 0.8662,
+      "step": 33504
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5275167427798255,
+      "learning_rate": 2.2383458215439513e-06,
+      "loss": 0.9697,
+      "step": 33505
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6978289539875053,
+      "learning_rate": 2.2378647243663563e-06,
+      "loss": 0.8297,
+      "step": 33506
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6158294111806852,
+      "learning_rate": 2.2373836723823714e-06,
+      "loss": 0.8168,
+      "step": 33507
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5979045549670656,
+      "learning_rate": 2.2369026655947988e-06,
+      "loss": 0.7739,
+      "step": 33508
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5879666463386188,
+      "learning_rate": 2.236421704006436e-06,
+      "loss": 0.8208,
+      "step": 33509
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7553290762170035,
+      "learning_rate": 2.2359407876200822e-06,
+      "loss": 0.8068,
+      "step": 33510
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4468560144314782,
+      "learning_rate": 2.235459916438538e-06,
+      "loss": 0.7753,
+      "step": 33511
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6029017292563381,
+      "learning_rate": 2.2349790904646085e-06,
+      "loss": 0.7986,
+      "step": 33512
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6104274712924438,
+      "learning_rate": 2.2344983097010854e-06,
+      "loss": 0.8868,
+      "step": 33513
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7222848818694547,
+      "learning_rate": 2.234017574150774e-06,
+      "loss": 0.8432,
+      "step": 33514
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.43515024206068,
+      "learning_rate": 2.2335368838164715e-06,
+      "loss": 0.7972,
+      "step": 33515
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.649021270015167,
+      "learning_rate": 2.2330562387009745e-06,
+      "loss": 0.9308,
+      "step": 33516
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5945103280308293,
+      "learning_rate": 2.2325756388070817e-06,
+      "loss": 0.718,
+      "step": 33517
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8177384668431311,
+      "learning_rate": 2.2320950841375956e-06,
+      "loss": 0.71,
+      "step": 33518
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.393548041146507,
+      "learning_rate": 2.2316145746953112e-06,
+      "loss": 0.7453,
+      "step": 33519
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7006583780431297,
+      "learning_rate": 2.231134110483023e-06,
+      "loss": 0.9752,
+      "step": 33520
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5998841267423372,
+      "learning_rate": 2.2306536915035326e-06,
+      "loss": 0.9501,
+      "step": 33521
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.679361292163013,
+      "learning_rate": 2.2301733177596407e-06,
+      "loss": 0.8382,
+      "step": 33522
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6592380587156688,
+      "learning_rate": 2.229692989254135e-06,
+      "loss": 0.8027,
+      "step": 33523
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.66474134310653,
+      "learning_rate": 2.229212705989816e-06,
+      "loss": 0.7915,
+      "step": 33524
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6985487665482237,
+      "learning_rate": 2.228732467969481e-06,
+      "loss": 0.8573,
+      "step": 33525
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5774387462422381,
+      "learning_rate": 2.228252275195929e-06,
+      "loss": 0.8593,
+      "step": 33526
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8880339062035352,
+      "learning_rate": 2.2277721276719532e-06,
+      "loss": 0.8006,
+      "step": 33527
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4792204737592154,
+      "learning_rate": 2.227292025400345e-06,
+      "loss": 0.7643,
+      "step": 33528
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7365192011378734,
+      "learning_rate": 2.226811968383906e-06,
+      "loss": 0.7093,
+      "step": 33529
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6450788497880466,
+      "learning_rate": 2.226331956625427e-06,
+      "loss": 0.828,
+      "step": 33530
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4444884085915581,
+      "learning_rate": 2.2258519901277066e-06,
+      "loss": 0.7406,
+      "step": 33531
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6297918621133907,
+      "learning_rate": 2.2253720688935344e-06,
+      "loss": 0.8739,
+      "step": 33532
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9697014032098954,
+      "learning_rate": 2.2248921929257106e-06,
+      "loss": 0.8439,
+      "step": 33533
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.82460899663233,
+      "learning_rate": 2.2244123622270218e-06,
+      "loss": 0.9061,
+      "step": 33534
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.757693157217772,
+      "learning_rate": 2.223932576800271e-06,
+      "loss": 0.8533,
+      "step": 33535
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3833795534566706,
+      "learning_rate": 2.2234528366482445e-06,
+      "loss": 0.8165,
+      "step": 33536
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5226192866101256,
+      "learning_rate": 2.2229731417737354e-06,
+      "loss": 0.8206,
+      "step": 33537
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4142320353519717,
+      "learning_rate": 2.2224934921795392e-06,
+      "loss": 0.8569,
+      "step": 33538
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.014435065394181,
+      "learning_rate": 2.2220138878684516e-06,
+      "loss": 0.8243,
+      "step": 33539
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5424074139145003,
+      "learning_rate": 2.221534328843261e-06,
+      "loss": 0.7876,
+      "step": 33540
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7815145396899923,
+      "learning_rate": 2.221054815106757e-06,
+      "loss": 0.6641,
+      "step": 33541
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4770086007361303,
+      "learning_rate": 2.2205753466617353e-06,
+      "loss": 0.869,
+      "step": 33542
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5156652736935488,
+      "learning_rate": 2.2200959235109886e-06,
+      "loss": 0.8063,
+      "step": 33543
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.651546854042159,
+      "learning_rate": 2.219616545657307e-06,
+      "loss": 0.7563,
+      "step": 33544
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9320351524818313,
+      "learning_rate": 2.2191372131034793e-06,
+      "loss": 0.8297,
+      "step": 33545
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5732515667309288,
+      "learning_rate": 2.2186579258522966e-06,
+      "loss": 0.7869,
+      "step": 33546
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.701586332645632,
+      "learning_rate": 2.2181786839065544e-06,
+      "loss": 0.8251,
+      "step": 33547
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6596293040870123,
+      "learning_rate": 2.2176994872690404e-06,
+      "loss": 0.8492,
+      "step": 33548
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.467111039581484,
+      "learning_rate": 2.2172203359425405e-06,
+      "loss": 0.7663,
+      "step": 33549
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3715700933557695,
+      "learning_rate": 2.2167412299298497e-06,
+      "loss": 0.6999,
+      "step": 33550
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8536148085891104,
+      "learning_rate": 2.2162621692337537e-06,
+      "loss": 0.9561,
+      "step": 33551
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3510486990344797,
+      "learning_rate": 2.2157831538570464e-06,
+      "loss": 0.8181,
+      "step": 33552
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6710304636787214,
+      "learning_rate": 2.2153041838025114e-06,
+      "loss": 0.7886,
+      "step": 33553
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5998429898448852,
+      "learning_rate": 2.2148252590729424e-06,
+      "loss": 0.8295,
+      "step": 33554
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7643951651832772,
+      "learning_rate": 2.2143463796711228e-06,
+      "loss": 0.7315,
+      "step": 33555
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7889645383327605,
+      "learning_rate": 2.213867545599846e-06,
+      "loss": 0.73,
+      "step": 33556
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6902398309485693,
+      "learning_rate": 2.2133887568618972e-06,
+      "loss": 0.9496,
+      "step": 33557
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5348826458156846,
+      "learning_rate": 2.2129100134600624e-06,
+      "loss": 0.8122,
+      "step": 33558
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6942484715068864,
+      "learning_rate": 2.21243131539713e-06,
+      "loss": 0.8337,
+      "step": 33559
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4641093577612931,
+      "learning_rate": 2.2119526626758914e-06,
+      "loss": 0.8074,
+      "step": 33560
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8540146810179843,
+      "learning_rate": 2.21147405529913e-06,
+      "loss": 0.7556,
+      "step": 33561
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7162082882244578,
+      "learning_rate": 2.2109954932696285e-06,
+      "loss": 0.8432,
+      "step": 33562
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5327360643947006,
+      "learning_rate": 2.210516976590179e-06,
+      "loss": 0.763,
+      "step": 33563
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7309356973299155,
+      "learning_rate": 2.2100385052635677e-06,
+      "loss": 0.8615,
+      "step": 33564
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9412070226756974,
+      "learning_rate": 2.209560079292579e-06,
+      "loss": 0.8327,
+      "step": 33565
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4515147637173125,
+      "learning_rate": 2.2090816986799934e-06,
+      "loss": 0.7179,
+      "step": 33566
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8145217073168047,
+      "learning_rate": 2.208603363428602e-06,
+      "loss": 0.7033,
+      "step": 33567
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6485415372247079,
+      "learning_rate": 2.2081250735411917e-06,
+      "loss": 0.886,
+      "step": 33568
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7430505194670143,
+      "learning_rate": 2.2076468290205433e-06,
+      "loss": 0.7355,
+      "step": 33569
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6998046242305895,
+      "learning_rate": 2.2071686298694393e-06,
+      "loss": 0.7871,
+      "step": 33570
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7821914004386333,
+      "learning_rate": 2.2066904760906704e-06,
+      "loss": 0.7233,
+      "step": 33571
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5532742275023326,
+      "learning_rate": 2.206212367687014e-06,
+      "loss": 0.7952,
+      "step": 33572
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7999794079287371,
+      "learning_rate": 2.205734304661259e-06,
+      "loss": 0.6673,
+      "step": 33573
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1577303364778704,
+      "learning_rate": 2.2052562870161832e-06,
+      "loss": 0.745,
+      "step": 33574
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6230481620402173,
+      "learning_rate": 2.2047783147545763e-06,
+      "loss": 0.8702,
+      "step": 33575
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4821261496286382,
+      "learning_rate": 2.2043003878792156e-06,
+      "loss": 0.8851,
+      "step": 33576
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3924149002778203,
+      "learning_rate": 2.2038225063928887e-06,
+      "loss": 0.7351,
+      "step": 33577
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.955737992577079,
+      "learning_rate": 2.203344670298375e-06,
+      "loss": 0.921,
+      "step": 33578
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7315071922271141,
+      "learning_rate": 2.202866879598454e-06,
+      "loss": 0.8282,
+      "step": 33579
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7706291572177457,
+      "learning_rate": 2.2023891342959114e-06,
+      "loss": 0.8727,
+      "step": 33580
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.646340984303448,
+      "learning_rate": 2.20191143439353e-06,
+      "loss": 0.8109,
+      "step": 33581
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7257030298203464,
+      "learning_rate": 2.20143377989409e-06,
+      "loss": 0.9345,
+      "step": 33582
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.545026824269654,
+      "learning_rate": 2.200956170800368e-06,
+      "loss": 0.744,
+      "step": 33583
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.181822317780578,
+      "learning_rate": 2.2004786071151486e-06,
+      "loss": 0.837,
+      "step": 33584
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.274835097055456,
+      "learning_rate": 2.2000010888412137e-06,
+      "loss": 0.7239,
+      "step": 33585
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5064878640003774,
+      "learning_rate": 2.1995236159813437e-06,
+      "loss": 0.9529,
+      "step": 33586
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6492805704316078,
+      "learning_rate": 2.199046188538313e-06,
+      "loss": 0.9093,
+      "step": 33587
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.508789472190252,
+      "learning_rate": 2.198568806514906e-06,
+      "loss": 0.7869,
+      "step": 33588
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5848775974203617,
+      "learning_rate": 2.198091469913902e-06,
+      "loss": 0.7835,
+      "step": 33589
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5493090853246128,
+      "learning_rate": 2.197614178738081e-06,
+      "loss": 0.9384,
+      "step": 33590
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0237483554234266,
+      "learning_rate": 2.197136932990218e-06,
+      "loss": 0.8356,
+      "step": 33591
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.092013876408737,
+      "learning_rate": 2.1966597326730944e-06,
+      "loss": 0.9724,
+      "step": 33592
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9180346712400707,
+      "learning_rate": 2.19618257778949e-06,
+      "loss": 0.7621,
+      "step": 33593
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5647903475109946,
+      "learning_rate": 2.1957054683421818e-06,
+      "loss": 0.8352,
+      "step": 33594
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3341424783099236,
+      "learning_rate": 2.1952284043339443e-06,
+      "loss": 0.9074,
+      "step": 33595
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.846322956927842,
+      "learning_rate": 2.194751385767561e-06,
+      "loss": 0.7854,
+      "step": 33596
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6054998649727013,
+      "learning_rate": 2.1942744126458027e-06,
+      "loss": 0.8258,
+      "step": 33597
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3876865120347226,
+      "learning_rate": 2.1937974849714505e-06,
+      "loss": 0.7991,
+      "step": 33598
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9465858187747713,
+      "learning_rate": 2.193320602747283e-06,
+      "loss": 0.7738,
+      "step": 33599
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7349526168340181,
+      "learning_rate": 2.192843765976075e-06,
+      "loss": 0.8629,
+      "step": 33600
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6362668711303558,
+      "learning_rate": 2.1923669746606004e-06,
+      "loss": 0.8508,
+      "step": 33601
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.505155953246246,
+      "learning_rate": 2.191890228803636e-06,
+      "loss": 0.856,
+      "step": 33602
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6276412042953492,
+      "learning_rate": 2.191413528407964e-06,
+      "loss": 0.9096,
+      "step": 33603
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.661258953845482,
+      "learning_rate": 2.1909368734763505e-06,
+      "loss": 0.9856,
+      "step": 33604
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4052376116695176,
+      "learning_rate": 2.1904602640115733e-06,
+      "loss": 0.9225,
+      "step": 33605
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.626936147945132,
+      "learning_rate": 2.189983700016409e-06,
+      "loss": 0.8997,
+      "step": 33606
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5651050280122256,
+      "learning_rate": 2.1895071814936386e-06,
+      "loss": 0.9001,
+      "step": 33607
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.908876989928664,
+      "learning_rate": 2.189030708446024e-06,
+      "loss": 0.8046,
+      "step": 33608
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5755785234568147,
+      "learning_rate": 2.188554280876346e-06,
+      "loss": 0.8715,
+      "step": 33609
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5509259575456913,
+      "learning_rate": 2.1880778987873806e-06,
+      "loss": 0.8567,
+      "step": 33610
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5303994509468237,
+      "learning_rate": 2.1876015621818956e-06,
+      "loss": 0.8652,
+      "step": 33611
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.601307147864735,
+      "learning_rate": 2.187125271062671e-06,
+      "loss": 0.9541,
+      "step": 33612
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7384548886790567,
+      "learning_rate": 2.1866490254324745e-06,
+      "loss": 0.6435,
+      "step": 33613
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.004280816490065,
+      "learning_rate": 2.186172825294083e-06,
+      "loss": 0.8477,
+      "step": 33614
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5101188043044877,
+      "learning_rate": 2.1856966706502647e-06,
+      "loss": 0.7889,
+      "step": 33615
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4389515976843053,
+      "learning_rate": 2.1852205615037968e-06,
+      "loss": 0.8413,
+      "step": 33616
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4171228472424455,
+      "learning_rate": 2.1847444978574485e-06,
+      "loss": 0.8954,
+      "step": 33617
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7630478821952296,
+      "learning_rate": 2.18426847971399e-06,
+      "loss": 0.7653,
+      "step": 33618
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7886280972754119,
+      "learning_rate": 2.1837925070761946e-06,
+      "loss": 0.7346,
+      "step": 33619
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7080291497677921,
+      "learning_rate": 2.183316579946836e-06,
+      "loss": 0.6661,
+      "step": 33620
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4655741864819594,
+      "learning_rate": 2.1828406983286832e-06,
+      "loss": 0.7771,
+      "step": 33621
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6348603945995965,
+      "learning_rate": 2.1823648622245043e-06,
+      "loss": 0.8093,
+      "step": 33622
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5309172416696137,
+      "learning_rate": 2.181889071637072e-06,
+      "loss": 0.8599,
+      "step": 33623
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7348279234748039,
+      "learning_rate": 2.1814133265691594e-06,
+      "loss": 0.8408,
+      "step": 33624
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5355468298560389,
+      "learning_rate": 2.1809376270235348e-06,
+      "loss": 0.846,
+      "step": 33625
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9519117897795237,
+      "learning_rate": 2.180461973002963e-06,
+      "loss": 0.8728,
+      "step": 33626
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7116527175637213,
+      "learning_rate": 2.179986364510217e-06,
+      "loss": 0.7703,
+      "step": 33627
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.748853386194565,
+      "learning_rate": 2.1795108015480694e-06,
+      "loss": 0.6726,
+      "step": 33628
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7031144198009194,
+      "learning_rate": 2.1790352841192863e-06,
+      "loss": 0.868,
+      "step": 33629
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6124891749999548,
+      "learning_rate": 2.1785598122266327e-06,
+      "loss": 0.7062,
+      "step": 33630
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4148264248675584,
+      "learning_rate": 2.178084385872882e-06,
+      "loss": 0.754,
+      "step": 33631
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7117479949357588,
+      "learning_rate": 2.177609005060799e-06,
+      "loss": 0.8106,
+      "step": 33632
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6526505153139914,
+      "learning_rate": 2.1771336697931544e-06,
+      "loss": 0.8178,
+      "step": 33633
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7023050975260179,
+      "learning_rate": 2.1766583800727128e-06,
+      "loss": 0.7919,
+      "step": 33634
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5463423004371788,
+      "learning_rate": 2.1761831359022446e-06,
+      "loss": 0.8269,
+      "step": 33635
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7585545522324354,
+      "learning_rate": 2.1757079372845125e-06,
+      "loss": 0.6868,
+      "step": 33636
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4787908184430703,
+      "learning_rate": 2.175232784222289e-06,
+      "loss": 0.7756,
+      "step": 33637
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.82374163631227,
+      "learning_rate": 2.1747576767183374e-06,
+      "loss": 0.7938,
+      "step": 33638
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8213186320922745,
+      "learning_rate": 2.1742826147754213e-06,
+      "loss": 0.6933,
+      "step": 33639
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.250887846465551,
+      "learning_rate": 2.173807598396309e-06,
+      "loss": 0.7746,
+      "step": 33640
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6034013980133457,
+      "learning_rate": 2.1733326275837708e-06,
+      "loss": 0.7751,
+      "step": 33641
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7286101888359624,
+      "learning_rate": 2.172857702340566e-06,
+      "loss": 0.864,
+      "step": 33642
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7113822862782069,
+      "learning_rate": 2.17238282266946e-06,
+      "loss": 0.9226,
+      "step": 33643
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.609441134786767,
+      "learning_rate": 2.17190798857322e-06,
+      "loss": 0.7904,
+      "step": 33644
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6762450813262673,
+      "learning_rate": 2.1714332000546114e-06,
+      "loss": 0.7145,
+      "step": 33645
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1709752420216635,
+      "learning_rate": 2.1709584571163976e-06,
+      "loss": 0.7573,
+      "step": 33646
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.534091769002438,
+      "learning_rate": 2.1704837597613406e-06,
+      "loss": 0.9101,
+      "step": 33647
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8667063298527753,
+      "learning_rate": 2.1700091079922047e-06,
+      "loss": 0.939,
+      "step": 33648
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5437129566382903,
+      "learning_rate": 2.1695345018117577e-06,
+      "loss": 0.8965,
+      "step": 33649
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7156386125204546,
+      "learning_rate": 2.16905994122276e-06,
+      "loss": 0.8787,
+      "step": 33650
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7141601006869114,
+      "learning_rate": 2.1685854262279714e-06,
+      "loss": 0.8958,
+      "step": 33651
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9347520248829024,
+      "learning_rate": 2.1681109568301617e-06,
+      "loss": 0.6468,
+      "step": 33652
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.701876779861756,
+      "learning_rate": 2.1676365330320857e-06,
+      "loss": 0.8383,
+      "step": 33653
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7563230315696415,
+      "learning_rate": 2.167162154836513e-06,
+      "loss": 0.6998,
+      "step": 33654
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5060951240085965,
+      "learning_rate": 2.1666878222461986e-06,
+      "loss": 0.7921,
+      "step": 33655
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.572727403148007,
+      "learning_rate": 2.1662135352639115e-06,
+      "loss": 0.8935,
+      "step": 33656
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4688862889111367,
+      "learning_rate": 2.1657392938924047e-06,
+      "loss": 0.8667,
+      "step": 33657
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.957672453652838,
+      "learning_rate": 2.165265098134448e-06,
+      "loss": 0.7377,
+      "step": 33658
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5451990750831541,
+      "learning_rate": 2.1647909479927987e-06,
+      "loss": 0.8503,
+      "step": 33659
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7787403025975073,
+      "learning_rate": 2.1643168434702142e-06,
+      "loss": 0.8882,
+      "step": 33660
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.767011305943306,
+      "learning_rate": 2.1638427845694574e-06,
+      "loss": 0.6749,
+      "step": 33661
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4834107082955297,
+      "learning_rate": 2.1633687712932917e-06,
+      "loss": 0.7392,
+      "step": 33662
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7807714021708783,
+      "learning_rate": 2.1628948036444743e-06,
+      "loss": 0.6852,
+      "step": 33663
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4594945027741382,
+      "learning_rate": 2.162420881625761e-06,
+      "loss": 0.7671,
+      "step": 33664
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4909935712713664,
+      "learning_rate": 2.161947005239915e-06,
+      "loss": 0.815,
+      "step": 33665
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5062315133877264,
+      "learning_rate": 2.1614731744896976e-06,
+      "loss": 0.7327,
+      "step": 33666
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.487016596827307,
+      "learning_rate": 2.160999389377866e-06,
+      "loss": 0.8245,
+      "step": 33667
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6910862935102111,
+      "learning_rate": 2.160525649907175e-06,
+      "loss": 0.8278,
+      "step": 33668
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6824106297310635,
+      "learning_rate": 2.1600519560803844e-06,
+      "loss": 1.0382,
+      "step": 33669
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7486290191119571,
+      "learning_rate": 2.1595783079002565e-06,
+      "loss": 0.8159,
+      "step": 33670
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5995571286742445,
+      "learning_rate": 2.159104705369546e-06,
+      "loss": 0.9068,
+      "step": 33671
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5757719677822588,
+      "learning_rate": 2.158631148491007e-06,
+      "loss": 0.9212,
+      "step": 33672
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.499896858173607,
+      "learning_rate": 2.158157637267404e-06,
+      "loss": 0.8157,
+      "step": 33673
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5937499372909454,
+      "learning_rate": 2.1576841717014864e-06,
+      "loss": 0.8148,
+      "step": 33674
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6816335532505484,
+      "learning_rate": 2.1572107517960174e-06,
+      "loss": 0.8319,
+      "step": 33675
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4613158050687027,
+      "learning_rate": 2.156737377553748e-06,
+      "loss": 0.7554,
+      "step": 33676
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5589321397184128,
+      "learning_rate": 2.1562640489774388e-06,
+      "loss": 0.8526,
+      "step": 33677
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4796119203604905,
+      "learning_rate": 2.1557907660698408e-06,
+      "loss": 0.871,
+      "step": 33678
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.567736964846587,
+      "learning_rate": 2.155317528833716e-06,
+      "loss": 0.9314,
+      "step": 33679
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4646658527712304,
+      "learning_rate": 2.1548443372718153e-06,
+      "loss": 0.8234,
+      "step": 33680
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4943908484219364,
+      "learning_rate": 2.1543711913868915e-06,
+      "loss": 0.824,
+      "step": 33681
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4707789553116457,
+      "learning_rate": 2.153898091181703e-06,
+      "loss": 0.8213,
+      "step": 33682
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5376690226477214,
+      "learning_rate": 2.153425036659007e-06,
+      "loss": 0.7695,
+      "step": 33683
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6314819766290067,
+      "learning_rate": 2.1529520278215543e-06,
+      "loss": 0.7572,
+      "step": 33684
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5690195084179157,
+      "learning_rate": 2.152479064672096e-06,
+      "loss": 0.834,
+      "step": 33685
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4596826474901698,
+      "learning_rate": 2.1520061472133903e-06,
+      "loss": 0.7803,
+      "step": 33686
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.703115221187424,
+      "learning_rate": 2.151533275448191e-06,
+      "loss": 0.8448,
+      "step": 33687
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5522566235994524,
+      "learning_rate": 2.151060449379251e-06,
+      "loss": 0.8698,
+      "step": 33688
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6972163119686217,
+      "learning_rate": 2.1505876690093176e-06,
+      "loss": 0.8683,
+      "step": 33689
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0384997377380785,
+      "learning_rate": 2.1501149343411486e-06,
+      "loss": 0.8374,
+      "step": 33690
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4744379886209313,
+      "learning_rate": 2.149642245377499e-06,
+      "loss": 0.8513,
+      "step": 33691
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3785940659803788,
+      "learning_rate": 2.1491696021211174e-06,
+      "loss": 0.8203,
+      "step": 33692
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6864032754152085,
+      "learning_rate": 2.1486970045747525e-06,
+      "loss": 0.7602,
+      "step": 33693
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5822157383382116,
+      "learning_rate": 2.148224452741162e-06,
+      "loss": 0.79,
+      "step": 33694
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.570720138709063,
+      "learning_rate": 2.1477519466230935e-06,
+      "loss": 0.8359,
+      "step": 33695
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4857785025466512,
+      "learning_rate": 2.147279486223297e-06,
+      "loss": 1.0153,
+      "step": 33696
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.731385194686662,
+      "learning_rate": 2.14680707154453e-06,
+      "loss": 0.7007,
+      "step": 33697
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4236443012804059,
+      "learning_rate": 2.1463347025895375e-06,
+      "loss": 0.9575,
+      "step": 33698
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6268130341898062,
+      "learning_rate": 2.1458623793610678e-06,
+      "loss": 0.9655,
+      "step": 33699
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3186587710299136,
+      "learning_rate": 2.145390101861875e-06,
+      "loss": 0.7952,
+      "step": 33700
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5780390901252013,
+      "learning_rate": 2.1449178700947125e-06,
+      "loss": 0.8345,
+      "step": 33701
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.578010049657664,
+      "learning_rate": 2.1444456840623195e-06,
+      "loss": 0.8262,
+      "step": 33702
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.855408041346746,
+      "learning_rate": 2.1439735437674504e-06,
+      "loss": 0.7768,
+      "step": 33703
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6909432130993782,
+      "learning_rate": 2.1435014492128547e-06,
+      "loss": 0.823,
+      "step": 33704
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.341964821767395,
+      "learning_rate": 2.1430294004012874e-06,
+      "loss": 0.6786,
+      "step": 33705
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4900303787490807,
+      "learning_rate": 2.142557397335484e-06,
+      "loss": 0.8675,
+      "step": 33706
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8786288347790946,
+      "learning_rate": 2.142085440018199e-06,
+      "loss": 0.7151,
+      "step": 33707
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9170626832254387,
+      "learning_rate": 2.1416135284521812e-06,
+      "loss": 0.7871,
+      "step": 33708
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.065044117508884,
+      "learning_rate": 2.1411416626401793e-06,
+      "loss": 0.8832,
+      "step": 33709
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6195296332799902,
+      "learning_rate": 2.1406698425849383e-06,
+      "loss": 0.9243,
+      "step": 33710
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8818359763188004,
+      "learning_rate": 2.140198068289204e-06,
+      "loss": 0.7526,
+      "step": 33711
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5891550430481833,
+      "learning_rate": 2.1397263397557266e-06,
+      "loss": 0.8777,
+      "step": 33712
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4550438435047637,
+      "learning_rate": 2.1392546569872498e-06,
+      "loss": 0.9066,
+      "step": 33713
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6178695310159816,
+      "learning_rate": 2.1387830199865224e-06,
+      "loss": 0.8118,
+      "step": 33714
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8472857517296537,
+      "learning_rate": 2.138311428756287e-06,
+      "loss": 1.0183,
+      "step": 33715
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9480173395009244,
+      "learning_rate": 2.137839883299294e-06,
+      "loss": 0.9202,
+      "step": 33716
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7258842200473343,
+      "learning_rate": 2.1373683836182835e-06,
+      "loss": 0.8964,
+      "step": 33717
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7777541426271235,
+      "learning_rate": 2.136896929716006e-06,
+      "loss": 0.7106,
+      "step": 33718
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4994127186301105,
+      "learning_rate": 2.1364255215952047e-06,
+      "loss": 0.8305,
+      "step": 33719
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.434779485066591,
+      "learning_rate": 2.1359541592586197e-06,
+      "loss": 0.8494,
+      "step": 33720
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.560983147450561,
+      "learning_rate": 2.1354828427090004e-06,
+      "loss": 0.8779,
+      "step": 33721
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6799591163311542,
+      "learning_rate": 2.135011571949093e-06,
+      "loss": 0.9472,
+      "step": 33722
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6281044451241247,
+      "learning_rate": 2.134540346981637e-06,
+      "loss": 0.9299,
+      "step": 33723
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5919661763909054,
+      "learning_rate": 2.134069167809375e-06,
+      "loss": 0.8581,
+      "step": 33724
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5019897944692187,
+      "learning_rate": 2.1335980344350527e-06,
+      "loss": 0.8944,
+      "step": 33725
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.576876386409334,
+      "learning_rate": 2.1331269468614156e-06,
+      "loss": 0.774,
+      "step": 33726
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.43157164691752,
+      "learning_rate": 2.1326559050912042e-06,
+      "loss": 0.7314,
+      "step": 33727
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5045485571084913,
+      "learning_rate": 2.1321849091271573e-06,
+      "loss": 0.8166,
+      "step": 33728
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.916269513575731,
+      "learning_rate": 2.1317139589720227e-06,
+      "loss": 0.8345,
+      "step": 33729
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.073827910455265,
+      "learning_rate": 2.1312430546285414e-06,
+      "loss": 0.8028,
+      "step": 33730
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8401260365696022,
+      "learning_rate": 2.1307721960994556e-06,
+      "loss": 0.8861,
+      "step": 33731
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6742115783359797,
+      "learning_rate": 2.130301383387502e-06,
+      "loss": 0.8124,
+      "step": 33732
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4756025292727941,
+      "learning_rate": 2.1298306164954275e-06,
+      "loss": 0.7359,
+      "step": 33733
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6744939158989045,
+      "learning_rate": 2.129359895425969e-06,
+      "loss": 0.7848,
+      "step": 33734
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4437260968833219,
+      "learning_rate": 2.128889220181871e-06,
+      "loss": 0.8216,
+      "step": 33735
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5360115091701951,
+      "learning_rate": 2.1284185907658695e-06,
+      "loss": 0.7933,
+      "step": 33736
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.147498838550252,
+      "learning_rate": 2.12794800718071e-06,
+      "loss": 0.845,
+      "step": 33737
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9611532805700307,
+      "learning_rate": 2.1274774694291267e-06,
+      "loss": 0.8665,
+      "step": 33738
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6438963621198797,
+      "learning_rate": 2.1270069775138647e-06,
+      "loss": 0.8827,
+      "step": 33739
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7724238276209197,
+      "learning_rate": 2.1265365314376598e-06,
+      "loss": 0.887,
+      "step": 33740
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.837872904961689,
+      "learning_rate": 2.1260661312032506e-06,
+      "loss": 0.9389,
+      "step": 33741
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5450760627321938,
+      "learning_rate": 2.125595776813376e-06,
+      "loss": 0.8301,
+      "step": 33742
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5259568329558184,
+      "learning_rate": 2.125125468270779e-06,
+      "loss": 0.8279,
+      "step": 33743
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6704047644863238,
+      "learning_rate": 2.124655205578194e-06,
+      "loss": 0.8562,
+      "step": 33744
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6718846399790266,
+      "learning_rate": 2.1241849887383583e-06,
+      "loss": 0.8198,
+      "step": 33745
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6675625701459695,
+      "learning_rate": 2.12371481775401e-06,
+      "loss": 0.8484,
+      "step": 33746
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.057876572262276,
+      "learning_rate": 2.1232446926278905e-06,
+      "loss": 0.7107,
+      "step": 33747
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.511359772770021,
+      "learning_rate": 2.1227746133627346e-06,
+      "loss": 0.8827,
+      "step": 33748
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6328946648248948,
+      "learning_rate": 2.1223045799612763e-06,
+      "loss": 0.8567,
+      "step": 33749
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5220273712404215,
+      "learning_rate": 2.121834592426254e-06,
+      "loss": 0.7157,
+      "step": 33750
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.704803107062752,
+      "learning_rate": 2.121364650760408e-06,
+      "loss": 0.8098,
+      "step": 33751
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7756132210702711,
+      "learning_rate": 2.1208947549664706e-06,
+      "loss": 0.715,
+      "step": 33752
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5139652767888512,
+      "learning_rate": 2.120424905047176e-06,
+      "loss": 0.8441,
+      "step": 33753
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.658917413402619,
+      "learning_rate": 2.1199551010052653e-06,
+      "loss": 0.7243,
+      "step": 33754
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5857045718362675,
+      "learning_rate": 2.119485342843468e-06,
+      "loss": 0.8509,
+      "step": 33755
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6549564231717844,
+      "learning_rate": 2.1190156305645247e-06,
+      "loss": 0.8463,
+      "step": 33756
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5190409542036354,
+      "learning_rate": 2.1185459641711646e-06,
+      "loss": 0.8688,
+      "step": 33757
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.779630662301675,
+      "learning_rate": 2.118076343666129e-06,
+      "loss": 0.905,
+      "step": 33758
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7356747317690218,
+      "learning_rate": 2.117606769052144e-06,
+      "loss": 0.8057,
+      "step": 33759
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7162500749809213,
+      "learning_rate": 2.1171372403319514e-06,
+      "loss": 0.8855,
+      "step": 33760
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6486343123748946,
+      "learning_rate": 2.116667757508282e-06,
+      "loss": 0.9309,
+      "step": 33761
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5034830093098144,
+      "learning_rate": 2.1161983205838645e-06,
+      "loss": 0.8822,
+      "step": 33762
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.99915299155344,
+      "learning_rate": 2.115728929561437e-06,
+      "loss": 0.885,
+      "step": 33763
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5850215135565082,
+      "learning_rate": 2.115259584443735e-06,
+      "loss": 0.83,
+      "step": 33764
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8035316820486658,
+      "learning_rate": 2.114790285233488e-06,
+      "loss": 0.6461,
+      "step": 33765
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9008650201696453,
+      "learning_rate": 2.114321031933425e-06,
+      "loss": 0.771,
+      "step": 33766
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.599116040724131,
+      "learning_rate": 2.1138518245462816e-06,
+      "loss": 0.7737,
+      "step": 33767
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.717350919098574,
+      "learning_rate": 2.113382663074792e-06,
+      "loss": 0.852,
+      "step": 33768
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6529121334261838,
+      "learning_rate": 2.1129135475216866e-06,
+      "loss": 0.7391,
+      "step": 33769
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5627194197266348,
+      "learning_rate": 2.1124444778896914e-06,
+      "loss": 0.7416,
+      "step": 33770
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5419351474045964,
+      "learning_rate": 2.111975454181543e-06,
+      "loss": 0.8937,
+      "step": 33771
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8229809159045126,
+      "learning_rate": 2.1115064763999726e-06,
+      "loss": 0.6949,
+      "step": 33772
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7394283464150742,
+      "learning_rate": 2.1110375445477093e-06,
+      "loss": 0.8957,
+      "step": 33773
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6286965595660987,
+      "learning_rate": 2.1105686586274797e-06,
+      "loss": 0.8905,
+      "step": 33774
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6800539074860026,
+      "learning_rate": 2.110099818642021e-06,
+      "loss": 0.7751,
+      "step": 33775
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8341770008435812,
+      "learning_rate": 2.1096310245940555e-06,
+      "loss": 0.8483,
+      "step": 33776
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6294947161295381,
+      "learning_rate": 2.1091622764863185e-06,
+      "loss": 0.8593,
+      "step": 33777
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5606030172760448,
+      "learning_rate": 2.1086935743215354e-06,
+      "loss": 0.8545,
+      "step": 33778
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7971555515603304,
+      "learning_rate": 2.1082249181024384e-06,
+      "loss": 0.8117,
+      "step": 33779
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4925086610212819,
+      "learning_rate": 2.107756307831752e-06,
+      "loss": 0.743,
+      "step": 33780
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7874745888364432,
+      "learning_rate": 2.1072877435122095e-06,
+      "loss": 0.7087,
+      "step": 33781
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4503230290334703,
+      "learning_rate": 2.106819225146538e-06,
+      "loss": 0.7844,
+      "step": 33782
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.657696739692737,
+      "learning_rate": 2.1063507527374594e-06,
+      "loss": 0.8691,
+      "step": 33783
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8310675049968965,
+      "learning_rate": 2.105882326287707e-06,
+      "loss": 0.9102,
+      "step": 33784
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.43083915589012,
+      "learning_rate": 2.105413945800007e-06,
+      "loss": 0.8382,
+      "step": 33785
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7955314450159897,
+      "learning_rate": 2.1049456112770907e-06,
+      "loss": 0.6872,
+      "step": 33786
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.923944390330958,
+      "learning_rate": 2.104477322721675e-06,
+      "loss": 0.8175,
+      "step": 33787
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6702172710690684,
+      "learning_rate": 2.104009080136493e-06,
+      "loss": 0.7806,
+      "step": 33788
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7803411991268083,
+      "learning_rate": 2.103540883524272e-06,
+      "loss": 0.6833,
+      "step": 33789
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5537415350239518,
+      "learning_rate": 2.1030727328877367e-06,
+      "loss": 0.809,
+      "step": 33790
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.712520443621245,
+      "learning_rate": 2.1026046282296085e-06,
+      "loss": 0.8234,
+      "step": 33791
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9823568366185975,
+      "learning_rate": 2.1021365695526176e-06,
+      "loss": 0.8395,
+      "step": 33792
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.94532465893551,
+      "learning_rate": 2.10166855685949e-06,
+      "loss": 0.9295,
+      "step": 33793
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.068159907612467,
+      "learning_rate": 2.1012005901529463e-06,
+      "loss": 0.8687,
+      "step": 33794
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9349290742978034,
+      "learning_rate": 2.1007326694357153e-06,
+      "loss": 0.7056,
+      "step": 33795
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.783606766351841,
+      "learning_rate": 2.1002647947105205e-06,
+      "loss": 0.7049,
+      "step": 33796
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4799714591786608,
+      "learning_rate": 2.099796965980082e-06,
+      "loss": 0.7794,
+      "step": 33797
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.631827497232184,
+      "learning_rate": 2.099329183247126e-06,
+      "loss": 0.7504,
+      "step": 33798
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.388811870870678,
+      "learning_rate": 2.0988614465143807e-06,
+      "loss": 0.7848,
+      "step": 33799
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.597658420922832,
+      "learning_rate": 2.0983937557845647e-06,
+      "loss": 0.8917,
+      "step": 33800
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7689823727349491,
+      "learning_rate": 2.0979261110604e-06,
+      "loss": 0.7368,
+      "step": 33801
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7906731351480976,
+      "learning_rate": 2.09745851234461e-06,
+      "loss": 0.8739,
+      "step": 33802
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8872050436460974,
+      "learning_rate": 2.0969909596399253e-06,
+      "loss": 0.8417,
+      "step": 33803
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7759442784255619,
+      "learning_rate": 2.0965234529490552e-06,
+      "loss": 0.703,
+      "step": 33804
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5161091063468075,
+      "learning_rate": 2.096055992274727e-06,
+      "loss": 0.8543,
+      "step": 33805
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6239302254363437,
+      "learning_rate": 2.0955885776196637e-06,
+      "loss": 0.8774,
+      "step": 33806
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7812495208099624,
+      "learning_rate": 2.09512120898659e-06,
+      "loss": 0.8482,
+      "step": 33807
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6760364481053462,
+      "learning_rate": 2.0946538863782216e-06,
+      "loss": 0.8401,
+      "step": 33808
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.035395698918301,
+      "learning_rate": 2.0941866097972797e-06,
+      "loss": 0.7564,
+      "step": 33809
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.580831654437184,
+      "learning_rate": 2.0937193792464883e-06,
+      "loss": 0.7688,
+      "step": 33810
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6699575371935838,
+      "learning_rate": 2.0932521947285635e-06,
+      "loss": 0.9922,
+      "step": 33811
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5994700620348226,
+      "learning_rate": 2.092785056246229e-06,
+      "loss": 0.8262,
+      "step": 33812
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6622296027807633,
+      "learning_rate": 2.092317963802202e-06,
+      "loss": 0.7714,
+      "step": 33813
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5809541986988107,
+      "learning_rate": 2.091850917399206e-06,
+      "loss": 0.9405,
+      "step": 33814
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5419268266396495,
+      "learning_rate": 2.091383917039954e-06,
+      "loss": 0.7953,
+      "step": 33815
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7456182371695338,
+      "learning_rate": 2.090916962727172e-06,
+      "loss": 0.6864,
+      "step": 33816
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5782136919116834,
+      "learning_rate": 2.0904500544635752e-06,
+      "loss": 0.9182,
+      "step": 33817
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6265393240696209,
+      "learning_rate": 2.08998319225188e-06,
+      "loss": 0.7491,
+      "step": 33818
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.027462153883574,
+      "learning_rate": 2.089516376094807e-06,
+      "loss": 0.8019,
+      "step": 33819
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7562410075928534,
+      "learning_rate": 2.089049605995076e-06,
+      "loss": 0.8451,
+      "step": 33820
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5999096253785352,
+      "learning_rate": 2.0885828819554035e-06,
+      "loss": 0.8241,
+      "step": 33821
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5964872836704722,
+      "learning_rate": 2.088116203978503e-06,
+      "loss": 0.9172,
+      "step": 33822
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6915345711278398,
+      "learning_rate": 2.087649572067095e-06,
+      "loss": 0.7567,
+      "step": 33823
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4980211913559884,
+      "learning_rate": 2.087182986223899e-06,
+      "loss": 0.7842,
+      "step": 33824
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8147802065958774,
+      "learning_rate": 2.0867164464516287e-06,
+      "loss": 0.7879,
+      "step": 33825
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3760957321268834,
+      "learning_rate": 2.086249952752998e-06,
+      "loss": 0.7978,
+      "step": 33826
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5152761754450816,
+      "learning_rate": 2.0857835051307253e-06,
+      "loss": 0.7536,
+      "step": 33827
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.849314428833928,
+      "learning_rate": 2.0853171035875296e-06,
+      "loss": 0.8481,
+      "step": 33828
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7773761262604117,
+      "learning_rate": 2.084850748126124e-06,
+      "loss": 0.661,
+      "step": 33829
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5653971657563024,
+      "learning_rate": 2.0843844387492197e-06,
+      "loss": 0.8792,
+      "step": 33830
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7582073905235616,
+      "learning_rate": 2.083918175459535e-06,
+      "loss": 0.9275,
+      "step": 33831
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.982146535682623,
+      "learning_rate": 2.083451958259788e-06,
+      "loss": 0.7651,
+      "step": 33832
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6162808092902206,
+      "learning_rate": 2.0829857871526904e-06,
+      "loss": 0.7635,
+      "step": 33833
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7393998444800438,
+      "learning_rate": 2.0825196621409526e-06,
+      "loss": 0.6778,
+      "step": 33834
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7687835865469785,
+      "learning_rate": 2.0820535832272947e-06,
+      "loss": 0.7552,
+      "step": 33835
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.727054632664574,
+      "learning_rate": 2.0815875504144257e-06,
+      "loss": 0.7079,
+      "step": 33836
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.088627792667368,
+      "learning_rate": 2.081121563705063e-06,
+      "loss": 0.8283,
+      "step": 33837
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7994782872858892,
+      "learning_rate": 2.080655623101915e-06,
+      "loss": 0.936,
+      "step": 33838
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6876111002456868,
+      "learning_rate": 2.0801897286077e-06,
+      "loss": 0.893,
+      "step": 33839
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8301046329241841,
+      "learning_rate": 2.0797238802251252e-06,
+      "loss": 0.8082,
+      "step": 33840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4247617007293376,
+      "learning_rate": 2.079258077956907e-06,
+      "loss": 0.9338,
+      "step": 33841
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5866873266659836,
+      "learning_rate": 2.078792321805757e-06,
+      "loss": 0.8355,
+      "step": 33842
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.648503938701592,
+      "learning_rate": 2.078326611774384e-06,
+      "loss": 0.8182,
+      "step": 33843
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.558716199085991,
+      "learning_rate": 2.0778609478655e-06,
+      "loss": 0.7399,
+      "step": 33844
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8157247092624371,
+      "learning_rate": 2.0773953300818204e-06,
+      "loss": 0.7366,
+      "step": 33845
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.406393210762271,
+      "learning_rate": 2.076929758426055e-06,
+      "loss": 0.914,
+      "step": 33846
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7489236395430076,
+      "learning_rate": 2.0764642329009087e-06,
+      "loss": 0.7482,
+      "step": 33847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.749358512434171,
+      "learning_rate": 2.0759987535090966e-06,
+      "loss": 0.8392,
+      "step": 33848
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8356243321093364,
+      "learning_rate": 2.075533320253331e-06,
+      "loss": 0.9551,
+      "step": 33849
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5277662532416245,
+      "learning_rate": 2.07506793313632e-06,
+      "loss": 0.822,
+      "step": 33850
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.511779664835189,
+      "learning_rate": 2.0746025921607695e-06,
+      "loss": 0.894,
+      "step": 33851
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.758552564741132,
+      "learning_rate": 2.0741372973293915e-06,
+      "loss": 0.9601,
+      "step": 33852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5959816783881784,
+      "learning_rate": 2.073672048644898e-06,
+      "loss": 0.7961,
+      "step": 33853
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4834525555486862,
+      "learning_rate": 2.0732068461099952e-06,
+      "loss": 0.8044,
+      "step": 33854
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.984066082832385,
+      "learning_rate": 2.07274168972739e-06,
+      "loss": 0.8026,
+      "step": 33855
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6002485718223594,
+      "learning_rate": 2.0722765794997947e-06,
+      "loss": 0.7684,
+      "step": 33856
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6012429602778278,
+      "learning_rate": 2.0718115154299113e-06,
+      "loss": 0.821,
+      "step": 33857
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8141614084322362,
+      "learning_rate": 2.0713464975204556e-06,
+      "loss": 0.8579,
+      "step": 33858
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.618577335563039,
+      "learning_rate": 2.0708815257741264e-06,
+      "loss": 0.8543,
+      "step": 33859
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7143337145803343,
+      "learning_rate": 2.0704166001936387e-06,
+      "loss": 0.9039,
+      "step": 33860
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9475845365689366,
+      "learning_rate": 2.0699517207816944e-06,
+      "loss": 0.813,
+      "step": 33861
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.259154121082319,
+      "learning_rate": 2.0694868875410036e-06,
+      "loss": 0.975,
+      "step": 33862
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6205459704055445,
+      "learning_rate": 2.0690221004742704e-06,
+      "loss": 0.7934,
+      "step": 33863
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5374357054597234,
+      "learning_rate": 2.068557359584199e-06,
+      "loss": 0.8621,
+      "step": 33864
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5495735442204117,
+      "learning_rate": 2.068092664873498e-06,
+      "loss": 0.9293,
+      "step": 33865
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7620688237471097,
+      "learning_rate": 2.067628016344875e-06,
+      "loss": 0.7032,
+      "step": 33866
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5205953193866244,
+      "learning_rate": 2.0671634140010333e-06,
+      "loss": 0.7428,
+      "step": 33867
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6655017833312666,
+      "learning_rate": 2.0666988578446757e-06,
+      "loss": 0.8613,
+      "step": 33868
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7496731307464402,
+      "learning_rate": 2.0662343478785074e-06,
+      "loss": 0.8684,
+      "step": 33869
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5382345026033828,
+      "learning_rate": 2.0657698841052386e-06,
+      "loss": 0.9147,
+      "step": 33870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5627495593398824,
+      "learning_rate": 2.0653054665275686e-06,
+      "loss": 0.848,
+      "step": 33871
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6824583363043226,
+      "learning_rate": 2.0648410951482e-06,
+      "loss": 0.7631,
+      "step": 33872
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7599267498169744,
+      "learning_rate": 2.064376769969839e-06,
+      "loss": 0.9165,
+      "step": 33873
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7432118448562488,
+      "learning_rate": 2.0639124909951913e-06,
+      "loss": 0.7413,
+      "step": 33874
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7504463895070626,
+      "learning_rate": 2.0634482582269578e-06,
+      "loss": 0.9499,
+      "step": 33875
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7766201446924361,
+      "learning_rate": 2.0629840716678384e-06,
+      "loss": 0.8358,
+      "step": 33876
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.425535257556808,
+      "learning_rate": 2.0625199313205403e-06,
+      "loss": 0.7103,
+      "step": 33877
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3889988925884575,
+      "learning_rate": 2.062055837187763e-06,
+      "loss": 0.8367,
+      "step": 33878
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.529710975377098,
+      "learning_rate": 2.061591789272208e-06,
+      "loss": 0.8049,
+      "step": 33879
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9191741591385116,
+      "learning_rate": 2.0611277875765835e-06,
+      "loss": 0.7346,
+      "step": 33880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.552344511154427,
+      "learning_rate": 2.060663832103584e-06,
+      "loss": 0.9849,
+      "step": 33881
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1755615391289105,
+      "learning_rate": 2.0601999228559134e-06,
+      "loss": 0.8348,
+      "step": 33882
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6055920240059396,
+      "learning_rate": 2.0597360598362705e-06,
+      "loss": 0.8819,
+      "step": 33883
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6860610992093952,
+      "learning_rate": 2.0592722430473645e-06,
+      "loss": 0.7057,
+      "step": 33884
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6672639175243256,
+      "learning_rate": 2.0588084724918834e-06,
+      "loss": 0.8492,
+      "step": 33885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7928572847466115,
+      "learning_rate": 2.058344748172534e-06,
+      "loss": 0.737,
+      "step": 33886
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.82332079976638,
+      "learning_rate": 2.057881070092015e-06,
+      "loss": 0.818,
+      "step": 33887
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7109039236971226,
+      "learning_rate": 2.0574174382530332e-06,
+      "loss": 0.8748,
+      "step": 33888
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.388265951529136,
+      "learning_rate": 2.0569538526582753e-06,
+      "loss": 0.7746,
+      "step": 33889
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4647033451860039,
+      "learning_rate": 2.0564903133104474e-06,
+      "loss": 0.7891,
+      "step": 33890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7896804539053155,
+      "learning_rate": 2.0560268202122514e-06,
+      "loss": 0.6745,
+      "step": 33891
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7681809778913693,
+      "learning_rate": 2.055563373366378e-06,
+      "loss": 0.817,
+      "step": 33892
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.738530300080714,
+      "learning_rate": 2.0550999727755327e-06,
+      "loss": 0.6706,
+      "step": 33893
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7180161116146417,
+      "learning_rate": 2.0546366184424094e-06,
+      "loss": 0.9012,
+      "step": 33894
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4511472641297463,
+      "learning_rate": 2.054173310369708e-06,
+      "loss": 0.8296,
+      "step": 33895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.577442674050949,
+      "learning_rate": 2.0537100485601238e-06,
+      "loss": 0.9,
+      "step": 33896
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.712054909978274,
+      "learning_rate": 2.053246833016358e-06,
+      "loss": 0.8999,
+      "step": 33897
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.605988924873107,
+      "learning_rate": 2.052783663741105e-06,
+      "loss": 0.7887,
+      "step": 33898
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4376071258821526,
+      "learning_rate": 2.052320540737058e-06,
+      "loss": 0.9279,
+      "step": 33899
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.505341129135447,
+      "learning_rate": 2.0518574640069178e-06,
+      "loss": 0.7769,
+      "step": 33900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8056119822878623,
+      "learning_rate": 2.051394433553382e-06,
+      "loss": 0.7316,
+      "step": 33901
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.692262506411833,
+      "learning_rate": 2.0509314493791444e-06,
+      "loss": 0.9189,
+      "step": 33902
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0664203238171353,
+      "learning_rate": 2.050468511486898e-06,
+      "loss": 0.9109,
+      "step": 33903
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6924724689589195,
+      "learning_rate": 2.0500056198793405e-06,
+      "loss": 0.7763,
+      "step": 33904
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.768634729397253,
+      "learning_rate": 2.04954277455917e-06,
+      "loss": 0.7297,
+      "step": 33905
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7093146339684677,
+      "learning_rate": 2.049079975529079e-06,
+      "loss": 0.8455,
+      "step": 33906
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6098322564351495,
+      "learning_rate": 2.0486172227917587e-06,
+      "loss": 0.7503,
+      "step": 33907
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4949540230089962,
+      "learning_rate": 2.0481545163499062e-06,
+      "loss": 0.8678,
+      "step": 33908
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6021996795587983,
+      "learning_rate": 2.0476918562062186e-06,
+      "loss": 0.775,
+      "step": 33909
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4454949025047426,
+      "learning_rate": 2.0472292423633857e-06,
+      "loss": 0.6978,
+      "step": 33910
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.512220394698581,
+      "learning_rate": 2.0467666748241e-06,
+      "loss": 0.7408,
+      "step": 33911
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6868249604531058,
+      "learning_rate": 2.0463041535910587e-06,
+      "loss": 0.8757,
+      "step": 33912
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4904907308659048,
+      "learning_rate": 2.045841678666951e-06,
+      "loss": 0.8273,
+      "step": 33913
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6985052522244137,
+      "learning_rate": 2.0453792500544734e-06,
+      "loss": 0.8394,
+      "step": 33914
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7808631128627328,
+      "learning_rate": 2.0449168677563136e-06,
+      "loss": 0.7225,
+      "step": 33915
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8148431951186468,
+      "learning_rate": 2.04445453177517e-06,
+      "loss": 0.9662,
+      "step": 33916
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.812226072449721,
+      "learning_rate": 2.043992242113727e-06,
+      "loss": 0.786,
+      "step": 33917
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6431362721955574,
+      "learning_rate": 2.0435299987746825e-06,
+      "loss": 0.8479,
+      "step": 33918
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.475449096109148,
+      "learning_rate": 2.0430678017607266e-06,
+      "loss": 0.7672,
+      "step": 33919
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.061332610143939,
+      "learning_rate": 2.0426056510745453e-06,
+      "loss": 0.8461,
+      "step": 33920
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5851829821537302,
+      "learning_rate": 2.0421435467188333e-06,
+      "loss": 0.8873,
+      "step": 33921
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5830634263153145,
+      "learning_rate": 2.041681488696283e-06,
+      "loss": 0.9882,
+      "step": 33922
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.572797232974539,
+      "learning_rate": 2.041219477009584e-06,
+      "loss": 0.9013,
+      "step": 33923
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7030497799430167,
+      "learning_rate": 2.040757511661421e-06,
+      "loss": 0.868,
+      "step": 33924
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5723834755330561,
+      "learning_rate": 2.0402955926544887e-06,
+      "loss": 0.9402,
+      "step": 33925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5492480044747163,
+      "learning_rate": 2.0398337199914774e-06,
+      "loss": 0.862,
+      "step": 33926
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7466510566637703,
+      "learning_rate": 2.039371893675075e-06,
+      "loss": 0.8103,
+      "step": 33927
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.976432529065184,
+      "learning_rate": 2.038910113707966e-06,
+      "loss": 0.8689,
+      "step": 33928
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8106305796227837,
+      "learning_rate": 2.0384483800928433e-06,
+      "loss": 0.8131,
+      "step": 33929
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4565033078029062,
+      "learning_rate": 2.0379866928323977e-06,
+      "loss": 0.8632,
+      "step": 33930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.64186562671414,
+      "learning_rate": 2.0375250519293145e-06,
+      "loss": 0.8837,
+      "step": 33931
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6864912283195719,
+      "learning_rate": 2.0370634573862778e-06,
+      "loss": 0.6564,
+      "step": 33932
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6622465880361803,
+      "learning_rate": 2.036601909205982e-06,
+      "loss": 0.8385,
+      "step": 33933
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.893273778284142,
+      "learning_rate": 2.0361404073911086e-06,
+      "loss": 0.8166,
+      "step": 33934
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.673220137444664,
+      "learning_rate": 2.035678951944349e-06,
+      "loss": 0.7223,
+      "step": 33935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.844232379670582,
+      "learning_rate": 2.0352175428683853e-06,
+      "loss": 0.9011,
+      "step": 33936
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8156058507565958,
+      "learning_rate": 2.03475618016591e-06,
+      "loss": 0.7917,
+      "step": 33937
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.698281342193231,
+      "learning_rate": 2.034294863839603e-06,
+      "loss": 0.8845,
+      "step": 33938
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.580485041108814,
+      "learning_rate": 2.033833593892156e-06,
+      "loss": 0.906,
+      "step": 33939
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6093975195658938,
+      "learning_rate": 2.033372370326252e-06,
+      "loss": 0.8892,
+      "step": 33940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7957773932736103,
+      "learning_rate": 2.032911193144573e-06,
+      "loss": 0.7986,
+      "step": 33941
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.291920873026613,
+      "learning_rate": 2.0324500623498076e-06,
+      "loss": 0.8471,
+      "step": 33942
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8086934851793894,
+      "learning_rate": 2.0319889779446432e-06,
+      "loss": 0.8286,
+      "step": 33943
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5273272374233406,
+      "learning_rate": 2.031527939931761e-06,
+      "loss": 0.8974,
+      "step": 33944
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7841962653766927,
+      "learning_rate": 2.0310669483138446e-06,
+      "loss": 0.7567,
+      "step": 33945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4946023497594711,
+      "learning_rate": 2.030606003093578e-06,
+      "loss": 0.8563,
+      "step": 33946
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5735415371678632,
+      "learning_rate": 2.0301451042736496e-06,
+      "loss": 0.7145,
+      "step": 33947
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.284843281385773,
+      "learning_rate": 2.0296842518567395e-06,
+      "loss": 0.8274,
+      "step": 33948
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.789044969935397,
+      "learning_rate": 2.0292234458455284e-06,
+      "loss": 0.6801,
+      "step": 33949
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.758279799346372,
+      "learning_rate": 2.0287626862427024e-06,
+      "loss": 0.7882,
+      "step": 33950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4813895807748514,
+      "learning_rate": 2.0283019730509456e-06,
+      "loss": 0.7502,
+      "step": 33951
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8736300732094424,
+      "learning_rate": 2.0278413062729394e-06,
+      "loss": 0.8339,
+      "step": 33952
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7498065200781965,
+      "learning_rate": 2.027380685911361e-06,
+      "loss": 0.7098,
+      "step": 33953
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7049243806549002,
+      "learning_rate": 2.0269201119688975e-06,
+      "loss": 1.0111,
+      "step": 33954
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7649166386938835,
+      "learning_rate": 2.0264595844482315e-06,
+      "loss": 0.7017,
+      "step": 33955
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5638942314195334,
+      "learning_rate": 2.025999103352042e-06,
+      "loss": 0.81,
+      "step": 33956
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.601563054541166,
+      "learning_rate": 2.025538668683008e-06,
+      "loss": 0.8129,
+      "step": 33957
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7480214591788259,
+      "learning_rate": 2.025078280443815e-06,
+      "loss": 0.712,
+      "step": 33958
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4944023868782013,
+      "learning_rate": 2.024617938637139e-06,
+      "loss": 0.8177,
+      "step": 33959
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5963586912208036,
+      "learning_rate": 2.024157643265665e-06,
+      "loss": 0.8636,
+      "step": 33960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.602687764213466,
+      "learning_rate": 2.023697394332067e-06,
+      "loss": 0.9241,
+      "step": 33961
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5399998566017468,
+      "learning_rate": 2.0232371918390305e-06,
+      "loss": 0.7237,
+      "step": 33962
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.69692400264193,
+      "learning_rate": 2.0227770357892307e-06,
+      "loss": 0.7337,
+      "step": 33963
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7655312229030654,
+      "learning_rate": 2.02231692618535e-06,
+      "loss": 0.6678,
+      "step": 33964
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.959704396782596,
+      "learning_rate": 2.021856863030067e-06,
+      "loss": 0.9114,
+      "step": 33965
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5866451517558524,
+      "learning_rate": 2.0213968463260556e-06,
+      "loss": 0.7104,
+      "step": 33966
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.634381643029579,
+      "learning_rate": 2.0209368760759983e-06,
+      "loss": 0.8719,
+      "step": 33967
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7717462911055125,
+      "learning_rate": 2.020476952282573e-06,
+      "loss": 0.7646,
+      "step": 33968
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.927921707191439,
+      "learning_rate": 2.0200170749484614e-06,
+      "loss": 0.7743,
+      "step": 33969
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7790306727817544,
+      "learning_rate": 2.019557244076332e-06,
+      "loss": 0.8519,
+      "step": 33970
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.67969952871138,
+      "learning_rate": 2.0190974596688673e-06,
+      "loss": 0.818,
+      "step": 33971
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5347069218727687,
+      "learning_rate": 2.018637721728747e-06,
+      "loss": 0.8316,
+      "step": 33972
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5948693179582236,
+      "learning_rate": 2.018178030258643e-06,
+      "loss": 0.9237,
+      "step": 33973
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5496129581830933,
+      "learning_rate": 2.0177183852612316e-06,
+      "loss": 0.9605,
+      "step": 33974
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7097082604215399,
+      "learning_rate": 2.017258786739191e-06,
+      "loss": 0.7084,
+      "step": 33975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9364448964328664,
+      "learning_rate": 2.0167992346952005e-06,
+      "loss": 0.8318,
+      "step": 33976
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9197980913080726,
+      "learning_rate": 2.016339729131929e-06,
+      "loss": 0.8786,
+      "step": 33977
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5904896835408133,
+      "learning_rate": 2.0158802700520576e-06,
+      "loss": 0.6718,
+      "step": 33978
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6951853191748925,
+      "learning_rate": 2.015420857458259e-06,
+      "loss": 0.737,
+      "step": 33979
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.455888086284673,
+      "learning_rate": 2.0149614913532055e-06,
+      "loss": 0.7158,
+      "step": 33980
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5058380445227886,
+      "learning_rate": 2.014502171739574e-06,
+      "loss": 0.7967,
+      "step": 33981
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6002050691635077,
+      "learning_rate": 2.014042898620042e-06,
+      "loss": 0.807,
+      "step": 33982
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5881855983830844,
+      "learning_rate": 2.013583671997281e-06,
+      "loss": 0.9403,
+      "step": 33983
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8574541950264588,
+      "learning_rate": 2.0131244918739612e-06,
+      "loss": 0.7563,
+      "step": 33984
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5162282794532247,
+      "learning_rate": 2.012665358252759e-06,
+      "loss": 0.8836,
+      "step": 33985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.997916737959878,
+      "learning_rate": 2.012206271136353e-06,
+      "loss": 0.8681,
+      "step": 33986
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.673286579516864,
+      "learning_rate": 2.0117472305274065e-06,
+      "loss": 0.9108,
+      "step": 33987
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5454445647150046,
+      "learning_rate": 2.0112882364285968e-06,
+      "loss": 0.7972,
+      "step": 33988
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6314774067444995,
+      "learning_rate": 2.010829288842595e-06,
+      "loss": 0.6715,
+      "step": 33989
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6335271642978595,
+      "learning_rate": 2.0103703877720783e-06,
+      "loss": 0.7233,
+      "step": 33990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.778747516287952,
+      "learning_rate": 2.009911533219714e-06,
+      "loss": 0.7242,
+      "step": 33991
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4009279554018277,
+      "learning_rate": 2.0094527251881713e-06,
+      "loss": 0.9123,
+      "step": 33992
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4826100276787146,
+      "learning_rate": 2.008993963680127e-06,
+      "loss": 0.8723,
+      "step": 33993
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.678888568736809,
+      "learning_rate": 2.008535248698248e-06,
+      "loss": 0.8345,
+      "step": 33994
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5020157549411448,
+      "learning_rate": 2.0080765802452097e-06,
+      "loss": 0.8337,
+      "step": 33995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7717096137631648,
+      "learning_rate": 2.0076179583236765e-06,
+      "loss": 0.7194,
+      "step": 33996
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.762061285238207,
+      "learning_rate": 2.0071593829363245e-06,
+      "loss": 0.7232,
+      "step": 33997
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.769583807029489,
+      "learning_rate": 2.006700854085819e-06,
+      "loss": 0.8528,
+      "step": 33998
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5794296473792593,
+      "learning_rate": 2.0062423717748336e-06,
+      "loss": 0.7974,
+      "step": 33999
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9438213979433496,
+      "learning_rate": 2.005783936006036e-06,
+      "loss": 0.8615,
+      "step": 34000
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3240947434555475,
+      "learning_rate": 2.0053255467820933e-06,
+      "loss": 0.7214,
+      "step": 34001
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5358202922188966,
+      "learning_rate": 2.004867204105676e-06,
+      "loss": 0.9446,
+      "step": 34002
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.614004375416998,
+      "learning_rate": 2.004408907979456e-06,
+      "loss": 0.8281,
+      "step": 34003
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4746776204627425,
+      "learning_rate": 2.0039506584060987e-06,
+      "loss": 0.8021,
+      "step": 34004
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8516271008454261,
+      "learning_rate": 2.0034924553882683e-06,
+      "loss": 0.8514,
+      "step": 34005
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5873804378743455,
+      "learning_rate": 2.0030342989286376e-06,
+      "loss": 0.8413,
+      "step": 34006
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6882808516824674,
+      "learning_rate": 2.0025761890298757e-06,
+      "loss": 0.8247,
+      "step": 34007
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4811016173173401,
+      "learning_rate": 2.0021181256946465e-06,
+      "loss": 0.9095,
+      "step": 34008
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4442276198698265,
+      "learning_rate": 2.001660108925616e-06,
+      "loss": 0.8815,
+      "step": 34009
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5203247802730069,
+      "learning_rate": 2.001202138725451e-06,
+      "loss": 0.7653,
+      "step": 34010
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.677741983179121,
+      "learning_rate": 2.0007442150968236e-06,
+      "loss": 0.7665,
+      "step": 34011
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.588622777531966,
+      "learning_rate": 2.000286338042395e-06,
+      "loss": 0.7835,
+      "step": 34012
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4959237079134793,
+      "learning_rate": 1.9998285075648304e-06,
+      "loss": 0.7884,
+      "step": 34013
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6868341498420898,
+      "learning_rate": 1.9993707236667992e-06,
+      "loss": 0.7316,
+      "step": 34014
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5635203430622593,
+      "learning_rate": 1.9989129863509617e-06,
+      "loss": 0.8233,
+      "step": 34015
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.423144906956763,
+      "learning_rate": 1.998455295619989e-06,
+      "loss": 0.7695,
+      "step": 34016
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7924526077345266,
+      "learning_rate": 1.9979976514765397e-06,
+      "loss": 0.7649,
+      "step": 34017
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7212585315736606,
+      "learning_rate": 1.9975400539232836e-06,
+      "loss": 0.8791,
+      "step": 34018
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6386079811272738,
+      "learning_rate": 1.9970825029628814e-06,
+      "loss": 0.9681,
+      "step": 34019
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.618135294819682,
+      "learning_rate": 1.996624998598e-06,
+      "loss": 0.9174,
+      "step": 34020
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5857310169129926,
+      "learning_rate": 1.996167540831303e-06,
+      "loss": 0.8655,
+      "step": 34021
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7542525925252053,
+      "learning_rate": 1.995710129665449e-06,
+      "loss": 0.646,
+      "step": 34022
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5241660908849934,
+      "learning_rate": 1.995252765103105e-06,
+      "loss": 0.7766,
+      "step": 34023
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4732331316111855,
+      "learning_rate": 1.9947954471469365e-06,
+      "loss": 0.8863,
+      "step": 34024
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8670161798808493,
+      "learning_rate": 1.9943381757996027e-06,
+      "loss": 0.8623,
+      "step": 34025
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.460488916637623,
+      "learning_rate": 1.9938809510637648e-06,
+      "loss": 0.812,
+      "step": 34026
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5992214960326911,
+      "learning_rate": 1.993423772942086e-06,
+      "loss": 0.8257,
+      "step": 34027
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7064409077772407,
+      "learning_rate": 1.992966641437232e-06,
+      "loss": 0.9164,
+      "step": 34028
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4883218366897162,
+      "learning_rate": 1.9925095565518605e-06,
+      "loss": 0.8597,
+      "step": 34029
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6902449349767383,
+      "learning_rate": 1.992052518288632e-06,
+      "loss": 0.8563,
+      "step": 34030
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7843612767802207,
+      "learning_rate": 1.9915955266502073e-06,
+      "loss": 0.7064,
+      "step": 34031
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6546920445622826,
+      "learning_rate": 1.9911385816392536e-06,
+      "loss": 0.7946,
+      "step": 34032
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5457881401214375,
+      "learning_rate": 1.990681683258425e-06,
+      "loss": 0.9218,
+      "step": 34033
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4312772554601187,
+      "learning_rate": 1.990224831510381e-06,
+      "loss": 0.8023,
+      "step": 34034
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.652537086205061,
+      "learning_rate": 1.989768026397787e-06,
+      "loss": 0.9486,
+      "step": 34035
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.688675659715597,
+      "learning_rate": 1.989311267923296e-06,
+      "loss": 0.8562,
+      "step": 34036
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6367118384249761,
+      "learning_rate": 1.9888545560895746e-06,
+      "loss": 0.7531,
+      "step": 34037
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7944161372406107,
+      "learning_rate": 1.988397890899275e-06,
+      "loss": 0.8311,
+      "step": 34038
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.763183932791619,
+      "learning_rate": 1.987941272355063e-06,
+      "loss": 0.8502,
+      "step": 34039
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4511638801139162,
+      "learning_rate": 1.98748470045959e-06,
+      "loss": 0.7063,
+      "step": 34040
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.69054511699244,
+      "learning_rate": 1.9870281752155208e-06,
+      "loss": 0.9924,
+      "step": 34041
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6230071180715542,
+      "learning_rate": 1.98657169662551e-06,
+      "loss": 0.8704,
+      "step": 34042
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7398327647911485,
+      "learning_rate": 1.9861152646922133e-06,
+      "loss": 0.6874,
+      "step": 34043
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0810990025361367,
+      "learning_rate": 1.9856588794182906e-06,
+      "loss": 0.8268,
+      "step": 34044
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7615784978660489,
+      "learning_rate": 1.985202540806401e-06,
+      "loss": 0.703,
+      "step": 34045
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.660602012982795,
+      "learning_rate": 1.9847462488592017e-06,
+      "loss": 0.8678,
+      "step": 34046
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8083101820378202,
+      "learning_rate": 1.9842900035793435e-06,
+      "loss": 0.8242,
+      "step": 34047
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5565859627481986,
+      "learning_rate": 1.983833804969486e-06,
+      "loss": 0.7851,
+      "step": 34048
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4028187360525224,
+      "learning_rate": 1.983377653032289e-06,
+      "loss": 0.9072,
+      "step": 34049
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8102603957610693,
+      "learning_rate": 1.9829215477704066e-06,
+      "loss": 0.7248,
+      "step": 34050
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4595658448877507,
+      "learning_rate": 1.9824654891864893e-06,
+      "loss": 0.9294,
+      "step": 34051
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4213032444678118,
+      "learning_rate": 1.982009477283197e-06,
+      "loss": 0.8416,
+      "step": 34052
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7461013258952268,
+      "learning_rate": 1.9815535120631846e-06,
+      "loss": 0.7168,
+      "step": 34053
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8263025187040045,
+      "learning_rate": 1.9810975935291076e-06,
+      "loss": 0.6981,
+      "step": 34054
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.736794224802001,
+      "learning_rate": 1.980641721683617e-06,
+      "loss": 0.767,
+      "step": 34055
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9557185865093605,
+      "learning_rate": 1.980185896529372e-06,
+      "loss": 0.8581,
+      "step": 34056
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2450502048025056,
+      "learning_rate": 1.9797301180690197e-06,
+      "loss": 0.8426,
+      "step": 34057
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6366149572228312,
+      "learning_rate": 1.9792743863052223e-06,
+      "loss": 0.7244,
+      "step": 34058
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5897423281072285,
+      "learning_rate": 1.9788187012406247e-06,
+      "loss": 0.9183,
+      "step": 34059
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6414246692716514,
+      "learning_rate": 1.978363062877887e-06,
+      "loss": 0.839,
+      "step": 34060
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4018450665071387,
+      "learning_rate": 1.9779074712196567e-06,
+      "loss": 0.8151,
+      "step": 34061
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4489502135659893,
+      "learning_rate": 1.9774519262685908e-06,
+      "loss": 0.8598,
+      "step": 34062
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5845094781573386,
+      "learning_rate": 1.97699642802734e-06,
+      "loss": 0.8822,
+      "step": 34063
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7863504266676469,
+      "learning_rate": 1.9765409764985543e-06,
+      "loss": 0.6546,
+      "step": 34064
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.737784690955662,
+      "learning_rate": 1.976085571684887e-06,
+      "loss": 0.6806,
+      "step": 34065
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4185465594843643,
+      "learning_rate": 1.9756302135889894e-06,
+      "loss": 0.7586,
+      "step": 34066
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5075741464102248,
+      "learning_rate": 1.975174902213518e-06,
+      "loss": 0.9279,
+      "step": 34067
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5933473916478733,
+      "learning_rate": 1.9747196375611143e-06,
+      "loss": 0.8585,
+      "step": 34068
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5075164857619432,
+      "learning_rate": 1.974264419634433e-06,
+      "loss": 0.8371,
+      "step": 34069
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3276987018038633,
+      "learning_rate": 1.973809248436128e-06,
+      "loss": 0.9951,
+      "step": 34070
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.425229937006496,
+      "learning_rate": 1.973354123968847e-06,
+      "loss": 0.7036,
+      "step": 34071
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7530001002149306,
+      "learning_rate": 1.972899046235237e-06,
+      "loss": 0.6865,
+      "step": 34072
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5649957684799058,
+      "learning_rate": 1.97244401523795e-06,
+      "loss": 0.7641,
+      "step": 34073
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6017961174135642,
+      "learning_rate": 1.9719890309796374e-06,
+      "loss": 0.967,
+      "step": 34074
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.541734538265921,
+      "learning_rate": 1.9715340934629446e-06,
+      "loss": 0.8047,
+      "step": 34075
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7660810117909924,
+      "learning_rate": 1.9710792026905243e-06,
+      "loss": 0.7115,
+      "step": 34076
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6441952542020914,
+      "learning_rate": 1.97062435866502e-06,
+      "loss": 0.9057,
+      "step": 34077
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5767910039484223,
+      "learning_rate": 1.970169561389086e-06,
+      "loss": 0.6569,
+      "step": 34078
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5349852478692043,
+      "learning_rate": 1.969714810865363e-06,
+      "loss": 0.7979,
+      "step": 34079
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5710675645854923,
+      "learning_rate": 1.969260107096507e-06,
+      "loss": 0.7593,
+      "step": 34080
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7622796746433871,
+      "learning_rate": 1.9688054500851596e-06,
+      "loss": 0.7093,
+      "step": 34081
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6018501021870535,
+      "learning_rate": 1.968350839833968e-06,
+      "loss": 0.6917,
+      "step": 34082
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4639184534559235,
+      "learning_rate": 1.96789627634558e-06,
+      "loss": 0.8462,
+      "step": 34083
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0495889301760086,
+      "learning_rate": 1.967441759622646e-06,
+      "loss": 0.8828,
+      "step": 34084
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7271065606575404,
+      "learning_rate": 1.9669872896678087e-06,
+      "loss": 0.7973,
+      "step": 34085
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8999661665534515,
+      "learning_rate": 1.9665328664837114e-06,
+      "loss": 0.7912,
+      "step": 34086
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.772418271357786,
+      "learning_rate": 1.9660784900730047e-06,
+      "loss": 0.7393,
+      "step": 34087
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4554782540755618,
+      "learning_rate": 1.9656241604383343e-06,
+      "loss": 0.7344,
+      "step": 34088
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1716813960475205,
+      "learning_rate": 1.9651698775823425e-06,
+      "loss": 0.7608,
+      "step": 34089
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5676433590110856,
+      "learning_rate": 1.964715641507674e-06,
+      "loss": 0.7983,
+      "step": 34090
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.619949477738865,
+      "learning_rate": 1.9642614522169754e-06,
+      "loss": 0.7889,
+      "step": 34091
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7364758813148121,
+      "learning_rate": 1.963807309712893e-06,
+      "loss": 0.7071,
+      "step": 34092
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.830966771431172,
+      "learning_rate": 1.9633532139980684e-06,
+      "loss": 0.7534,
+      "step": 34093
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4855529075686769,
+      "learning_rate": 1.962899165075143e-06,
+      "loss": 0.7663,
+      "step": 34094
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7165329316496722,
+      "learning_rate": 1.9624451629467658e-06,
+      "loss": 0.8798,
+      "step": 34095
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.447463653695832,
+      "learning_rate": 1.9619912076155757e-06,
+      "loss": 0.9099,
+      "step": 34096
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.774388877023464,
+      "learning_rate": 1.9615372990842184e-06,
+      "loss": 0.7376,
+      "step": 34097
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4988158619318488,
+      "learning_rate": 1.9610834373553347e-06,
+      "loss": 0.7938,
+      "step": 34098
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7479402706902742,
+      "learning_rate": 1.9606296224315713e-06,
+      "loss": 0.8041,
+      "step": 34099
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7672436159394063,
+      "learning_rate": 1.9601758543155636e-06,
+      "loss": 0.7807,
+      "step": 34100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8369369764454078,
+      "learning_rate": 1.9597221330099613e-06,
+      "loss": 0.9736,
+      "step": 34101
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5836499900225114,
+      "learning_rate": 1.9592684585174014e-06,
+      "loss": 0.8445,
+      "step": 34102
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.730389418301245,
+      "learning_rate": 1.958814830840523e-06,
+      "loss": 0.81,
+      "step": 34103
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9258076552811174,
+      "learning_rate": 1.9583612499819713e-06,
+      "loss": 0.8456,
+      "step": 34104
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.79436477094197,
+      "learning_rate": 1.9579077159443893e-06,
+      "loss": 0.6865,
+      "step": 34105
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9521781895325365,
+      "learning_rate": 1.957454228730414e-06,
+      "loss": 0.8071,
+      "step": 34106
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5820979681376888,
+      "learning_rate": 1.9570007883426845e-06,
+      "loss": 0.9371,
+      "step": 34107
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5596782937167002,
+      "learning_rate": 1.956547394783842e-06,
+      "loss": 0.7129,
+      "step": 34108
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.483283062572781,
+      "learning_rate": 1.9560940480565304e-06,
+      "loss": 0.7901,
+      "step": 34109
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7761042992431557,
+      "learning_rate": 1.9556407481633854e-06,
+      "loss": 0.7942,
+      "step": 34110
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0968435638166736,
+      "learning_rate": 1.9551874951070438e-06,
+      "loss": 0.6907,
+      "step": 34111
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5771537202526644,
+      "learning_rate": 1.954734288890149e-06,
+      "loss": 0.9021,
+      "step": 34112
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7835884436866476,
+      "learning_rate": 1.9542811295153396e-06,
+      "loss": 0.8624,
+      "step": 34113
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6953320582106004,
+      "learning_rate": 1.953828016985253e-06,
+      "loss": 0.8495,
+      "step": 34114
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.65725044227734,
+      "learning_rate": 1.9533749513025234e-06,
+      "loss": 0.7153,
+      "step": 34115
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.235671025744072,
+      "learning_rate": 1.9529219324697966e-06,
+      "loss": 0.9595,
+      "step": 34116
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5660739299832172,
+      "learning_rate": 1.9524689604897028e-06,
+      "loss": 0.8354,
+      "step": 34117
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6594708621439611,
+      "learning_rate": 1.9520160353648852e-06,
+      "loss": 0.7378,
+      "step": 34118
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0933516520100395,
+      "learning_rate": 1.951563157097975e-06,
+      "loss": 0.8359,
+      "step": 34119
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4345205436653339,
+      "learning_rate": 1.9511103256916154e-06,
+      "loss": 0.8929,
+      "step": 34120
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7233172524289482,
+      "learning_rate": 1.9506575411484373e-06,
+      "loss": 0.9164,
+      "step": 34121
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5567622717937024,
+      "learning_rate": 1.9502048034710818e-06,
+      "loss": 0.8523,
+      "step": 34122
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5954102882149772,
+      "learning_rate": 1.9497521126621823e-06,
+      "loss": 0.8151,
+      "step": 34123
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.606721395209238,
+      "learning_rate": 1.9492994687243715e-06,
+      "loss": 0.7946,
+      "step": 34124
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7398451586918123,
+      "learning_rate": 1.9488468716602895e-06,
+      "loss": 0.7514,
+      "step": 34125
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.619955713362595,
+      "learning_rate": 1.948394321472571e-06,
+      "loss": 0.8778,
+      "step": 34126
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6885492433015041,
+      "learning_rate": 1.9479418181638508e-06,
+      "loss": 0.7256,
+      "step": 34127
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4412447318275106,
+      "learning_rate": 1.9474893617367597e-06,
+      "loss": 0.7954,
+      "step": 34128
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4386773404312905,
+      "learning_rate": 1.947036952193935e-06,
+      "loss": 0.8758,
+      "step": 34129
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.720702256588522,
+      "learning_rate": 1.946584589538013e-06,
+      "loss": 0.8734,
+      "step": 34130
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5066298524194763,
+      "learning_rate": 1.946132273771625e-06,
+      "loss": 0.7531,
+      "step": 34131
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4537494880398307,
+      "learning_rate": 1.9456800048974023e-06,
+      "loss": 0.7174,
+      "step": 34132
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7836269788908566,
+      "learning_rate": 1.945227782917981e-06,
+      "loss": 0.7501,
+      "step": 34133
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4546229613062918,
+      "learning_rate": 1.944775607835996e-06,
+      "loss": 0.715,
+      "step": 34134
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6195773115075223,
+      "learning_rate": 1.944323479654078e-06,
+      "loss": 0.8913,
+      "step": 34135
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7761834928707503,
+      "learning_rate": 1.9438713983748556e-06,
+      "loss": 0.7889,
+      "step": 34136
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7408566143153477,
+      "learning_rate": 1.943419364000968e-06,
+      "loss": 0.821,
+      "step": 34137
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.607578446513134,
+      "learning_rate": 1.94296737653504e-06,
+      "loss": 0.7669,
+      "step": 34138
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6383614105331383,
+      "learning_rate": 1.9425154359797103e-06,
+      "loss": 0.7409,
+      "step": 34139
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6550390216796595,
+      "learning_rate": 1.9420635423376033e-06,
+      "loss": 0.9873,
+      "step": 34140
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.532729610193189,
+      "learning_rate": 1.941611695611356e-06,
+      "loss": 0.7551,
+      "step": 34141
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6184201667912868,
+      "learning_rate": 1.941159895803595e-06,
+      "loss": 0.8767,
+      "step": 34142
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.672921663644073,
+      "learning_rate": 1.940708142916955e-06,
+      "loss": 0.7283,
+      "step": 34143
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8376498215193873,
+      "learning_rate": 1.9402564369540634e-06,
+      "loss": 0.9488,
+      "step": 34144
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5900404318255201,
+      "learning_rate": 1.939804777917548e-06,
+      "loss": 0.8763,
+      "step": 34145
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4392119589621983,
+      "learning_rate": 1.9393531658100395e-06,
+      "loss": 0.9212,
+      "step": 34146
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4877233084522556,
+      "learning_rate": 1.938901600634173e-06,
+      "loss": 0.7011,
+      "step": 34147
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5005391701267412,
+      "learning_rate": 1.9384500823925732e-06,
+      "loss": 0.8339,
+      "step": 34148
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5444630451172707,
+      "learning_rate": 1.9379986110878666e-06,
+      "loss": 1.0112,
+      "step": 34149
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.785267107369238,
+      "learning_rate": 1.9375471867226846e-06,
+      "loss": 0.643,
+      "step": 34150
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8579368221741608,
+      "learning_rate": 1.9370958092996582e-06,
+      "loss": 0.8781,
+      "step": 34151
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.569828161272494,
+      "learning_rate": 1.9366444788214123e-06,
+      "loss": 0.8417,
+      "step": 34152
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.809605467170953,
+      "learning_rate": 1.9361931952905723e-06,
+      "loss": 0.7145,
+      "step": 34153
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7611266017364433,
+      "learning_rate": 1.935741958709768e-06,
+      "loss": 0.7145,
+      "step": 34154
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.550761787947929,
+      "learning_rate": 1.9352907690816313e-06,
+      "loss": 0.7908,
+      "step": 34155
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7384990664499225,
+      "learning_rate": 1.9348396264087845e-06,
+      "loss": 0.7699,
+      "step": 34156
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0068073406409557,
+      "learning_rate": 1.9343885306938514e-06,
+      "loss": 0.8383,
+      "step": 34157
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8849275301413737,
+      "learning_rate": 1.933937481939464e-06,
+      "loss": 0.8768,
+      "step": 34158
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5511774400720144,
+      "learning_rate": 1.9334864801482446e-06,
+      "loss": 0.8863,
+      "step": 34159
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6533915939641624,
+      "learning_rate": 1.9330355253228237e-06,
+      "loss": 0.8166,
+      "step": 34160
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5839618261488255,
+      "learning_rate": 1.93258461746582e-06,
+      "loss": 0.8041,
+      "step": 34161
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.805229185286669,
+      "learning_rate": 1.932133756579867e-06,
+      "loss": 0.73,
+      "step": 34162
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7141249323281826,
+      "learning_rate": 1.931682942667582e-06,
+      "loss": 0.7397,
+      "step": 34163
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6615482641757928,
+      "learning_rate": 1.931232175731593e-06,
+      "loss": 0.8527,
+      "step": 34164
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6712733415766654,
+      "learning_rate": 1.9307814557745307e-06,
+      "loss": 0.8386,
+      "step": 34165
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3869723110436383,
+      "learning_rate": 1.930330782799008e-06,
+      "loss": 0.5701,
+      "step": 34166
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6664985325324562,
+      "learning_rate": 1.9298801568076543e-06,
+      "loss": 0.9211,
+      "step": 34167
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.845529317357237,
+      "learning_rate": 1.9294295778030924e-06,
+      "loss": 0.7241,
+      "step": 34168
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.489651456209226,
+      "learning_rate": 1.928979045787953e-06,
+      "loss": 0.8062,
+      "step": 34169
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.500538069485248,
+      "learning_rate": 1.9285285607648476e-06,
+      "loss": 0.8794,
+      "step": 34170
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6474186740495593,
+      "learning_rate": 1.928078122736403e-06,
+      "loss": 0.8196,
+      "step": 34171
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.548534057937897,
+      "learning_rate": 1.927627731705247e-06,
+      "loss": 0.763,
+      "step": 34172
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4567767382697387,
+      "learning_rate": 1.9271773876739953e-06,
+      "loss": 0.8655,
+      "step": 34173
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.527643813790304,
+      "learning_rate": 1.926727090645275e-06,
+      "loss": 0.7829,
+      "step": 34174
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.045348399071866,
+      "learning_rate": 1.9262768406217026e-06,
+      "loss": 0.7334,
+      "step": 34175
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.618664442638559,
+      "learning_rate": 1.9258266376059053e-06,
+      "loss": 0.8467,
+      "step": 34176
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.68835841444524,
+      "learning_rate": 1.9253764816004993e-06,
+      "loss": 0.7177,
+      "step": 34177
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.420273720210473,
+      "learning_rate": 1.9249263726081103e-06,
+      "loss": 0.8987,
+      "step": 34178
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7307895132431637,
+      "learning_rate": 1.924476310631357e-06,
+      "loss": 0.7738,
+      "step": 34179
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3908309439542865,
+      "learning_rate": 1.924026295672856e-06,
+      "loss": 0.9183,
+      "step": 34180
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7791685059110403,
+      "learning_rate": 1.9235763277352304e-06,
+      "loss": 0.8468,
+      "step": 34181
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5482291303764153,
+      "learning_rate": 1.9231264068211033e-06,
+      "loss": 0.7434,
+      "step": 34182
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8571669283748458,
+      "learning_rate": 1.922676532933092e-06,
+      "loss": 0.8184,
+      "step": 34183
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4899183836431056,
+      "learning_rate": 1.922226706073812e-06,
+      "loss": 0.7641,
+      "step": 34184
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8216941853368944,
+      "learning_rate": 1.921776926245885e-06,
+      "loss": 0.738,
+      "step": 34185
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8207237429585463,
+      "learning_rate": 1.921327193451934e-06,
+      "loss": 0.7593,
+      "step": 34186
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4325963533920683,
+      "learning_rate": 1.920877507694573e-06,
+      "loss": 0.775,
+      "step": 34187
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6067260812063093,
+      "learning_rate": 1.920427868976418e-06,
+      "loss": 0.8284,
+      "step": 34188
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6253578960078525,
+      "learning_rate": 1.9199782773000908e-06,
+      "loss": 0.8105,
+      "step": 34189
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6773034680956322,
+      "learning_rate": 1.91952873266821e-06,
+      "loss": 0.9981,
+      "step": 34190
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4956689553784586,
+      "learning_rate": 1.919079235083391e-06,
+      "loss": 0.8099,
+      "step": 34191
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6166529463773986,
+      "learning_rate": 1.9186297845482495e-06,
+      "loss": 0.8562,
+      "step": 34192
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7163693434100833,
+      "learning_rate": 1.9181803810654053e-06,
+      "loss": 0.683,
+      "step": 34193
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7647418189357048,
+      "learning_rate": 1.9177310246374725e-06,
+      "loss": 0.6725,
+      "step": 34194
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.123339390222834,
+      "learning_rate": 1.9172817152670696e-06,
+      "loss": 0.8562,
+      "step": 34195
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8042299562727895,
+      "learning_rate": 1.9168324529568093e-06,
+      "loss": 0.7738,
+      "step": 34196
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8516771731447998,
+      "learning_rate": 1.9163832377093126e-06,
+      "loss": 0.9039,
+      "step": 34197
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7091295079185405,
+      "learning_rate": 1.915934069527189e-06,
+      "loss": 0.8402,
+      "step": 34198
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0759029458027487,
+      "learning_rate": 1.91548494841306e-06,
+      "loss": 0.7828,
+      "step": 34199
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.531822823041901,
+      "learning_rate": 1.915035874369533e-06,
+      "loss": 0.7812,
+      "step": 34200
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.631781737122454,
+      "learning_rate": 1.91458684739923e-06,
+      "loss": 0.7979,
+      "step": 34201
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7352743676864049,
+      "learning_rate": 1.9141378675047608e-06,
+      "loss": 0.6701,
+      "step": 34202
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7409833663685631,
+      "learning_rate": 1.9136889346887423e-06,
+      "loss": 0.6657,
+      "step": 34203
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5008325151662665,
+      "learning_rate": 1.9132400489537872e-06,
+      "loss": 1.0069,
+      "step": 34204
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4749615687605833,
+      "learning_rate": 1.9127912103025058e-06,
+      "loss": 0.8179,
+      "step": 34205
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5705802432013363,
+      "learning_rate": 1.912342418737515e-06,
+      "loss": 0.7638,
+      "step": 34206
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.649133882333506,
+      "learning_rate": 1.911893674261429e-06,
+      "loss": 0.9789,
+      "step": 34207
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7149186735231983,
+      "learning_rate": 1.9114449768768584e-06,
+      "loss": 0.8798,
+      "step": 34208
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.715450701621632,
+      "learning_rate": 1.910996326586414e-06,
+      "loss": 0.6388,
+      "step": 34209
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.431132190954369,
+      "learning_rate": 1.9105477233927105e-06,
+      "loss": 0.8298,
+      "step": 34210
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7419901749046269,
+      "learning_rate": 1.9100991672983616e-06,
+      "loss": 0.6373,
+      "step": 34211
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9804610194270738,
+      "learning_rate": 1.909650658305976e-06,
+      "loss": 0.7169,
+      "step": 34212
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6699538584415332,
+      "learning_rate": 1.909202196418163e-06,
+      "loss": 0.9114,
+      "step": 34213
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.629073834006765,
+      "learning_rate": 1.908753781637537e-06,
+      "loss": 0.9108,
+      "step": 34214
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6314009360985708,
+      "learning_rate": 1.9083054139667103e-06,
+      "loss": 0.6916,
+      "step": 34215
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.410542635482456,
+      "learning_rate": 1.9078570934082907e-06,
+      "loss": 0.7503,
+      "step": 34216
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8328095852492505,
+      "learning_rate": 1.9074088199648866e-06,
+      "loss": 0.8708,
+      "step": 34217
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7381857955466615,
+      "learning_rate": 1.9069605936391133e-06,
+      "loss": 0.7242,
+      "step": 34218
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6648485287948371,
+      "learning_rate": 1.9065124144335756e-06,
+      "loss": 0.8423,
+      "step": 34219
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.234423512014171,
+      "learning_rate": 1.9060642823508869e-06,
+      "loss": 0.8107,
+      "step": 34220
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.768709176546675,
+      "learning_rate": 1.9056161973936516e-06,
+      "loss": 0.7139,
+      "step": 34221
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5139302910094947,
+      "learning_rate": 1.9051681595644844e-06,
+      "loss": 0.8026,
+      "step": 34222
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5697101932779003,
+      "learning_rate": 1.904720168865989e-06,
+      "loss": 0.8451,
+      "step": 34223
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5895632204830628,
+      "learning_rate": 1.9042722253007772e-06,
+      "loss": 0.8231,
+      "step": 34224
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5512785874898791,
+      "learning_rate": 1.9038243288714563e-06,
+      "loss": 0.8828,
+      "step": 34225
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8011250029584954,
+      "learning_rate": 1.903376479580631e-06,
+      "loss": 0.6683,
+      "step": 34226
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6096248029749287,
+      "learning_rate": 1.9029286774309108e-06,
+      "loss": 0.8245,
+      "step": 34227
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7381917617829225,
+      "learning_rate": 1.9024809224249052e-06,
+      "loss": 0.7109,
+      "step": 34228
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.626366245588864,
+      "learning_rate": 1.9020332145652197e-06,
+      "loss": 0.8218,
+      "step": 34229
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4991318075050868,
+      "learning_rate": 1.9015855538544592e-06,
+      "loss": 0.8222,
+      "step": 34230
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5286536068118903,
+      "learning_rate": 1.90113794029523e-06,
+      "loss": 0.8262,
+      "step": 34231
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.707539108084783,
+      "learning_rate": 1.9006903738901427e-06,
+      "loss": 1.0223,
+      "step": 34232
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7528408914627137,
+      "learning_rate": 1.9002428546418005e-06,
+      "loss": 0.7486,
+      "step": 34233
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4050030676865095,
+      "learning_rate": 1.8997953825528071e-06,
+      "loss": 0.7954,
+      "step": 34234
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6278782952565733,
+      "learning_rate": 1.8993479576257678e-06,
+      "loss": 0.6934,
+      "step": 34235
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8167923626684004,
+      "learning_rate": 1.898900579863293e-06,
+      "loss": 0.772,
+      "step": 34236
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7060986318828415,
+      "learning_rate": 1.898453249267983e-06,
+      "loss": 0.897,
+      "step": 34237
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.580588940696066,
+      "learning_rate": 1.898005965842441e-06,
+      "loss": 0.8449,
+      "step": 34238
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7514694965618196,
+      "learning_rate": 1.8975587295892761e-06,
+      "loss": 0.7639,
+      "step": 34239
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9583963417444041,
+      "learning_rate": 1.8971115405110863e-06,
+      "loss": 0.8047,
+      "step": 34240
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8138148795810327,
+      "learning_rate": 1.8966643986104816e-06,
+      "loss": 0.9409,
+      "step": 34241
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7658793802621103,
+      "learning_rate": 1.896217303890059e-06,
+      "loss": 0.6436,
+      "step": 34242
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8083138464450969,
+      "learning_rate": 1.8957702563524282e-06,
+      "loss": 0.6684,
+      "step": 34243
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5630116980841413,
+      "learning_rate": 1.8953232560001855e-06,
+      "loss": 0.9292,
+      "step": 34244
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.775859600575519,
+      "learning_rate": 1.8948763028359396e-06,
+      "loss": 0.7172,
+      "step": 34245
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2922322514658537,
+      "learning_rate": 1.8944293968622896e-06,
+      "loss": 0.667,
+      "step": 34246
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.816424532286867,
+      "learning_rate": 1.8939825380818354e-06,
+      "loss": 0.7843,
+      "step": 34247
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6175835954074576,
+      "learning_rate": 1.893535726497182e-06,
+      "loss": 0.7032,
+      "step": 34248
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5066999725916346,
+      "learning_rate": 1.8930889621109295e-06,
+      "loss": 0.8227,
+      "step": 34249
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.864170133801029,
+      "learning_rate": 1.8926422449256842e-06,
+      "loss": 0.9619,
+      "step": 34250
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4728745848579523,
+      "learning_rate": 1.892195574944038e-06,
+      "loss": 0.75,
+      "step": 34251
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.499878997321908,
+      "learning_rate": 1.8917489521685961e-06,
+      "loss": 0.8645,
+      "step": 34252
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7094077087669755,
+      "learning_rate": 1.891302376601961e-06,
+      "loss": 0.9301,
+      "step": 34253
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8025246235416489,
+      "learning_rate": 1.890855848246731e-06,
+      "loss": 0.7946,
+      "step": 34254
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5549435732250818,
+      "learning_rate": 1.8904093671055023e-06,
+      "loss": 0.8917,
+      "step": 34255
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4901220562821185,
+      "learning_rate": 1.8899629331808788e-06,
+      "loss": 0.7578,
+      "step": 34256
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5615654583955525,
+      "learning_rate": 1.8895165464754605e-06,
+      "loss": 0.6845,
+      "step": 34257
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5203186934613042,
+      "learning_rate": 1.8890702069918442e-06,
+      "loss": 0.8822,
+      "step": 34258
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5525138100029152,
+      "learning_rate": 1.888623914732628e-06,
+      "loss": 0.8747,
+      "step": 34259
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5999220703762367,
+      "learning_rate": 1.8881776697004128e-06,
+      "loss": 0.7616,
+      "step": 34260
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7250769206816612,
+      "learning_rate": 1.8877314718977923e-06,
+      "loss": 0.699,
+      "step": 34261
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5467667941615408,
+      "learning_rate": 1.887285321327369e-06,
+      "loss": 0.8565,
+      "step": 34262
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7522827817249035,
+      "learning_rate": 1.8868392179917404e-06,
+      "loss": 0.6873,
+      "step": 34263
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6728774170414158,
+      "learning_rate": 1.8863931618935028e-06,
+      "loss": 1.0395,
+      "step": 34264
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.590932712154965,
+      "learning_rate": 1.8859471530352502e-06,
+      "loss": 0.9156,
+      "step": 34265
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.554066345721741,
+      "learning_rate": 1.885501191419582e-06,
+      "loss": 0.9158,
+      "step": 34266
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5044870061265494,
+      "learning_rate": 1.8850552770491004e-06,
+      "loss": 0.8652,
+      "step": 34267
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7274104541719264,
+      "learning_rate": 1.8846094099263911e-06,
+      "loss": 0.7695,
+      "step": 34268
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6117379948929689,
+      "learning_rate": 1.8841635900540544e-06,
+      "loss": 0.9073,
+      "step": 34269
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5386890917741711,
+      "learning_rate": 1.8837178174346882e-06,
+      "loss": 0.8977,
+      "step": 34270
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5046798226762512,
+      "learning_rate": 1.883272092070888e-06,
+      "loss": 0.8922,
+      "step": 34271
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8416952641180593,
+      "learning_rate": 1.8828264139652474e-06,
+      "loss": 0.7758,
+      "step": 34272
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5037812913923736,
+      "learning_rate": 1.882380783120359e-06,
+      "loss": 0.8349,
+      "step": 34273
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8110086438328041,
+      "learning_rate": 1.8819351995388214e-06,
+      "loss": 0.7798,
+      "step": 34274
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5883912713714072,
+      "learning_rate": 1.881489663223225e-06,
+      "loss": 0.8358,
+      "step": 34275
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5901626304573553,
+      "learning_rate": 1.8810441741761687e-06,
+      "loss": 0.7346,
+      "step": 34276
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4410686557534518,
+      "learning_rate": 1.8805987324002418e-06,
+      "loss": 0.639,
+      "step": 34277
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8901052699629919,
+      "learning_rate": 1.8801533378980407e-06,
+      "loss": 0.8384,
+      "step": 34278
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7275922133805204,
+      "learning_rate": 1.879707990672156e-06,
+      "loss": 0.7059,
+      "step": 34279
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.613397292714456,
+      "learning_rate": 1.8792626907251855e-06,
+      "loss": 0.8036,
+      "step": 34280
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5321313858179324,
+      "learning_rate": 1.878817438059718e-06,
+      "loss": 0.7466,
+      "step": 34281
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.580599584789532,
+      "learning_rate": 1.8783722326783427e-06,
+      "loss": 0.889,
+      "step": 34282
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6907451560367226,
+      "learning_rate": 1.8779270745836576e-06,
+      "loss": 0.6912,
+      "step": 34283
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.600370000198625,
+      "learning_rate": 1.8774819637782538e-06,
+      "loss": 0.8259,
+      "step": 34284
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6441748278781951,
+      "learning_rate": 1.8770369002647215e-06,
+      "loss": 0.8711,
+      "step": 34285
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6123557975103422,
+      "learning_rate": 1.87659188404565e-06,
+      "loss": 0.8487,
+      "step": 34286
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.381076544960243,
+      "learning_rate": 1.8761469151236322e-06,
+      "loss": 0.8089,
+      "step": 34287
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5134110201749433,
+      "learning_rate": 1.875701993501261e-06,
+      "loss": 0.9137,
+      "step": 34288
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.66292183288147,
+      "learning_rate": 1.8752571191811253e-06,
+      "loss": 0.8152,
+      "step": 34289
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6433239466812697,
+      "learning_rate": 1.874812292165813e-06,
+      "loss": 0.8719,
+      "step": 34290
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7098129381746714,
+      "learning_rate": 1.8743675124579152e-06,
+      "loss": 0.8639,
+      "step": 34291
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.710221305521041,
+      "learning_rate": 1.8739227800600246e-06,
+      "loss": 0.6884,
+      "step": 34292
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9708936932640362,
+      "learning_rate": 1.873478094974729e-06,
+      "loss": 0.9549,
+      "step": 34293
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4117358440056014,
+      "learning_rate": 1.8730334572046138e-06,
+      "loss": 0.9382,
+      "step": 34294
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4880772812876069,
+      "learning_rate": 1.872588866752273e-06,
+      "loss": 0.91,
+      "step": 34295
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3286179635532167,
+      "learning_rate": 1.8721443236202907e-06,
+      "loss": 0.8535,
+      "step": 34296
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7689722763933524,
+      "learning_rate": 1.8716998278112597e-06,
+      "loss": 0.8768,
+      "step": 34297
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7009103439487026,
+      "learning_rate": 1.8712553793277643e-06,
+      "loss": 0.8265,
+      "step": 34298
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5911623753023798,
+      "learning_rate": 1.8708109781723949e-06,
+      "loss": 0.774,
+      "step": 34299
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5700967675944317,
+      "learning_rate": 1.8703666243477347e-06,
+      "loss": 0.8307,
+      "step": 34300
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4577868794689555,
+      "learning_rate": 1.8699223178563775e-06,
+      "loss": 0.9407,
+      "step": 34301
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6942411297092255,
+      "learning_rate": 1.8694780587009065e-06,
+      "loss": 0.907,
+      "step": 34302
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6168819824533172,
+      "learning_rate": 1.8690338468839053e-06,
+      "loss": 0.8646,
+      "step": 34303
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7843955100752626,
+      "learning_rate": 1.868589682407963e-06,
+      "loss": 0.8794,
+      "step": 34304
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6262306890244072,
+      "learning_rate": 1.8681455652756685e-06,
+      "loss": 0.8216,
+      "step": 34305
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8407058484396797,
+      "learning_rate": 1.8677014954896055e-06,
+      "loss": 0.707,
+      "step": 34306
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4907187931885069,
+      "learning_rate": 1.8672574730523563e-06,
+      "loss": 0.8125,
+      "step": 34307
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6087152954145167,
+      "learning_rate": 1.8668134979665086e-06,
+      "loss": 0.7721,
+      "step": 34308
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.758551188800284,
+      "learning_rate": 1.8663695702346508e-06,
+      "loss": 0.95,
+      "step": 34309
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9784826789790804,
+      "learning_rate": 1.8659256898593648e-06,
+      "loss": 0.7734,
+      "step": 34310
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5309069967423314,
+      "learning_rate": 1.8654818568432308e-06,
+      "loss": 0.6518,
+      "step": 34311
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4705614221609262,
+      "learning_rate": 1.8650380711888372e-06,
+      "loss": 0.835,
+      "step": 34312
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7700310312132321,
+      "learning_rate": 1.8645943328987704e-06,
+      "loss": 0.7125,
+      "step": 34313
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6523352990824887,
+      "learning_rate": 1.8641506419756105e-06,
+      "loss": 0.8409,
+      "step": 34314
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.684502271024444,
+      "learning_rate": 1.8637069984219391e-06,
+      "loss": 0.8402,
+      "step": 34315
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.001299813482209,
+      "learning_rate": 1.863263402240344e-06,
+      "loss": 0.8598,
+      "step": 34316
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.661213536217143,
+      "learning_rate": 1.862819853433403e-06,
+      "loss": 0.7383,
+      "step": 34317
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.55930666172582,
+      "learning_rate": 1.862376352003703e-06,
+      "loss": 0.7784,
+      "step": 34318
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5195293930996652,
+      "learning_rate": 1.8619328979538221e-06,
+      "loss": 0.8673,
+      "step": 34319
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.656678543108583,
+      "learning_rate": 1.8614894912863469e-06,
+      "loss": 0.8884,
+      "step": 34320
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6178432844964428,
+      "learning_rate": 1.8610461320038541e-06,
+      "loss": 0.9511,
+      "step": 34321
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.642411674423005,
+      "learning_rate": 1.8606028201089298e-06,
+      "loss": 0.9971,
+      "step": 34322
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7458525489459111,
+      "learning_rate": 1.860159555604153e-06,
+      "loss": 0.9115,
+      "step": 34323
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5842136661085202,
+      "learning_rate": 1.8597163384921013e-06,
+      "loss": 0.8488,
+      "step": 34324
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7072586557611942,
+      "learning_rate": 1.8592731687753585e-06,
+      "loss": 0.8953,
+      "step": 34325
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7374362648131805,
+      "learning_rate": 1.8588300464565068e-06,
+      "loss": 0.8863,
+      "step": 34326
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4459835017226599,
+      "learning_rate": 1.8583869715381242e-06,
+      "loss": 0.7983,
+      "step": 34327
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.459929296539786,
+      "learning_rate": 1.8579439440227886e-06,
+      "loss": 0.9206,
+      "step": 34328
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.447364774312403,
+      "learning_rate": 1.8575009639130803e-06,
+      "loss": 0.9147,
+      "step": 34329
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7735493328248678,
+      "learning_rate": 1.8570580312115815e-06,
+      "loss": 0.7304,
+      "step": 34330
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6626758615163073,
+      "learning_rate": 1.8566151459208692e-06,
+      "loss": 0.7867,
+      "step": 34331
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5212378276693572,
+      "learning_rate": 1.856172308043519e-06,
+      "loss": 0.7855,
+      "step": 34332
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5366827313683,
+      "learning_rate": 1.855729517582112e-06,
+      "loss": 0.8573,
+      "step": 34333
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6373730140035607,
+      "learning_rate": 1.855286774539229e-06,
+      "loss": 0.8358,
+      "step": 34334
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4373489685947571,
+      "learning_rate": 1.854844078917445e-06,
+      "loss": 0.895,
+      "step": 34335
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5337818604340527,
+      "learning_rate": 1.8544014307193348e-06,
+      "loss": 0.7486,
+      "step": 34336
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8122187500570042,
+      "learning_rate": 1.8539588299474775e-06,
+      "loss": 0.8118,
+      "step": 34337
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4393325287327123,
+      "learning_rate": 1.8535162766044544e-06,
+      "loss": 0.8345,
+      "step": 34338
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6950406273381358,
+      "learning_rate": 1.853073770692838e-06,
+      "loss": 0.9374,
+      "step": 34339
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8726662106904548,
+      "learning_rate": 1.852631312215204e-06,
+      "loss": 0.9933,
+      "step": 34340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4937484435713437,
+      "learning_rate": 1.8521889011741312e-06,
+      "loss": 0.7675,
+      "step": 34341
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5485149160207363,
+      "learning_rate": 1.851746537572192e-06,
+      "loss": 0.9311,
+      "step": 34342
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6422364496551676,
+      "learning_rate": 1.8513042214119669e-06,
+      "loss": 0.868,
+      "step": 34343
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7897154590134268,
+      "learning_rate": 1.8508619526960258e-06,
+      "loss": 0.8938,
+      "step": 34344
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.475037209232914,
+      "learning_rate": 1.850419731426949e-06,
+      "loss": 0.7653,
+      "step": 34345
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5591627236211825,
+      "learning_rate": 1.8499775576073064e-06,
+      "loss": 0.8856,
+      "step": 34346
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.043286089785884,
+      "learning_rate": 1.849535431239674e-06,
+      "loss": 0.8711,
+      "step": 34347
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6654115279508817,
+      "learning_rate": 1.8490933523266318e-06,
+      "loss": 0.7778,
+      "step": 34348
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6610638206882882,
+      "learning_rate": 1.8486513208707446e-06,
+      "loss": 0.7917,
+      "step": 34349
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7802843200389689,
+      "learning_rate": 1.848209336874589e-06,
+      "loss": 0.9543,
+      "step": 34350
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.548036072290487,
+      "learning_rate": 1.8477674003407398e-06,
+      "loss": 0.9136,
+      "step": 34351
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5682350757589272,
+      "learning_rate": 1.8473255112717747e-06,
+      "loss": 0.7589,
+      "step": 34352
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5096320092231763,
+      "learning_rate": 1.8468836696702574e-06,
+      "loss": 0.8683,
+      "step": 34353
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0952406967500288,
+      "learning_rate": 1.8464418755387635e-06,
+      "loss": 0.7423,
+      "step": 34354
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.623886583913754,
+      "learning_rate": 1.846000128879869e-06,
+      "loss": 0.8324,
+      "step": 34355
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.550087203389436,
+      "learning_rate": 1.8455584296961415e-06,
+      "loss": 0.8662,
+      "step": 34356
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7041327979680525,
+      "learning_rate": 1.8451167779901568e-06,
+      "loss": 0.6943,
+      "step": 34357
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6079906035840852,
+      "learning_rate": 1.8446751737644808e-06,
+      "loss": 0.7528,
+      "step": 34358
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4465838036358603,
+      "learning_rate": 1.8442336170216912e-06,
+      "loss": 0.831,
+      "step": 34359
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5113837949316773,
+      "learning_rate": 1.8437921077643516e-06,
+      "loss": 0.8437,
+      "step": 34360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4578339018440913,
+      "learning_rate": 1.8433506459950412e-06,
+      "loss": 0.8092,
+      "step": 34361
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.884566526338616,
+      "learning_rate": 1.8429092317163244e-06,
+      "loss": 0.6127,
+      "step": 34362
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.811761207043115,
+      "learning_rate": 1.8424678649307702e-06,
+      "loss": 0.9317,
+      "step": 34363
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4568161824265946,
+      "learning_rate": 1.8420265456409502e-06,
+      "loss": 0.7426,
+      "step": 34364
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0983919831999325,
+      "learning_rate": 1.8415852738494378e-06,
+      "loss": 0.9267,
+      "step": 34365
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.472722862968552,
+      "learning_rate": 1.8411440495587985e-06,
+      "loss": 0.9875,
+      "step": 34366
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8787456573811019,
+      "learning_rate": 1.8407028727715992e-06,
+      "loss": 0.7885,
+      "step": 34367
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8229431661947068,
+      "learning_rate": 1.8402617434904103e-06,
+      "loss": 0.7441,
+      "step": 34368
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.753618290017804,
+      "learning_rate": 1.839820661717804e-06,
+      "loss": 0.728,
+      "step": 34369
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.743642032133667,
+      "learning_rate": 1.8393796274563458e-06,
+      "loss": 0.6837,
+      "step": 34370
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4642035648237253,
+      "learning_rate": 1.8389386407085997e-06,
+      "loss": 0.9242,
+      "step": 34371
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4857305254368198,
+      "learning_rate": 1.8384977014771365e-06,
+      "loss": 0.7334,
+      "step": 34372
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.600115719974973,
+      "learning_rate": 1.8380568097645268e-06,
+      "loss": 0.7878,
+      "step": 34373
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5679885381784058,
+      "learning_rate": 1.837615965573334e-06,
+      "loss": 0.8332,
+      "step": 34374
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4565168963700235,
+      "learning_rate": 1.8371751689061223e-06,
+      "loss": 0.7632,
+      "step": 34375
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5511076670396737,
+      "learning_rate": 1.8367344197654647e-06,
+      "loss": 0.9497,
+      "step": 34376
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5253655488298175,
+      "learning_rate": 1.8362937181539198e-06,
+      "loss": 0.6618,
+      "step": 34377
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0822370733821796,
+      "learning_rate": 1.8358530640740612e-06,
+      "loss": 0.8161,
+      "step": 34378
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.542377344286076,
+      "learning_rate": 1.8354124575284471e-06,
+      "loss": 0.7901,
+      "step": 34379
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7545943236674664,
+      "learning_rate": 1.8349718985196496e-06,
+      "loss": 0.7709,
+      "step": 34380
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5080571717112654,
+      "learning_rate": 1.834531387050229e-06,
+      "loss": 0.7217,
+      "step": 34381
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7629984787042621,
+      "learning_rate": 1.8340909231227532e-06,
+      "loss": 0.7112,
+      "step": 34382
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.847249095869871,
+      "learning_rate": 1.8336505067397858e-06,
+      "loss": 0.8968,
+      "step": 34383
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6781510401140904,
+      "learning_rate": 1.833210137903888e-06,
+      "loss": 0.9825,
+      "step": 34384
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5471424506906863,
+      "learning_rate": 1.8327698166176266e-06,
+      "loss": 0.8892,
+      "step": 34385
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8709685752189815,
+      "learning_rate": 1.8323295428835674e-06,
+      "loss": 0.792,
+      "step": 34386
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5515669974287896,
+      "learning_rate": 1.8318893167042717e-06,
+      "loss": 0.9376,
+      "step": 34387
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5462012827385887,
+      "learning_rate": 1.8314491380822985e-06,
+      "loss": 0.9027,
+      "step": 34388
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9464637480837728,
+      "learning_rate": 1.8310090070202158e-06,
+      "loss": 0.8631,
+      "step": 34389
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.199587729443989,
+      "learning_rate": 1.830568923520587e-06,
+      "loss": 0.8105,
+      "step": 34390
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5472501083358972,
+      "learning_rate": 1.8301288875859724e-06,
+      "loss": 0.8306,
+      "step": 34391
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3823350998708934,
+      "learning_rate": 1.829688899218932e-06,
+      "loss": 0.8284,
+      "step": 34392
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.033636931311446,
+      "learning_rate": 1.8292489584220286e-06,
+      "loss": 0.8272,
+      "step": 34393
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6003453647659591,
+      "learning_rate": 1.8288090651978274e-06,
+      "loss": 0.7955,
+      "step": 34394
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.003056245784043,
+      "learning_rate": 1.8283692195488877e-06,
+      "loss": 0.8808,
+      "step": 34395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7288977960319126,
+      "learning_rate": 1.8279294214777653e-06,
+      "loss": 0.6903,
+      "step": 34396
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3508700023699778,
+      "learning_rate": 1.8274896709870294e-06,
+      "loss": 0.6931,
+      "step": 34397
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.634529591245397,
+      "learning_rate": 1.827049968079232e-06,
+      "loss": 0.8463,
+      "step": 34398
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.028751086732328,
+      "learning_rate": 1.8266103127569401e-06,
+      "loss": 0.8615,
+      "step": 34399
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5599374431930089,
+      "learning_rate": 1.8261707050227084e-06,
+      "loss": 0.7707,
+      "step": 34400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.629082254150847,
+      "learning_rate": 1.8257311448791015e-06,
+      "loss": 0.7488,
+      "step": 34401
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3951579400131469,
+      "learning_rate": 1.825291632328673e-06,
+      "loss": 0.8526,
+      "step": 34402
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9581342133428428,
+      "learning_rate": 1.8248521673739872e-06,
+      "loss": 0.8514,
+      "step": 34403
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6984498135347144,
+      "learning_rate": 1.8244127500176e-06,
+      "loss": 0.7392,
+      "step": 34404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5970057876115669,
+      "learning_rate": 1.8239733802620674e-06,
+      "loss": 0.9457,
+      "step": 34405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.639024443302211,
+      "learning_rate": 1.8235340581099503e-06,
+      "loss": 0.8834,
+      "step": 34406
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7154049113023646,
+      "learning_rate": 1.823094783563809e-06,
+      "loss": 0.9387,
+      "step": 34407
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.494193319952871,
+      "learning_rate": 1.8226555566261995e-06,
+      "loss": 0.8575,
+      "step": 34408
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5520066679511917,
+      "learning_rate": 1.8222163772996749e-06,
+      "loss": 0.773,
+      "step": 34409
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6055396367394206,
+      "learning_rate": 1.8217772455867965e-06,
+      "loss": 0.8369,
+      "step": 34410
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6172958938926758,
+      "learning_rate": 1.8213381614901215e-06,
+      "loss": 0.8274,
+      "step": 34411
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7214777896925888,
+      "learning_rate": 1.8208991250122054e-06,
+      "loss": 0.8196,
+      "step": 34412
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9825663421166249,
+      "learning_rate": 1.8204601361556019e-06,
+      "loss": 0.8318,
+      "step": 34413
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.410746119006178,
+      "learning_rate": 1.8200211949228686e-06,
+      "loss": 0.8657,
+      "step": 34414
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4070623318208715,
+      "learning_rate": 1.8195823013165648e-06,
+      "loss": 0.9394,
+      "step": 34415
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8173023321326935,
+      "learning_rate": 1.819143455339243e-06,
+      "loss": 0.8203,
+      "step": 34416
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5079947701205314,
+      "learning_rate": 1.818704656993454e-06,
+      "loss": 0.8158,
+      "step": 34417
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6039168390285035,
+      "learning_rate": 1.8182659062817609e-06,
+      "loss": 0.8822,
+      "step": 34418
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.617077647284175,
+      "learning_rate": 1.8178272032067102e-06,
+      "loss": 0.8821,
+      "step": 34419
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4881054086477763,
+      "learning_rate": 1.8173885477708631e-06,
+      "loss": 0.9586,
+      "step": 34420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7499720906535341,
+      "learning_rate": 1.8169499399767676e-06,
+      "loss": 0.7047,
+      "step": 34421
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4886907466821728,
+      "learning_rate": 1.8165113798269828e-06,
+      "loss": 0.8392,
+      "step": 34422
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.62015558373562,
+      "learning_rate": 1.8160728673240568e-06,
+      "loss": 0.7582,
+      "step": 34423
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5753379080335723,
+      "learning_rate": 1.8156344024705485e-06,
+      "loss": 0.9096,
+      "step": 34424
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7727140091291635,
+      "learning_rate": 1.8151959852690083e-06,
+      "loss": 0.899,
+      "step": 34425
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7368938274892796,
+      "learning_rate": 1.814757615721985e-06,
+      "loss": 0.788,
+      "step": 34426
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.55753086876335,
+      "learning_rate": 1.8143192938320353e-06,
+      "loss": 0.8143,
+      "step": 34427
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5023870494530598,
+      "learning_rate": 1.813881019601712e-06,
+      "loss": 0.7443,
+      "step": 34428
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3909678043757936,
+      "learning_rate": 1.8134427930335651e-06,
+      "loss": 0.8518,
+      "step": 34429
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3395573511972665,
+      "learning_rate": 1.8130046141301426e-06,
+      "loss": 0.6806,
+      "step": 34430
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5435506400987227,
+      "learning_rate": 1.8125664828940005e-06,
+      "loss": 0.8132,
+      "step": 34431
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8181188473237213,
+      "learning_rate": 1.8121283993276906e-06,
+      "loss": 0.8665,
+      "step": 34432
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5979726307686177,
+      "learning_rate": 1.8116903634337612e-06,
+      "loss": 0.7908,
+      "step": 34433
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5414446411016918,
+      "learning_rate": 1.8112523752147615e-06,
+      "loss": 0.7859,
+      "step": 34434
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5647239883845852,
+      "learning_rate": 1.8108144346732414e-06,
+      "loss": 0.8091,
+      "step": 34435
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.675870754630046,
+      "learning_rate": 1.810376541811756e-06,
+      "loss": 0.8766,
+      "step": 34436
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.678545806915165,
+      "learning_rate": 1.8099386966328503e-06,
+      "loss": 0.7057,
+      "step": 34437
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2279323653490084,
+      "learning_rate": 1.8095008991390728e-06,
+      "loss": 0.8314,
+      "step": 34438
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9398614232241806,
+      "learning_rate": 1.8090631493329759e-06,
+      "loss": 0.7459,
+      "step": 34439
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7386809312950142,
+      "learning_rate": 1.8086254472171039e-06,
+      "loss": 0.728,
+      "step": 34440
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.478435093782218,
+      "learning_rate": 1.8081877927940107e-06,
+      "loss": 0.9723,
+      "step": 34441
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4605021566791208,
+      "learning_rate": 1.8077501860662395e-06,
+      "loss": 0.7576,
+      "step": 34442
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8711119859615364,
+      "learning_rate": 1.8073126270363428e-06,
+      "loss": 0.8333,
+      "step": 34443
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9201149635112413,
+      "learning_rate": 1.8068751157068632e-06,
+      "loss": 0.6869,
+      "step": 34444
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7804392457114592,
+      "learning_rate": 1.8064376520803495e-06,
+      "loss": 0.6983,
+      "step": 34445
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3742570419617097,
+      "learning_rate": 1.8060002361593565e-06,
+      "loss": 0.8209,
+      "step": 34446
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7891927475479799,
+      "learning_rate": 1.8055628679464177e-06,
+      "loss": 0.6735,
+      "step": 34447
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7700366764088963,
+      "learning_rate": 1.8051255474440876e-06,
+      "loss": 0.6479,
+      "step": 34448
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.89827931916302,
+      "learning_rate": 1.8046882746549099e-06,
+      "loss": 0.8504,
+      "step": 34449
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.838259376981991,
+      "learning_rate": 1.8042510495814369e-06,
+      "loss": 0.7095,
+      "step": 34450
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.525406138804314,
+      "learning_rate": 1.8038138722262045e-06,
+      "loss": 0.8672,
+      "step": 34451
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4814521221873185,
+      "learning_rate": 1.8033767425917625e-06,
+      "loss": 0.7103,
+      "step": 34452
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6073447256546165,
+      "learning_rate": 1.802939660680656e-06,
+      "loss": 0.8516,
+      "step": 34453
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7418601826777083,
+      "learning_rate": 1.8025026264954326e-06,
+      "loss": 0.6723,
+      "step": 34454
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7186571765178786,
+      "learning_rate": 1.8020656400386338e-06,
+      "loss": 0.858,
+      "step": 34455
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5626522983334339,
+      "learning_rate": 1.801628701312802e-06,
+      "loss": 0.9465,
+      "step": 34456
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4649000089888662,
+      "learning_rate": 1.801191810320486e-06,
+      "loss": 0.9394,
+      "step": 34457
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8309454607015558,
+      "learning_rate": 1.8007549670642244e-06,
+      "loss": 0.9089,
+      "step": 34458
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5724071169502893,
+      "learning_rate": 1.8003181715465657e-06,
+      "loss": 0.8824,
+      "step": 34459
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4670254507650038,
+      "learning_rate": 1.799881423770049e-06,
+      "loss": 0.8553,
+      "step": 34460
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.531121478819501,
+      "learning_rate": 1.7994447237372204e-06,
+      "loss": 0.8196,
+      "step": 34461
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.752089539568956,
+      "learning_rate": 1.7990080714506175e-06,
+      "loss": 0.7073,
+      "step": 34462
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0721697383737596,
+      "learning_rate": 1.7985714669127897e-06,
+      "loss": 0.9613,
+      "step": 34463
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9842786274504103,
+      "learning_rate": 1.7981349101262746e-06,
+      "loss": 0.8041,
+      "step": 34464
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6480048678150365,
+      "learning_rate": 1.7976984010936128e-06,
+      "loss": 0.8105,
+      "step": 34465
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4573114987409945,
+      "learning_rate": 1.7972619398173474e-06,
+      "loss": 0.9498,
+      "step": 34466
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0520412088153286,
+      "learning_rate": 1.7968255263000235e-06,
+      "loss": 0.9107,
+      "step": 34467
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.731580664063065,
+      "learning_rate": 1.7963891605441775e-06,
+      "loss": 0.884,
+      "step": 34468
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8149690400807,
+      "learning_rate": 1.7959528425523488e-06,
+      "loss": 0.7947,
+      "step": 34469
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4739852947085437,
+      "learning_rate": 1.7955165723270796e-06,
+      "loss": 0.8119,
+      "step": 34470
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7637967503550879,
+      "learning_rate": 1.7950803498709124e-06,
+      "loss": 0.7042,
+      "step": 34471
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6666112183294834,
+      "learning_rate": 1.7946441751863864e-06,
+      "loss": 0.711,
+      "step": 34472
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8333060413077464,
+      "learning_rate": 1.794208048276036e-06,
+      "loss": 0.7314,
+      "step": 34473
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7513593901571609,
+      "learning_rate": 1.793771969142405e-06,
+      "loss": 0.7265,
+      "step": 34474
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.648882075251698,
+      "learning_rate": 1.7933359377880344e-06,
+      "loss": 0.912,
+      "step": 34475
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6371708139437702,
+      "learning_rate": 1.79289995421546e-06,
+      "loss": 0.8562,
+      "step": 34476
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.55439224055385,
+      "learning_rate": 1.7924640184272169e-06,
+      "loss": 0.7396,
+      "step": 34477
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8214169817360752,
+      "learning_rate": 1.7920281304258502e-06,
+      "loss": 0.8036,
+      "step": 34478
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5580778454413826,
+      "learning_rate": 1.7915922902138916e-06,
+      "loss": 0.8256,
+      "step": 34479
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8397082159096296,
+      "learning_rate": 1.7911564977938845e-06,
+      "loss": 0.822,
+      "step": 34480
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7688270276839148,
+      "learning_rate": 1.7907207531683601e-06,
+      "loss": 0.7086,
+      "step": 34481
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1732454256384335,
+      "learning_rate": 1.790285056339861e-06,
+      "loss": 0.7131,
+      "step": 34482
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3944573742635178,
+      "learning_rate": 1.7898494073109184e-06,
+      "loss": 0.76,
+      "step": 34483
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8344550553603444,
+      "learning_rate": 1.7894138060840749e-06,
+      "loss": 0.7071,
+      "step": 34484
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5466657709879832,
+      "learning_rate": 1.7889782526618637e-06,
+      "loss": 0.7583,
+      "step": 34485
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5436563447606118,
+      "learning_rate": 1.7885427470468185e-06,
+      "loss": 0.8678,
+      "step": 34486
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6184965937790254,
+      "learning_rate": 1.7881072892414763e-06,
+      "loss": 0.7642,
+      "step": 34487
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7923931114112994,
+      "learning_rate": 1.7876718792483772e-06,
+      "loss": 0.672,
+      "step": 34488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6620979146892894,
+      "learning_rate": 1.7872365170700511e-06,
+      "loss": 0.8965,
+      "step": 34489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5156362552699592,
+      "learning_rate": 1.786801202709032e-06,
+      "loss": 0.8985,
+      "step": 34490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7131493093344314,
+      "learning_rate": 1.7863659361678566e-06,
+      "loss": 0.8446,
+      "step": 34491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.711062496183376,
+      "learning_rate": 1.7859307174490614e-06,
+      "loss": 0.6831,
+      "step": 34492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5579075595888414,
+      "learning_rate": 1.7854955465551783e-06,
+      "loss": 0.8813,
+      "step": 34493
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4820139838202586,
+      "learning_rate": 1.7850604234887382e-06,
+      "loss": 0.9163,
+      "step": 34494
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8036969722261127,
+      "learning_rate": 1.7846253482522768e-06,
+      "loss": 0.869,
+      "step": 34495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8325672807299243,
+      "learning_rate": 1.7841903208483302e-06,
+      "loss": 0.8023,
+      "step": 34496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7247482140100048,
+      "learning_rate": 1.7837553412794295e-06,
+      "loss": 0.9976,
+      "step": 34497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7772566878661311,
+      "learning_rate": 1.7833204095481027e-06,
+      "loss": 0.7923,
+      "step": 34498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4660274701650975,
+      "learning_rate": 1.7828855256568889e-06,
+      "loss": 0.796,
+      "step": 34499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8015291703245468,
+      "learning_rate": 1.782450689608315e-06,
+      "loss": 0.7317,
+      "step": 34500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.624467816514216,
+      "learning_rate": 1.7820159014049165e-06,
+      "loss": 0.8433,
+      "step": 34501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5444248549079658,
+      "learning_rate": 1.7815811610492217e-06,
+      "loss": 0.882,
+      "step": 34502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7302934592785746,
+      "learning_rate": 1.781146468543765e-06,
+      "loss": 0.6913,
+      "step": 34503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8892229516960228,
+      "learning_rate": 1.7807118238910725e-06,
+      "loss": 0.7541,
+      "step": 34504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5434001727456306,
+      "learning_rate": 1.7802772270936818e-06,
+      "loss": 0.9033,
+      "step": 34505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5289408453870752,
+      "learning_rate": 1.779842678154119e-06,
+      "loss": 0.6692,
+      "step": 34506
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7170574090021296,
+      "learning_rate": 1.7794081770749116e-06,
+      "loss": 0.9521,
+      "step": 34507
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7292554905475523,
+      "learning_rate": 1.7789737238585925e-06,
+      "loss": 0.8021,
+      "step": 34508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5295948617891897,
+      "learning_rate": 1.7785393185076939e-06,
+      "loss": 0.8038,
+      "step": 34509
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4907345828515732,
+      "learning_rate": 1.7781049610247425e-06,
+      "loss": 0.8663,
+      "step": 34510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.742557126557966,
+      "learning_rate": 1.7776706514122644e-06,
+      "loss": 0.8931,
+      "step": 34511
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4599223989404584,
+      "learning_rate": 1.7772363896727907e-06,
+      "loss": 0.9067,
+      "step": 34512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4781963545468577,
+      "learning_rate": 1.7768021758088517e-06,
+      "loss": 0.6884,
+      "step": 34513
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4311682591794646,
+      "learning_rate": 1.7763680098229752e-06,
+      "loss": 0.9999,
+      "step": 34514
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6053678568682608,
+      "learning_rate": 1.775933891717685e-06,
+      "loss": 0.7963,
+      "step": 34515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.719119878902611,
+      "learning_rate": 1.775499821495511e-06,
+      "loss": 0.8127,
+      "step": 34516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.974804098725573,
+      "learning_rate": 1.7750657991589827e-06,
+      "loss": 0.8999,
+      "step": 34517
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7547221295296362,
+      "learning_rate": 1.7746318247106265e-06,
+      "loss": 0.8471,
+      "step": 34518
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7581893608144954,
+      "learning_rate": 1.7741978981529651e-06,
+      "loss": 0.6964,
+      "step": 34519
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.807006855554728,
+      "learning_rate": 1.7737640194885297e-06,
+      "loss": 0.8351,
+      "step": 34520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.772729275032229,
+      "learning_rate": 1.773330188719843e-06,
+      "loss": 0.9685,
+      "step": 34521
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6806743924276761,
+      "learning_rate": 1.7728964058494336e-06,
+      "loss": 0.7706,
+      "step": 34522
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4829649976324755,
+      "learning_rate": 1.7724626708798253e-06,
+      "loss": 0.8142,
+      "step": 34523
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7433314147090295,
+      "learning_rate": 1.7720289838135452e-06,
+      "loss": 0.9078,
+      "step": 34524
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5157089797345427,
+      "learning_rate": 1.7715953446531153e-06,
+      "loss": 0.866,
+      "step": 34525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6084017796994143,
+      "learning_rate": 1.771161753401065e-06,
+      "loss": 0.7833,
+      "step": 34526
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.358143004238984,
+      "learning_rate": 1.7707282100599154e-06,
+      "loss": 0.6937,
+      "step": 34527
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7625251561748063,
+      "learning_rate": 1.7702947146321904e-06,
+      "loss": 0.8329,
+      "step": 34528
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9993683374672284,
+      "learning_rate": 1.769861267120414e-06,
+      "loss": 0.7528,
+      "step": 34529
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6102899422814578,
+      "learning_rate": 1.7694278675271115e-06,
+      "loss": 0.8339,
+      "step": 34530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7607537357215552,
+      "learning_rate": 1.7689945158548094e-06,
+      "loss": 0.7637,
+      "step": 34531
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7873510530176329,
+      "learning_rate": 1.7685612121060235e-06,
+      "loss": 0.8775,
+      "step": 34532
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7710300725749356,
+      "learning_rate": 1.7681279562832798e-06,
+      "loss": 0.727,
+      "step": 34533
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3191034646729136,
+      "learning_rate": 1.7676947483891039e-06,
+      "loss": 0.8126,
+      "step": 34534
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7525699561444203,
+      "learning_rate": 1.7672615884260159e-06,
+      "loss": 0.7292,
+      "step": 34535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8144240549595199,
+      "learning_rate": 1.7668284763965337e-06,
+      "loss": 0.9706,
+      "step": 34536
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9915543542049097,
+      "learning_rate": 1.7663954123031834e-06,
+      "loss": 0.8276,
+      "step": 34537
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4800124739940244,
+      "learning_rate": 1.7659623961484883e-06,
+      "loss": 0.7775,
+      "step": 34538
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9603445661148586,
+      "learning_rate": 1.7655294279349666e-06,
+      "loss": 0.9106,
+      "step": 34539
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5165641689710785,
+      "learning_rate": 1.7650965076651383e-06,
+      "loss": 0.8482,
+      "step": 34540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6367805138454856,
+      "learning_rate": 1.764663635341527e-06,
+      "loss": 0.862,
+      "step": 34541
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6777973835417397,
+      "learning_rate": 1.7642308109666483e-06,
+      "loss": 0.805,
+      "step": 34542
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.381102475888645,
+      "learning_rate": 1.763798034543025e-06,
+      "loss": 0.8691,
+      "step": 34543
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7573546390147545,
+      "learning_rate": 1.7633653060731803e-06,
+      "loss": 0.8207,
+      "step": 34544
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.352616341583527,
+      "learning_rate": 1.7629326255596302e-06,
+      "loss": 0.8701,
+      "step": 34545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5865138704883375,
+      "learning_rate": 1.7624999930048913e-06,
+      "loss": 0.6807,
+      "step": 34546
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6503662835179842,
+      "learning_rate": 1.7620674084114865e-06,
+      "loss": 0.931,
+      "step": 34547
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6699659738030042,
+      "learning_rate": 1.7616348717819376e-06,
+      "loss": 0.8364,
+      "step": 34548
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7818047064828346,
+      "learning_rate": 1.761202383118754e-06,
+      "loss": 0.7239,
+      "step": 34549
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8455052224281456,
+      "learning_rate": 1.7607699424244583e-06,
+      "loss": 0.8278,
+      "step": 34550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6502023732374287,
+      "learning_rate": 1.7603375497015684e-06,
+      "loss": 0.8464,
+      "step": 34551
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9045849609097254,
+      "learning_rate": 1.7599052049526045e-06,
+      "loss": 0.7983,
+      "step": 34552
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6451086975943108,
+      "learning_rate": 1.759472908180081e-06,
+      "loss": 0.8019,
+      "step": 34553
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4616793426240662,
+      "learning_rate": 1.7590406593865128e-06,
+      "loss": 0.793,
+      "step": 34554
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6110308383568555,
+      "learning_rate": 1.7586084585744213e-06,
+      "loss": 0.8046,
+      "step": 34555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1459593774900787,
+      "learning_rate": 1.7581763057463187e-06,
+      "loss": 0.7637,
+      "step": 34556
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6108187142075014,
+      "learning_rate": 1.757744200904724e-06,
+      "loss": 0.9637,
+      "step": 34557
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7397876866553585,
+      "learning_rate": 1.7573121440521513e-06,
+      "loss": 0.7972,
+      "step": 34558
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6279534411542362,
+      "learning_rate": 1.756880135191118e-06,
+      "loss": 0.8276,
+      "step": 34559
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.204664441641471,
+      "learning_rate": 1.756448174324137e-06,
+      "loss": 0.835,
+      "step": 34560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7279391749467745,
+      "learning_rate": 1.7560162614537257e-06,
+      "loss": 0.8802,
+      "step": 34561
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7277540703154541,
+      "learning_rate": 1.7555843965823992e-06,
+      "loss": 0.7643,
+      "step": 34562
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6427504530455177,
+      "learning_rate": 1.7551525797126668e-06,
+      "loss": 0.7902,
+      "step": 34563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6338779704911275,
+      "learning_rate": 1.7547208108470471e-06,
+      "loss": 0.8304,
+      "step": 34564
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5021922793438132,
+      "learning_rate": 1.7542890899880548e-06,
+      "loss": 0.8557,
+      "step": 34565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4831804105241535,
+      "learning_rate": 1.7538574171382028e-06,
+      "loss": 0.8399,
+      "step": 34566
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4618452014477101,
+      "learning_rate": 1.7534257923000008e-06,
+      "loss": 0.8996,
+      "step": 34567
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3387460275778933,
+      "learning_rate": 1.752994215475965e-06,
+      "loss": 0.898,
+      "step": 34568
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.276892066960434,
+      "learning_rate": 1.7525626866686107e-06,
+      "loss": 0.7904,
+      "step": 34569
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5382725898513494,
+      "learning_rate": 1.7521312058804462e-06,
+      "loss": 0.7203,
+      "step": 34570
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7316565440579373,
+      "learning_rate": 1.751699773113984e-06,
+      "loss": 0.7129,
+      "step": 34571
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7361475273401463,
+      "learning_rate": 1.7512683883717363e-06,
+      "loss": 0.6905,
+      "step": 34572
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.973648166352599,
+      "learning_rate": 1.7508370516562179e-06,
+      "loss": 0.9045,
+      "step": 34573
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0667019076891573,
+      "learning_rate": 1.7504057629699389e-06,
+      "loss": 0.8044,
+      "step": 34574
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9855297361551107,
+      "learning_rate": 1.7499745223154053e-06,
+      "loss": 0.8488,
+      "step": 34575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6057545290362727,
+      "learning_rate": 1.7495433296951325e-06,
+      "loss": 0.9535,
+      "step": 34576
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6060219042947481,
+      "learning_rate": 1.7491121851116332e-06,
+      "loss": 0.6577,
+      "step": 34577
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7645747420924798,
+      "learning_rate": 1.7486810885674154e-06,
+      "loss": 0.738,
+      "step": 34578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6014545324687273,
+      "learning_rate": 1.7482500400649849e-06,
+      "loss": 0.8416,
+      "step": 34579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3853907509568133,
+      "learning_rate": 1.7478190396068584e-06,
+      "loss": 0.6813,
+      "step": 34580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.69380794646228,
+      "learning_rate": 1.7473880871955395e-06,
+      "loss": 0.9055,
+      "step": 34581
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5123112868421278,
+      "learning_rate": 1.746957182833543e-06,
+      "loss": 0.7294,
+      "step": 34582
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.508293145391352,
+      "learning_rate": 1.7465263265233712e-06,
+      "loss": 0.917,
+      "step": 34583
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5577294812001403,
+      "learning_rate": 1.7460955182675387e-06,
+      "loss": 0.9299,
+      "step": 34584
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4840960800543648,
+      "learning_rate": 1.7456647580685492e-06,
+      "loss": 0.9131,
+      "step": 34585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.705732399377629,
+      "learning_rate": 1.7452340459289152e-06,
+      "loss": 0.8049,
+      "step": 34586
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7038016116605186,
+      "learning_rate": 1.7448033818511422e-06,
+      "loss": 0.8525,
+      "step": 34587
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8397793592314768,
+      "learning_rate": 1.744372765837734e-06,
+      "loss": 0.8004,
+      "step": 34588
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0234465604035217,
+      "learning_rate": 1.743942197891202e-06,
+      "loss": 0.8324,
+      "step": 34589
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8187516064532816,
+      "learning_rate": 1.743511678014055e-06,
+      "loss": 0.7218,
+      "step": 34590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4717384495850305,
+      "learning_rate": 1.7430812062087965e-06,
+      "loss": 0.8181,
+      "step": 34591
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6700360763306699,
+      "learning_rate": 1.7426507824779303e-06,
+      "loss": 0.8298,
+      "step": 34592
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9013245350766237,
+      "learning_rate": 1.7422204068239667e-06,
+      "loss": 0.7667,
+      "step": 34593
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5790642696229584,
+      "learning_rate": 1.7417900792494126e-06,
+      "loss": 0.8934,
+      "step": 34594
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1794891673098853,
+      "learning_rate": 1.7413597997567699e-06,
+      "loss": 0.7964,
+      "step": 34595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5662191640759373,
+      "learning_rate": 1.740929568348544e-06,
+      "loss": 0.7439,
+      "step": 34596
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.574720651656396,
+      "learning_rate": 1.74049938502724e-06,
+      "loss": 0.8529,
+      "step": 34597
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.580347751345335,
+      "learning_rate": 1.7400692497953665e-06,
+      "loss": 0.8811,
+      "step": 34598
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7048065700638992,
+      "learning_rate": 1.7396391626554254e-06,
+      "loss": 0.8783,
+      "step": 34599
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.780692744276207,
+      "learning_rate": 1.7392091236099173e-06,
+      "loss": 0.7237,
+      "step": 34600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6338120583302795,
+      "learning_rate": 1.7387791326613511e-06,
+      "loss": 0.883,
+      "step": 34601
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.669585365607309,
+      "learning_rate": 1.7383491898122262e-06,
+      "loss": 0.8164,
+      "step": 34602
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5980409851760604,
+      "learning_rate": 1.7379192950650504e-06,
+      "loss": 0.8791,
+      "step": 34603
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5420257467581024,
+      "learning_rate": 1.7374894484223215e-06,
+      "loss": 0.8197,
+      "step": 34604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6691391743111716,
+      "learning_rate": 1.7370596498865467e-06,
+      "loss": 0.8832,
+      "step": 34605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5877744137578227,
+      "learning_rate": 1.736629899460225e-06,
+      "loss": 0.86,
+      "step": 34606
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9179901657018419,
+      "learning_rate": 1.736200197145863e-06,
+      "loss": 0.8418,
+      "step": 34607
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8454084696342878,
+      "learning_rate": 1.7357705429459592e-06,
+      "loss": 0.8123,
+      "step": 34608
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7880067119281998,
+      "learning_rate": 1.7353409368630126e-06,
+      "loss": 0.7363,
+      "step": 34609
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6363968991858548,
+      "learning_rate": 1.7349113788995288e-06,
+      "loss": 0.8458,
+      "step": 34610
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7380464052355623,
+      "learning_rate": 1.734481869058009e-06,
+      "loss": 0.7186,
+      "step": 34611
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.67560542096746,
+      "learning_rate": 1.7340524073409537e-06,
+      "loss": 0.7998,
+      "step": 34612
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7911356104439385,
+      "learning_rate": 1.7336229937508586e-06,
+      "loss": 0.7212,
+      "step": 34613
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7920098339897715,
+      "learning_rate": 1.7331936282902284e-06,
+      "loss": 0.7375,
+      "step": 34614
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6921023612749082,
+      "learning_rate": 1.7327643109615644e-06,
+      "loss": 0.7766,
+      "step": 34615
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5088026399140404,
+      "learning_rate": 1.7323350417673634e-06,
+      "loss": 0.8194,
+      "step": 34616
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.624998184260256,
+      "learning_rate": 1.7319058207101237e-06,
+      "loss": 0.9058,
+      "step": 34617
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6911081802598036,
+      "learning_rate": 1.7314766477923461e-06,
+      "loss": 0.779,
+      "step": 34618
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5129042150680607,
+      "learning_rate": 1.7310475230165302e-06,
+      "loss": 0.8167,
+      "step": 34619
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6510106669580518,
+      "learning_rate": 1.7306184463851749e-06,
+      "loss": 0.7956,
+      "step": 34620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4650885034571635,
+      "learning_rate": 1.7301894179007738e-06,
+      "loss": 0.7645,
+      "step": 34621
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5888083240386774,
+      "learning_rate": 1.7297604375658317e-06,
+      "loss": 0.7616,
+      "step": 34622
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5177315427076696,
+      "learning_rate": 1.7293315053828385e-06,
+      "loss": 0.8503,
+      "step": 34623
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6735782567544613,
+      "learning_rate": 1.7289026213542992e-06,
+      "loss": 0.8751,
+      "step": 34624
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5026909717110173,
+      "learning_rate": 1.728473785482705e-06,
+      "loss": 0.7348,
+      "step": 34625
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.493449063481272,
+      "learning_rate": 1.7280449977705572e-06,
+      "loss": 0.7967,
+      "step": 34626
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6202429448628053,
+      "learning_rate": 1.727616258220348e-06,
+      "loss": 0.8306,
+      "step": 34627
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.534352866476453,
+      "learning_rate": 1.727187566834576e-06,
+      "loss": 0.7968,
+      "step": 34628
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7571543801662919,
+      "learning_rate": 1.7267589236157422e-06,
+      "loss": 0.8304,
+      "step": 34629
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4499191778482619,
+      "learning_rate": 1.7263303285663324e-06,
+      "loss": 0.8188,
+      "step": 34630
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7645268605268867,
+      "learning_rate": 1.7259017816888469e-06,
+      "loss": 0.695,
+      "step": 34631
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5774850172480304,
+      "learning_rate": 1.72547328298578e-06,
+      "loss": 0.8216,
+      "step": 34632
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5889831266133254,
+      "learning_rate": 1.7250448324596325e-06,
+      "loss": 0.7971,
+      "step": 34633
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7148898446630785,
+      "learning_rate": 1.7246164301128888e-06,
+      "loss": 0.9587,
+      "step": 34634
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6050439851729137,
+      "learning_rate": 1.724188075948049e-06,
+      "loss": 1.0039,
+      "step": 34635
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.674004567652839,
+      "learning_rate": 1.7237597699676079e-06,
+      "loss": 0.8074,
+      "step": 34636
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7602164989694995,
+      "learning_rate": 1.723331512174058e-06,
+      "loss": 0.7637,
+      "step": 34637
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7589611735235925,
+      "learning_rate": 1.7229033025698894e-06,
+      "loss": 0.6619,
+      "step": 34638
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8757213470774279,
+      "learning_rate": 1.7224751411575991e-06,
+      "loss": 0.8107,
+      "step": 34639
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6704375997209417,
+      "learning_rate": 1.7220470279396806e-06,
+      "loss": 0.9473,
+      "step": 34640
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0298445494667203,
+      "learning_rate": 1.721618962918623e-06,
+      "loss": 0.8278,
+      "step": 34641
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5732986259815067,
+      "learning_rate": 1.7211909460969233e-06,
+      "loss": 0.9476,
+      "step": 34642
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6534394540018,
+      "learning_rate": 1.7207629774770718e-06,
+      "loss": 0.8006,
+      "step": 34643
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.60494540219461,
+      "learning_rate": 1.7203350570615563e-06,
+      "loss": 0.8201,
+      "step": 34644
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7683233299667767,
+      "learning_rate": 1.7199071848528703e-06,
+      "loss": 0.6935,
+      "step": 34645
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.46459247293592,
+      "learning_rate": 1.7194793608535099e-06,
+      "loss": 0.9318,
+      "step": 34646
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5168360914543957,
+      "learning_rate": 1.7190515850659616e-06,
+      "loss": 0.9074,
+      "step": 34647
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5699890813237374,
+      "learning_rate": 1.718623857492715e-06,
+      "loss": 0.9335,
+      "step": 34648
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5672026839133117,
+      "learning_rate": 1.718196178136261e-06,
+      "loss": 0.8811,
+      "step": 34649
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7664497419246596,
+      "learning_rate": 1.7177685469990946e-06,
+      "loss": 0.8362,
+      "step": 34650
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7981280228293951,
+      "learning_rate": 1.7173409640837002e-06,
+      "loss": 0.6919,
+      "step": 34651
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5501809930724293,
+      "learning_rate": 1.7169134293925682e-06,
+      "loss": 0.7287,
+      "step": 34652
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6010291632213702,
+      "learning_rate": 1.7164859429281876e-06,
+      "loss": 0.8485,
+      "step": 34653
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2124270671161126,
+      "learning_rate": 1.716058504693051e-06,
+      "loss": 0.8051,
+      "step": 34654
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4879509182103876,
+      "learning_rate": 1.7156311146896442e-06,
+      "loss": 0.8644,
+      "step": 34655
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5548897861649087,
+      "learning_rate": 1.715203772920453e-06,
+      "loss": 0.7926,
+      "step": 34656
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.06013996413842,
+      "learning_rate": 1.7147764793879718e-06,
+      "loss": 0.7493,
+      "step": 34657
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5872070660974282,
+      "learning_rate": 1.7143492340946821e-06,
+      "loss": 0.7558,
+      "step": 34658
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4321724894866752,
+      "learning_rate": 1.7139220370430765e-06,
+      "loss": 0.8789,
+      "step": 34659
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4875150404086648,
+      "learning_rate": 1.7134948882356383e-06,
+      "loss": 0.9263,
+      "step": 34660
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5487433579040717,
+      "learning_rate": 1.7130677876748592e-06,
+      "loss": 0.8393,
+      "step": 34661
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7100686685997206,
+      "learning_rate": 1.7126407353632191e-06,
+      "loss": 0.7614,
+      "step": 34662
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7281134117915582,
+      "learning_rate": 1.7122137313032116e-06,
+      "loss": 0.8895,
+      "step": 34663
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7950321054765328,
+      "learning_rate": 1.7117867754973206e-06,
+      "loss": 0.7956,
+      "step": 34664
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5534696957394634,
+      "learning_rate": 1.711359867948027e-06,
+      "loss": 0.9158,
+      "step": 34665
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8708066806108041,
+      "learning_rate": 1.7109330086578213e-06,
+      "loss": 0.8314,
+      "step": 34666
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0650792197532573,
+      "learning_rate": 1.7105061976291893e-06,
+      "loss": 0.7791,
+      "step": 34667
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5937301350943849,
+      "learning_rate": 1.7100794348646155e-06,
+      "loss": 0.748,
+      "step": 34668
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7638768034677891,
+      "learning_rate": 1.7096527203665814e-06,
+      "loss": 0.7336,
+      "step": 34669
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.517947959301209,
+      "learning_rate": 1.7092260541375738e-06,
+      "loss": 0.8189,
+      "step": 34670
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0810280397037966,
+      "learning_rate": 1.7087994361800787e-06,
+      "loss": 0.7068,
+      "step": 34671
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4801354467135497,
+      "learning_rate": 1.7083728664965793e-06,
+      "loss": 0.9725,
+      "step": 34672
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7386343041259442,
+      "learning_rate": 1.707946345089555e-06,
+      "loss": 0.7956,
+      "step": 34673
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8718883955710834,
+      "learning_rate": 1.7075198719614927e-06,
+      "loss": 0.8958,
+      "step": 34674
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.665117514085628,
+      "learning_rate": 1.707093447114877e-06,
+      "loss": 0.8345,
+      "step": 34675
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5602359398180918,
+      "learning_rate": 1.7066670705521893e-06,
+      "loss": 0.8971,
+      "step": 34676
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.769026945112901,
+      "learning_rate": 1.7062407422759085e-06,
+      "loss": 0.9079,
+      "step": 34677
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7369844764652741,
+      "learning_rate": 1.705814462288522e-06,
+      "loss": 0.6884,
+      "step": 34678
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5511552025141058,
+      "learning_rate": 1.7053882305925085e-06,
+      "loss": 0.7305,
+      "step": 34679
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4945475404822293,
+      "learning_rate": 1.7049620471903517e-06,
+      "loss": 0.6729,
+      "step": 34680
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4422737819398568,
+      "learning_rate": 1.7045359120845307e-06,
+      "loss": 0.8001,
+      "step": 34681
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5784757569276135,
+      "learning_rate": 1.70410982527753e-06,
+      "loss": 0.8064,
+      "step": 34682
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.002649256015159,
+      "learning_rate": 1.7036837867718259e-06,
+      "loss": 0.9588,
+      "step": 34683
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6116685010738676,
+      "learning_rate": 1.7032577965699037e-06,
+      "loss": 0.9676,
+      "step": 34684
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5319302287842647,
+      "learning_rate": 1.7028318546742418e-06,
+      "loss": 0.8026,
+      "step": 34685
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6683185261982907,
+      "learning_rate": 1.7024059610873167e-06,
+      "loss": 0.8655,
+      "step": 34686
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.44101611591744,
+      "learning_rate": 1.7019801158116112e-06,
+      "loss": 0.8597,
+      "step": 34687
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.414052142934681,
+      "learning_rate": 1.7015543188496065e-06,
+      "loss": 0.8533,
+      "step": 34688
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5270690198886978,
+      "learning_rate": 1.7011285702037804e-06,
+      "loss": 0.7571,
+      "step": 34689
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3332268145499402,
+      "learning_rate": 1.7007028698766082e-06,
+      "loss": 0.7461,
+      "step": 34690
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.857952643458814,
+      "learning_rate": 1.7002772178705717e-06,
+      "loss": 0.8518,
+      "step": 34691
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6225966408876689,
+      "learning_rate": 1.6998516141881517e-06,
+      "loss": 0.8572,
+      "step": 34692
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5410478309831959,
+      "learning_rate": 1.6994260588318235e-06,
+      "loss": 0.7934,
+      "step": 34693
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1411121610901707,
+      "learning_rate": 1.699000551804062e-06,
+      "loss": 0.8114,
+      "step": 34694
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.441591621690123,
+      "learning_rate": 1.6985750931073476e-06,
+      "loss": 0.7991,
+      "step": 34695
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6678330247192037,
+      "learning_rate": 1.698149682744159e-06,
+      "loss": 0.8552,
+      "step": 34696
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7744888581920113,
+      "learning_rate": 1.6977243207169714e-06,
+      "loss": 0.8658,
+      "step": 34697
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4812625941163227,
+      "learning_rate": 1.697299007028259e-06,
+      "loss": 0.7359,
+      "step": 34698
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8423242718054693,
+      "learning_rate": 1.6968737416805003e-06,
+      "loss": 0.679,
+      "step": 34699
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2313953213548965,
+      "learning_rate": 1.696448524676173e-06,
+      "loss": 0.8837,
+      "step": 34700
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6185882942623033,
+      "learning_rate": 1.6960233560177531e-06,
+      "loss": 0.8567,
+      "step": 34701
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7272933324732394,
+      "learning_rate": 1.6955982357077106e-06,
+      "loss": 0.8761,
+      "step": 34702
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5837256961776804,
+      "learning_rate": 1.6951731637485269e-06,
+      "loss": 0.8296,
+      "step": 34703
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5058974527448892,
+      "learning_rate": 1.6947481401426712e-06,
+      "loss": 0.9107,
+      "step": 34704
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7431139491451618,
+      "learning_rate": 1.6943231648926251e-06,
+      "loss": 0.7172,
+      "step": 34705
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0121030577708043,
+      "learning_rate": 1.693898238000856e-06,
+      "loss": 0.7123,
+      "step": 34706
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5344301667268399,
+      "learning_rate": 1.6934733594698426e-06,
+      "loss": 0.8381,
+      "step": 34707
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4370077741444351,
+      "learning_rate": 1.693048529302056e-06,
+      "loss": 0.8834,
+      "step": 34708
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4092537876529532,
+      "learning_rate": 1.692623747499973e-06,
+      "loss": 0.7604,
+      "step": 34709
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8538661941401704,
+      "learning_rate": 1.6921990140660638e-06,
+      "loss": 0.8474,
+      "step": 34710
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7918606803574604,
+      "learning_rate": 1.6917743290027999e-06,
+      "loss": 0.9561,
+      "step": 34711
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4919057091300885,
+      "learning_rate": 1.691349692312657e-06,
+      "loss": 0.798,
+      "step": 34712
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5486886108133457,
+      "learning_rate": 1.6909251039981056e-06,
+      "loss": 0.7839,
+      "step": 34713
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7413202642301406,
+      "learning_rate": 1.6905005640616245e-06,
+      "loss": 0.7953,
+      "step": 34714
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6543457999962228,
+      "learning_rate": 1.6900760725056742e-06,
+      "loss": 0.8422,
+      "step": 34715
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6271883049251001,
+      "learning_rate": 1.6896516293327326e-06,
+      "loss": 0.7186,
+      "step": 34716
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6113787361165615,
+      "learning_rate": 1.6892272345452732e-06,
+      "loss": 0.7774,
+      "step": 34717
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9692278243831722,
+      "learning_rate": 1.688802888145763e-06,
+      "loss": 0.7847,
+      "step": 34718
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8698681590854003,
+      "learning_rate": 1.6883785901366712e-06,
+      "loss": 0.9958,
+      "step": 34719
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4378975474280808,
+      "learning_rate": 1.6879543405204712e-06,
+      "loss": 0.7128,
+      "step": 34720
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5516100706096028,
+      "learning_rate": 1.6875301392996346e-06,
+      "loss": 0.8439,
+      "step": 34721
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9291102642570286,
+      "learning_rate": 1.6871059864766293e-06,
+      "loss": 0.7703,
+      "step": 34722
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.053178633999048,
+      "learning_rate": 1.6866818820539232e-06,
+      "loss": 0.8641,
+      "step": 34723
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6782148044078968,
+      "learning_rate": 1.686257826033988e-06,
+      "loss": 0.8852,
+      "step": 34724
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.819233433259811,
+      "learning_rate": 1.68583381841929e-06,
+      "loss": 0.7487,
+      "step": 34725
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5448099187265898,
+      "learning_rate": 1.6854098592123002e-06,
+      "loss": 0.8317,
+      "step": 34726
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.504664577766812,
+      "learning_rate": 1.6849859484154885e-06,
+      "loss": 0.7503,
+      "step": 34727
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4509233999484448,
+      "learning_rate": 1.6845620860313216e-06,
+      "loss": 0.728,
+      "step": 34728
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5415749811396156,
+      "learning_rate": 1.6841382720622646e-06,
+      "loss": 0.747,
+      "step": 34729
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.827684459647484,
+      "learning_rate": 1.6837145065107862e-06,
+      "loss": 0.8884,
+      "step": 34730
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6660483788802707,
+      "learning_rate": 1.6832907893793604e-06,
+      "loss": 0.9307,
+      "step": 34731
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.880162786591324,
+      "learning_rate": 1.682867120670444e-06,
+      "loss": 0.8513,
+      "step": 34732
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7590347099381266,
+      "learning_rate": 1.6824435003865081e-06,
+      "loss": 0.6773,
+      "step": 34733
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5771858000168053,
+      "learning_rate": 1.6820199285300199e-06,
+      "loss": 0.6614,
+      "step": 34734
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.574626547675088,
+      "learning_rate": 1.6815964051034472e-06,
+      "loss": 0.8029,
+      "step": 34735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.760700104653416,
+      "learning_rate": 1.6811729301092528e-06,
+      "loss": 0.6844,
+      "step": 34736
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5188119830938607,
+      "learning_rate": 1.6807495035499022e-06,
+      "loss": 0.8625,
+      "step": 34737
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5381572238168337,
+      "learning_rate": 1.6803261254278635e-06,
+      "loss": 1.0057,
+      "step": 34738
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.837923680624787,
+      "learning_rate": 1.6799027957455983e-06,
+      "loss": 0.8411,
+      "step": 34739
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6337409184100806,
+      "learning_rate": 1.6794795145055753e-06,
+      "loss": 0.8104,
+      "step": 34740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8979053418637801,
+      "learning_rate": 1.6790562817102541e-06,
+      "loss": 0.7866,
+      "step": 34741
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5255186149002768,
+      "learning_rate": 1.6786330973621045e-06,
+      "loss": 0.8345,
+      "step": 34742
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5583680097982335,
+      "learning_rate": 1.6782099614635848e-06,
+      "loss": 0.7074,
+      "step": 34743
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5379011970979442,
+      "learning_rate": 1.6777868740171633e-06,
+      "loss": 0.8213,
+      "step": 34744
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.349572791230794,
+      "learning_rate": 1.677363835025302e-06,
+      "loss": 0.688,
+      "step": 34745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.636505222596208,
+      "learning_rate": 1.6769408444904612e-06,
+      "loss": 0.73,
+      "step": 34746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.671530231118078,
+      "learning_rate": 1.6765179024151047e-06,
+      "loss": 0.873,
+      "step": 34747
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7388149049407036,
+      "learning_rate": 1.6760950088016991e-06,
+      "loss": 0.981,
+      "step": 34748
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5566785142842134,
+      "learning_rate": 1.675672163652704e-06,
+      "loss": 1.0037,
+      "step": 34749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4294043711753757,
+      "learning_rate": 1.6752493669705782e-06,
+      "loss": 0.8251,
+      "step": 34750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.550752727413799,
+      "learning_rate": 1.6748266187577866e-06,
+      "loss": 0.8209,
+      "step": 34751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5339329643590542,
+      "learning_rate": 1.6744039190167926e-06,
+      "loss": 0.8768,
+      "step": 34752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5931249766484947,
+      "learning_rate": 1.6739812677500543e-06,
+      "loss": 0.8736,
+      "step": 34753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5238178951630357,
+      "learning_rate": 1.673558664960031e-06,
+      "loss": 0.652,
+      "step": 34754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4504839025600982,
+      "learning_rate": 1.673136110649185e-06,
+      "loss": 0.8806,
+      "step": 34755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3442154710755545,
+      "learning_rate": 1.6727136048199788e-06,
+      "loss": 0.697,
+      "step": 34756
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.055911679995969,
+      "learning_rate": 1.6722911474748715e-06,
+      "loss": 0.9108,
+      "step": 34757
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5565157441840902,
+      "learning_rate": 1.671868738616319e-06,
+      "loss": 0.7715,
+      "step": 34758
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.562547920207476,
+      "learning_rate": 1.6714463782467848e-06,
+      "loss": 0.7189,
+      "step": 34759
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4238585149704985,
+      "learning_rate": 1.671024066368725e-06,
+      "loss": 0.8352,
+      "step": 34760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6828435601918377,
+      "learning_rate": 1.6706018029846016e-06,
+      "loss": 0.7578,
+      "step": 34761
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8456189269143823,
+      "learning_rate": 1.6701795880968696e-06,
+      "loss": 0.7137,
+      "step": 34762
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.724749438961912,
+      "learning_rate": 1.6697574217079926e-06,
+      "loss": 0.7988,
+      "step": 34763
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6815113004988367,
+      "learning_rate": 1.6693353038204218e-06,
+      "loss": 0.6779,
+      "step": 34764
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.60974483825143,
+      "learning_rate": 1.668913234436621e-06,
+      "loss": 0.8429,
+      "step": 34765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.629037307551531,
+      "learning_rate": 1.6684912135590447e-06,
+      "loss": 0.7736,
+      "step": 34766
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.713225319684264,
+      "learning_rate": 1.6680692411901478e-06,
+      "loss": 0.8989,
+      "step": 34767
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6972861274335829,
+      "learning_rate": 1.6676473173323892e-06,
+      "loss": 0.8062,
+      "step": 34768
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.609611124593736,
+      "learning_rate": 1.6672254419882295e-06,
+      "loss": 0.906,
+      "step": 34769
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.587204307606214,
+      "learning_rate": 1.6668036151601197e-06,
+      "loss": 0.8619,
+      "step": 34770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4930754196926135,
+      "learning_rate": 1.6663818368505168e-06,
+      "loss": 0.8254,
+      "step": 34771
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.624605062155383,
+      "learning_rate": 1.6659601070618758e-06,
+      "loss": 0.8414,
+      "step": 34772
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9404069805472528,
+      "learning_rate": 1.6655384257966566e-06,
+      "loss": 0.8065,
+      "step": 34773
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5917136473523652,
+      "learning_rate": 1.665116793057312e-06,
+      "loss": 0.8254,
+      "step": 34774
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.535241480607739,
+      "learning_rate": 1.6646952088462932e-06,
+      "loss": 0.8103,
+      "step": 34775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9258373980263235,
+      "learning_rate": 1.664273673166058e-06,
+      "loss": 0.8363,
+      "step": 34776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.513318235748683,
+      "learning_rate": 1.6638521860190625e-06,
+      "loss": 0.8707,
+      "step": 34777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7896342069102642,
+      "learning_rate": 1.6634307474077594e-06,
+      "loss": 0.745,
+      "step": 34778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5619995185874052,
+      "learning_rate": 1.6630093573345985e-06,
+      "loss": 0.7524,
+      "step": 34779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4115572404538477,
+      "learning_rate": 1.6625880158020403e-06,
+      "loss": 0.8604,
+      "step": 34780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5508097069218378,
+      "learning_rate": 1.6621667228125305e-06,
+      "loss": 0.8548,
+      "step": 34781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0179388046938813,
+      "learning_rate": 1.6617454783685283e-06,
+      "loss": 0.8456,
+      "step": 34782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5058607492304565,
+      "learning_rate": 1.6613242824724818e-06,
+      "loss": 0.6861,
+      "step": 34783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4878071698849675,
+      "learning_rate": 1.660903135126848e-06,
+      "loss": 0.7926,
+      "step": 34784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5577846273049978,
+      "learning_rate": 1.6604820363340735e-06,
+      "loss": 0.9345,
+      "step": 34785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6862983271577257,
+      "learning_rate": 1.6600609860966155e-06,
+      "loss": 0.8478,
+      "step": 34786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8995100623334111,
+      "learning_rate": 1.659639984416922e-06,
+      "loss": 0.8083,
+      "step": 34787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7958677080284493,
+      "learning_rate": 1.6592190312974433e-06,
+      "loss": 0.7318,
+      "step": 34788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5580266409861605,
+      "learning_rate": 1.6587981267406328e-06,
+      "loss": 0.8695,
+      "step": 34789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5176381526899954,
+      "learning_rate": 1.658377270748941e-06,
+      "loss": 0.7732,
+      "step": 34790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5703624241754726,
+      "learning_rate": 1.6579564633248191e-06,
+      "loss": 0.8604,
+      "step": 34791
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3947363899249587,
+      "learning_rate": 1.657535704470713e-06,
+      "loss": 0.7526,
+      "step": 34792
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5381131019161407,
+      "learning_rate": 1.6571149941890753e-06,
+      "loss": 0.7958,
+      "step": 34793
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.562605219870766,
+      "learning_rate": 1.656694332482357e-06,
+      "loss": 0.9503,
+      "step": 34794
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7647910114593381,
+      "learning_rate": 1.6562737193530065e-06,
+      "loss": 0.7437,
+      "step": 34795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4452538156999761,
+      "learning_rate": 1.6558531548034696e-06,
+      "loss": 0.7628,
+      "step": 34796
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7031357435837153,
+      "learning_rate": 1.6554326388361963e-06,
+      "loss": 0.7289,
+      "step": 34797
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4823251892651872,
+      "learning_rate": 1.6550121714536393e-06,
+      "loss": 0.7871,
+      "step": 34798
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6865878556548604,
+      "learning_rate": 1.6545917526582434e-06,
+      "loss": 0.8354,
+      "step": 34799
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5646909018187882,
+      "learning_rate": 1.6541713824524542e-06,
+      "loss": 0.7457,
+      "step": 34800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7134706349810956,
+      "learning_rate": 1.653751060838723e-06,
+      "loss": 0.6844,
+      "step": 34801
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0349292503829526,
+      "learning_rate": 1.653330787819495e-06,
+      "loss": 0.7826,
+      "step": 34802
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.584647571092365,
+      "learning_rate": 1.6529105633972176e-06,
+      "loss": 0.6565,
+      "step": 34803
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.855195477731076,
+      "learning_rate": 1.652490387574337e-06,
+      "loss": 0.8808,
+      "step": 34804
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4614351666114402,
+      "learning_rate": 1.6520702603533012e-06,
+      "loss": 0.9221,
+      "step": 34805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9262283792650765,
+      "learning_rate": 1.6516501817365538e-06,
+      "loss": 0.8147,
+      "step": 34806
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7194422907494777,
+      "learning_rate": 1.651230151726545e-06,
+      "loss": 0.8298,
+      "step": 34807
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7513262899897637,
+      "learning_rate": 1.6508101703257161e-06,
+      "loss": 0.7967,
+      "step": 34808
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7562057834120515,
+      "learning_rate": 1.6503902375365121e-06,
+      "loss": 0.7529,
+      "step": 34809
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4767404705362788,
+      "learning_rate": 1.6499703533613788e-06,
+      "loss": 0.7173,
+      "step": 34810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6509102935706925,
+      "learning_rate": 1.6495505178027626e-06,
+      "loss": 0.8348,
+      "step": 34811
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5573116588984586,
+      "learning_rate": 1.649130730863111e-06,
+      "loss": 0.7174,
+      "step": 34812
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5007434750196913,
+      "learning_rate": 1.6487109925448597e-06,
+      "loss": 0.7267,
+      "step": 34813
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6668347218407076,
+      "learning_rate": 1.6482913028504577e-06,
+      "loss": 1.0444,
+      "step": 34814
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4650777384716622,
+      "learning_rate": 1.6478716617823488e-06,
+      "loss": 0.842,
+      "step": 34815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4340240293390765,
+      "learning_rate": 1.6474520693429763e-06,
+      "loss": 0.7661,
+      "step": 34816
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.52548507214186,
+      "learning_rate": 1.6470325255347798e-06,
+      "loss": 0.847,
+      "step": 34817
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5789887620442196,
+      "learning_rate": 1.6466130303602047e-06,
+      "loss": 0.8608,
+      "step": 34818
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5719079896834114,
+      "learning_rate": 1.6461935838216947e-06,
+      "loss": 0.7193,
+      "step": 34819
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7504812614048775,
+      "learning_rate": 1.6457741859216903e-06,
+      "loss": 1.0147,
+      "step": 34820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5916984743620854,
+      "learning_rate": 1.6453548366626316e-06,
+      "loss": 0.8042,
+      "step": 34821
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3958660250602652,
+      "learning_rate": 1.6449355360469621e-06,
+      "loss": 0.7741,
+      "step": 34822
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7380185011736576,
+      "learning_rate": 1.6445162840771256e-06,
+      "loss": 0.7212,
+      "step": 34823
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8760613177231813,
+      "learning_rate": 1.644097080755558e-06,
+      "loss": 0.8434,
+      "step": 34824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5653768904848446,
+      "learning_rate": 1.6436779260847058e-06,
+      "loss": 0.8176,
+      "step": 34825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7331740204174717,
+      "learning_rate": 1.6432588200670053e-06,
+      "loss": 0.9105,
+      "step": 34826
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.589164464269215,
+      "learning_rate": 1.6428397627048953e-06,
+      "loss": 0.8818,
+      "step": 34827
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.627719194104652,
+      "learning_rate": 1.6424207540008185e-06,
+      "loss": 0.8307,
+      "step": 34828
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6688541656679203,
+      "learning_rate": 1.6420017939572164e-06,
+      "loss": 0.7829,
+      "step": 34829
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6186133431429703,
+      "learning_rate": 1.6415828825765256e-06,
+      "loss": 0.8271,
+      "step": 34830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.702525766720937,
+      "learning_rate": 1.6411640198611834e-06,
+      "loss": 0.745,
+      "step": 34831
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8222946974925707,
+      "learning_rate": 1.6407452058136298e-06,
+      "loss": 0.7809,
+      "step": 34832
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5248887897118721,
+      "learning_rate": 1.6403264404363073e-06,
+      "loss": 0.8487,
+      "step": 34833
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8396951246714877,
+      "learning_rate": 1.6399077237316507e-06,
+      "loss": 0.7748,
+      "step": 34834
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.69833404854066,
+      "learning_rate": 1.639489055702096e-06,
+      "loss": 0.9443,
+      "step": 34835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5519098107226166,
+      "learning_rate": 1.6390704363500832e-06,
+      "loss": 0.8811,
+      "step": 34836
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6841600950213351,
+      "learning_rate": 1.6386518656780504e-06,
+      "loss": 0.6763,
+      "step": 34837
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4484915467217607,
+      "learning_rate": 1.6382333436884345e-06,
+      "loss": 0.8901,
+      "step": 34838
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7639162416688978,
+      "learning_rate": 1.6378148703836695e-06,
+      "loss": 0.6028,
+      "step": 34839
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4735796567984847,
+      "learning_rate": 1.637396445766195e-06,
+      "loss": 0.8289,
+      "step": 34840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4795126210704275,
+      "learning_rate": 1.6369780698384442e-06,
+      "loss": 0.8208,
+      "step": 34841
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.498152974718157,
+      "learning_rate": 1.6365597426028569e-06,
+      "loss": 0.7675,
+      "step": 34842
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.708398299847132,
+      "learning_rate": 1.6361414640618645e-06,
+      "loss": 0.8828,
+      "step": 34843
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5495773336758285,
+      "learning_rate": 1.6357232342179063e-06,
+      "loss": 0.7751,
+      "step": 34844
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5175531369469348,
+      "learning_rate": 1.6353050530734139e-06,
+      "loss": 0.8579,
+      "step": 34845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6949640259898175,
+      "learning_rate": 1.634886920630826e-06,
+      "loss": 0.8753,
+      "step": 34846
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6093092945382819,
+      "learning_rate": 1.6344688368925743e-06,
+      "loss": 0.8126,
+      "step": 34847
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7829188904595603,
+      "learning_rate": 1.6340508018610913e-06,
+      "loss": 0.6692,
+      "step": 34848
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6290112693554581,
+      "learning_rate": 1.6336328155388138e-06,
+      "loss": 0.8823,
+      "step": 34849
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7534959103266216,
+      "learning_rate": 1.6332148779281765e-06,
+      "loss": 0.639,
+      "step": 34850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6674694481551113,
+      "learning_rate": 1.6327969890316109e-06,
+      "loss": 0.7971,
+      "step": 34851
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6606533827527552,
+      "learning_rate": 1.6323791488515472e-06,
+      "loss": 0.8987,
+      "step": 34852
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7235932249340021,
+      "learning_rate": 1.6319613573904226e-06,
+      "loss": 0.6827,
+      "step": 34853
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7846464422762074,
+      "learning_rate": 1.6315436146506702e-06,
+      "loss": 0.7617,
+      "step": 34854
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.468286739365478,
+      "learning_rate": 1.6311259206347207e-06,
+      "loss": 0.7199,
+      "step": 34855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8830952651824049,
+      "learning_rate": 1.6307082753450022e-06,
+      "loss": 0.7972,
+      "step": 34856
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6613848307294565,
+      "learning_rate": 1.6302906787839501e-06,
+      "loss": 0.842,
+      "step": 34857
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6521682471501655,
+      "learning_rate": 1.6298731309539984e-06,
+      "loss": 0.6864,
+      "step": 34858
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.658271011545625,
+      "learning_rate": 1.629455631857575e-06,
+      "loss": 0.9622,
+      "step": 34859
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4250868238486563,
+      "learning_rate": 1.629038181497109e-06,
+      "loss": 0.6991,
+      "step": 34860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.610935093384068,
+      "learning_rate": 1.6286207798750342e-06,
+      "loss": 0.8398,
+      "step": 34861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6244102151213309,
+      "learning_rate": 1.6282034269937785e-06,
+      "loss": 0.9136,
+      "step": 34862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7945443034080053,
+      "learning_rate": 1.6277861228557745e-06,
+      "loss": 1.0057,
+      "step": 34863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8250826504082464,
+      "learning_rate": 1.627368867463448e-06,
+      "loss": 0.7212,
+      "step": 34864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7567034815044643,
+      "learning_rate": 1.6269516608192338e-06,
+      "loss": 0.7022,
+      "step": 34865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8894152403608335,
+      "learning_rate": 1.6265345029255542e-06,
+      "loss": 0.7933,
+      "step": 34866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6430717859633572,
+      "learning_rate": 1.6261173937848451e-06,
+      "loss": 0.8427,
+      "step": 34867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.593819772082617,
+      "learning_rate": 1.6257003333995315e-06,
+      "loss": 0.7326,
+      "step": 34868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7044544424691186,
+      "learning_rate": 1.6252833217720386e-06,
+      "loss": 0.9042,
+      "step": 34869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5338954216758833,
+      "learning_rate": 1.6248663589047985e-06,
+      "loss": 0.9373,
+      "step": 34870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4949078504425592,
+      "learning_rate": 1.6244494448002402e-06,
+      "loss": 0.7201,
+      "step": 34871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5119907419741532,
+      "learning_rate": 1.624032579460788e-06,
+      "loss": 0.6791,
+      "step": 34872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8169421920030726,
+      "learning_rate": 1.6236157628888694e-06,
+      "loss": 0.6923,
+      "step": 34873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5259351291979617,
+      "learning_rate": 1.6231989950869099e-06,
+      "loss": 0.8501,
+      "step": 34874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8353897764330889,
+      "learning_rate": 1.6227822760573408e-06,
+      "loss": 0.7857,
+      "step": 34875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4295832200287457,
+      "learning_rate": 1.6223656058025861e-06,
+      "loss": 0.7348,
+      "step": 34876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5514456508201109,
+      "learning_rate": 1.621948984325068e-06,
+      "loss": 0.9092,
+      "step": 34877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5633619072167053,
+      "learning_rate": 1.6215324116272158e-06,
+      "loss": 0.928,
+      "step": 34878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5283505795531687,
+      "learning_rate": 1.6211158877114574e-06,
+      "loss": 0.7787,
+      "step": 34879
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8558992222807997,
+      "learning_rate": 1.6206994125802144e-06,
+      "loss": 0.7894,
+      "step": 34880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4298410956671268,
+      "learning_rate": 1.6202829862359093e-06,
+      "loss": 0.7721,
+      "step": 34881
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4654410087591605,
+      "learning_rate": 1.6198666086809733e-06,
+      "loss": 0.8022,
+      "step": 34882
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6039696445717875,
+      "learning_rate": 1.6194502799178236e-06,
+      "loss": 0.8038,
+      "step": 34883
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4148670080661816,
+      "learning_rate": 1.6190339999488902e-06,
+      "loss": 0.7704,
+      "step": 34884
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4891137983308096,
+      "learning_rate": 1.6186177687765926e-06,
+      "loss": 0.9074,
+      "step": 34885
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7204537524545302,
+      "learning_rate": 1.6182015864033574e-06,
+      "loss": 0.7826,
+      "step": 34886
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8531955039892265,
+      "learning_rate": 1.617785452831604e-06,
+      "loss": 0.8138,
+      "step": 34887
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8549015449829191,
+      "learning_rate": 1.6173693680637592e-06,
+      "loss": 0.8113,
+      "step": 34888
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4316920706701064,
+      "learning_rate": 1.6169533321022446e-06,
+      "loss": 0.8188,
+      "step": 34889
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7315562833793992,
+      "learning_rate": 1.6165373449494793e-06,
+      "loss": 0.8764,
+      "step": 34890
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.507934067072229,
+      "learning_rate": 1.6161214066078878e-06,
+      "loss": 0.7916,
+      "step": 34891
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8268100618669108,
+      "learning_rate": 1.615705517079894e-06,
+      "loss": 0.8161,
+      "step": 34892
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9107641532810127,
+      "learning_rate": 1.615289676367917e-06,
+      "loss": 0.8053,
+      "step": 34893
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7387712246504425,
+      "learning_rate": 1.614873884474376e-06,
+      "loss": 0.7946,
+      "step": 34894
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.99119320862021,
+      "learning_rate": 1.6144581414016936e-06,
+      "loss": 0.8034,
+      "step": 34895
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.729528908468285,
+      "learning_rate": 1.614042447152293e-06,
+      "loss": 0.9153,
+      "step": 34896
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7263133807320248,
+      "learning_rate": 1.6136268017285928e-06,
+      "loss": 0.7863,
+      "step": 34897
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7502145520669926,
+      "learning_rate": 1.6132112051330095e-06,
+      "loss": 0.7075,
+      "step": 34898
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8133000802550205,
+      "learning_rate": 1.612795657367966e-06,
+      "loss": 0.762,
+      "step": 34899
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7711395598521953,
+      "learning_rate": 1.6123801584358845e-06,
+      "loss": 0.7027,
+      "step": 34900
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5986060795022707,
+      "learning_rate": 1.6119647083391798e-06,
+      "loss": 0.8565,
+      "step": 34901
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5504758556503817,
+      "learning_rate": 1.6115493070802712e-06,
+      "loss": 0.9094,
+      "step": 34902
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.590534731130395,
+      "learning_rate": 1.61113395466158e-06,
+      "loss": 0.9065,
+      "step": 34903
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9953794000868597,
+      "learning_rate": 1.61071865108552e-06,
+      "loss": 0.9158,
+      "step": 34904
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5988322148672967,
+      "learning_rate": 1.6103033963545156e-06,
+      "loss": 0.9372,
+      "step": 34905
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5627587764998023,
+      "learning_rate": 1.6098881904709774e-06,
+      "loss": 0.7056,
+      "step": 34906
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6008292906380934,
+      "learning_rate": 1.6094730334373287e-06,
+      "loss": 0.8225,
+      "step": 34907
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7767490088151525,
+      "learning_rate": 1.6090579252559834e-06,
+      "loss": 0.8365,
+      "step": 34908
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5662162941422038,
+      "learning_rate": 1.6086428659293585e-06,
+      "loss": 0.8739,
+      "step": 34909
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6727891675213045,
+      "learning_rate": 1.6082278554598763e-06,
+      "loss": 0.7895,
+      "step": 34910
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4651246323306604,
+      "learning_rate": 1.6078128938499427e-06,
+      "loss": 0.8977,
+      "step": 34911
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0795111275042424,
+      "learning_rate": 1.6073979811019803e-06,
+      "loss": 0.7607,
+      "step": 34912
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6128502162358354,
+      "learning_rate": 1.6069831172184035e-06,
+      "loss": 0.8674,
+      "step": 34913
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.666365377761105,
+      "learning_rate": 1.6065683022016343e-06,
+      "loss": 0.7492,
+      "step": 34914
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.426859822100825,
+      "learning_rate": 1.6061535360540758e-06,
+      "loss": 0.7571,
+      "step": 34915
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4958924292370142,
+      "learning_rate": 1.6057388187781486e-06,
+      "loss": 0.7859,
+      "step": 34916
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.793237471525641,
+      "learning_rate": 1.6053241503762707e-06,
+      "loss": 0.7849,
+      "step": 34917
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.705004768794769,
+      "learning_rate": 1.6049095308508521e-06,
+      "loss": 0.6818,
+      "step": 34918
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8568016924186237,
+      "learning_rate": 1.6044949602043069e-06,
+      "loss": 0.883,
+      "step": 34919
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.810536409741184,
+      "learning_rate": 1.6040804384390496e-06,
+      "loss": 0.8315,
+      "step": 34920
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5451296006097857,
+      "learning_rate": 1.603665965557497e-06,
+      "loss": 0.7619,
+      "step": 34921
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5545941367791898,
+      "learning_rate": 1.6032515415620564e-06,
+      "loss": 0.9271,
+      "step": 34922
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5659013687405647,
+      "learning_rate": 1.6028371664551456e-06,
+      "loss": 0.8019,
+      "step": 34923
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5472754001037792,
+      "learning_rate": 1.6024228402391762e-06,
+      "loss": 0.8903,
+      "step": 34924
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7248693887792814,
+      "learning_rate": 1.6020085629165572e-06,
+      "loss": 0.8547,
+      "step": 34925
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6411275305110933,
+      "learning_rate": 1.6015943344897022e-06,
+      "loss": 0.7796,
+      "step": 34926
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7037855025491604,
+      "learning_rate": 1.6011801549610273e-06,
+      "loss": 0.8966,
+      "step": 34927
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0792201861613484,
+      "learning_rate": 1.6007660243329405e-06,
+      "loss": 0.7945,
+      "step": 34928
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6970027017507492,
+      "learning_rate": 1.6003519426078507e-06,
+      "loss": 0.8288,
+      "step": 34929
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.435892190058005,
+      "learning_rate": 1.5999379097881706e-06,
+      "loss": 0.7789,
+      "step": 34930
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7369583498673283,
+      "learning_rate": 1.599523925876314e-06,
+      "loss": 0.6658,
+      "step": 34931
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5528085367018218,
+      "learning_rate": 1.5991099908746887e-06,
+      "loss": 0.777,
+      "step": 34932
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9088207589501087,
+      "learning_rate": 1.598696104785702e-06,
+      "loss": 0.7879,
+      "step": 34933
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6081632062199869,
+      "learning_rate": 1.5982822676117672e-06,
+      "loss": 0.7622,
+      "step": 34934
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4338769232909243,
+      "learning_rate": 1.5978684793552946e-06,
+      "loss": 0.8887,
+      "step": 34935
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5562712006913868,
+      "learning_rate": 1.5974547400186914e-06,
+      "loss": 0.9293,
+      "step": 34936
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1190363599028186,
+      "learning_rate": 1.5970410496043653e-06,
+      "loss": 0.9053,
+      "step": 34937
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7658989588143528,
+      "learning_rate": 1.5966274081147293e-06,
+      "loss": 0.8517,
+      "step": 34938
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8493494472522265,
+      "learning_rate": 1.5962138155521855e-06,
+      "loss": 0.8501,
+      "step": 34939
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.571123607057389,
+      "learning_rate": 1.595800271919149e-06,
+      "loss": 0.7489,
+      "step": 34940
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.847499366797101,
+      "learning_rate": 1.595386777218021e-06,
+      "loss": 0.8209,
+      "step": 34941
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8270216350490418,
+      "learning_rate": 1.594973331451214e-06,
+      "loss": 0.7405,
+      "step": 34942
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.605511076933658,
+      "learning_rate": 1.5945599346211315e-06,
+      "loss": 0.787,
+      "step": 34943
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4752211276384297,
+      "learning_rate": 1.5941465867301842e-06,
+      "loss": 0.7121,
+      "step": 34944
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7321071016912823,
+      "learning_rate": 1.5937332877807753e-06,
+      "loss": 0.8082,
+      "step": 34945
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7482653111131462,
+      "learning_rate": 1.5933200377753145e-06,
+      "loss": 0.7014,
+      "step": 34946
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4788538732573333,
+      "learning_rate": 1.5929068367162026e-06,
+      "loss": 0.8581,
+      "step": 34947
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2462257363872236,
+      "learning_rate": 1.5924936846058515e-06,
+      "loss": 0.9161,
+      "step": 34948
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.509193543965051,
+      "learning_rate": 1.5920805814466643e-06,
+      "loss": 0.7926,
+      "step": 34949
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.686556004963282,
+      "learning_rate": 1.5916675272410442e-06,
+      "loss": 0.7748,
+      "step": 34950
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6610430168494814,
+      "learning_rate": 1.5912545219913966e-06,
+      "loss": 0.8277,
+      "step": 34951
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.467068665730009,
+      "learning_rate": 1.5908415657001297e-06,
+      "loss": 0.8035,
+      "step": 34952
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.451435318859382,
+      "learning_rate": 1.590428658369646e-06,
+      "loss": 0.766,
+      "step": 34953
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.697148932944595,
+      "learning_rate": 1.5900158000023469e-06,
+      "loss": 0.7563,
+      "step": 34954
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5352370850598058,
+      "learning_rate": 1.5896029906006383e-06,
+      "loss": 0.9674,
+      "step": 34955
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.912641684858911,
+      "learning_rate": 1.589190230166926e-06,
+      "loss": 0.704,
+      "step": 34956
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.568933622584719,
+      "learning_rate": 1.5887775187036113e-06,
+      "loss": 0.7753,
+      "step": 34957
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5006087970407518,
+      "learning_rate": 1.5883648562130937e-06,
+      "loss": 0.9051,
+      "step": 34958
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6368758559824725,
+      "learning_rate": 1.587952242697779e-06,
+      "loss": 0.7203,
+      "step": 34959
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.698042876106373,
+      "learning_rate": 1.5875396781600727e-06,
+      "loss": 0.7115,
+      "step": 34960
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6465856607337008,
+      "learning_rate": 1.5871271626023733e-06,
+      "loss": 0.8245,
+      "step": 34961
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6451901052383715,
+      "learning_rate": 1.5867146960270806e-06,
+      "loss": 0.763,
+      "step": 34962
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5529329243084573,
+      "learning_rate": 1.5863022784366012e-06,
+      "loss": 0.8993,
+      "step": 34963
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.267125607307456,
+      "learning_rate": 1.5858899098333314e-06,
+      "loss": 0.8596,
+      "step": 34964
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8132884275377507,
+      "learning_rate": 1.5854775902196762e-06,
+      "loss": 0.6845,
+      "step": 34965
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5511805181740113,
+      "learning_rate": 1.5850653195980325e-06,
+      "loss": 0.7257,
+      "step": 34966
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4245592396582236,
+      "learning_rate": 1.5846530979708053e-06,
+      "loss": 0.7584,
+      "step": 34967
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4678672803080002,
+      "learning_rate": 1.584240925340389e-06,
+      "loss": 0.7437,
+      "step": 34968
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5822088742140068,
+      "learning_rate": 1.5838288017091885e-06,
+      "loss": 0.9338,
+      "step": 34969
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6084220675824097,
+      "learning_rate": 1.583416727079602e-06,
+      "loss": 0.7703,
+      "step": 34970
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9682827998728445,
+      "learning_rate": 1.583004701454025e-06,
+      "loss": 0.7982,
+      "step": 34971
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.573454770714319,
+      "learning_rate": 1.5825927248348604e-06,
+      "loss": 0.7805,
+      "step": 34972
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7953941297753369,
+      "learning_rate": 1.5821807972245073e-06,
+      "loss": 0.731,
+      "step": 34973
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6554979095126003,
+      "learning_rate": 1.5817689186253627e-06,
+      "loss": 0.8445,
+      "step": 34974
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7430028082927034,
+      "learning_rate": 1.5813570890398234e-06,
+      "loss": 0.7112,
+      "step": 34975
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7270665671234062,
+      "learning_rate": 1.5809453084702875e-06,
+      "loss": 0.9835,
+      "step": 34976
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7586873184661047,
+      "learning_rate": 1.5805335769191555e-06,
+      "loss": 0.8481,
+      "step": 34977
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8360395981580093,
+      "learning_rate": 1.580121894388823e-06,
+      "loss": 0.9493,
+      "step": 34978
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5775025202962325,
+      "learning_rate": 1.5797102608816838e-06,
+      "loss": 0.7758,
+      "step": 34979
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6220409932090571,
+      "learning_rate": 1.5792986764001372e-06,
+      "loss": 0.8358,
+      "step": 34980
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7920827021965878,
+      "learning_rate": 1.5788871409465822e-06,
+      "loss": 0.7373,
+      "step": 34981
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1689316699710224,
+      "learning_rate": 1.5784756545234126e-06,
+      "loss": 0.649,
+      "step": 34982
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7252247005003079,
+      "learning_rate": 1.5780642171330217e-06,
+      "loss": 0.8422,
+      "step": 34983
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5737262730804389,
+      "learning_rate": 1.5776528287778093e-06,
+      "loss": 0.9794,
+      "step": 34984
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5103660750968275,
+      "learning_rate": 1.5772414894601663e-06,
+      "loss": 0.9325,
+      "step": 34985
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6364826654524447,
+      "learning_rate": 1.576830199182493e-06,
+      "loss": 0.7936,
+      "step": 34986
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7979102664834087,
+      "learning_rate": 1.5764189579471778e-06,
+      "loss": 0.7099,
+      "step": 34987
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6450880274852722,
+      "learning_rate": 1.5760077657566208e-06,
+      "loss": 0.7263,
+      "step": 34988
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3917719346875768,
+      "learning_rate": 1.5755966226132114e-06,
+      "loss": 0.8345,
+      "step": 34989
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7345759337411142,
+      "learning_rate": 1.5751855285193474e-06,
+      "loss": 0.8298,
+      "step": 34990
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9948623577807745,
+      "learning_rate": 1.5747744834774215e-06,
+      "loss": 0.9349,
+      "step": 34991
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5838612915138723,
+      "learning_rate": 1.5743634874898228e-06,
+      "loss": 0.8201,
+      "step": 34992
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6221997378179196,
+      "learning_rate": 1.5739525405589474e-06,
+      "loss": 0.7215,
+      "step": 34993
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4829764608233118,
+      "learning_rate": 1.5735416426871886e-06,
+      "loss": 0.7483,
+      "step": 34994
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5406775380172943,
+      "learning_rate": 1.5731307938769426e-06,
+      "loss": 0.753,
+      "step": 34995
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.697787576969082,
+      "learning_rate": 1.5727199941305914e-06,
+      "loss": 0.824,
+      "step": 34996
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6139939729200763,
+      "learning_rate": 1.5723092434505327e-06,
+      "loss": 0.7302,
+      "step": 34997
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8324967094199087,
+      "learning_rate": 1.5718985418391609e-06,
+      "loss": 0.7259,
+      "step": 34998
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.630410614345829,
+      "learning_rate": 1.5714878892988638e-06,
+      "loss": 0.792,
+      "step": 34999
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.975408644892472,
+      "learning_rate": 1.57107728583203e-06,
+      "loss": 0.7645,
+      "step": 35000
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5371102761697066,
+      "learning_rate": 1.5706667314410528e-06,
+      "loss": 0.729,
+      "step": 35001
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.501275475120577,
+      "learning_rate": 1.570256226128325e-06,
+      "loss": 0.8309,
+      "step": 35002
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7496960275687417,
+      "learning_rate": 1.5698457698962344e-06,
+      "loss": 0.7537,
+      "step": 35003
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6631887929838867,
+      "learning_rate": 1.569435362747168e-06,
+      "loss": 0.936,
+      "step": 35004
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6565979528155332,
+      "learning_rate": 1.5690250046835209e-06,
+      "loss": 0.9811,
+      "step": 35005
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5174365925026054,
+      "learning_rate": 1.5686146957076776e-06,
+      "loss": 0.7486,
+      "step": 35006
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7337412613380754,
+      "learning_rate": 1.5682044358220272e-06,
+      "loss": 0.7441,
+      "step": 35007
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7050737311488493,
+      "learning_rate": 1.5677942250289634e-06,
+      "loss": 0.8603,
+      "step": 35008
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8122131865720585,
+      "learning_rate": 1.5673840633308713e-06,
+      "loss": 0.8204,
+      "step": 35009
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7557275161377275,
+      "learning_rate": 1.5669739507301362e-06,
+      "loss": 0.6185,
+      "step": 35010
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.644265465038154,
+      "learning_rate": 1.5665638872291488e-06,
+      "loss": 0.9235,
+      "step": 35011
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7109267536434497,
+      "learning_rate": 1.5661538728303016e-06,
+      "loss": 0.6504,
+      "step": 35012
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6539893121278773,
+      "learning_rate": 1.5657439075359716e-06,
+      "loss": 0.7785,
+      "step": 35013
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5844452091846275,
+      "learning_rate": 1.5653339913485498e-06,
+      "loss": 0.8287,
+      "step": 35014
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5315384667596783,
+      "learning_rate": 1.564924124270425e-06,
+      "loss": 0.7919,
+      "step": 35015
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.853033708918298,
+      "learning_rate": 1.564514306303987e-06,
+      "loss": 0.809,
+      "step": 35016
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6992899583825978,
+      "learning_rate": 1.5641045374516117e-06,
+      "loss": 0.8726,
+      "step": 35017
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6105503795382332,
+      "learning_rate": 1.563694817715692e-06,
+      "loss": 0.9304,
+      "step": 35018
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8060989310562097,
+      "learning_rate": 1.5632851470986132e-06,
+      "loss": 0.6741,
+      "step": 35019
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6950742723205796,
+      "learning_rate": 1.5628755256027573e-06,
+      "loss": 0.6758,
+      "step": 35020
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.953360580273725,
+      "learning_rate": 1.5624659532305142e-06,
+      "loss": 0.9033,
+      "step": 35021
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4890803155302068,
+      "learning_rate": 1.5620564299842623e-06,
+      "loss": 0.8308,
+      "step": 35022
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.478456636129854,
+      "learning_rate": 1.5616469558663927e-06,
+      "loss": 0.8496,
+      "step": 35023
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3486149007536934,
+      "learning_rate": 1.5612375308792826e-06,
+      "loss": 0.801,
+      "step": 35024
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4712398606256802,
+      "learning_rate": 1.5608281550253224e-06,
+      "loss": 0.8635,
+      "step": 35025
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6543027051965917,
+      "learning_rate": 1.5604188283068921e-06,
+      "loss": 0.7655,
+      "step": 35026
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5642673974177246,
+      "learning_rate": 1.5600095507263725e-06,
+      "loss": 0.704,
+      "step": 35027
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7571058542407676,
+      "learning_rate": 1.55960032228615e-06,
+      "loss": 0.8253,
+      "step": 35028
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.597898164187338,
+      "learning_rate": 1.5591911429886087e-06,
+      "loss": 0.8663,
+      "step": 35029
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5134444652938086,
+      "learning_rate": 1.5587820128361297e-06,
+      "loss": 0.8279,
+      "step": 35030
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8886846092397895,
+      "learning_rate": 1.5583729318310902e-06,
+      "loss": 0.8486,
+      "step": 35031
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.772807025486586,
+      "learning_rate": 1.5579638999758772e-06,
+      "loss": 0.9134,
+      "step": 35032
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.545355374128367,
+      "learning_rate": 1.5575549172728733e-06,
+      "loss": 0.8988,
+      "step": 35033
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8612362995906517,
+      "learning_rate": 1.5571459837244585e-06,
+      "loss": 0.8394,
+      "step": 35034
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.658235725945467,
+      "learning_rate": 1.556737099333009e-06,
+      "loss": 0.8626,
+      "step": 35035
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5540611349264666,
+      "learning_rate": 1.5563282641009092e-06,
+      "loss": 0.6617,
+      "step": 35036
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5524374202557838,
+      "learning_rate": 1.555919478030542e-06,
+      "loss": 0.8228,
+      "step": 35037
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4795416094296385,
+      "learning_rate": 1.5555107411242854e-06,
+      "loss": 0.8948,
+      "step": 35038
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8087427655715441,
+      "learning_rate": 1.5551020533845162e-06,
+      "loss": 0.6808,
+      "step": 35039
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.11580560406824,
+      "learning_rate": 1.5546934148136196e-06,
+      "loss": 0.7915,
+      "step": 35040
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9461758082863156,
+      "learning_rate": 1.5542848254139674e-06,
+      "loss": 0.9067,
+      "step": 35041
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5847318030698687,
+      "learning_rate": 1.5538762851879464e-06,
+      "loss": 0.7715,
+      "step": 35042
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3565330292488142,
+      "learning_rate": 1.5534677941379295e-06,
+      "loss": 0.9034,
+      "step": 35043
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6860048285712463,
+      "learning_rate": 1.5530593522662995e-06,
+      "loss": 0.828,
+      "step": 35044
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4386500034417917,
+      "learning_rate": 1.5526509595754291e-06,
+      "loss": 0.845,
+      "step": 35045
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6829692832617436,
+      "learning_rate": 1.5522426160677029e-06,
+      "loss": 0.9536,
+      "step": 35046
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4403529013760135,
+      "learning_rate": 1.5518343217454934e-06,
+      "loss": 0.7891,
+      "step": 35047
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5452004993666504,
+      "learning_rate": 1.5514260766111777e-06,
+      "loss": 0.8752,
+      "step": 35048
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.688391666063098,
+      "learning_rate": 1.5510178806671328e-06,
+      "loss": 0.8418,
+      "step": 35049
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7507437917045978,
+      "learning_rate": 1.5506097339157399e-06,
+      "loss": 0.8091,
+      "step": 35050
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.188755538820646,
+      "learning_rate": 1.550201636359373e-06,
+      "loss": 0.8237,
+      "step": 35051
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.636200536361201,
+      "learning_rate": 1.5497935880004034e-06,
+      "loss": 0.9305,
+      "step": 35052
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6912989306692836,
+      "learning_rate": 1.5493855888412123e-06,
+      "loss": 0.7954,
+      "step": 35053
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.656805199630053,
+      "learning_rate": 1.5489776388841748e-06,
+      "loss": 0.8455,
+      "step": 35054
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7083435440888028,
+      "learning_rate": 1.5485697381316656e-06,
+      "loss": 0.8618,
+      "step": 35055
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9960497033145381,
+      "learning_rate": 1.548161886586057e-06,
+      "loss": 0.7853,
+      "step": 35056
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7737584171625131,
+      "learning_rate": 1.5477540842497251e-06,
+      "loss": 0.8349,
+      "step": 35057
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5396729997778325,
+      "learning_rate": 1.5473463311250481e-06,
+      "loss": 0.8333,
+      "step": 35058
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7459562589064004,
+      "learning_rate": 1.5469386272143961e-06,
+      "loss": 0.7673,
+      "step": 35059
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4707765470856118,
+      "learning_rate": 1.5465309725201416e-06,
+      "loss": 0.7999,
+      "step": 35060
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.645364664870184,
+      "learning_rate": 1.5461233670446618e-06,
+      "loss": 0.8294,
+      "step": 35061
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.619701177452632,
+      "learning_rate": 1.5457158107903269e-06,
+      "loss": 0.7689,
+      "step": 35062
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5542989040499866,
+      "learning_rate": 1.5453083037595128e-06,
+      "loss": 0.8862,
+      "step": 35063
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.486092974847252,
+      "learning_rate": 1.5449008459545877e-06,
+      "loss": 0.8367,
+      "step": 35064
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7513377452483766,
+      "learning_rate": 1.5444934373779296e-06,
+      "loss": 0.6699,
+      "step": 35065
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5896251248283524,
+      "learning_rate": 1.5440860780319055e-06,
+      "loss": 0.8387,
+      "step": 35066
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6384232189461279,
+      "learning_rate": 1.5436787679188913e-06,
+      "loss": 0.9545,
+      "step": 35067
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4921359424958693,
+      "learning_rate": 1.5432715070412552e-06,
+      "loss": 0.7997,
+      "step": 35068
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.678218317227419,
+      "learning_rate": 1.5428642954013706e-06,
+      "loss": 0.6766,
+      "step": 35069
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7758083745822217,
+      "learning_rate": 1.5424571330016059e-06,
+      "loss": 0.7611,
+      "step": 35070
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4376155375030741,
+      "learning_rate": 1.5420500198443367e-06,
+      "loss": 0.8464,
+      "step": 35071
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6675328401115899,
+      "learning_rate": 1.541642955931929e-06,
+      "loss": 0.8115,
+      "step": 35072
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5842550856629434,
+      "learning_rate": 1.541235941266751e-06,
+      "loss": 0.883,
+      "step": 35073
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6655106708791707,
+      "learning_rate": 1.5408289758511763e-06,
+      "loss": 0.7071,
+      "step": 35074
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7183427101291364,
+      "learning_rate": 1.5404220596875752e-06,
+      "loss": 0.7148,
+      "step": 35075
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3588522439970538,
+      "learning_rate": 1.5400151927783158e-06,
+      "loss": 0.8035,
+      "step": 35076
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4381207939385994,
+      "learning_rate": 1.5396083751257628e-06,
+      "loss": 0.8802,
+      "step": 35077
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8303668433458917,
+      "learning_rate": 1.5392016067322891e-06,
+      "loss": 0.8072,
+      "step": 35078
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6030426266312374,
+      "learning_rate": 1.5387948876002645e-06,
+      "loss": 0.7662,
+      "step": 35079
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8142665023705453,
+      "learning_rate": 1.5383882177320553e-06,
+      "loss": 0.7352,
+      "step": 35080
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7804512560037762,
+      "learning_rate": 1.537981597130026e-06,
+      "loss": 0.8567,
+      "step": 35081
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3725375732122473,
+      "learning_rate": 1.537575025796547e-06,
+      "loss": 0.7639,
+      "step": 35082
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.556106545724761,
+      "learning_rate": 1.5371685037339878e-06,
+      "loss": 0.8143,
+      "step": 35083
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8719623785289562,
+      "learning_rate": 1.536762030944713e-06,
+      "loss": 0.7662,
+      "step": 35084
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8697578763837595,
+      "learning_rate": 1.536355607431086e-06,
+      "loss": 0.876,
+      "step": 35085
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7523437355466043,
+      "learning_rate": 1.5359492331954795e-06,
+      "loss": 0.8286,
+      "step": 35086
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6981363398395615,
+      "learning_rate": 1.5355429082402528e-06,
+      "loss": 0.8427,
+      "step": 35087
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.512417443655776,
+      "learning_rate": 1.5351366325677785e-06,
+      "loss": 0.7842,
+      "step": 35088
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5108927736151778,
+      "learning_rate": 1.5347304061804158e-06,
+      "loss": 0.7065,
+      "step": 35089
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6844194342591325,
+      "learning_rate": 1.5343242290805348e-06,
+      "loss": 0.8786,
+      "step": 35090
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7188043576639991,
+      "learning_rate": 1.5339181012704974e-06,
+      "loss": 0.6336,
+      "step": 35091
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6064891732925684,
+      "learning_rate": 1.5335120227526678e-06,
+      "loss": 0.778,
+      "step": 35092
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6100454657082113,
+      "learning_rate": 1.5331059935294168e-06,
+      "loss": 0.7848,
+      "step": 35093
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6846318902460915,
+      "learning_rate": 1.532700013603098e-06,
+      "loss": 0.8826,
+      "step": 35094
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8225678872832853,
+      "learning_rate": 1.5322940829760802e-06,
+      "loss": 0.9872,
+      "step": 35095
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.564025505568297,
+      "learning_rate": 1.5318882016507286e-06,
+      "loss": 0.9157,
+      "step": 35096
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6387190010828572,
+      "learning_rate": 1.531482369629408e-06,
+      "loss": 0.7805,
+      "step": 35097
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6186427480030583,
+      "learning_rate": 1.531076586914474e-06,
+      "loss": 0.8925,
+      "step": 35098
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4603087407450432,
+      "learning_rate": 1.5306708535082937e-06,
+      "loss": 0.7597,
+      "step": 35099
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7348742469376577,
+      "learning_rate": 1.530265169413231e-06,
+      "loss": 0.6715,
+      "step": 35100
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.821815932468281,
+      "learning_rate": 1.5298595346316457e-06,
+      "loss": 0.817,
+      "step": 35101
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0336631230125493,
+      "learning_rate": 1.5294539491658978e-06,
+      "loss": 0.7794,
+      "step": 35102
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4197661160767223,
+      "learning_rate": 1.5290484130183515e-06,
+      "loss": 0.7468,
+      "step": 35103
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.58676002306853,
+      "learning_rate": 1.5286429261913682e-06,
+      "loss": 0.7992,
+      "step": 35104
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7298407166515686,
+      "learning_rate": 1.5282374886873063e-06,
+      "loss": 0.7455,
+      "step": 35105
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7419094028584983,
+      "learning_rate": 1.5278321005085296e-06,
+      "loss": 0.8156,
+      "step": 35106
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.102157684840973,
+      "learning_rate": 1.5274267616573968e-06,
+      "loss": 0.9229,
+      "step": 35107
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5895579188808489,
+      "learning_rate": 1.5270214721362653e-06,
+      "loss": 0.8048,
+      "step": 35108
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7676001806391344,
+      "learning_rate": 1.5266162319474974e-06,
+      "loss": 0.775,
+      "step": 35109
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5542103017662918,
+      "learning_rate": 1.5262110410934561e-06,
+      "loss": 0.7785,
+      "step": 35110
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7481490342427584,
+      "learning_rate": 1.5258058995764946e-06,
+      "loss": 0.7778,
+      "step": 35111
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6116702723655143,
+      "learning_rate": 1.5254008073989734e-06,
+      "loss": 0.9295,
+      "step": 35112
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8662656739101153,
+      "learning_rate": 1.5249957645632508e-06,
+      "loss": 0.8466,
+      "step": 35113
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.652769826953919,
+      "learning_rate": 1.5245907710716912e-06,
+      "loss": 0.787,
+      "step": 35114
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6058282441709564,
+      "learning_rate": 1.524185826926643e-06,
+      "loss": 0.8226,
+      "step": 35115
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7309367182714024,
+      "learning_rate": 1.5237809321304674e-06,
+      "loss": 0.9334,
+      "step": 35116
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6231340217786325,
+      "learning_rate": 1.523376086685523e-06,
+      "loss": 0.8986,
+      "step": 35117
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.927286422175019,
+      "learning_rate": 1.5229712905941696e-06,
+      "loss": 0.6711,
+      "step": 35118
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6306780177488,
+      "learning_rate": 1.52256654385876e-06,
+      "loss": 0.8548,
+      "step": 35119
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.198471713083713,
+      "learning_rate": 1.5221618464816501e-06,
+      "loss": 0.7882,
+      "step": 35120
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6305567057756467,
+      "learning_rate": 1.5217571984652012e-06,
+      "loss": 0.7451,
+      "step": 35121
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5833694530766955,
+      "learning_rate": 1.5213525998117628e-06,
+      "loss": 0.9588,
+      "step": 35122
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6927395046516955,
+      "learning_rate": 1.5209480505236963e-06,
+      "loss": 0.853,
+      "step": 35123
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8067308507600999,
+      "learning_rate": 1.5205435506033529e-06,
+      "loss": 0.7821,
+      "step": 35124
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5084815449573972,
+      "learning_rate": 1.5201391000530908e-06,
+      "loss": 0.8428,
+      "step": 35125
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5643376661704946,
+      "learning_rate": 1.5197346988752615e-06,
+      "loss": 0.7573,
+      "step": 35126
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9573718697064788,
+      "learning_rate": 1.5193303470722243e-06,
+      "loss": 0.9286,
+      "step": 35127
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7859549705326383,
+      "learning_rate": 1.5189260446463306e-06,
+      "loss": 0.798,
+      "step": 35128
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5196457206605913,
+      "learning_rate": 1.5185217915999328e-06,
+      "loss": 0.8148,
+      "step": 35129
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.404745543439277,
+      "learning_rate": 1.518117587935385e-06,
+      "loss": 0.7989,
+      "step": 35130
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5009325513866667,
+      "learning_rate": 1.517713433655046e-06,
+      "loss": 0.764,
+      "step": 35131
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.500002738415713,
+      "learning_rate": 1.5173093287612628e-06,
+      "loss": 0.8382,
+      "step": 35132
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1069850132138965,
+      "learning_rate": 1.5169052732563895e-06,
+      "loss": 0.7902,
+      "step": 35133
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5689752482254238,
+      "learning_rate": 1.5165012671427793e-06,
+      "loss": 0.9392,
+      "step": 35134
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4911147832472555,
+      "learning_rate": 1.516097310422786e-06,
+      "loss": 0.8446,
+      "step": 35135
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6428166979092909,
+      "learning_rate": 1.5156934030987603e-06,
+      "loss": 0.8104,
+      "step": 35136
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9063511267351454,
+      "learning_rate": 1.515289545173052e-06,
+      "loss": 0.7784,
+      "step": 35137
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.610701958194524,
+      "learning_rate": 1.5148857366480141e-06,
+      "loss": 0.8783,
+      "step": 35138
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.356842478825016,
+      "learning_rate": 1.514481977525999e-06,
+      "loss": 0.8585,
+      "step": 35139
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6197478854372476,
+      "learning_rate": 1.514078267809358e-06,
+      "loss": 0.8332,
+      "step": 35140
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.591068446457414,
+      "learning_rate": 1.513674607500436e-06,
+      "loss": 0.7611,
+      "step": 35141
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7447335681879962,
+      "learning_rate": 1.5132709966015901e-06,
+      "loss": 0.678,
+      "step": 35142
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5389303025285919,
+      "learning_rate": 1.512867435115165e-06,
+      "loss": 0.7997,
+      "step": 35143
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4307596754487557,
+      "learning_rate": 1.5124639230435157e-06,
+      "loss": 0.8072,
+      "step": 35144
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5438463237875721,
+      "learning_rate": 1.5120604603889855e-06,
+      "loss": 0.8764,
+      "step": 35145
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.669053146217471,
+      "learning_rate": 1.5116570471539294e-06,
+      "loss": 0.8907,
+      "step": 35146
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7630407726721751,
+      "learning_rate": 1.5112536833406898e-06,
+      "loss": 0.7133,
+      "step": 35147
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4758106706461736,
+      "learning_rate": 1.5108503689516218e-06,
+      "loss": 0.841,
+      "step": 35148
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.82757716613104,
+      "learning_rate": 1.5104471039890711e-06,
+      "loss": 0.7549,
+      "step": 35149
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5772220689604113,
+      "learning_rate": 1.5100438884553826e-06,
+      "loss": 0.7979,
+      "step": 35150
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6978836626088287,
+      "learning_rate": 1.5096407223529053e-06,
+      "loss": 0.8349,
+      "step": 35151
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.57146199730502,
+      "learning_rate": 1.5092376056839908e-06,
+      "loss": 0.8806,
+      "step": 35152
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.753533577749429,
+      "learning_rate": 1.5088345384509838e-06,
+      "loss": 0.7854,
+      "step": 35153
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7582025452344898,
+      "learning_rate": 1.5084315206562272e-06,
+      "loss": 0.6769,
+      "step": 35154
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3847355531213636,
+      "learning_rate": 1.508028552302071e-06,
+      "loss": 0.8865,
+      "step": 35155
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4196223220277575,
+      "learning_rate": 1.5076256333908635e-06,
+      "loss": 0.7295,
+      "step": 35156
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.500665950451119,
+      "learning_rate": 1.5072227639249482e-06,
+      "loss": 0.8847,
+      "step": 35157
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5295127028998132,
+      "learning_rate": 1.5068199439066678e-06,
+      "loss": 0.7596,
+      "step": 35158
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7809781301142533,
+      "learning_rate": 1.5064171733383715e-06,
+      "loss": 0.7651,
+      "step": 35159
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7692035596037894,
+      "learning_rate": 1.5060144522224052e-06,
+      "loss": 0.929,
+      "step": 35160
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8160216835143108,
+      "learning_rate": 1.5056117805611115e-06,
+      "loss": 0.6959,
+      "step": 35161
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8319492129498287,
+      "learning_rate": 1.505209158356834e-06,
+      "loss": 0.714,
+      "step": 35162
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4572099842385444,
+      "learning_rate": 1.50480658561192e-06,
+      "loss": 0.7335,
+      "step": 35163
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6763616386956133,
+      "learning_rate": 1.5044040623287092e-06,
+      "loss": 0.826,
+      "step": 35164
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6588286927878595,
+      "learning_rate": 1.5040015885095494e-06,
+      "loss": 0.6951,
+      "step": 35165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.633767510558133,
+      "learning_rate": 1.5035991641567803e-06,
+      "loss": 0.8668,
+      "step": 35166
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5992381849263113,
+      "learning_rate": 1.5031967892727494e-06,
+      "loss": 0.6692,
+      "step": 35167
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5303614695257461,
+      "learning_rate": 1.5027944638597947e-06,
+      "loss": 0.7849,
+      "step": 35168
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6314351500612294,
+      "learning_rate": 1.502392187920263e-06,
+      "loss": 0.9401,
+      "step": 35169
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5199402345359099,
+      "learning_rate": 1.5019899614564937e-06,
+      "loss": 0.8478,
+      "step": 35170
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7747383552862694,
+      "learning_rate": 1.5015877844708272e-06,
+      "loss": 0.8106,
+      "step": 35171
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6409079492548897,
+      "learning_rate": 1.501185656965607e-06,
+      "loss": 0.9248,
+      "step": 35172
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7596495910058663,
+      "learning_rate": 1.500783578943177e-06,
+      "loss": 0.6894,
+      "step": 35173
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6907278169350377,
+      "learning_rate": 1.500381550405876e-06,
+      "loss": 0.8082,
+      "step": 35174
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8000235552548314,
+      "learning_rate": 1.4999795713560428e-06,
+      "loss": 0.9604,
+      "step": 35175
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6326881586041209,
+      "learning_rate": 1.4995776417960184e-06,
+      "loss": 0.6937,
+      "step": 35176
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9409699979742128,
+      "learning_rate": 1.4991757617281466e-06,
+      "loss": 0.8958,
+      "step": 35177
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6474859875129604,
+      "learning_rate": 1.4987739311547656e-06,
+      "loss": 0.9312,
+      "step": 35178
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6882394005602939,
+      "learning_rate": 1.4983721500782122e-06,
+      "loss": 0.782,
+      "step": 35179
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4719778711358766,
+      "learning_rate": 1.497970418500827e-06,
+      "loss": 0.7922,
+      "step": 35180
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.654842968896338,
+      "learning_rate": 1.4975687364249513e-06,
+      "loss": 0.908,
+      "step": 35181
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7464125507420387,
+      "learning_rate": 1.4971671038529235e-06,
+      "loss": 0.8854,
+      "step": 35182
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.064029792176515,
+      "learning_rate": 1.496765520787078e-06,
+      "loss": 0.7462,
+      "step": 35183
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5961714878178546,
+      "learning_rate": 1.4963639872297586e-06,
+      "loss": 0.8812,
+      "step": 35184
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.738603046954485,
+      "learning_rate": 1.4959625031832969e-06,
+      "loss": 0.8877,
+      "step": 35185
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5617421125433701,
+      "learning_rate": 1.4955610686500365e-06,
+      "loss": 0.759,
+      "step": 35186
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6712463148868373,
+      "learning_rate": 1.4951596836323102e-06,
+      "loss": 0.7534,
+      "step": 35187
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9208783856292206,
+      "learning_rate": 1.494758348132459e-06,
+      "loss": 0.8333,
+      "step": 35188
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4272994616255443,
+      "learning_rate": 1.4943570621528148e-06,
+      "loss": 0.9183,
+      "step": 35189
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.07034573611255,
+      "learning_rate": 1.4939558256957166e-06,
+      "loss": 0.7332,
+      "step": 35190
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.63635294939852,
+      "learning_rate": 1.4935546387635047e-06,
+      "loss": 0.9253,
+      "step": 35191
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6844718548432622,
+      "learning_rate": 1.4931535013585075e-06,
+      "loss": 0.7401,
+      "step": 35192
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8232329395142144,
+      "learning_rate": 1.4927524134830629e-06,
+      "loss": 0.8081,
+      "step": 35193
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6430562705418874,
+      "learning_rate": 1.492351375139507e-06,
+      "loss": 0.7709,
+      "step": 35194
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7916047210229122,
+      "learning_rate": 1.4919503863301799e-06,
+      "loss": 0.8349,
+      "step": 35195
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.585173276575349,
+      "learning_rate": 1.4915494470574065e-06,
+      "loss": 0.8238,
+      "step": 35196
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7090896308876382,
+      "learning_rate": 1.4911485573235262e-06,
+      "loss": 0.9332,
+      "step": 35197
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6834930459434787,
+      "learning_rate": 1.4907477171308727e-06,
+      "loss": 0.9318,
+      "step": 35198
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7635913945004142,
+      "learning_rate": 1.4903469264817838e-06,
+      "loss": 0.9548,
+      "step": 35199
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9215681385593528,
+      "learning_rate": 1.4899461853785857e-06,
+      "loss": 0.7659,
+      "step": 35200
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7443222423103835,
+      "learning_rate": 1.4895454938236142e-06,
+      "loss": 0.8599,
+      "step": 35201
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8388644236846732,
+      "learning_rate": 1.4891448518192065e-06,
+      "loss": 0.8779,
+      "step": 35202
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.521950199341215,
+      "learning_rate": 1.4887442593676893e-06,
+      "loss": 0.7669,
+      "step": 35203
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4702086244642891,
+      "learning_rate": 1.4883437164713999e-06,
+      "loss": 0.8018,
+      "step": 35204
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5032233329896698,
+      "learning_rate": 1.4879432231326663e-06,
+      "loss": 0.8534,
+      "step": 35205
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4599836533511168,
+      "learning_rate": 1.4875427793538233e-06,
+      "loss": 0.843,
+      "step": 35206
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6020571596791244,
+      "learning_rate": 1.487142385137199e-06,
+      "loss": 0.806,
+      "step": 35207
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5396446148731344,
+      "learning_rate": 1.4867420404851306e-06,
+      "loss": 0.8867,
+      "step": 35208
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5949470822853569,
+      "learning_rate": 1.486341745399944e-06,
+      "loss": 0.8456,
+      "step": 35209
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6393143299871373,
+      "learning_rate": 1.4859414998839694e-06,
+      "loss": 0.887,
+      "step": 35210
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7146656255908033,
+      "learning_rate": 1.4855413039395383e-06,
+      "loss": 0.7608,
+      "step": 35211
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6371666834568452,
+      "learning_rate": 1.4851411575689845e-06,
+      "loss": 0.8016,
+      "step": 35212
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.676075715464574,
+      "learning_rate": 1.4847410607746337e-06,
+      "loss": 0.8148,
+      "step": 35213
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5255661319223697,
+      "learning_rate": 1.4843410135588155e-06,
+      "loss": 0.7051,
+      "step": 35214
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5247806316093127,
+      "learning_rate": 1.4839410159238587e-06,
+      "loss": 0.7921,
+      "step": 35215
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7588548801194408,
+      "learning_rate": 1.4835410678720963e-06,
+      "loss": 0.879,
+      "step": 35216
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5238396632061872,
+      "learning_rate": 1.483141169405854e-06,
+      "loss": 0.7973,
+      "step": 35217
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5119331330910235,
+      "learning_rate": 1.4827413205274565e-06,
+      "loss": 0.9249,
+      "step": 35218
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.741475709425425,
+      "learning_rate": 1.4823415212392378e-06,
+      "loss": 0.6841,
+      "step": 35219
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.760381419163689,
+      "learning_rate": 1.4819417715435247e-06,
+      "loss": 0.8044,
+      "step": 35220
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5072941138945517,
+      "learning_rate": 1.481542071442643e-06,
+      "loss": 0.7831,
+      "step": 35221
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7110517606361737,
+      "learning_rate": 1.481142420938918e-06,
+      "loss": 0.8892,
+      "step": 35222
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6343827982384214,
+      "learning_rate": 1.4807428200346817e-06,
+      "loss": 0.8559,
+      "step": 35223
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6041115262015193,
+      "learning_rate": 1.480343268732255e-06,
+      "loss": 0.8962,
+      "step": 35224
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8177633656570058,
+      "learning_rate": 1.4799437670339701e-06,
+      "loss": 0.7103,
+      "step": 35225
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5734764554007736,
+      "learning_rate": 1.4795443149421473e-06,
+      "loss": 0.8893,
+      "step": 35226
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4741368336356881,
+      "learning_rate": 1.4791449124591173e-06,
+      "loss": 0.7873,
+      "step": 35227
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.289193105587639,
+      "learning_rate": 1.4787455595872002e-06,
+      "loss": 0.8814,
+      "step": 35228
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4937926901366176,
+      "learning_rate": 1.4783462563287277e-06,
+      "loss": 0.6255,
+      "step": 35229
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5661914429849957,
+      "learning_rate": 1.477947002686021e-06,
+      "loss": 0.8921,
+      "step": 35230
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7085460222420383,
+      "learning_rate": 1.477547798661403e-06,
+      "loss": 0.8279,
+      "step": 35231
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.48195987682822,
+      "learning_rate": 1.477148644257198e-06,
+      "loss": 0.8044,
+      "step": 35232
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6950507818972975,
+      "learning_rate": 1.4767495394757359e-06,
+      "loss": 0.8166,
+      "step": 35233
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.825057899645854,
+      "learning_rate": 1.4763504843193366e-06,
+      "loss": 0.7937,
+      "step": 35234
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4371089355586542,
+      "learning_rate": 1.4759514787903195e-06,
+      "loss": 0.6947,
+      "step": 35235
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5036667916082094,
+      "learning_rate": 1.4755525228910128e-06,
+      "loss": 0.6994,
+      "step": 35236
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4511984146680055,
+      "learning_rate": 1.4751536166237402e-06,
+      "loss": 0.8183,
+      "step": 35237
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.858001282508594,
+      "learning_rate": 1.4747547599908218e-06,
+      "loss": 0.8486,
+      "step": 35238
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6324869086790719,
+      "learning_rate": 1.4743559529945784e-06,
+      "loss": 0.7907,
+      "step": 35239
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.856523852578237,
+      "learning_rate": 1.4739571956373333e-06,
+      "loss": 0.8988,
+      "step": 35240
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8662079878007243,
+      "learning_rate": 1.4735584879214116e-06,
+      "loss": 0.8072,
+      "step": 35241
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5742063826531896,
+      "learning_rate": 1.4731598298491313e-06,
+      "loss": 0.882,
+      "step": 35242
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4439550095167029,
+      "learning_rate": 1.4727612214228115e-06,
+      "loss": 0.8756,
+      "step": 35243
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9810058441787612,
+      "learning_rate": 1.4723626626447784e-06,
+      "loss": 0.8905,
+      "step": 35244
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6155962976939926,
+      "learning_rate": 1.471964153517348e-06,
+      "loss": 0.8394,
+      "step": 35245
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.42908103300342,
+      "learning_rate": 1.4715656940428437e-06,
+      "loss": 0.789,
+      "step": 35246
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8877603352218526,
+      "learning_rate": 1.4711672842235814e-06,
+      "loss": 0.7956,
+      "step": 35247
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6144456277090458,
+      "learning_rate": 1.470768924061886e-06,
+      "loss": 0.8849,
+      "step": 35248
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7832900669942109,
+      "learning_rate": 1.4703706135600715e-06,
+      "loss": 0.7578,
+      "step": 35249
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.678265386322677,
+      "learning_rate": 1.4699723527204623e-06,
+      "loss": 0.884,
+      "step": 35250
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7915823672642153,
+      "learning_rate": 1.4695741415453757e-06,
+      "loss": 0.927,
+      "step": 35251
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5708073166809695,
+      "learning_rate": 1.4691759800371252e-06,
+      "loss": 0.9784,
+      "step": 35252
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.409957048455459,
+      "learning_rate": 1.4687778681980335e-06,
+      "loss": 0.6553,
+      "step": 35253
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7257104903692557,
+      "learning_rate": 1.468379806030421e-06,
+      "loss": 0.8724,
+      "step": 35254
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7712572134099323,
+      "learning_rate": 1.4679817935366014e-06,
+      "loss": 0.8048,
+      "step": 35255
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.807493330423973,
+      "learning_rate": 1.4675838307188916e-06,
+      "loss": 0.8582,
+      "step": 35256
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.610233586272662,
+      "learning_rate": 1.4671859175796087e-06,
+      "loss": 0.8599,
+      "step": 35257
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6535242782791482,
+      "learning_rate": 1.4667880541210744e-06,
+      "loss": 0.9695,
+      "step": 35258
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5734438801115984,
+      "learning_rate": 1.4663902403456009e-06,
+      "loss": 0.768,
+      "step": 35259
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5545422780320297,
+      "learning_rate": 1.4659924762555033e-06,
+      "loss": 0.9106,
+      "step": 35260
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8629016616715428,
+      "learning_rate": 1.4655947618530998e-06,
+      "loss": 0.7903,
+      "step": 35261
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0409952584190494,
+      "learning_rate": 1.465197097140707e-06,
+      "loss": 0.7521,
+      "step": 35262
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.592932642850189,
+      "learning_rate": 1.464799482120639e-06,
+      "loss": 0.8115,
+      "step": 35263
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4979168717842106,
+      "learning_rate": 1.4644019167952084e-06,
+      "loss": 0.8215,
+      "step": 35264
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6304444561167675,
+      "learning_rate": 1.4640044011667343e-06,
+      "loss": 0.7935,
+      "step": 35265
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.744514203050781,
+      "learning_rate": 1.463606935237526e-06,
+      "loss": 0.6638,
+      "step": 35266
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6218210033374303,
+      "learning_rate": 1.4632095190099039e-06,
+      "loss": 0.8126,
+      "step": 35267
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.52252638946752,
+      "learning_rate": 1.462812152486176e-06,
+      "loss": 0.8172,
+      "step": 35268
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7068257225169428,
+      "learning_rate": 1.4624148356686619e-06,
+      "loss": 0.8366,
+      "step": 35269
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.262256540670229,
+      "learning_rate": 1.4620175685596683e-06,
+      "loss": 0.7429,
+      "step": 35270
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.058310726786244,
+      "learning_rate": 1.4616203511615134e-06,
+      "loss": 0.9385,
+      "step": 35271
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6352596213447914,
+      "learning_rate": 1.461223183476509e-06,
+      "loss": 0.9269,
+      "step": 35272
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7844433895951193,
+      "learning_rate": 1.4608260655069639e-06,
+      "loss": 0.8426,
+      "step": 35273
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7493194986319351,
+      "learning_rate": 1.4604289972551933e-06,
+      "loss": 0.9274,
+      "step": 35274
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5592586652558953,
+      "learning_rate": 1.4600319787235073e-06,
+      "loss": 0.7579,
+      "step": 35275
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6497107298500806,
+      "learning_rate": 1.4596350099142243e-06,
+      "loss": 1.0266,
+      "step": 35276
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5413617859507107,
+      "learning_rate": 1.4592380908296455e-06,
+      "loss": 0.8142,
+      "step": 35277
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5539036345226118,
+      "learning_rate": 1.458841221472087e-06,
+      "loss": 0.7199,
+      "step": 35278
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6436990525695172,
+      "learning_rate": 1.4584444018438616e-06,
+      "loss": 0.8066,
+      "step": 35279
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4012022027202953,
+      "learning_rate": 1.4580476319472758e-06,
+      "loss": 0.7838,
+      "step": 35280
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5095538256264112,
+      "learning_rate": 1.4576509117846394e-06,
+      "loss": 0.8899,
+      "step": 35281
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6623208374271594,
+      "learning_rate": 1.4572542413582635e-06,
+      "loss": 0.8795,
+      "step": 35282
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.875798429331673,
+      "learning_rate": 1.456857620670461e-06,
+      "loss": 0.8853,
+      "step": 35283
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6965287254884847,
+      "learning_rate": 1.4564610497235377e-06,
+      "loss": 0.7721,
+      "step": 35284
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8942747335236179,
+      "learning_rate": 1.4560645285198006e-06,
+      "loss": 0.8291,
+      "step": 35285
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5638404553729488,
+      "learning_rate": 1.4556680570615623e-06,
+      "loss": 0.8326,
+      "step": 35286
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9882918694788465,
+      "learning_rate": 1.4552716353511286e-06,
+      "loss": 0.8142,
+      "step": 35287
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7416065344070536,
+      "learning_rate": 1.454875263390808e-06,
+      "loss": 0.6716,
+      "step": 35288
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.836480210165693,
+      "learning_rate": 1.4544789411829118e-06,
+      "loss": 0.8447,
+      "step": 35289
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.483267224726074,
+      "learning_rate": 1.454082668729745e-06,
+      "loss": 0.865,
+      "step": 35290
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7178068799886133,
+      "learning_rate": 1.4536864460336108e-06,
+      "loss": 0.8018,
+      "step": 35291
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6146231480641717,
+      "learning_rate": 1.4532902730968213e-06,
+      "loss": 0.8091,
+      "step": 35292
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9983371372178473,
+      "learning_rate": 1.4528941499216853e-06,
+      "loss": 0.855,
+      "step": 35293
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6566135095725094,
+      "learning_rate": 1.4524980765105013e-06,
+      "loss": 0.7435,
+      "step": 35294
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7562051598078334,
+      "learning_rate": 1.4521020528655794e-06,
+      "loss": 0.8577,
+      "step": 35295
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.563004972975194,
+      "learning_rate": 1.4517060789892257e-06,
+      "loss": 0.7673,
+      "step": 35296
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9154504163413597,
+      "learning_rate": 1.4513101548837505e-06,
+      "loss": 0.7846,
+      "step": 35297
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5582567929623434,
+      "learning_rate": 1.4509142805514486e-06,
+      "loss": 0.6898,
+      "step": 35298
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6281280029215641,
+      "learning_rate": 1.4505184559946306e-06,
+      "loss": 0.9166,
+      "step": 35299
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7325237851570198,
+      "learning_rate": 1.4501226812156032e-06,
+      "loss": 0.7352,
+      "step": 35300
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6375177755072403,
+      "learning_rate": 1.4497269562166672e-06,
+      "loss": 0.7645,
+      "step": 35301
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.785988885387799,
+      "learning_rate": 1.4493312810001293e-06,
+      "loss": 0.9068,
+      "step": 35302
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4046011372626532,
+      "learning_rate": 1.4489356555682887e-06,
+      "loss": 0.8223,
+      "step": 35303
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4924202330733825,
+      "learning_rate": 1.448540079923455e-06,
+      "loss": 0.8766,
+      "step": 35304
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8346255881907073,
+      "learning_rate": 1.4481445540679261e-06,
+      "loss": 0.6597,
+      "step": 35305
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.748104374655128,
+      "learning_rate": 1.4477490780040093e-06,
+      "loss": 0.8317,
+      "step": 35306
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8539758119275082,
+      "learning_rate": 1.4473536517340047e-06,
+      "loss": 0.74,
+      "step": 35307
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4634249254234222,
+      "learning_rate": 1.446958275260213e-06,
+      "loss": 0.76,
+      "step": 35308
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7645377167165512,
+      "learning_rate": 1.4465629485849376e-06,
+      "loss": 0.8328,
+      "step": 35309
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.422921939342699,
+      "learning_rate": 1.4461676717104834e-06,
+      "loss": 0.8592,
+      "step": 35310
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7404229239515496,
+      "learning_rate": 1.4457724446391486e-06,
+      "loss": 0.7871,
+      "step": 35311
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.972720039364337,
+      "learning_rate": 1.4453772673732336e-06,
+      "loss": 0.7445,
+      "step": 35312
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4619438021480304,
+      "learning_rate": 1.4449821399150398e-06,
+      "loss": 0.806,
+      "step": 35313
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9525057062968012,
+      "learning_rate": 1.44458706226687e-06,
+      "loss": 0.8862,
+      "step": 35314
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8885645848863934,
+      "learning_rate": 1.4441920344310235e-06,
+      "loss": 0.8675,
+      "step": 35315
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8506760470550114,
+      "learning_rate": 1.443797056409798e-06,
+      "loss": 0.8281,
+      "step": 35316
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.533666815256227,
+      "learning_rate": 1.4434021282054943e-06,
+      "loss": 0.9361,
+      "step": 35317
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.123694943424594,
+      "learning_rate": 1.443007249820415e-06,
+      "loss": 0.859,
+      "step": 35318
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7170033951815438,
+      "learning_rate": 1.442612421256856e-06,
+      "loss": 0.7502,
+      "step": 35319
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.570783184756032,
+      "learning_rate": 1.4422176425171153e-06,
+      "loss": 0.8344,
+      "step": 35320
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.60727652016603,
+      "learning_rate": 1.4418229136034922e-06,
+      "loss": 0.7257,
+      "step": 35321
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7330516089256802,
+      "learning_rate": 1.4414282345182885e-06,
+      "loss": 0.972,
+      "step": 35322
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.675349175160879,
+      "learning_rate": 1.4410336052637985e-06,
+      "loss": 0.6523,
+      "step": 35323
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4308215490360865,
+      "learning_rate": 1.4406390258423186e-06,
+      "loss": 0.7144,
+      "step": 35324
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7064386201507673,
+      "learning_rate": 1.4402444962561502e-06,
+      "loss": 0.8205,
+      "step": 35325
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.07222789029567,
+      "learning_rate": 1.4398500165075857e-06,
+      "loss": 0.8139,
+      "step": 35326
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7521398436272904,
+      "learning_rate": 1.4394555865989279e-06,
+      "loss": 0.8333,
+      "step": 35327
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7301928139192048,
+      "learning_rate": 1.439061206532466e-06,
+      "loss": 0.6776,
+      "step": 35328
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6003065387797688,
+      "learning_rate": 1.4386668763105039e-06,
+      "loss": 0.7564,
+      "step": 35329
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5107742557685275,
+      "learning_rate": 1.4382725959353305e-06,
+      "loss": 0.7865,
+      "step": 35330
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5502116568710522,
+      "learning_rate": 1.4378783654092476e-06,
+      "loss": 0.9071,
+      "step": 35331
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6013956909493283,
+      "learning_rate": 1.4374841847345478e-06,
+      "loss": 0.7966,
+      "step": 35332
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7776709866116716,
+      "learning_rate": 1.4370900539135225e-06,
+      "loss": 0.8532,
+      "step": 35333
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6401374184789241,
+      "learning_rate": 1.4366959729484708e-06,
+      "loss": 0.8264,
+      "step": 35334
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7696526227156815,
+      "learning_rate": 1.4363019418416878e-06,
+      "loss": 0.7426,
+      "step": 35335
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3446156278864057,
+      "learning_rate": 1.4359079605954673e-06,
+      "loss": 0.7937,
+      "step": 35336
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.502291075582562,
+      "learning_rate": 1.4355140292120983e-06,
+      "loss": 0.8882,
+      "step": 35337
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7434013761401093,
+      "learning_rate": 1.4351201476938792e-06,
+      "loss": 0.7093,
+      "step": 35338
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8457578121609854,
+      "learning_rate": 1.4347263160431036e-06,
+      "loss": 0.9107,
+      "step": 35339
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7664569618322306,
+      "learning_rate": 1.434332534262064e-06,
+      "loss": 0.6617,
+      "step": 35340
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5295541749420114,
+      "learning_rate": 1.4339388023530498e-06,
+      "loss": 0.9279,
+      "step": 35341
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5519460527502602,
+      "learning_rate": 1.433545120318355e-06,
+      "loss": 0.8354,
+      "step": 35342
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.655781242029953,
+      "learning_rate": 1.4331514881602759e-06,
+      "loss": 0.758,
+      "step": 35343
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6525668875018542,
+      "learning_rate": 1.4327579058811003e-06,
+      "loss": 0.895,
+      "step": 35344
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7992464268849415,
+      "learning_rate": 1.4323643734831182e-06,
+      "loss": 0.9397,
+      "step": 35345
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6966191783092541,
+      "learning_rate": 1.4319708909686259e-06,
+      "loss": 0.9162,
+      "step": 35346
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5789906637771989,
+      "learning_rate": 1.43157745833991e-06,
+      "loss": 0.8231,
+      "step": 35347
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7084849435737333,
+      "learning_rate": 1.4311840755992634e-06,
+      "loss": 0.786,
+      "step": 35348
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4412785187344581,
+      "learning_rate": 1.4307907427489753e-06,
+      "loss": 0.9822,
+      "step": 35349
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.470784442662976,
+      "learning_rate": 1.4303974597913383e-06,
+      "loss": 0.7562,
+      "step": 35350
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5963792960901604,
+      "learning_rate": 1.4300042267286374e-06,
+      "loss": 0.7823,
+      "step": 35351
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4830638571036656,
+      "learning_rate": 1.4296110435631683e-06,
+      "loss": 0.799,
+      "step": 35352
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8434616409717443,
+      "learning_rate": 1.429217910297217e-06,
+      "loss": 0.7524,
+      "step": 35353
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4297914383760462,
+      "learning_rate": 1.4288248269330695e-06,
+      "loss": 0.8054,
+      "step": 35354
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7388384207864438,
+      "learning_rate": 1.4284317934730175e-06,
+      "loss": 0.7395,
+      "step": 35355
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6476921348844833,
+      "learning_rate": 1.428038809919352e-06,
+      "loss": 0.915,
+      "step": 35356
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5068197142490372,
+      "learning_rate": 1.4276458762743583e-06,
+      "loss": 0.8461,
+      "step": 35357
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6821798381582989,
+      "learning_rate": 1.4272529925403221e-06,
+      "loss": 0.8876,
+      "step": 35358
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5970236492713807,
+      "learning_rate": 1.426860158719533e-06,
+      "loss": 0.8334,
+      "step": 35359
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0582603502511767,
+      "learning_rate": 1.426467374814281e-06,
+      "loss": 0.7622,
+      "step": 35360
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0581922003581297,
+      "learning_rate": 1.426074640826849e-06,
+      "loss": 0.8188,
+      "step": 35361
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5429414058251196,
+      "learning_rate": 1.4256819567595238e-06,
+      "loss": 0.796,
+      "step": 35362
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7057732214528643,
+      "learning_rate": 1.4252893226145924e-06,
+      "loss": 0.6834,
+      "step": 35363
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5099199656133568,
+      "learning_rate": 1.4248967383943435e-06,
+      "loss": 0.9265,
+      "step": 35364
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2881343483357197,
+      "learning_rate": 1.4245042041010604e-06,
+      "loss": 0.8635,
+      "step": 35365
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.557140768594496,
+      "learning_rate": 1.424111719737027e-06,
+      "loss": 0.8805,
+      "step": 35366
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7671771226921176,
+      "learning_rate": 1.4237192853045323e-06,
+      "loss": 0.693,
+      "step": 35367
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6785625449007273,
+      "learning_rate": 1.423326900805857e-06,
+      "loss": 0.9792,
+      "step": 35368
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4454679776829984,
+      "learning_rate": 1.4229345662432903e-06,
+      "loss": 0.7302,
+      "step": 35369
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8384655071311093,
+      "learning_rate": 1.4225422816191115e-06,
+      "loss": 0.8899,
+      "step": 35370
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7597882300321142,
+      "learning_rate": 1.4221500469356086e-06,
+      "loss": 0.7726,
+      "step": 35371
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.007672190170206,
+      "learning_rate": 1.4217578621950623e-06,
+      "loss": 0.8893,
+      "step": 35372
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7181051913476963,
+      "learning_rate": 1.4213657273997571e-06,
+      "loss": 0.8686,
+      "step": 35373
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.490097970402172,
+      "learning_rate": 1.4209736425519827e-06,
+      "loss": 0.8664,
+      "step": 35374
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7123870951245803,
+      "learning_rate": 1.420581607654009e-06,
+      "loss": 0.7917,
+      "step": 35375
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8785620846186966,
+      "learning_rate": 1.420189622708127e-06,
+      "loss": 0.9427,
+      "step": 35376
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.723094924760353,
+      "learning_rate": 1.4197976877166165e-06,
+      "loss": 0.8391,
+      "step": 35377
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6410635800661368,
+      "learning_rate": 1.4194058026817658e-06,
+      "loss": 0.8085,
+      "step": 35378
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8948808108829882,
+      "learning_rate": 1.4190139676058457e-06,
+      "loss": 0.7933,
+      "step": 35379
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9300239393131084,
+      "learning_rate": 1.4186221824911428e-06,
+      "loss": 0.8174,
+      "step": 35380
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6036227966138317,
+      "learning_rate": 1.418230447339941e-06,
+      "loss": 0.8311,
+      "step": 35381
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6031258551456062,
+      "learning_rate": 1.4178387621545187e-06,
+      "loss": 0.7123,
+      "step": 35382
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4448755391901809,
+      "learning_rate": 1.4174471269371537e-06,
+      "loss": 0.8071,
+      "step": 35383
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.785266375346856,
+      "learning_rate": 1.417055541690129e-06,
+      "loss": 0.762,
+      "step": 35384
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4855332705074535,
+      "learning_rate": 1.4166640064157267e-06,
+      "loss": 0.8958,
+      "step": 35385
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7272402390596062,
+      "learning_rate": 1.416272521116221e-06,
+      "loss": 0.7397,
+      "step": 35386
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4327676972692232,
+      "learning_rate": 1.4158810857938964e-06,
+      "loss": 0.8189,
+      "step": 35387
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6247073452883818,
+      "learning_rate": 1.4154897004510304e-06,
+      "loss": 0.94,
+      "step": 35388
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7423544831365106,
+      "learning_rate": 1.4150983650898987e-06,
+      "loss": 0.7139,
+      "step": 35389
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5985988275328573,
+      "learning_rate": 1.4147070797127826e-06,
+      "loss": 0.8771,
+      "step": 35390
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5568245942416272,
+      "learning_rate": 1.4143158443219617e-06,
+      "loss": 0.8295,
+      "step": 35391
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.097656822321057,
+      "learning_rate": 1.413924658919712e-06,
+      "loss": 0.7223,
+      "step": 35392
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9004348416675676,
+      "learning_rate": 1.4135335235083092e-06,
+      "loss": 0.8281,
+      "step": 35393
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6727400163955408,
+      "learning_rate": 1.4131424380900327e-06,
+      "loss": 0.9271,
+      "step": 35394
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6693060907453323,
+      "learning_rate": 1.412751402667164e-06,
+      "loss": 0.916,
+      "step": 35395
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6269099059244791,
+      "learning_rate": 1.4123604172419714e-06,
+      "loss": 0.8577,
+      "step": 35396
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6345936823540985,
+      "learning_rate": 1.4119694818167351e-06,
+      "loss": 0.7895,
+      "step": 35397
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8447817421024304,
+      "learning_rate": 1.4115785963937302e-06,
+      "loss": 0.8018,
+      "step": 35398
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.627706046917281,
+      "learning_rate": 1.4111877609752367e-06,
+      "loss": 0.8494,
+      "step": 35399
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.699238707025119,
+      "learning_rate": 1.4107969755635277e-06,
+      "loss": 0.8478,
+      "step": 35400
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7971890908082965,
+      "learning_rate": 1.4104062401608753e-06,
+      "loss": 0.6984,
+      "step": 35401
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6470175764756143,
+      "learning_rate": 1.4100155547695592e-06,
+      "loss": 0.8184,
+      "step": 35402
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5137829711799373,
+      "learning_rate": 1.4096249193918509e-06,
+      "loss": 0.8209,
+      "step": 35403
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6715786750465254,
+      "learning_rate": 1.4092343340300273e-06,
+      "loss": 0.8167,
+      "step": 35404
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5226170505581276,
+      "learning_rate": 1.408843798686359e-06,
+      "loss": 0.8365,
+      "step": 35405
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.44748487212268,
+      "learning_rate": 1.408453313363124e-06,
+      "loss": 0.7469,
+      "step": 35406
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5895338263468397,
+      "learning_rate": 1.4080628780625927e-06,
+      "loss": 0.783,
+      "step": 35407
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9060802846774951,
+      "learning_rate": 1.407672492787041e-06,
+      "loss": 0.7273,
+      "step": 35408
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.754375186624267,
+      "learning_rate": 1.4072821575387407e-06,
+      "loss": 0.769,
+      "step": 35409
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.028173210535953,
+      "learning_rate": 1.406891872319961e-06,
+      "loss": 0.8431,
+      "step": 35410
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6837286784973275,
+      "learning_rate": 1.4065016371329775e-06,
+      "loss": 0.8211,
+      "step": 35411
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7608847461546457,
+      "learning_rate": 1.4061114519800655e-06,
+      "loss": 0.8137,
+      "step": 35412
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4925603665000449,
+      "learning_rate": 1.405721316863492e-06,
+      "loss": 0.8064,
+      "step": 35413
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6723744571096573,
+      "learning_rate": 1.4053312317855284e-06,
+      "loss": 0.8179,
+      "step": 35414
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7059320660353097,
+      "learning_rate": 1.4049411967484472e-06,
+      "loss": 0.8222,
+      "step": 35415
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6669568758259858,
+      "learning_rate": 1.4045512117545224e-06,
+      "loss": 0.8088,
+      "step": 35416
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5443543995337399,
+      "learning_rate": 1.4041612768060208e-06,
+      "loss": 0.937,
+      "step": 35417
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5468239672342357,
+      "learning_rate": 1.4037713919052121e-06,
+      "loss": 0.8912,
+      "step": 35418
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6060342307838122,
+      "learning_rate": 1.4033815570543673e-06,
+      "loss": 0.8147,
+      "step": 35419
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8082214340150214,
+      "learning_rate": 1.4029917722557597e-06,
+      "loss": 0.7232,
+      "step": 35420
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3297536316070837,
+      "learning_rate": 1.4026020375116545e-06,
+      "loss": 0.7599,
+      "step": 35421
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5973067476989706,
+      "learning_rate": 1.4022123528243204e-06,
+      "loss": 0.9039,
+      "step": 35422
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5050426782866426,
+      "learning_rate": 1.4018227181960308e-06,
+      "loss": 0.7004,
+      "step": 35423
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7526149839446735,
+      "learning_rate": 1.4014331336290487e-06,
+      "loss": 0.7148,
+      "step": 35424
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6517059689371991,
+      "learning_rate": 1.4010435991256477e-06,
+      "loss": 0.773,
+      "step": 35425
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6270479070228399,
+      "learning_rate": 1.4006541146880914e-06,
+      "loss": 0.8719,
+      "step": 35426
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5344989297517286,
+      "learning_rate": 1.4002646803186514e-06,
+      "loss": 0.9089,
+      "step": 35427
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5678699068727577,
+      "learning_rate": 1.3998752960195905e-06,
+      "loss": 0.8215,
+      "step": 35428
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6701160914886515,
+      "learning_rate": 1.39948596179318e-06,
+      "loss": 0.9075,
+      "step": 35429
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6726137790362847,
+      "learning_rate": 1.3990966776416859e-06,
+      "loss": 0.7577,
+      "step": 35430
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7867734333064154,
+      "learning_rate": 1.398707443567372e-06,
+      "loss": 0.9041,
+      "step": 35431
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7757936217014789,
+      "learning_rate": 1.3983182595725054e-06,
+      "loss": 0.689,
+      "step": 35432
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.779614085852157,
+      "learning_rate": 1.3979291256593564e-06,
+      "loss": 0.7678,
+      "step": 35433
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.608184032178308,
+      "learning_rate": 1.3975400418301866e-06,
+      "loss": 0.7892,
+      "step": 35434
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.157191998292082,
+      "learning_rate": 1.3971510080872608e-06,
+      "loss": 0.7934,
+      "step": 35435
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7714946491154864,
+      "learning_rate": 1.396762024432845e-06,
+      "loss": 0.9424,
+      "step": 35436
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4680094756575752,
+      "learning_rate": 1.3963730908692065e-06,
+      "loss": 0.854,
+      "step": 35437
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.207718747272585,
+      "learning_rate": 1.3959842073986085e-06,
+      "loss": 0.9375,
+      "step": 35438
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.770319803747664,
+      "learning_rate": 1.3955953740233107e-06,
+      "loss": 0.9484,
+      "step": 35439
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6328784260761944,
+      "learning_rate": 1.3952065907455825e-06,
+      "loss": 1.0303,
+      "step": 35440
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6991230095002352,
+      "learning_rate": 1.3948178575676862e-06,
+      "loss": 0.8287,
+      "step": 35441
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7399599546156083,
+      "learning_rate": 1.394429174491886e-06,
+      "loss": 0.6681,
+      "step": 35442
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9209887453834802,
+      "learning_rate": 1.3940405415204417e-06,
+      "loss": 0.9777,
+      "step": 35443
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7164876404757063,
+      "learning_rate": 1.3936519586556163e-06,
+      "loss": 0.7308,
+      "step": 35444
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7000221862913685,
+      "learning_rate": 1.393263425899678e-06,
+      "loss": 0.8749,
+      "step": 35445
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5059902330126556,
+      "learning_rate": 1.3928749432548837e-06,
+      "loss": 0.877,
+      "step": 35446
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.388952219112882,
+      "learning_rate": 1.3924865107234942e-06,
+      "loss": 0.931,
+      "step": 35447
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.660345885686713,
+      "learning_rate": 1.3920981283077762e-06,
+      "loss": 0.7777,
+      "step": 35448
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.021417174630642,
+      "learning_rate": 1.3917097960099856e-06,
+      "loss": 0.8196,
+      "step": 35449
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5819486126244762,
+      "learning_rate": 1.3913215138323877e-06,
+      "loss": 0.9121,
+      "step": 35450
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.491571237229409,
+      "learning_rate": 1.3909332817772402e-06,
+      "loss": 0.81,
+      "step": 35451
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5050064401999874,
+      "learning_rate": 1.3905450998468062e-06,
+      "loss": 0.892,
+      "step": 35452
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.708551018913534,
+      "learning_rate": 1.3901569680433425e-06,
+      "loss": 0.8775,
+      "step": 35453
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.271974365722867,
+      "learning_rate": 1.3897688863691128e-06,
+      "loss": 0.8461,
+      "step": 35454
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.691081614256209,
+      "learning_rate": 1.3893808548263742e-06,
+      "loss": 0.8112,
+      "step": 35455
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5804992271100626,
+      "learning_rate": 1.388992873417384e-06,
+      "loss": 0.7851,
+      "step": 35456
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5873325215425487,
+      "learning_rate": 1.3886049421444048e-06,
+      "loss": 0.842,
+      "step": 35457
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5958075189381102,
+      "learning_rate": 1.3882170610096946e-06,
+      "loss": 0.7003,
+      "step": 35458
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7574928127099738,
+      "learning_rate": 1.3878292300155115e-06,
+      "loss": 0.8501,
+      "step": 35459
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7738582641237729,
+      "learning_rate": 1.387441449164112e-06,
+      "loss": 0.6806,
+      "step": 35460
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5765980121529934,
+      "learning_rate": 1.3870537184577538e-06,
+      "loss": 0.9374,
+      "step": 35461
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4831564883816388,
+      "learning_rate": 1.3866660378986985e-06,
+      "loss": 0.6831,
+      "step": 35462
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7781903011440439,
+      "learning_rate": 1.386278407489201e-06,
+      "loss": 0.7252,
+      "step": 35463
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5015167955389204,
+      "learning_rate": 1.3858908272315153e-06,
+      "loss": 0.7402,
+      "step": 35464
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5179130495756223,
+      "learning_rate": 1.3855032971278991e-06,
+      "loss": 0.8943,
+      "step": 35465
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5093203981682841,
+      "learning_rate": 1.3851158171806134e-06,
+      "loss": 0.8108,
+      "step": 35466
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6634932287968678,
+      "learning_rate": 1.3847283873919103e-06,
+      "loss": 0.8903,
+      "step": 35467
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8635505975734898,
+      "learning_rate": 1.384341007764044e-06,
+      "loss": 0.7893,
+      "step": 35468
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5901589862209615,
+      "learning_rate": 1.3839536782992746e-06,
+      "loss": 0.8796,
+      "step": 35469
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9018078387170059,
+      "learning_rate": 1.3835663989998528e-06,
+      "loss": 0.795,
+      "step": 35470
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6971213624422852,
+      "learning_rate": 1.3831791698680342e-06,
+      "loss": 0.7772,
+      "step": 35471
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.712097002197623,
+      "learning_rate": 1.3827919909060772e-06,
+      "loss": 0.7135,
+      "step": 35472
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9777779215327895,
+      "learning_rate": 1.3824048621162334e-06,
+      "loss": 0.7645,
+      "step": 35473
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8650277257644765,
+      "learning_rate": 1.3820177835007541e-06,
+      "loss": 0.7772,
+      "step": 35474
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6650629362089588,
+      "learning_rate": 1.3816307550618968e-06,
+      "loss": 0.7694,
+      "step": 35475
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3222635468949107,
+      "learning_rate": 1.3812437768019172e-06,
+      "loss": 0.8595,
+      "step": 35476
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7272616061377677,
+      "learning_rate": 1.3808568487230611e-06,
+      "loss": 0.9682,
+      "step": 35477
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6773668864465812,
+      "learning_rate": 1.3804699708275847e-06,
+      "loss": 0.7743,
+      "step": 35478
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.518946395675502,
+      "learning_rate": 1.3800831431177419e-06,
+      "loss": 0.6526,
+      "step": 35479
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9182486791137907,
+      "learning_rate": 1.3796963655957872e-06,
+      "loss": 0.7231,
+      "step": 35480
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6250765364180706,
+      "learning_rate": 1.3793096382639658e-06,
+      "loss": 0.8522,
+      "step": 35481
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.76907668503072,
+      "learning_rate": 1.3789229611245335e-06,
+      "loss": 0.7467,
+      "step": 35482
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.576733893021136,
+      "learning_rate": 1.378536334179743e-06,
+      "loss": 0.8647,
+      "step": 35483
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7868514689472834,
+      "learning_rate": 1.3781497574318402e-06,
+      "loss": 0.7814,
+      "step": 35484
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9995181650074134,
+      "learning_rate": 1.3777632308830824e-06,
+      "loss": 0.8391,
+      "step": 35485
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6126353484116138,
+      "learning_rate": 1.3773767545357141e-06,
+      "loss": 0.8552,
+      "step": 35486
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.435147656408095,
+      "learning_rate": 1.3769903283919905e-06,
+      "loss": 0.7692,
+      "step": 35487
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5366852489412035,
+      "learning_rate": 1.3766039524541575e-06,
+      "loss": 0.796,
+      "step": 35488
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.701502543358083,
+      "learning_rate": 1.3762176267244686e-06,
+      "loss": 0.8143,
+      "step": 35489
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8158832503392757,
+      "learning_rate": 1.3758313512051702e-06,
+      "loss": 0.7766,
+      "step": 35490
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7529119395053375,
+      "learning_rate": 1.3754451258985102e-06,
+      "loss": 0.7331,
+      "step": 35491
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6063463671556355,
+      "learning_rate": 1.375058950806739e-06,
+      "loss": 0.7707,
+      "step": 35492
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7651538427268871,
+      "learning_rate": 1.3746728259321085e-06,
+      "loss": 0.8211,
+      "step": 35493
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7836430221436385,
+      "learning_rate": 1.374286751276862e-06,
+      "loss": 0.7485,
+      "step": 35494
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6566907580468713,
+      "learning_rate": 1.373900726843247e-06,
+      "loss": 0.7812,
+      "step": 35495
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0935092061574894,
+      "learning_rate": 1.3735147526335135e-06,
+      "loss": 0.9234,
+      "step": 35496
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.550846507568028,
+      "learning_rate": 1.37312882864991e-06,
+      "loss": 0.8952,
+      "step": 35497
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6689182568291214,
+      "learning_rate": 1.3727429548946825e-06,
+      "loss": 0.8987,
+      "step": 35498
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7975152076353018,
+      "learning_rate": 1.3723571313700734e-06,
+      "loss": 0.728,
+      "step": 35499
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7699873684864824,
+      "learning_rate": 1.3719713580783334e-06,
+      "loss": 0.7466,
+      "step": 35500
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6509324001024515,
+      "learning_rate": 1.3715856350217095e-06,
+      "loss": 0.7725,
+      "step": 35501
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.544408545629751,
+      "learning_rate": 1.3711999622024463e-06,
+      "loss": 0.8224,
+      "step": 35502
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6183582645059569,
+      "learning_rate": 1.3708143396227857e-06,
+      "loss": 0.8569,
+      "step": 35503
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8073790722747016,
+      "learning_rate": 1.3704287672849793e-06,
+      "loss": 0.7068,
+      "step": 35504
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6869896932322044,
+      "learning_rate": 1.370043245191266e-06,
+      "loss": 0.7626,
+      "step": 35505
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.746167995990665,
+      "learning_rate": 1.3696577733438943e-06,
+      "loss": 0.8041,
+      "step": 35506
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4254003487720448,
+      "learning_rate": 1.3692723517451067e-06,
+      "loss": 0.8282,
+      "step": 35507
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.642782887467167,
+      "learning_rate": 1.3688869803971493e-06,
+      "loss": 0.7663,
+      "step": 35508
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9926370557316133,
+      "learning_rate": 1.3685016593022627e-06,
+      "loss": 0.808,
+      "step": 35509
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7489831515255996,
+      "learning_rate": 1.3681163884626935e-06,
+      "loss": 0.6914,
+      "step": 35510
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.489006883341014,
+      "learning_rate": 1.3677311678806838e-06,
+      "loss": 0.7843,
+      "step": 35511
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.401568046134534,
+      "learning_rate": 1.3673459975584746e-06,
+      "loss": 0.7906,
+      "step": 35512
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.594453797151353,
+      "learning_rate": 1.3669608774983102e-06,
+      "loss": 0.9115,
+      "step": 35513
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.707541701221588,
+      "learning_rate": 1.3665758077024338e-06,
+      "loss": 0.8318,
+      "step": 35514
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.620956337247361,
+      "learning_rate": 1.3661907881730874e-06,
+      "loss": 0.8458,
+      "step": 35515
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8516283800524664,
+      "learning_rate": 1.365805818912509e-06,
+      "loss": 0.8,
+      "step": 35516
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.641451706445098,
+      "learning_rate": 1.3654208999229423e-06,
+      "loss": 0.816,
+      "step": 35517
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6043252209241117,
+      "learning_rate": 1.3650360312066314e-06,
+      "loss": 0.751,
+      "step": 35518
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7383388466213527,
+      "learning_rate": 1.364651212765814e-06,
+      "loss": 0.7735,
+      "step": 35519
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.738232268088011,
+      "learning_rate": 1.3642664446027298e-06,
+      "loss": 0.7992,
+      "step": 35520
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7124166643391243,
+      "learning_rate": 1.3638817267196203e-06,
+      "loss": 0.9513,
+      "step": 35521
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.452755376613697,
+      "learning_rate": 1.3634970591187269e-06,
+      "loss": 0.732,
+      "step": 35522
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7448662501562376,
+      "learning_rate": 1.363112441802289e-06,
+      "loss": 0.7418,
+      "step": 35523
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.63125871170655,
+      "learning_rate": 1.3627278747725415e-06,
+      "loss": 0.8853,
+      "step": 35524
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6219498245023105,
+      "learning_rate": 1.3623433580317303e-06,
+      "loss": 0.7535,
+      "step": 35525
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6084389014877911,
+      "learning_rate": 1.3619588915820869e-06,
+      "loss": 0.9224,
+      "step": 35526
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6034125721547494,
+      "learning_rate": 1.3615744754258575e-06,
+      "loss": 0.8458,
+      "step": 35527
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.619255808496766,
+      "learning_rate": 1.3611901095652736e-06,
+      "loss": 0.7565,
+      "step": 35528
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5595565222681071,
+      "learning_rate": 1.3608057940025775e-06,
+      "loss": 0.7406,
+      "step": 35529
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5800355316945822,
+      "learning_rate": 1.3604215287400036e-06,
+      "loss": 0.7722,
+      "step": 35530
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8244811259340674,
+      "learning_rate": 1.3600373137797917e-06,
+      "loss": 0.8372,
+      "step": 35531
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.631592420730289,
+      "learning_rate": 1.3596531491241793e-06,
+      "loss": 0.7112,
+      "step": 35532
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7317042560085542,
+      "learning_rate": 1.3592690347754e-06,
+      "loss": 0.6974,
+      "step": 35533
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9661432242254473,
+      "learning_rate": 1.358884970735691e-06,
+      "loss": 0.8098,
+      "step": 35534
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.814115747545173,
+      "learning_rate": 1.3585009570072917e-06,
+      "loss": 1.001,
+      "step": 35535
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6114252874131065,
+      "learning_rate": 1.3581169935924355e-06,
+      "loss": 0.9264,
+      "step": 35536
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.580715392757994,
+      "learning_rate": 1.3577330804933563e-06,
+      "loss": 0.7394,
+      "step": 35537
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7792423268304103,
+      "learning_rate": 1.3573492177122916e-06,
+      "loss": 0.8046,
+      "step": 35538
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2260281932122172,
+      "learning_rate": 1.356965405251478e-06,
+      "loss": 0.8545,
+      "step": 35539
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4289730296318188,
+      "learning_rate": 1.3565816431131474e-06,
+      "loss": 0.8184,
+      "step": 35540
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5132996734304556,
+      "learning_rate": 1.3561979312995333e-06,
+      "loss": 0.7851,
+      "step": 35541
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9435979825301535,
+      "learning_rate": 1.3558142698128719e-06,
+      "loss": 0.8581,
+      "step": 35542
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8194828722770273,
+      "learning_rate": 1.355430658655398e-06,
+      "loss": 0.6451,
+      "step": 35543
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6805927035927444,
+      "learning_rate": 1.355047097829344e-06,
+      "loss": 0.7746,
+      "step": 35544
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7950408430462134,
+      "learning_rate": 1.3546635873369396e-06,
+      "loss": 0.7289,
+      "step": 35545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4603001300592624,
+      "learning_rate": 1.354280127180424e-06,
+      "loss": 0.8342,
+      "step": 35546
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.63278181159005,
+      "learning_rate": 1.3538967173620243e-06,
+      "loss": 0.7397,
+      "step": 35547
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5898530084635643,
+      "learning_rate": 1.3535133578839777e-06,
+      "loss": 0.8428,
+      "step": 35548
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5885574968648846,
+      "learning_rate": 1.3531300487485111e-06,
+      "loss": 0.8762,
+      "step": 35549
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8176466196448067,
+      "learning_rate": 1.3527467899578617e-06,
+      "loss": 0.7615,
+      "step": 35550
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.559513736476766,
+      "learning_rate": 1.3523635815142556e-06,
+      "loss": 0.8185,
+      "step": 35551
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.380545065729994,
+      "learning_rate": 1.3519804234199286e-06,
+      "loss": 0.8752,
+      "step": 35552
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7751102441515159,
+      "learning_rate": 1.351597315677109e-06,
+      "loss": 0.757,
+      "step": 35553
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.63910363840714,
+      "learning_rate": 1.351214258288026e-06,
+      "loss": 0.7959,
+      "step": 35554
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8353837022112811,
+      "learning_rate": 1.3508312512549127e-06,
+      "loss": 0.7316,
+      "step": 35555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4495577203324417,
+      "learning_rate": 1.3504482945799991e-06,
+      "loss": 0.8596,
+      "step": 35556
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6911100971680488,
+      "learning_rate": 1.3500653882655146e-06,
+      "loss": 0.7734,
+      "step": 35557
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8325018058737068,
+      "learning_rate": 1.3496825323136852e-06,
+      "loss": 0.909,
+      "step": 35558
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7632722014739891,
+      "learning_rate": 1.349299726726744e-06,
+      "loss": 0.6882,
+      "step": 35559
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5201692430669151,
+      "learning_rate": 1.3489169715069195e-06,
+      "loss": 0.8484,
+      "step": 35560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5339801617115665,
+      "learning_rate": 1.3485342666564394e-06,
+      "loss": 0.8549,
+      "step": 35561
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.71665762981755,
+      "learning_rate": 1.3481516121775295e-06,
+      "loss": 0.8091,
+      "step": 35562
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6058756630675095,
+      "learning_rate": 1.3477690080724203e-06,
+      "loss": 0.965,
+      "step": 35563
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7476513175070761,
+      "learning_rate": 1.3473864543433423e-06,
+      "loss": 0.7214,
+      "step": 35564
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7282429484094164,
+      "learning_rate": 1.3470039509925192e-06,
+      "loss": 0.6978,
+      "step": 35565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7058059500855776,
+      "learning_rate": 1.346621498022176e-06,
+      "loss": 0.8903,
+      "step": 35566
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.906596820763718,
+      "learning_rate": 1.3462390954345416e-06,
+      "loss": 0.8,
+      "step": 35567
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6975202708404122,
+      "learning_rate": 1.3458567432318458e-06,
+      "loss": 0.8448,
+      "step": 35568
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6383783173913893,
+      "learning_rate": 1.3454744414163101e-06,
+      "loss": 0.8597,
+      "step": 35569
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5987413825150807,
+      "learning_rate": 1.3450921899901637e-06,
+      "loss": 0.8269,
+      "step": 35570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7325952335417714,
+      "learning_rate": 1.3447099889556302e-06,
+      "loss": 0.8982,
+      "step": 35571
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4296414885859132,
+      "learning_rate": 1.3443278383149338e-06,
+      "loss": 0.8342,
+      "step": 35572
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7211423966228003,
+      "learning_rate": 1.3439457380703015e-06,
+      "loss": 0.6746,
+      "step": 35573
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3787946868311485,
+      "learning_rate": 1.3435636882239588e-06,
+      "loss": 0.7914,
+      "step": 35574
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.432019443308508,
+      "learning_rate": 1.3431816887781302e-06,
+      "loss": 0.7122,
+      "step": 35575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7372459146225064,
+      "learning_rate": 1.3427997397350356e-06,
+      "loss": 0.731,
+      "step": 35576
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.488652012525969,
+      "learning_rate": 1.3424178410969013e-06,
+      "loss": 0.7746,
+      "step": 35577
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7424665810804782,
+      "learning_rate": 1.3420359928659564e-06,
+      "loss": 0.7091,
+      "step": 35578
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4643001607762465,
+      "learning_rate": 1.341654195044414e-06,
+      "loss": 0.8553,
+      "step": 35579
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7462764071141004,
+      "learning_rate": 1.341272447634503e-06,
+      "loss": 0.724,
+      "step": 35580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5276822443309852,
+      "learning_rate": 1.340890750638444e-06,
+      "loss": 0.8732,
+      "step": 35581
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7546343786700012,
+      "learning_rate": 1.3405091040584628e-06,
+      "loss": 0.7052,
+      "step": 35582
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.689746839047995,
+      "learning_rate": 1.34012750789678e-06,
+      "loss": 0.9543,
+      "step": 35583
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.640272807099489,
+      "learning_rate": 1.339745962155613e-06,
+      "loss": 0.8708,
+      "step": 35584
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7716479316514175,
+      "learning_rate": 1.3393644668371897e-06,
+      "loss": 0.6777,
+      "step": 35585
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7275472059106463,
+      "learning_rate": 1.3389830219437272e-06,
+      "loss": 0.8938,
+      "step": 35586
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4910706064766723,
+      "learning_rate": 1.3386016274774482e-06,
+      "loss": 0.8455,
+      "step": 35587
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7164267629359735,
+      "learning_rate": 1.338220283440571e-06,
+      "loss": 0.7552,
+      "step": 35588
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5929311962932806,
+      "learning_rate": 1.3378389898353206e-06,
+      "loss": 0.9091,
+      "step": 35589
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.369340939006641,
+      "learning_rate": 1.3374577466639105e-06,
+      "loss": 0.7332,
+      "step": 35590
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7245829480618106,
+      "learning_rate": 1.337076553928567e-06,
+      "loss": 0.675,
+      "step": 35591
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7801986134662559,
+      "learning_rate": 1.3366954116315057e-06,
+      "loss": 0.7367,
+      "step": 35592
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4960240821228241,
+      "learning_rate": 1.3363143197749451e-06,
+      "loss": 0.777,
+      "step": 35593
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6726266572298902,
+      "learning_rate": 1.3359332783611045e-06,
+      "loss": 0.8382,
+      "step": 35594
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4320681653292555,
+      "learning_rate": 1.3355522873922068e-06,
+      "loss": 0.8609,
+      "step": 35595
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5526611111783672,
+      "learning_rate": 1.3351713468704663e-06,
+      "loss": 0.7977,
+      "step": 35596
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.498078583985729,
+      "learning_rate": 1.3347904567980984e-06,
+      "loss": 0.8103,
+      "step": 35597
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6660587788328114,
+      "learning_rate": 1.3344096171773246e-06,
+      "loss": 0.7719,
+      "step": 35598
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.652303845723252,
+      "learning_rate": 1.3340288280103641e-06,
+      "loss": 0.8044,
+      "step": 35599
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.911135398005811,
+      "learning_rate": 1.333648089299431e-06,
+      "loss": 0.7231,
+      "step": 35600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5199984447687482,
+      "learning_rate": 1.3332674010467395e-06,
+      "loss": 0.7626,
+      "step": 35601
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4794984732783905,
+      "learning_rate": 1.3328867632545094e-06,
+      "loss": 0.8235,
+      "step": 35602
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5699491469684461,
+      "learning_rate": 1.3325061759249591e-06,
+      "loss": 0.7988,
+      "step": 35603
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4795988978405432,
+      "learning_rate": 1.332125639060302e-06,
+      "loss": 0.9278,
+      "step": 35604
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5409208999962865,
+      "learning_rate": 1.3317451526627512e-06,
+      "loss": 0.8786,
+      "step": 35605
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4995243158961973,
+      "learning_rate": 1.3313647167345266e-06,
+      "loss": 0.7947,
+      "step": 35606
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.72244466710968,
+      "learning_rate": 1.330984331277838e-06,
+      "loss": 0.8857,
+      "step": 35607
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.715532746979391,
+      "learning_rate": 1.330603996294907e-06,
+      "loss": 0.7997,
+      "step": 35608
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.571648077253885,
+      "learning_rate": 1.3302237117879413e-06,
+      "loss": 0.6794,
+      "step": 35609
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5456366003608433,
+      "learning_rate": 1.3298434777591596e-06,
+      "loss": 0.8894,
+      "step": 35610
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7719915920139564,
+      "learning_rate": 1.3294632942107733e-06,
+      "loss": 0.7058,
+      "step": 35611
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9334425195612885,
+      "learning_rate": 1.3290831611449973e-06,
+      "loss": 0.7793,
+      "step": 35612
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5690165524746926,
+      "learning_rate": 1.3287030785640454e-06,
+      "loss": 0.6991,
+      "step": 35613
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6161376659921685,
+      "learning_rate": 1.3283230464701268e-06,
+      "loss": 0.787,
+      "step": 35614
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4296491890084009,
+      "learning_rate": 1.3279430648654567e-06,
+      "loss": 0.8894,
+      "step": 35615
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.464165730123241,
+      "learning_rate": 1.327563133752251e-06,
+      "loss": 0.762,
+      "step": 35616
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8543109651454723,
+      "learning_rate": 1.3271832531327166e-06,
+      "loss": 0.8732,
+      "step": 35617
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2664232512730442,
+      "learning_rate": 1.3268034230090664e-06,
+      "loss": 0.8422,
+      "step": 35618
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3613402665433876,
+      "learning_rate": 1.326423643383512e-06,
+      "loss": 0.8383,
+      "step": 35619
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8618584077453375,
+      "learning_rate": 1.326043914258267e-06,
+      "loss": 0.7997,
+      "step": 35620
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3518252517861693,
+      "learning_rate": 1.325664235635541e-06,
+      "loss": 0.7186,
+      "step": 35621
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.966894800004457,
+      "learning_rate": 1.3252846075175407e-06,
+      "loss": 0.9059,
+      "step": 35622
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0781956383563003,
+      "learning_rate": 1.3249050299064815e-06,
+      "loss": 0.8944,
+      "step": 35623
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6308199630403672,
+      "learning_rate": 1.3245255028045722e-06,
+      "loss": 0.768,
+      "step": 35624
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.98709276792244,
+      "learning_rate": 1.3241460262140237e-06,
+      "loss": 0.7879,
+      "step": 35625
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.768846675881824,
+      "learning_rate": 1.323766600137041e-06,
+      "loss": 0.9169,
+      "step": 35626
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.524652250303493,
+      "learning_rate": 1.3233872245758373e-06,
+      "loss": 0.8366,
+      "step": 35627
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6033663116437906,
+      "learning_rate": 1.323007899532619e-06,
+      "loss": 0.8874,
+      "step": 35628
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6214293310853154,
+      "learning_rate": 1.3226286250095978e-06,
+      "loss": 0.8507,
+      "step": 35629
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9690023696110202,
+      "learning_rate": 1.322249401008977e-06,
+      "loss": 0.7579,
+      "step": 35630
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7740478754978681,
+      "learning_rate": 1.3218702275329708e-06,
+      "loss": 0.874,
+      "step": 35631
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.585381731969283,
+      "learning_rate": 1.321491104583781e-06,
+      "loss": 0.8748,
+      "step": 35632
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.653343532870117,
+      "learning_rate": 1.3211120321636194e-06,
+      "loss": 0.9315,
+      "step": 35633
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7941809715643326,
+      "learning_rate": 1.3207330102746907e-06,
+      "loss": 1.0032,
+      "step": 35634
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0553700339589223,
+      "learning_rate": 1.3203540389192014e-06,
+      "loss": 0.8541,
+      "step": 35635
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5783109884325373,
+      "learning_rate": 1.319975118099358e-06,
+      "loss": 0.8517,
+      "step": 35636
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9635174712939967,
+      "learning_rate": 1.3195962478173695e-06,
+      "loss": 0.9547,
+      "step": 35637
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5888767720957984,
+      "learning_rate": 1.3192174280754411e-06,
+      "loss": 0.8696,
+      "step": 35638
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9231566161321133,
+      "learning_rate": 1.318838658875774e-06,
+      "loss": 0.7557,
+      "step": 35639
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7807250072220313,
+      "learning_rate": 1.3184599402205766e-06,
+      "loss": 0.9094,
+      "step": 35640
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.049506389215925,
+      "learning_rate": 1.318081272112056e-06,
+      "loss": 0.8738,
+      "step": 35641
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7691774308071134,
+      "learning_rate": 1.3177026545524152e-06,
+      "loss": 0.7185,
+      "step": 35642
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7362589074416497,
+      "learning_rate": 1.3173240875438565e-06,
+      "loss": 0.7432,
+      "step": 35643
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8155581256437,
+      "learning_rate": 1.3169455710885847e-06,
+      "loss": 0.7769,
+      "step": 35644
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5172981275424902,
+      "learning_rate": 1.3165671051888073e-06,
+      "loss": 0.7255,
+      "step": 35645
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6425612897731674,
+      "learning_rate": 1.316188689846727e-06,
+      "loss": 0.7711,
+      "step": 35646
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6719315190098112,
+      "learning_rate": 1.3158103250645416e-06,
+      "loss": 0.8585,
+      "step": 35647
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6397408844313743,
+      "learning_rate": 1.3154320108444597e-06,
+      "loss": 0.8275,
+      "step": 35648
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7335244433550895,
+      "learning_rate": 1.3150537471886814e-06,
+      "loss": 0.7251,
+      "step": 35649
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7409787094548248,
+      "learning_rate": 1.314675534099411e-06,
+      "loss": 0.7734,
+      "step": 35650
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1729317183390098,
+      "learning_rate": 1.3142973715788477e-06,
+      "loss": 0.8071,
+      "step": 35651
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7878642609048518,
+      "learning_rate": 1.313919259629196e-06,
+      "loss": 0.793,
+      "step": 35652
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.855227234577425,
+      "learning_rate": 1.3135411982526546e-06,
+      "loss": 0.8774,
+      "step": 35653
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8109287708710202,
+      "learning_rate": 1.3131631874514272e-06,
+      "loss": 0.7628,
+      "step": 35654
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.936295309835491,
+      "learning_rate": 1.3127852272277187e-06,
+      "loss": 0.8755,
+      "step": 35655
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7809905243970765,
+      "learning_rate": 1.3124073175837193e-06,
+      "loss": 0.7057,
+      "step": 35656
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7460087131988917,
+      "learning_rate": 1.3120294585216353e-06,
+      "loss": 0.7015,
+      "step": 35657
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7670565411073194,
+      "learning_rate": 1.311651650043666e-06,
+      "loss": 0.7393,
+      "step": 35658
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.474706903052446,
+      "learning_rate": 1.3112738921520164e-06,
+      "loss": 0.621,
+      "step": 35659
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5426289596220972,
+      "learning_rate": 1.3108961848488767e-06,
+      "loss": 0.8218,
+      "step": 35660
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4566933713176555,
+      "learning_rate": 1.3105185281364507e-06,
+      "loss": 0.8686,
+      "step": 35661
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5289646377900548,
+      "learning_rate": 1.310140922016938e-06,
+      "loss": 0.7729,
+      "step": 35662
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5672507539417941,
+      "learning_rate": 1.309763366492537e-06,
+      "loss": 0.7645,
+      "step": 35663
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1961845783729417,
+      "learning_rate": 1.3093858615654421e-06,
+      "loss": 0.8283,
+      "step": 35664
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5149116643630074,
+      "learning_rate": 1.309008407237854e-06,
+      "loss": 0.8373,
+      "step": 35665
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6459819582647908,
+      "learning_rate": 1.3086310035119732e-06,
+      "loss": 0.8347,
+      "step": 35666
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5702566769597521,
+      "learning_rate": 1.3082536503899922e-06,
+      "loss": 0.8145,
+      "step": 35667
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4306894691305854,
+      "learning_rate": 1.3078763478741118e-06,
+      "loss": 0.7721,
+      "step": 35668
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.236299785674953,
+      "learning_rate": 1.3074990959665279e-06,
+      "loss": 0.7993,
+      "step": 35669
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6431899986432907,
+      "learning_rate": 1.307121894669433e-06,
+      "loss": 0.7562,
+      "step": 35670
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7686478275242254,
+      "learning_rate": 1.3067447439850279e-06,
+      "loss": 0.9273,
+      "step": 35671
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8662027805662624,
+      "learning_rate": 1.306367643915508e-06,
+      "loss": 0.7805,
+      "step": 35672
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5836034327287687,
+      "learning_rate": 1.3059905944630691e-06,
+      "loss": 0.7888,
+      "step": 35673
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6111653618898143,
+      "learning_rate": 1.3056135956299032e-06,
+      "loss": 0.8842,
+      "step": 35674
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5773210067568355,
+      "learning_rate": 1.3052366474182065e-06,
+      "loss": 0.8139,
+      "step": 35675
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5629933335780628,
+      "learning_rate": 1.3048597498301808e-06,
+      "loss": 0.9145,
+      "step": 35676
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4448431968530306,
+      "learning_rate": 1.3044829028680084e-06,
+      "loss": 0.8089,
+      "step": 35677
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.466446104268479,
+      "learning_rate": 1.30410610653389e-06,
+      "loss": 0.8438,
+      "step": 35678
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8230217488228189,
+      "learning_rate": 1.3037293608300195e-06,
+      "loss": 0.8227,
+      "step": 35679
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5139320466231363,
+      "learning_rate": 1.3033526657585915e-06,
+      "loss": 0.9552,
+      "step": 35680
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4340200442051005,
+      "learning_rate": 1.302976021321798e-06,
+      "loss": 0.8035,
+      "step": 35681
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5829559675236347,
+      "learning_rate": 1.302599427521829e-06,
+      "loss": 0.8512,
+      "step": 35682
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7199012473278041,
+      "learning_rate": 1.302222884360882e-06,
+      "loss": 0.8375,
+      "step": 35683
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.624907351735355,
+      "learning_rate": 1.301846391841145e-06,
+      "loss": 0.917,
+      "step": 35684
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.764890230975145,
+      "learning_rate": 1.3014699499648142e-06,
+      "loss": 0.7056,
+      "step": 35685
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6616693853514914,
+      "learning_rate": 1.3010935587340767e-06,
+      "loss": 0.8718,
+      "step": 35686
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6503418419508467,
+      "learning_rate": 1.3007172181511296e-06,
+      "loss": 0.7798,
+      "step": 35687
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3240411337060243,
+      "learning_rate": 1.3003409282181579e-06,
+      "loss": 0.9573,
+      "step": 35688
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7699198686690563,
+      "learning_rate": 1.2999646889373585e-06,
+      "loss": 0.7304,
+      "step": 35689
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5655035769984746,
+      "learning_rate": 1.2995885003109166e-06,
+      "loss": 0.7928,
+      "step": 35690
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4001638699909702,
+      "learning_rate": 1.2992123623410268e-06,
+      "loss": 0.8124,
+      "step": 35691
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7660291975792549,
+      "learning_rate": 1.2988362750298754e-06,
+      "loss": 0.7075,
+      "step": 35692
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7321770374835699,
+      "learning_rate": 1.2984602383796551e-06,
+      "loss": 0.8743,
+      "step": 35693
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5995098668083418,
+      "learning_rate": 1.2980842523925553e-06,
+      "loss": 0.6909,
+      "step": 35694
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.603088648741776,
+      "learning_rate": 1.2977083170707616e-06,
+      "loss": 0.7998,
+      "step": 35695
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5918652312288581,
+      "learning_rate": 1.2973324324164649e-06,
+      "loss": 0.8403,
+      "step": 35696
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7047425634215054,
+      "learning_rate": 1.2969565984318567e-06,
+      "loss": 0.7995,
+      "step": 35697
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8547547912720614,
+      "learning_rate": 1.2965808151191229e-06,
+      "loss": 0.8527,
+      "step": 35698
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.605154044645456,
+      "learning_rate": 1.2962050824804472e-06,
+      "loss": 0.9571,
+      "step": 35699
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8274391271750199,
+      "learning_rate": 1.2958294005180228e-06,
+      "loss": 0.9064,
+      "step": 35700
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.568729113975638,
+      "learning_rate": 1.2954537692340363e-06,
+      "loss": 0.7858,
+      "step": 35701
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5253494402102206,
+      "learning_rate": 1.295078188630674e-06,
+      "loss": 0.7202,
+      "step": 35702
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7852922523965646,
+      "learning_rate": 1.2947026587101208e-06,
+      "loss": 0.9431,
+      "step": 35703
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1072608501871297,
+      "learning_rate": 1.2943271794745649e-06,
+      "loss": 0.7335,
+      "step": 35704
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8364875477074938,
+      "learning_rate": 1.2939517509261934e-06,
+      "loss": 0.8031,
+      "step": 35705
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5509160660179067,
+      "learning_rate": 1.2935763730671923e-06,
+      "loss": 0.8525,
+      "step": 35706
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5504400388538562,
+      "learning_rate": 1.2932010458997434e-06,
+      "loss": 0.8196,
+      "step": 35707
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.917860023569752,
+      "learning_rate": 1.2928257694260359e-06,
+      "loss": 0.7918,
+      "step": 35708
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4817383186014055,
+      "learning_rate": 1.2924505436482527e-06,
+      "loss": 0.8758,
+      "step": 35709
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5079383800640354,
+      "learning_rate": 1.2920753685685805e-06,
+      "loss": 0.9816,
+      "step": 35710
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5577905865195607,
+      "learning_rate": 1.2917002441892002e-06,
+      "loss": 0.8614,
+      "step": 35711
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6498660363780049,
+      "learning_rate": 1.2913251705122998e-06,
+      "loss": 0.9606,
+      "step": 35712
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6317318728009942,
+      "learning_rate": 1.2909501475400598e-06,
+      "loss": 0.8325,
+      "step": 35713
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7652664089440704,
+      "learning_rate": 1.2905751752746676e-06,
+      "loss": 0.879,
+      "step": 35714
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5632022541493256,
+      "learning_rate": 1.2902002537183044e-06,
+      "loss": 0.8654,
+      "step": 35715
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9294939892955214,
+      "learning_rate": 1.28982538287315e-06,
+      "loss": 0.731,
+      "step": 35716
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.534219537012723,
+      "learning_rate": 1.28945056274139e-06,
+      "loss": 0.8006,
+      "step": 35717
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8372302123623703,
+      "learning_rate": 1.2890757933252084e-06,
+      "loss": 0.7656,
+      "step": 35718
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6233684298586903,
+      "learning_rate": 1.2887010746267859e-06,
+      "loss": 0.8923,
+      "step": 35719
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7186363637914283,
+      "learning_rate": 1.2883264066483015e-06,
+      "loss": 0.7042,
+      "step": 35720
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5196938257658876,
+      "learning_rate": 1.287951789391938e-06,
+      "loss": 0.7354,
+      "step": 35721
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5957207252900156,
+      "learning_rate": 1.2875772228598805e-06,
+      "loss": 0.8004,
+      "step": 35722
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5826249615316368,
+      "learning_rate": 1.2872027070543057e-06,
+      "loss": 0.9501,
+      "step": 35723
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0204657409056206,
+      "learning_rate": 1.2868282419773936e-06,
+      "loss": 0.8789,
+      "step": 35724
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6027608738438404,
+      "learning_rate": 1.2864538276313265e-06,
+      "loss": 0.8763,
+      "step": 35725
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6755032232489802,
+      "learning_rate": 1.286079464018285e-06,
+      "loss": 0.8508,
+      "step": 35726
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6695330974528917,
+      "learning_rate": 1.2857051511404484e-06,
+      "loss": 0.8024,
+      "step": 35727
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6513576102994205,
+      "learning_rate": 1.2853308889999927e-06,
+      "loss": 0.6461,
+      "step": 35728
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6695855732545573,
+      "learning_rate": 1.2849566775991029e-06,
+      "loss": 0.8163,
+      "step": 35729
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5568057011639265,
+      "learning_rate": 1.2845825169399506e-06,
+      "loss": 0.8498,
+      "step": 35730
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.47359813153026,
+      "learning_rate": 1.2842084070247217e-06,
+      "loss": 0.7229,
+      "step": 35731
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6393860230986095,
+      "learning_rate": 1.283834347855588e-06,
+      "loss": 0.9172,
+      "step": 35732
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.107943885595552,
+      "learning_rate": 1.2834603394347323e-06,
+      "loss": 0.8611,
+      "step": 35733
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7996388562270502,
+      "learning_rate": 1.2830863817643279e-06,
+      "loss": 0.7408,
+      "step": 35734
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0310385592415963,
+      "learning_rate": 1.2827124748465569e-06,
+      "loss": 0.7873,
+      "step": 35735
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6133025855435663,
+      "learning_rate": 1.282338618683594e-06,
+      "loss": 0.8836,
+      "step": 35736
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7052521370016855,
+      "learning_rate": 1.281964813277612e-06,
+      "loss": 0.8524,
+      "step": 35737
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5411178543588793,
+      "learning_rate": 1.2815910586307923e-06,
+      "loss": 0.81,
+      "step": 35738
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6297063971137458,
+      "learning_rate": 1.281217354745311e-06,
+      "loss": 0.8419,
+      "step": 35739
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4437218206446367,
+      "learning_rate": 1.2808437016233433e-06,
+      "loss": 0.8497,
+      "step": 35740
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.833310077932388,
+      "learning_rate": 1.2804700992670616e-06,
+      "loss": 0.8586,
+      "step": 35741
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6645175494667124,
+      "learning_rate": 1.2800965476786431e-06,
+      "loss": 0.8067,
+      "step": 35742
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7862013111972044,
+      "learning_rate": 1.279723046860266e-06,
+      "loss": 0.7765,
+      "step": 35743
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5525043816168582,
+      "learning_rate": 1.2793495968141023e-06,
+      "loss": 0.7872,
+      "step": 35744
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6596241110413266,
+      "learning_rate": 1.2789761975423231e-06,
+      "loss": 0.8869,
+      "step": 35745
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6503727326819904,
+      "learning_rate": 1.2786028490471058e-06,
+      "loss": 0.7451,
+      "step": 35746
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8132069071837371,
+      "learning_rate": 1.2782295513306264e-06,
+      "loss": 0.7609,
+      "step": 35747
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4976702715298376,
+      "learning_rate": 1.2778563043950553e-06,
+      "loss": 0.8275,
+      "step": 35748
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6203479302530488,
+      "learning_rate": 1.2774831082425644e-06,
+      "loss": 0.8917,
+      "step": 35749
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5799841833532204,
+      "learning_rate": 1.2771099628753304e-06,
+      "loss": 0.7533,
+      "step": 35750
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5375319710395012,
+      "learning_rate": 1.2767368682955216e-06,
+      "loss": 0.708,
+      "step": 35751
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.581688925325074,
+      "learning_rate": 1.2763638245053134e-06,
+      "loss": 0.6965,
+      "step": 35752
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4755218066564708,
+      "learning_rate": 1.275990831506878e-06,
+      "loss": 0.8099,
+      "step": 35753
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.690171648836501,
+      "learning_rate": 1.2756178893023862e-06,
+      "loss": 0.8172,
+      "step": 35754
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7273075381313614,
+      "learning_rate": 1.2752449978940062e-06,
+      "loss": 0.9119,
+      "step": 35755
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5002558396374461,
+      "learning_rate": 1.274872157283913e-06,
+      "loss": 0.8781,
+      "step": 35756
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5335466000153657,
+      "learning_rate": 1.2744993674742812e-06,
+      "loss": 0.8311,
+      "step": 35757
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7845994827427827,
+      "learning_rate": 1.2741266284672704e-06,
+      "loss": 0.8366,
+      "step": 35758
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6625752658988806,
+      "learning_rate": 1.273753940265059e-06,
+      "loss": 0.7879,
+      "step": 35759
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9784036068141069,
+      "learning_rate": 1.2733813028698138e-06,
+      "loss": 1.0288,
+      "step": 35760
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9399274047938704,
+      "learning_rate": 1.273008716283709e-06,
+      "loss": 0.7081,
+      "step": 35761
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.557209213447683,
+      "learning_rate": 1.2726361805089072e-06,
+      "loss": 0.7455,
+      "step": 35762
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7003978054248077,
+      "learning_rate": 1.272263695547581e-06,
+      "loss": 0.7787,
+      "step": 35763
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6655682016649274,
+      "learning_rate": 1.2718912614018996e-06,
+      "loss": 0.9203,
+      "step": 35764
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3763985631001316,
+      "learning_rate": 1.2715188780740295e-06,
+      "loss": 0.9556,
+      "step": 35765
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.195851869457485,
+      "learning_rate": 1.271146545566141e-06,
+      "loss": 0.7485,
+      "step": 35766
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8684071407554355,
+      "learning_rate": 1.2707742638803987e-06,
+      "loss": 0.7768,
+      "step": 35767
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.470019814274826,
+      "learning_rate": 1.2704020330189759e-06,
+      "loss": 0.8629,
+      "step": 35768
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.548523792455915,
+      "learning_rate": 1.2700298529840327e-06,
+      "loss": 0.9607,
+      "step": 35769
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.838254010408392,
+      "learning_rate": 1.2696577237777418e-06,
+      "loss": 0.9457,
+      "step": 35770
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5233763711537185,
+      "learning_rate": 1.2692856454022684e-06,
+      "loss": 0.8598,
+      "step": 35771
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7680370765873652,
+      "learning_rate": 1.268913617859775e-06,
+      "loss": 0.8062,
+      "step": 35772
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6890456315111306,
+      "learning_rate": 1.268541641152431e-06,
+      "loss": 0.8406,
+      "step": 35773
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5188518598114562,
+      "learning_rate": 1.2681697152824045e-06,
+      "loss": 0.7576,
+      "step": 35774
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6259535768533826,
+      "learning_rate": 1.2677978402518576e-06,
+      "loss": 0.7497,
+      "step": 35775
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7225064866848248,
+      "learning_rate": 1.2674260160629547e-06,
+      "loss": 0.8853,
+      "step": 35776
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3755480212416964,
+      "learning_rate": 1.2670542427178623e-06,
+      "loss": 0.8009,
+      "step": 35777
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5307849613810098,
+      "learning_rate": 1.266682520218747e-06,
+      "loss": 0.7817,
+      "step": 35778
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6463863028708154,
+      "learning_rate": 1.266310848567771e-06,
+      "loss": 0.8394,
+      "step": 35779
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4397743746459275,
+      "learning_rate": 1.2659392277670956e-06,
+      "loss": 0.8165,
+      "step": 35780
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.599892804690698,
+      "learning_rate": 1.265567657818888e-06,
+      "loss": 0.8209,
+      "step": 35781
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4146376985702962,
+      "learning_rate": 1.265196138725312e-06,
+      "loss": 0.845,
+      "step": 35782
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.331501075052715,
+      "learning_rate": 1.2648246704885303e-06,
+      "loss": 0.7523,
+      "step": 35783
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.722973366533524,
+      "learning_rate": 1.2644532531107024e-06,
+      "loss": 0.7851,
+      "step": 35784
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.449871806745659,
+      "learning_rate": 1.2640818865939952e-06,
+      "loss": 0.8002,
+      "step": 35785
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.675564166364995,
+      "learning_rate": 1.2637105709405662e-06,
+      "loss": 0.6926,
+      "step": 35786
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.726945596020389,
+      "learning_rate": 1.2633393061525834e-06,
+      "loss": 0.7079,
+      "step": 35787
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.575925857689508,
+      "learning_rate": 1.262968092232203e-06,
+      "loss": 0.9147,
+      "step": 35788
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0339707606880872,
+      "learning_rate": 1.2625969291815899e-06,
+      "loss": 0.8175,
+      "step": 35789
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4757283822148135,
+      "learning_rate": 1.2622258170029023e-06,
+      "loss": 0.937,
+      "step": 35790
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6911385865700577,
+      "learning_rate": 1.2618547556983042e-06,
+      "loss": 0.8773,
+      "step": 35791
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6031466259442315,
+      "learning_rate": 1.2614837452699536e-06,
+      "loss": 0.8323,
+      "step": 35792
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.663954545210885,
+      "learning_rate": 1.26111278572001e-06,
+      "loss": 0.7953,
+      "step": 35793
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8159333744790507,
+      "learning_rate": 1.2607418770506342e-06,
+      "loss": 0.7243,
+      "step": 35794
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5822974158088599,
+      "learning_rate": 1.2603710192639874e-06,
+      "loss": 0.9122,
+      "step": 35795
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6838982308940884,
+      "learning_rate": 1.2600002123622279e-06,
+      "loss": 0.9791,
+      "step": 35796
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7555234199666305,
+      "learning_rate": 1.259629456347512e-06,
+      "loss": 0.9311,
+      "step": 35797
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.658884329247299,
+      "learning_rate": 1.2592587512220011e-06,
+      "loss": 0.7633,
+      "step": 35798
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4568094181029532,
+      "learning_rate": 1.2588880969878547e-06,
+      "loss": 0.7461,
+      "step": 35799
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.599295409999182,
+      "learning_rate": 1.2585174936472288e-06,
+      "loss": 0.8685,
+      "step": 35800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4287686474011927,
+      "learning_rate": 1.2581469412022806e-06,
+      "loss": 0.896,
+      "step": 35801
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.479232892874715,
+      "learning_rate": 1.2577764396551673e-06,
+      "loss": 0.887,
+      "step": 35802
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7703194026388093,
+      "learning_rate": 1.2574059890080502e-06,
+      "loss": 0.6957,
+      "step": 35803
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.746930735002354,
+      "learning_rate": 1.2570355892630825e-06,
+      "loss": 0.9715,
+      "step": 35804
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6325032199946798,
+      "learning_rate": 1.25666524042242e-06,
+      "loss": 0.8543,
+      "step": 35805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6137772994748993,
+      "learning_rate": 1.2562949424882232e-06,
+      "loss": 0.7667,
+      "step": 35806
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8354519745628948,
+      "learning_rate": 1.2559246954626424e-06,
+      "loss": 0.853,
+      "step": 35807
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5629271335834904,
+      "learning_rate": 1.2555544993478386e-06,
+      "loss": 0.7427,
+      "step": 35808
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6021182968660812,
+      "learning_rate": 1.255184354145963e-06,
+      "loss": 0.7568,
+      "step": 35809
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5906281954585908,
+      "learning_rate": 1.254814259859175e-06,
+      "loss": 0.9071,
+      "step": 35810
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.517606538199138,
+      "learning_rate": 1.2544442164896254e-06,
+      "loss": 0.8722,
+      "step": 35811
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6747533970741904,
+      "learning_rate": 1.254074224039471e-06,
+      "loss": 0.8648,
+      "step": 35812
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.74248592431922,
+      "learning_rate": 1.2537042825108647e-06,
+      "loss": 0.7869,
+      "step": 35813
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6452629142351354,
+      "learning_rate": 1.2533343919059626e-06,
+      "loss": 0.7893,
+      "step": 35814
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1806282160040347,
+      "learning_rate": 1.2529645522269151e-06,
+      "loss": 0.8912,
+      "step": 35815
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0761260935581753,
+      "learning_rate": 1.2525947634758784e-06,
+      "loss": 0.7801,
+      "step": 35816
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5829774221679216,
+      "learning_rate": 1.252225025655005e-06,
+      "loss": 0.7627,
+      "step": 35817
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5090121051897731,
+      "learning_rate": 1.2518553387664456e-06,
+      "loss": 0.7645,
+      "step": 35818
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4589128945929526,
+      "learning_rate": 1.251485702812353e-06,
+      "loss": 0.8074,
+      "step": 35819
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7489984707964652,
+      "learning_rate": 1.2511161177948827e-06,
+      "loss": 0.7573,
+      "step": 35820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5910668305043723,
+      "learning_rate": 1.2507465837161848e-06,
+      "loss": 0.8898,
+      "step": 35821
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6792175467402726,
+      "learning_rate": 1.2503771005784071e-06,
+      "loss": 0.8294,
+      "step": 35822
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5355628266113408,
+      "learning_rate": 1.2500076683837048e-06,
+      "loss": 0.8949,
+      "step": 35823
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7224568387724999,
+      "learning_rate": 1.2496382871342295e-06,
+      "loss": 0.8839,
+      "step": 35824
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5902218593289321,
+      "learning_rate": 1.2492689568321314e-06,
+      "loss": 0.7646,
+      "step": 35825
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7622192947592763,
+      "learning_rate": 1.248899677479557e-06,
+      "loss": 0.6992,
+      "step": 35826
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7301706476841077,
+      "learning_rate": 1.2485304490786587e-06,
+      "loss": 0.758,
+      "step": 35827
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7341118685581922,
+      "learning_rate": 1.2481612716315905e-06,
+      "loss": 0.8935,
+      "step": 35828
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9690088388464353,
+      "learning_rate": 1.2477921451404973e-06,
+      "loss": 0.9184,
+      "step": 35829
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6910738011403512,
+      "learning_rate": 1.2474230696075262e-06,
+      "loss": 0.9232,
+      "step": 35830
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4469478473837865,
+      "learning_rate": 1.2470540450348323e-06,
+      "loss": 0.7413,
+      "step": 35831
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5635623853602123,
+      "learning_rate": 1.2466850714245581e-06,
+      "loss": 0.7577,
+      "step": 35832
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5891749133699693,
+      "learning_rate": 1.2463161487788578e-06,
+      "loss": 0.8204,
+      "step": 35833
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.276343171137933,
+      "learning_rate": 1.2459472770998727e-06,
+      "loss": 0.7887,
+      "step": 35834
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4370024275731588,
+      "learning_rate": 1.2455784563897577e-06,
+      "loss": 0.8329,
+      "step": 35835
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5233439079177937,
+      "learning_rate": 1.2452096866506536e-06,
+      "loss": 0.909,
+      "step": 35836
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9632438577883466,
+      "learning_rate": 1.2448409678847117e-06,
+      "loss": 0.8554,
+      "step": 35837
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5270746044929344,
+      "learning_rate": 1.2444723000940785e-06,
+      "loss": 0.6752,
+      "step": 35838
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8196348125837591,
+      "learning_rate": 1.2441036832808973e-06,
+      "loss": 0.7656,
+      "step": 35839
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7734175185298076,
+      "learning_rate": 1.2437351174473167e-06,
+      "loss": 0.9048,
+      "step": 35840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7109302456706883,
+      "learning_rate": 1.2433666025954815e-06,
+      "loss": 0.7561,
+      "step": 35841
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.717672452674541,
+      "learning_rate": 1.2429981387275435e-06,
+      "loss": 0.8034,
+      "step": 35842
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.818094380649587,
+      "learning_rate": 1.2426297258456388e-06,
+      "loss": 0.7992,
+      "step": 35843
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1203637404886555,
+      "learning_rate": 1.2422613639519154e-06,
+      "loss": 0.8153,
+      "step": 35844
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5309272927343993,
+      "learning_rate": 1.2418930530485219e-06,
+      "loss": 0.7674,
+      "step": 35845
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.423720679219886,
+      "learning_rate": 1.2415247931375996e-06,
+      "loss": 0.8046,
+      "step": 35846
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7984638967869957,
+      "learning_rate": 1.2411565842212902e-06,
+      "loss": 0.7199,
+      "step": 35847
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8341703916309955,
+      "learning_rate": 1.24078842630174e-06,
+      "loss": 0.9506,
+      "step": 35848
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6388433773605608,
+      "learning_rate": 1.240420319381096e-06,
+      "loss": 0.9692,
+      "step": 35849
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5899752313112057,
+      "learning_rate": 1.2400522634614953e-06,
+      "loss": 0.7833,
+      "step": 35850
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7445170676558295,
+      "learning_rate": 1.2396842585450852e-06,
+      "loss": 0.7262,
+      "step": 35851
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6424138224440963,
+      "learning_rate": 1.2393163046340073e-06,
+      "loss": 0.7105,
+      "step": 35852
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5988779319551074,
+      "learning_rate": 1.238948401730402e-06,
+      "loss": 0.8138,
+      "step": 35853
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.579406687090863,
+      "learning_rate": 1.2385805498364124e-06,
+      "loss": 0.8968,
+      "step": 35854
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7137563252525772,
+      "learning_rate": 1.238212748954183e-06,
+      "loss": 0.8669,
+      "step": 35855
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.768568122171053,
+      "learning_rate": 1.2378449990858522e-06,
+      "loss": 0.7164,
+      "step": 35856
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6908205136899344,
+      "learning_rate": 1.2374773002335605e-06,
+      "loss": 0.7631,
+      "step": 35857
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9765919303693749,
+      "learning_rate": 1.2371096523994498e-06,
+      "loss": 0.8654,
+      "step": 35858
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5332056502986569,
+      "learning_rate": 1.2367420555856646e-06,
+      "loss": 0.8025,
+      "step": 35859
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4549986069108318,
+      "learning_rate": 1.236374509794338e-06,
+      "loss": 0.7422,
+      "step": 35860
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.452088648761995,
+      "learning_rate": 1.2360070150276137e-06,
+      "loss": 0.7443,
+      "step": 35861
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.896955600822458,
+      "learning_rate": 1.2356395712876312e-06,
+      "loss": 0.7848,
+      "step": 35862
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8290316422503163,
+      "learning_rate": 1.235272178576531e-06,
+      "loss": 0.6739,
+      "step": 35863
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4913964907308201,
+      "learning_rate": 1.2349048368964522e-06,
+      "loss": 0.8911,
+      "step": 35864
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.436748289061442,
+      "learning_rate": 1.234537546249529e-06,
+      "loss": 0.9417,
+      "step": 35865
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0277711609192535,
+      "learning_rate": 1.2341703066379073e-06,
+      "loss": 0.7348,
+      "step": 35866
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5513572841592873,
+      "learning_rate": 1.2338031180637179e-06,
+      "loss": 0.7869,
+      "step": 35867
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.546161670015788,
+      "learning_rate": 1.233435980529104e-06,
+      "loss": 0.876,
+      "step": 35868
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2528922705479255,
+      "learning_rate": 1.2330688940362002e-06,
+      "loss": 0.8736,
+      "step": 35869
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4666692241541897,
+      "learning_rate": 1.2327018585871463e-06,
+      "loss": 0.844,
+      "step": 35870
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7135471806347184,
+      "learning_rate": 1.2323348741840756e-06,
+      "loss": 0.8117,
+      "step": 35871
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7162196380494108,
+      "learning_rate": 1.2319679408291296e-06,
+      "loss": 0.7265,
+      "step": 35872
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8683879600069058,
+      "learning_rate": 1.231601058524442e-06,
+      "loss": 0.7455,
+      "step": 35873
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5445424868056925,
+      "learning_rate": 1.2312342272721467e-06,
+      "loss": 0.7639,
+      "step": 35874
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5818754041403993,
+      "learning_rate": 1.2308674470743819e-06,
+      "loss": 0.8656,
+      "step": 35875
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.901413575572964,
+      "learning_rate": 1.2305007179332851e-06,
+      "loss": 0.8537,
+      "step": 35876
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5610917652802818,
+      "learning_rate": 1.2301340398509899e-06,
+      "loss": 0.9158,
+      "step": 35877
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4267093136078544,
+      "learning_rate": 1.2297674128296288e-06,
+      "loss": 0.8007,
+      "step": 35878
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7089253763505325,
+      "learning_rate": 1.2294008368713385e-06,
+      "loss": 0.7042,
+      "step": 35879
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7763230558663969,
+      "learning_rate": 1.2290343119782545e-06,
+      "loss": 0.7142,
+      "step": 35880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.627927455857739,
+      "learning_rate": 1.2286678381525108e-06,
+      "loss": 0.7525,
+      "step": 35881
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6022635695977667,
+      "learning_rate": 1.2283014153962368e-06,
+      "loss": 0.769,
+      "step": 35882
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4791874977836652,
+      "learning_rate": 1.227935043711569e-06,
+      "loss": 0.8198,
+      "step": 35883
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3257742385364149,
+      "learning_rate": 1.2275687231006428e-06,
+      "loss": 0.7922,
+      "step": 35884
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7556582593831223,
+      "learning_rate": 1.227202453565589e-06,
+      "loss": 0.8772,
+      "step": 35885
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4406936474976837,
+      "learning_rate": 1.2268362351085383e-06,
+      "loss": 0.8848,
+      "step": 35886
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4944387590960262,
+      "learning_rate": 1.2264700677316266e-06,
+      "loss": 0.7997,
+      "step": 35887
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5188783453272336,
+      "learning_rate": 1.226103951436981e-06,
+      "loss": 0.953,
+      "step": 35888
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4254143533511454,
+      "learning_rate": 1.2257378862267388e-06,
+      "loss": 0.7805,
+      "step": 35889
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8730366381695123,
+      "learning_rate": 1.225371872103026e-06,
+      "loss": 0.8843,
+      "step": 35890
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5308249562797736,
+      "learning_rate": 1.2250059090679779e-06,
+      "loss": 0.8385,
+      "step": 35891
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5573311175311202,
+      "learning_rate": 1.2246399971237221e-06,
+      "loss": 0.8731,
+      "step": 35892
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0455762374729516,
+      "learning_rate": 1.224274136272392e-06,
+      "loss": 0.7988,
+      "step": 35893
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6828616559202494,
+      "learning_rate": 1.2239083265161166e-06,
+      "loss": 0.837,
+      "step": 35894
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6185573067929595,
+      "learning_rate": 1.2235425678570234e-06,
+      "loss": 0.7634,
+      "step": 35895
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7386117797403666,
+      "learning_rate": 1.2231768602972439e-06,
+      "loss": 0.8506,
+      "step": 35896
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7957437513560992,
+      "learning_rate": 1.2228112038389085e-06,
+      "loss": 0.7954,
+      "step": 35897
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9497848268754305,
+      "learning_rate": 1.2224455984841466e-06,
+      "loss": 0.8336,
+      "step": 35898
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6633463249962634,
+      "learning_rate": 1.2220800442350823e-06,
+      "loss": 0.8079,
+      "step": 35899
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.742425222369663,
+      "learning_rate": 1.2217145410938468e-06,
+      "loss": 0.8365,
+      "step": 35900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5673060158643004,
+      "learning_rate": 1.2213490890625711e-06,
+      "loss": 0.7843,
+      "step": 35901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5174509927098878,
+      "learning_rate": 1.2209836881433802e-06,
+      "loss": 0.7999,
+      "step": 35902
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.802273708219121,
+      "learning_rate": 1.2206183383383997e-06,
+      "loss": 0.7163,
+      "step": 35903
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4853671225225609,
+      "learning_rate": 1.2202530396497592e-06,
+      "loss": 0.7917,
+      "step": 35904
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7870719300253086,
+      "learning_rate": 1.2198877920795859e-06,
+      "loss": 0.8024,
+      "step": 35905
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5088690823720536,
+      "learning_rate": 1.2195225956300073e-06,
+      "loss": 0.815,
+      "step": 35906
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5014846981875822,
+      "learning_rate": 1.2191574503031445e-06,
+      "loss": 0.7228,
+      "step": 35907
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0575372753560033,
+      "learning_rate": 1.2187923561011295e-06,
+      "loss": 0.7572,
+      "step": 35908
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7162425382723785,
+      "learning_rate": 1.2184273130260837e-06,
+      "loss": 0.6926,
+      "step": 35909
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6654906030836587,
+      "learning_rate": 1.2180623210801357e-06,
+      "loss": 1.0042,
+      "step": 35910
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7805265456878328,
+      "learning_rate": 1.2176973802654068e-06,
+      "loss": 0.8462,
+      "step": 35911
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5205005674750018,
+      "learning_rate": 1.2173324905840278e-06,
+      "loss": 0.8021,
+      "step": 35912
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6727801070788435,
+      "learning_rate": 1.2169676520381168e-06,
+      "loss": 0.7717,
+      "step": 35913
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.472367141090244,
+      "learning_rate": 1.2166028646298022e-06,
+      "loss": 0.7763,
+      "step": 35914
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1453979854315786,
+      "learning_rate": 1.2162381283612067e-06,
+      "loss": 0.8956,
+      "step": 35915
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.930337350948767,
+      "learning_rate": 1.2158734432344521e-06,
+      "loss": 0.8705,
+      "step": 35916
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7078284144746467,
+      "learning_rate": 1.2155088092516632e-06,
+      "loss": 0.7256,
+      "step": 35917
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.751296052261681,
+      "learning_rate": 1.215144226414965e-06,
+      "loss": 0.8041,
+      "step": 35918
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7691544982089724,
+      "learning_rate": 1.214779694726479e-06,
+      "loss": 0.9269,
+      "step": 35919
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3915664826957506,
+      "learning_rate": 1.2144152141883247e-06,
+      "loss": 0.7984,
+      "step": 35920
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7084704038357144,
+      "learning_rate": 1.2140507848026261e-06,
+      "loss": 0.8334,
+      "step": 35921
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6276342728949702,
+      "learning_rate": 1.213686406571507e-06,
+      "loss": 0.8497,
+      "step": 35922
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.624404000287472,
+      "learning_rate": 1.2133220794970879e-06,
+      "loss": 0.8283,
+      "step": 35923
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5863732965199373,
+      "learning_rate": 1.2129578035814881e-06,
+      "loss": 0.7996,
+      "step": 35924
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.669715345937702,
+      "learning_rate": 1.2125935788268296e-06,
+      "loss": 0.8629,
+      "step": 35925
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7317951036103324,
+      "learning_rate": 1.2122294052352347e-06,
+      "loss": 0.7027,
+      "step": 35926
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5209673064241604,
+      "learning_rate": 1.211865282808823e-06,
+      "loss": 0.8674,
+      "step": 35927
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5600791132462495,
+      "learning_rate": 1.2115012115497115e-06,
+      "loss": 0.9572,
+      "step": 35928
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.448625751858975,
+      "learning_rate": 1.2111371914600245e-06,
+      "loss": 0.8738,
+      "step": 35929
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.733985105300419,
+      "learning_rate": 1.2107732225418766e-06,
+      "loss": 0.779,
+      "step": 35930
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4825056839862816,
+      "learning_rate": 1.210409304797392e-06,
+      "loss": 0.8145,
+      "step": 35931
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7643810716322981,
+      "learning_rate": 1.2100454382286853e-06,
+      "loss": 0.7284,
+      "step": 35932
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.500725232918982,
+      "learning_rate": 1.2096816228378794e-06,
+      "loss": 0.7816,
+      "step": 35933
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6216330898919464,
+      "learning_rate": 1.209317858627087e-06,
+      "loss": 0.8328,
+      "step": 35934
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.56320762003941,
+      "learning_rate": 1.2089541455984321e-06,
+      "loss": 0.7879,
+      "step": 35935
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5710461750077391,
+      "learning_rate": 1.2085904837540275e-06,
+      "loss": 0.7311,
+      "step": 35936
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7030260884830326,
+      "learning_rate": 1.2082268730959935e-06,
+      "loss": 0.8418,
+      "step": 35937
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4374317449064673,
+      "learning_rate": 1.207863313626445e-06,
+      "loss": 0.8831,
+      "step": 35938
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9131375391982075,
+      "learning_rate": 1.2074998053474984e-06,
+      "loss": 0.8466,
+      "step": 35939
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6074248628520487,
+      "learning_rate": 1.2071363482612775e-06,
+      "loss": 0.7918,
+      "step": 35940
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.756060720319017,
+      "learning_rate": 1.2067729423698871e-06,
+      "loss": 0.7336,
+      "step": 35941
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5237412719521453,
+      "learning_rate": 1.206409587675449e-06,
+      "loss": 0.8921,
+      "step": 35942
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5490488823537933,
+      "learning_rate": 1.2060462841800779e-06,
+      "loss": 0.7903,
+      "step": 35943
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.673761873386997,
+      "learning_rate": 1.2056830318858947e-06,
+      "loss": 0.711,
+      "step": 35944
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8623118514206736,
+      "learning_rate": 1.2053198307950042e-06,
+      "loss": 0.8327,
+      "step": 35945
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6223468110292651,
+      "learning_rate": 1.2049566809095259e-06,
+      "loss": 0.7302,
+      "step": 35946
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9705546492578754,
+      "learning_rate": 1.2045935822315768e-06,
+      "loss": 0.7353,
+      "step": 35947
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7267531039868782,
+      "learning_rate": 1.2042305347632655e-06,
+      "loss": 0.7133,
+      "step": 35948
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6033102558426786,
+      "learning_rate": 1.2038675385067112e-06,
+      "loss": 0.866,
+      "step": 35949
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7483924842821242,
+      "learning_rate": 1.2035045934640233e-06,
+      "loss": 0.7373,
+      "step": 35950
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5354913850480212,
+      "learning_rate": 1.2031416996373179e-06,
+      "loss": 0.8256,
+      "step": 35951
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.445080047235658,
+      "learning_rate": 1.2027788570287047e-06,
+      "loss": 0.9922,
+      "step": 35952
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5434342827745182,
+      "learning_rate": 1.2024160656403006e-06,
+      "loss": 0.8692,
+      "step": 35953
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7779294544925095,
+      "learning_rate": 1.202053325474215e-06,
+      "loss": 0.8737,
+      "step": 35954
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.715480804031948,
+      "learning_rate": 1.2016906365325586e-06,
+      "loss": 0.7491,
+      "step": 35955
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4617711403181795,
+      "learning_rate": 1.2013279988174454e-06,
+      "loss": 0.8134,
+      "step": 35956
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.578638893414353,
+      "learning_rate": 1.2009654123309878e-06,
+      "loss": 0.9183,
+      "step": 35957
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1039328765515553,
+      "learning_rate": 1.2006028770752942e-06,
+      "loss": 0.8022,
+      "step": 35958
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6567145313253513,
+      "learning_rate": 1.2002403930524752e-06,
+      "loss": 0.8135,
+      "step": 35959
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.713801717372318,
+      "learning_rate": 1.1998779602646438e-06,
+      "loss": 0.8542,
+      "step": 35960
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7646953787514564,
+      "learning_rate": 1.199515578713909e-06,
+      "loss": 0.7015,
+      "step": 35961
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6549040417008503,
+      "learning_rate": 1.1991532484023816e-06,
+      "loss": 0.7445,
+      "step": 35962
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.442769899264862,
+      "learning_rate": 1.1987909693321687e-06,
+      "loss": 0.8692,
+      "step": 35963
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.678246109846165,
+      "learning_rate": 1.1984287415053807e-06,
+      "loss": 0.8415,
+      "step": 35964
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.721987908948511,
+      "learning_rate": 1.1980665649241297e-06,
+      "loss": 0.927,
+      "step": 35965
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6731206858691692,
+      "learning_rate": 1.1977044395905214e-06,
+      "loss": 0.9868,
+      "step": 35966
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.626127548965105,
+      "learning_rate": 1.1973423655066619e-06,
+      "loss": 0.7754,
+      "step": 35967
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7175803432027326,
+      "learning_rate": 1.1969803426746651e-06,
+      "loss": 0.7195,
+      "step": 35968
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.789399540554973,
+      "learning_rate": 1.1966183710966329e-06,
+      "loss": 0.9349,
+      "step": 35969
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6866414534321343,
+      "learning_rate": 1.1962564507746776e-06,
+      "loss": 0.9258,
+      "step": 35970
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5259489776543964,
+      "learning_rate": 1.1958945817109035e-06,
+      "loss": 0.7654,
+      "step": 35971
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4845391807766142,
+      "learning_rate": 1.1955327639074198e-06,
+      "loss": 0.7841,
+      "step": 35972
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9624845123642058,
+      "learning_rate": 1.1951709973663294e-06,
+      "loss": 0.8179,
+      "step": 35973
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.656755504583977,
+      "learning_rate": 1.194809282089744e-06,
+      "loss": 0.9028,
+      "step": 35974
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5079331437955459,
+      "learning_rate": 1.194447618079766e-06,
+      "loss": 0.9098,
+      "step": 35975
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7549741474748286,
+      "learning_rate": 1.1940860053384994e-06,
+      "loss": 0.8405,
+      "step": 35976
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5365721323283328,
+      "learning_rate": 1.1937244438680529e-06,
+      "loss": 0.8343,
+      "step": 35977
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6763569900948267,
+      "learning_rate": 1.1933629336705322e-06,
+      "loss": 0.7329,
+      "step": 35978
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5921069763908802,
+      "learning_rate": 1.19300147474804e-06,
+      "loss": 0.7522,
+      "step": 35979
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.70653585920572,
+      "learning_rate": 1.1926400671026805e-06,
+      "loss": 0.7268,
+      "step": 35980
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4969415947796783,
+      "learning_rate": 1.1922787107365585e-06,
+      "loss": 0.6981,
+      "step": 35981
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6583774496816657,
+      "learning_rate": 1.19191740565178e-06,
+      "loss": 0.8287,
+      "step": 35982
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4971615238990597,
+      "learning_rate": 1.1915561518504482e-06,
+      "loss": 0.8252,
+      "step": 35983
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.58462506183592,
+      "learning_rate": 1.191194949334662e-06,
+      "loss": 0.8366,
+      "step": 35984
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6727246449044024,
+      "learning_rate": 1.1908337981065276e-06,
+      "loss": 0.9051,
+      "step": 35985
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4294495211280325,
+      "learning_rate": 1.1904726981681502e-06,
+      "loss": 0.8908,
+      "step": 35986
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7770488613098847,
+      "learning_rate": 1.1901116495216303e-06,
+      "loss": 0.8239,
+      "step": 35987
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6102347636857717,
+      "learning_rate": 1.189750652169066e-06,
+      "loss": 0.7881,
+      "step": 35988
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5075835890135836,
+      "learning_rate": 1.189389706112566e-06,
+      "loss": 0.8894,
+      "step": 35989
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8348931337105494,
+      "learning_rate": 1.189028811354226e-06,
+      "loss": 0.8854,
+      "step": 35990
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.702712978988661,
+      "learning_rate": 1.188667967896152e-06,
+      "loss": 0.8131,
+      "step": 35991
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.976449525178539,
+      "learning_rate": 1.1883071757404408e-06,
+      "loss": 0.9375,
+      "step": 35992
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5537092821496326,
+      "learning_rate": 1.1879464348891967e-06,
+      "loss": 0.6888,
+      "step": 35993
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.417238672223317,
+      "learning_rate": 1.1875857453445162e-06,
+      "loss": 0.8753,
+      "step": 35994
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4844170419329517,
+      "learning_rate": 1.1872251071085038e-06,
+      "loss": 0.7296,
+      "step": 35995
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4906061444519796,
+      "learning_rate": 1.1868645201832563e-06,
+      "loss": 0.7434,
+      "step": 35996
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7868233109615874,
+      "learning_rate": 1.1865039845708725e-06,
+      "loss": 0.7685,
+      "step": 35997
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2057332829663965,
+      "learning_rate": 1.1861435002734523e-06,
+      "loss": 0.8949,
+      "step": 35998
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4409737810464136,
+      "learning_rate": 1.1857830672930982e-06,
+      "loss": 0.872,
+      "step": 35999
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.419738103176738,
+      "learning_rate": 1.185422685631904e-06,
+      "loss": 0.8675,
+      "step": 36000
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7566647040948227,
+      "learning_rate": 1.1850623552919683e-06,
+      "loss": 0.7615,
+      "step": 36001
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.532866181434204,
+      "learning_rate": 1.1847020762753903e-06,
+      "loss": 0.86,
+      "step": 36002
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7263166807839798,
+      "learning_rate": 1.1843418485842696e-06,
+      "loss": 0.8879,
+      "step": 36003
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6523572780501183,
+      "learning_rate": 1.183981672220702e-06,
+      "loss": 0.8493,
+      "step": 36004
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.119056014891217,
+      "learning_rate": 1.1836215471867819e-06,
+      "loss": 0.7623,
+      "step": 36005
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.588735831091023,
+      "learning_rate": 1.1832614734846082e-06,
+      "loss": 0.8961,
+      "step": 36006
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.768540543369847,
+      "learning_rate": 1.1829014511162794e-06,
+      "loss": 0.8764,
+      "step": 36007
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.727114185542334,
+      "learning_rate": 1.1825414800838908e-06,
+      "loss": 0.6907,
+      "step": 36008
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7596073500803608,
+      "learning_rate": 1.1821815603895336e-06,
+      "loss": 0.6701,
+      "step": 36009
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6577702633101212,
+      "learning_rate": 1.1818216920353109e-06,
+      "loss": 0.8949,
+      "step": 36010
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4847021029147913,
+      "learning_rate": 1.181461875023311e-06,
+      "loss": 0.7505,
+      "step": 36011
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6301865701770295,
+      "learning_rate": 1.1811021093556341e-06,
+      "loss": 0.8243,
+      "step": 36012
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.681653522641092,
+      "learning_rate": 1.180742395034371e-06,
+      "loss": 0.8706,
+      "step": 36013
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8013098169471513,
+      "learning_rate": 1.180382732061619e-06,
+      "loss": 0.8831,
+      "step": 36014
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7990412981629184,
+      "learning_rate": 1.1800231204394696e-06,
+      "loss": 0.6941,
+      "step": 36015
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8509512400157913,
+      "learning_rate": 1.1796635601700201e-06,
+      "loss": 0.9081,
+      "step": 36016
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.694965578209171,
+      "learning_rate": 1.1793040512553623e-06,
+      "loss": 0.8954,
+      "step": 36017
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5899654451761518,
+      "learning_rate": 1.1789445936975862e-06,
+      "loss": 0.9084,
+      "step": 36018
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.148521297917602,
+      "learning_rate": 1.1785851874987874e-06,
+      "loss": 0.9195,
+      "step": 36019
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2420040367443335,
+      "learning_rate": 1.1782258326610606e-06,
+      "loss": 0.868,
+      "step": 36020
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.582341112186919,
+      "learning_rate": 1.1778665291864955e-06,
+      "loss": 0.8341,
+      "step": 36021
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6166287659829373,
+      "learning_rate": 1.1775072770771833e-06,
+      "loss": 0.883,
+      "step": 36022
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4729627706775825,
+      "learning_rate": 1.177148076335216e-06,
+      "loss": 0.7425,
+      "step": 36023
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1720693737606758,
+      "learning_rate": 1.1767889269626887e-06,
+      "loss": 0.8274,
+      "step": 36024
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4248658170636723,
+      "learning_rate": 1.1764298289616893e-06,
+      "loss": 0.7496,
+      "step": 36025
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8117198423282409,
+      "learning_rate": 1.1760707823343065e-06,
+      "loss": 0.703,
+      "step": 36026
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5430541544151364,
+      "learning_rate": 1.1757117870826328e-06,
+      "loss": 0.892,
+      "step": 36027
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.423343136662556,
+      "learning_rate": 1.1753528432087624e-06,
+      "loss": 0.818,
+      "step": 36028
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6679716298824716,
+      "learning_rate": 1.174993950714781e-06,
+      "loss": 0.8137,
+      "step": 36029
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6627645790719316,
+      "learning_rate": 1.1746351096027763e-06,
+      "loss": 0.7502,
+      "step": 36030
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6439166462710766,
+      "learning_rate": 1.174276319874842e-06,
+      "loss": 0.8214,
+      "step": 36031
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5489754935925455,
+      "learning_rate": 1.173917581533064e-06,
+      "loss": 0.8052,
+      "step": 36032
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4983083589514137,
+      "learning_rate": 1.1735588945795307e-06,
+      "loss": 0.7995,
+      "step": 36033
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7651242215280724,
+      "learning_rate": 1.1732002590163349e-06,
+      "loss": 0.7069,
+      "step": 36034
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6963269785227038,
+      "learning_rate": 1.1728416748455619e-06,
+      "loss": 0.9347,
+      "step": 36035
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7501490479662749,
+      "learning_rate": 1.1724831420692973e-06,
+      "loss": 0.7296,
+      "step": 36036
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.709840391894655,
+      "learning_rate": 1.1721246606896297e-06,
+      "loss": 0.7082,
+      "step": 36037
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5281642458033853,
+      "learning_rate": 1.171766230708652e-06,
+      "loss": 0.8764,
+      "step": 36038
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6037899153653115,
+      "learning_rate": 1.1714078521284422e-06,
+      "loss": 0.8987,
+      "step": 36039
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8243466597157634,
+      "learning_rate": 1.1710495249510911e-06,
+      "loss": 0.8643,
+      "step": 36040
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6337827932407947,
+      "learning_rate": 1.1706912491786837e-06,
+      "loss": 0.8953,
+      "step": 36041
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5018490897441974,
+      "learning_rate": 1.1703330248133128e-06,
+      "loss": 0.9353,
+      "step": 36042
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6090806263718525,
+      "learning_rate": 1.169974851857053e-06,
+      "loss": 0.7581,
+      "step": 36043
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9705760996865598,
+      "learning_rate": 1.1696167303119964e-06,
+      "loss": 0.824,
+      "step": 36044
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.651268921455138,
+      "learning_rate": 1.169258660180228e-06,
+      "loss": 0.9008,
+      "step": 36045
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.251840011179552,
+      "learning_rate": 1.1689006414638294e-06,
+      "loss": 0.7271,
+      "step": 36046
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.618204787913225,
+      "learning_rate": 1.1685426741648897e-06,
+      "loss": 0.7856,
+      "step": 36047
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4631752698268885,
+      "learning_rate": 1.168184758285489e-06,
+      "loss": 0.7023,
+      "step": 36048
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6443611937572513,
+      "learning_rate": 1.1678268938277137e-06,
+      "loss": 0.6627,
+      "step": 36049
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6617388519960308,
+      "learning_rate": 1.167469080793645e-06,
+      "loss": 0.9009,
+      "step": 36050
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4784958557512067,
+      "learning_rate": 1.1671113191853688e-06,
+      "loss": 0.8349,
+      "step": 36051
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5380298277293407,
+      "learning_rate": 1.166753609004968e-06,
+      "loss": 0.8087,
+      "step": 36052
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6739543175686558,
+      "learning_rate": 1.1663959502545218e-06,
+      "loss": 0.8594,
+      "step": 36053
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7823072393095512,
+      "learning_rate": 1.1660383429361155e-06,
+      "loss": 0.7229,
+      "step": 36054
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4719964573955553,
+      "learning_rate": 1.1656807870518316e-06,
+      "loss": 0.7451,
+      "step": 36055
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4651098671227787,
+      "learning_rate": 1.165323282603752e-06,
+      "loss": 0.8397,
+      "step": 36056
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5773415224010345,
+      "learning_rate": 1.1649658295939548e-06,
+      "loss": 0.878,
+      "step": 36057
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2600743591511137,
+      "learning_rate": 1.164608428024524e-06,
+      "loss": 0.6644,
+      "step": 36058
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6221570265432887,
+      "learning_rate": 1.1642510778975425e-06,
+      "loss": 0.9346,
+      "step": 36059
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6910241235793526,
+      "learning_rate": 1.1638937792150883e-06,
+      "loss": 0.9536,
+      "step": 36060
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5241524622254907,
+      "learning_rate": 1.1635365319792402e-06,
+      "loss": 0.7783,
+      "step": 36061
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5319135651939146,
+      "learning_rate": 1.1631793361920795e-06,
+      "loss": 0.8713,
+      "step": 36062
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6702982766932644,
+      "learning_rate": 1.1628221918556882e-06,
+      "loss": 0.7941,
+      "step": 36063
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9873593192974002,
+      "learning_rate": 1.1624650989721443e-06,
+      "loss": 0.824,
+      "step": 36064
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6401994812755825,
+      "learning_rate": 1.162108057543524e-06,
+      "loss": 0.8612,
+      "step": 36065
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9687761911001096,
+      "learning_rate": 1.1617510675719091e-06,
+      "loss": 0.77,
+      "step": 36066
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.795893303163469,
+      "learning_rate": 1.16139412905938e-06,
+      "loss": 0.7634,
+      "step": 36067
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6472974273971694,
+      "learning_rate": 1.1610372420080118e-06,
+      "loss": 0.9473,
+      "step": 36068
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5689323683211513,
+      "learning_rate": 1.1606804064198817e-06,
+      "loss": 0.7847,
+      "step": 36069
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6937008038331614,
+      "learning_rate": 1.1603236222970704e-06,
+      "loss": 0.8379,
+      "step": 36070
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7565199660022965,
+      "learning_rate": 1.1599668896416516e-06,
+      "loss": 0.6389,
+      "step": 36071
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.767555234711981,
+      "learning_rate": 1.159610208455707e-06,
+      "loss": 0.7419,
+      "step": 36072
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.813326156973262,
+      "learning_rate": 1.1592535787413074e-06,
+      "loss": 0.7283,
+      "step": 36073
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3662385439325027,
+      "learning_rate": 1.158897000500535e-06,
+      "loss": 0.8108,
+      "step": 36074
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7455238131451583,
+      "learning_rate": 1.158540473735461e-06,
+      "loss": 0.8155,
+      "step": 36075
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5941470527725272,
+      "learning_rate": 1.1581839984481657e-06,
+      "loss": 1.0075,
+      "step": 36076
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6507092781243113,
+      "learning_rate": 1.157827574640723e-06,
+      "loss": 0.8978,
+      "step": 36077
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4291034590815552,
+      "learning_rate": 1.1574712023152058e-06,
+      "loss": 0.806,
+      "step": 36078
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.525642831545718,
+      "learning_rate": 1.1571148814736898e-06,
+      "loss": 0.8093,
+      "step": 36079
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.759681517760132,
+      "learning_rate": 1.1567586121182529e-06,
+      "loss": 0.8067,
+      "step": 36080
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6515445343292745,
+      "learning_rate": 1.1564023942509662e-06,
+      "loss": 0.7068,
+      "step": 36081
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6507653278872045,
+      "learning_rate": 1.1560462278739037e-06,
+      "loss": 0.8768,
+      "step": 36082
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.487981114710897,
+      "learning_rate": 1.1556901129891396e-06,
+      "loss": 0.7623,
+      "step": 36083
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5709319765814647,
+      "learning_rate": 1.1553340495987497e-06,
+      "loss": 0.9039,
+      "step": 36084
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.59182379709336,
+      "learning_rate": 1.154978037704806e-06,
+      "loss": 0.8867,
+      "step": 36085
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9499649102382675,
+      "learning_rate": 1.1546220773093774e-06,
+      "loss": 0.8575,
+      "step": 36086
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7024345230542335,
+      "learning_rate": 1.1542661684145396e-06,
+      "loss": 0.8858,
+      "step": 36087
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4653976947123932,
+      "learning_rate": 1.153910311022367e-06,
+      "loss": 0.7099,
+      "step": 36088
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0994723049321995,
+      "learning_rate": 1.1535545051349284e-06,
+      "loss": 0.7974,
+      "step": 36089
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1455052415632045,
+      "learning_rate": 1.1531987507542953e-06,
+      "loss": 0.8832,
+      "step": 36090
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8445412439070713,
+      "learning_rate": 1.1528430478825414e-06,
+      "loss": 0.6873,
+      "step": 36091
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7329603191684032,
+      "learning_rate": 1.1524873965217342e-06,
+      "loss": 0.6594,
+      "step": 36092
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7560466479224087,
+      "learning_rate": 1.1521317966739486e-06,
+      "loss": 0.6592,
+      "step": 36093
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4560858146436426,
+      "learning_rate": 1.1517762483412509e-06,
+      "loss": 0.7412,
+      "step": 36094
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.639957689081237,
+      "learning_rate": 1.1514207515257147e-06,
+      "loss": 0.8039,
+      "step": 36095
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8164879349189516,
+      "learning_rate": 1.1510653062294075e-06,
+      "loss": 0.8058,
+      "step": 36096
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6498139396496732,
+      "learning_rate": 1.1507099124544008e-06,
+      "loss": 0.8908,
+      "step": 36097
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.852941877827986,
+      "learning_rate": 1.1503545702027619e-06,
+      "loss": 0.8511,
+      "step": 36098
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8739128010802839,
+      "learning_rate": 1.1499992794765592e-06,
+      "loss": 0.7492,
+      "step": 36099
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.571349166485334,
+      "learning_rate": 1.149644040277863e-06,
+      "loss": 0.8302,
+      "step": 36100
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5355376524795055,
+      "learning_rate": 1.149288852608743e-06,
+      "loss": 0.8241,
+      "step": 36101
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5683874117573868,
+      "learning_rate": 1.1489337164712644e-06,
+      "loss": 0.8753,
+      "step": 36102
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7382686672632612,
+      "learning_rate": 1.1485786318674951e-06,
+      "loss": 0.8353,
+      "step": 36103
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6627408580054768,
+      "learning_rate": 1.1482235987995027e-06,
+      "loss": 0.7621,
+      "step": 36104
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7046299013909985,
+      "learning_rate": 1.1478686172693576e-06,
+      "loss": 0.9744,
+      "step": 36105
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5596210982455005,
+      "learning_rate": 1.1475136872791226e-06,
+      "loss": 0.842,
+      "step": 36106
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8791487080804519,
+      "learning_rate": 1.1471588088308649e-06,
+      "loss": 0.7126,
+      "step": 36107
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6093816134798111,
+      "learning_rate": 1.1468039819266508e-06,
+      "loss": 0.8887,
+      "step": 36108
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5099185684777001,
+      "learning_rate": 1.1464492065685483e-06,
+      "loss": 0.8316,
+      "step": 36109
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6212920084545581,
+      "learning_rate": 1.1460944827586228e-06,
+      "loss": 0.8253,
+      "step": 36110
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5622995167315048,
+      "learning_rate": 1.1457398104989347e-06,
+      "loss": 0.8841,
+      "step": 36111
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6710437164085492,
+      "learning_rate": 1.1453851897915557e-06,
+      "loss": 0.8806,
+      "step": 36112
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.215604897068542,
+      "learning_rate": 1.1450306206385453e-06,
+      "loss": 0.8392,
+      "step": 36113
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7515377115622264,
+      "learning_rate": 1.1446761030419718e-06,
+      "loss": 0.7507,
+      "step": 36114
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5292442692280919,
+      "learning_rate": 1.1443216370038968e-06,
+      "loss": 0.7636,
+      "step": 36115
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.629658118803454,
+      "learning_rate": 1.1439672225263853e-06,
+      "loss": 0.7884,
+      "step": 36116
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4735853472218632,
+      "learning_rate": 1.1436128596114992e-06,
+      "loss": 0.8704,
+      "step": 36117
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.879045722325479,
+      "learning_rate": 1.1432585482613045e-06,
+      "loss": 0.9221,
+      "step": 36118
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5557730090911097,
+      "learning_rate": 1.142904288477863e-06,
+      "loss": 0.7346,
+      "step": 36119
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7760669059501586,
+      "learning_rate": 1.142550080263234e-06,
+      "loss": 0.7145,
+      "step": 36120
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5749252890359862,
+      "learning_rate": 1.1421959236194835e-06,
+      "loss": 0.8116,
+      "step": 36121
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.913965644836988,
+      "learning_rate": 1.1418418185486725e-06,
+      "loss": 0.9302,
+      "step": 36122
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.836920821800045,
+      "learning_rate": 1.1414877650528667e-06,
+      "loss": 0.9786,
+      "step": 36123
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5394420293111322,
+      "learning_rate": 1.1411337631341202e-06,
+      "loss": 0.7423,
+      "step": 36124
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.541497309151508,
+      "learning_rate": 1.140779812794497e-06,
+      "loss": 0.8814,
+      "step": 36125
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5949281738000027,
+      "learning_rate": 1.1404259140360606e-06,
+      "loss": 0.8197,
+      "step": 36126
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5958546995440928,
+      "learning_rate": 1.140072066860869e-06,
+      "loss": 0.8406,
+      "step": 36127
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.71390619374512,
+      "learning_rate": 1.139718271270982e-06,
+      "loss": 0.8836,
+      "step": 36128
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.505466541095713,
+      "learning_rate": 1.1393645272684595e-06,
+      "loss": 0.9101,
+      "step": 36129
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5573269281190631,
+      "learning_rate": 1.139010834855364e-06,
+      "loss": 0.8701,
+      "step": 36130
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.393732393829399,
+      "learning_rate": 1.138657194033751e-06,
+      "loss": 0.8138,
+      "step": 36131
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8636148351143624,
+      "learning_rate": 1.1383036048056828e-06,
+      "loss": 0.7964,
+      "step": 36132
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6227629942782118,
+      "learning_rate": 1.137950067173217e-06,
+      "loss": 0.8422,
+      "step": 36133
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9835497126629302,
+      "learning_rate": 1.1375965811384093e-06,
+      "loss": 0.9704,
+      "step": 36134
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.464466033119904,
+      "learning_rate": 1.1372431467033208e-06,
+      "loss": 0.9358,
+      "step": 36135
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5118374971946937,
+      "learning_rate": 1.1368897638700105e-06,
+      "loss": 0.7382,
+      "step": 36136
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0682105297404814,
+      "learning_rate": 1.1365364326405325e-06,
+      "loss": 0.7758,
+      "step": 36137
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6222118974056292,
+      "learning_rate": 1.1361831530169453e-06,
+      "loss": 0.8227,
+      "step": 36138
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8157255195385092,
+      "learning_rate": 1.1358299250013049e-06,
+      "loss": 0.6797,
+      "step": 36139
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6778246985332648,
+      "learning_rate": 1.1354767485956741e-06,
+      "loss": 0.8704,
+      "step": 36140
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5712185663866725,
+      "learning_rate": 1.135123623802099e-06,
+      "loss": 0.7601,
+      "step": 36141
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6515233816103285,
+      "learning_rate": 1.1347705506226415e-06,
+      "loss": 0.68,
+      "step": 36142
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.769115609019467,
+      "learning_rate": 1.1344175290593563e-06,
+      "loss": 0.6771,
+      "step": 36143
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8557169038022742,
+      "learning_rate": 1.1340645591143007e-06,
+      "loss": 0.7501,
+      "step": 36144
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.064570584664468,
+      "learning_rate": 1.1337116407895289e-06,
+      "loss": 0.6951,
+      "step": 36145
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6144330684251396,
+      "learning_rate": 1.1333587740870921e-06,
+      "loss": 0.9172,
+      "step": 36146
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5727688913155518,
+      "learning_rate": 1.1330059590090502e-06,
+      "loss": 0.9292,
+      "step": 36147
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8039297953176356,
+      "learning_rate": 1.1326531955574526e-06,
+      "loss": 0.8145,
+      "step": 36148
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5901813559229934,
+      "learning_rate": 1.1323004837343576e-06,
+      "loss": 0.8825,
+      "step": 36149
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6430745810819751,
+      "learning_rate": 1.1319478235418146e-06,
+      "loss": 0.8576,
+      "step": 36150
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4245554579878446,
+      "learning_rate": 1.131595214981881e-06,
+      "loss": 0.7449,
+      "step": 36151
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7967110467305344,
+      "learning_rate": 1.131242658056606e-06,
+      "loss": 0.7209,
+      "step": 36152
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5976484445778207,
+      "learning_rate": 1.130890152768046e-06,
+      "loss": 0.696,
+      "step": 36153
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.892979010002226,
+      "learning_rate": 1.1305376991182516e-06,
+      "loss": 0.8722,
+      "step": 36154
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6948434821198795,
+      "learning_rate": 1.130185297109272e-06,
+      "loss": 0.9063,
+      "step": 36155
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8212583799575709,
+      "learning_rate": 1.1298329467431624e-06,
+      "loss": 0.7463,
+      "step": 36156
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5463026607418904,
+      "learning_rate": 1.1294806480219744e-06,
+      "loss": 0.8534,
+      "step": 36157
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.546171751372806,
+      "learning_rate": 1.1291284009477598e-06,
+      "loss": 0.7321,
+      "step": 36158
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0382420188379133,
+      "learning_rate": 1.128776205522566e-06,
+      "loss": 0.8107,
+      "step": 36159
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6176225398247654,
+      "learning_rate": 1.1284240617484455e-06,
+      "loss": 0.877,
+      "step": 36160
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9183859040077176,
+      "learning_rate": 1.12807196962745e-06,
+      "loss": 0.7819,
+      "step": 36161
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9353867982875794,
+      "learning_rate": 1.1277199291616291e-06,
+      "loss": 0.7686,
+      "step": 36162
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5770966140637193,
+      "learning_rate": 1.127367940353029e-06,
+      "loss": 0.8841,
+      "step": 36163
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.65342189812037,
+      "learning_rate": 1.1270160032037024e-06,
+      "loss": 1.0171,
+      "step": 36164
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7808834790317056,
+      "learning_rate": 1.1266641177157e-06,
+      "loss": 0.8577,
+      "step": 36165
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7657383926328023,
+      "learning_rate": 1.126312283891069e-06,
+      "loss": 0.6963,
+      "step": 36166
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5922652214148256,
+      "learning_rate": 1.1259605017318543e-06,
+      "loss": 0.9267,
+      "step": 36167
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6040069056037254,
+      "learning_rate": 1.1256087712401087e-06,
+      "loss": 0.8403,
+      "step": 36168
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1423791952742324,
+      "learning_rate": 1.1252570924178763e-06,
+      "loss": 0.7449,
+      "step": 36169
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5253191950215905,
+      "learning_rate": 1.1249054652672097e-06,
+      "loss": 0.7534,
+      "step": 36170
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.528509182369417,
+      "learning_rate": 1.124553889790151e-06,
+      "loss": 0.748,
+      "step": 36171
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4720044221150819,
+      "learning_rate": 1.1242023659887513e-06,
+      "loss": 0.7777,
+      "step": 36172
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6666368260830604,
+      "learning_rate": 1.1238508938650527e-06,
+      "loss": 0.7211,
+      "step": 36173
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.550868789036216,
+      "learning_rate": 1.123499473421108e-06,
+      "loss": 0.8041,
+      "step": 36174
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.54320413874109,
+      "learning_rate": 1.1231481046589587e-06,
+      "loss": 0.9499,
+      "step": 36175
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4510355874333685,
+      "learning_rate": 1.1227967875806489e-06,
+      "loss": 0.8008,
+      "step": 36176
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6458520117012578,
+      "learning_rate": 1.122445522188228e-06,
+      "loss": 0.7713,
+      "step": 36177
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5663292108486742,
+      "learning_rate": 1.122094308483741e-06,
+      "loss": 0.8768,
+      "step": 36178
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5871468631289103,
+      "learning_rate": 1.121743146469232e-06,
+      "loss": 0.7713,
+      "step": 36179
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5166151422617566,
+      "learning_rate": 1.1213920361467422e-06,
+      "loss": 0.902,
+      "step": 36180
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.787626338797401,
+      "learning_rate": 1.1210409775183207e-06,
+      "loss": 0.7709,
+      "step": 36181
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.91596883927881,
+      "learning_rate": 1.1206899705860097e-06,
+      "loss": 0.761,
+      "step": 36182
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5918483754322195,
+      "learning_rate": 1.1203390153518546e-06,
+      "loss": 0.7972,
+      "step": 36183
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5054748631553303,
+      "learning_rate": 1.1199881118178934e-06,
+      "loss": 0.9418,
+      "step": 36184
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6400640558666277,
+      "learning_rate": 1.1196372599861738e-06,
+      "loss": 0.9082,
+      "step": 36185
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4635926324215196,
+      "learning_rate": 1.1192864598587395e-06,
+      "loss": 0.8329,
+      "step": 36186
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7292111640396834,
+      "learning_rate": 1.1189357114376308e-06,
+      "loss": 0.6705,
+      "step": 36187
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5207319718403332,
+      "learning_rate": 1.1185850147248879e-06,
+      "loss": 0.8139,
+      "step": 36188
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9680415243024807,
+      "learning_rate": 1.118234369722555e-06,
+      "loss": 0.8413,
+      "step": 36189
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4878220300536447,
+      "learning_rate": 1.1178837764326766e-06,
+      "loss": 0.8114,
+      "step": 36190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6802065645411652,
+      "learning_rate": 1.1175332348572898e-06,
+      "loss": 0.8287,
+      "step": 36191
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3156444929777316,
+      "learning_rate": 1.1171827449984341e-06,
+      "loss": 0.8302,
+      "step": 36192
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6524309846960175,
+      "learning_rate": 1.1168323068581555e-06,
+      "loss": 0.8765,
+      "step": 36193
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5185043926511603,
+      "learning_rate": 1.1164819204384891e-06,
+      "loss": 0.8856,
+      "step": 36194
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7738796783695299,
+      "learning_rate": 1.11613158574148e-06,
+      "loss": 0.8679,
+      "step": 36195
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.578160648596896,
+      "learning_rate": 1.1157813027691632e-06,
+      "loss": 0.9437,
+      "step": 36196
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.607438342799049,
+      "learning_rate": 1.1154310715235827e-06,
+      "loss": 0.8372,
+      "step": 36197
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5892607324649166,
+      "learning_rate": 1.1150808920067723e-06,
+      "loss": 0.8372,
+      "step": 36198
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4946776298286324,
+      "learning_rate": 1.1147307642207771e-06,
+      "loss": 0.8219,
+      "step": 36199
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1841106101550007,
+      "learning_rate": 1.1143806881676311e-06,
+      "loss": 0.8699,
+      "step": 36200
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7321657564798858,
+      "learning_rate": 1.1140306638493725e-06,
+      "loss": 0.8257,
+      "step": 36201
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6375633951595576,
+      "learning_rate": 1.113680691268041e-06,
+      "loss": 0.8931,
+      "step": 36202
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.692003756426832,
+      "learning_rate": 1.1133307704256758e-06,
+      "loss": 0.9466,
+      "step": 36203
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.75623962213823,
+      "learning_rate": 1.112980901324312e-06,
+      "loss": 0.7117,
+      "step": 36204
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.418521104057378,
+      "learning_rate": 1.112631083965985e-06,
+      "loss": 0.8589,
+      "step": 36205
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5451125377776864,
+      "learning_rate": 1.1122813183527337e-06,
+      "loss": 0.8285,
+      "step": 36206
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9670598829605506,
+      "learning_rate": 1.111931604486597e-06,
+      "loss": 0.7727,
+      "step": 36207
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5715279970946583,
+      "learning_rate": 1.1115819423696073e-06,
+      "loss": 0.7635,
+      "step": 36208
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.83788049905508,
+      "learning_rate": 1.1112323320037998e-06,
+      "loss": 0.8688,
+      "step": 36209
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6174408842578396,
+      "learning_rate": 1.110882773391212e-06,
+      "loss": 0.9488,
+      "step": 36210
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.409616995265784,
+      "learning_rate": 1.1105332665338808e-06,
+      "loss": 0.6645,
+      "step": 36211
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5122303766136687,
+      "learning_rate": 1.1101838114338393e-06,
+      "loss": 0.9536,
+      "step": 36212
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7623385691837492,
+      "learning_rate": 1.10983440809312e-06,
+      "loss": 0.6826,
+      "step": 36213
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6589832531327016,
+      "learning_rate": 1.1094850565137616e-06,
+      "loss": 0.8822,
+      "step": 36214
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3649977903808543,
+      "learning_rate": 1.1091357566977934e-06,
+      "loss": 0.7523,
+      "step": 36215
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5500701924079072,
+      "learning_rate": 1.108786508647255e-06,
+      "loss": 0.8826,
+      "step": 36216
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6325839934549582,
+      "learning_rate": 1.1084373123641734e-06,
+      "loss": 0.8938,
+      "step": 36217
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4302839187906498,
+      "learning_rate": 1.1080881678505862e-06,
+      "loss": 0.8798,
+      "step": 36218
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4460688161116018,
+      "learning_rate": 1.1077390751085237e-06,
+      "loss": 0.8396,
+      "step": 36219
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.74977406155915,
+      "learning_rate": 1.107390034140019e-06,
+      "loss": 0.8854,
+      "step": 36220
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.032854565307446,
+      "learning_rate": 1.1070410449471103e-06,
+      "loss": 0.6914,
+      "step": 36221
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5597706754413048,
+      "learning_rate": 1.1066921075318183e-06,
+      "loss": 0.7675,
+      "step": 36222
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1217454668307507,
+      "learning_rate": 1.106343221896181e-06,
+      "loss": 0.9445,
+      "step": 36223
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5180586193404324,
+      "learning_rate": 1.1059943880422297e-06,
+      "loss": 0.8631,
+      "step": 36224
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7231112189720872,
+      "learning_rate": 1.1056456059719989e-06,
+      "loss": 0.7156,
+      "step": 36225
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7570780667051915,
+      "learning_rate": 1.1052968756875104e-06,
+      "loss": 0.908,
+      "step": 36226
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.710960235023472,
+      "learning_rate": 1.1049481971907993e-06,
+      "loss": 0.8921,
+      "step": 36227
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5282525978954018,
+      "learning_rate": 1.1045995704838997e-06,
+      "loss": 0.8529,
+      "step": 36228
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4356770441828945,
+      "learning_rate": 1.104250995568834e-06,
+      "loss": 0.7898,
+      "step": 36229
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8014464289692296,
+      "learning_rate": 1.1039024724476376e-06,
+      "loss": 0.7321,
+      "step": 36230
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6186279314452792,
+      "learning_rate": 1.1035540011223355e-06,
+      "loss": 0.7894,
+      "step": 36231
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7791264371693503,
+      "learning_rate": 1.1032055815949616e-06,
+      "loss": 0.8672,
+      "step": 36232
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3235595802832192,
+      "learning_rate": 1.1028572138675387e-06,
+      "loss": 0.9175,
+      "step": 36233
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8098479891268442,
+      "learning_rate": 1.1025088979420995e-06,
+      "loss": 0.7278,
+      "step": 36234
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.78128127164223,
+      "learning_rate": 1.1021606338206714e-06,
+      "loss": 0.8594,
+      "step": 36235
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4097284056764685,
+      "learning_rate": 1.1018124215052783e-06,
+      "loss": 0.8101,
+      "step": 36236
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5771910112129484,
+      "learning_rate": 1.1014642609979498e-06,
+      "loss": 0.7923,
+      "step": 36237
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7386461018355663,
+      "learning_rate": 1.1011161523007163e-06,
+      "loss": 0.7809,
+      "step": 36238
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5890757783047935,
+      "learning_rate": 1.100768095415603e-06,
+      "loss": 0.848,
+      "step": 36239
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8990829049309588,
+      "learning_rate": 1.1004200903446316e-06,
+      "loss": 0.9206,
+      "step": 36240
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.484542144110394,
+      "learning_rate": 1.1000721370898326e-06,
+      "loss": 0.7504,
+      "step": 36241
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6465184196120606,
+      "learning_rate": 1.0997242356532335e-06,
+      "loss": 0.9094,
+      "step": 36242
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.662427704866162,
+      "learning_rate": 1.099376386036858e-06,
+      "loss": 0.8308,
+      "step": 36243
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4276116702227999,
+      "learning_rate": 1.0990285882427286e-06,
+      "loss": 0.7802,
+      "step": 36244
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5478727476508365,
+      "learning_rate": 1.0986808422728722e-06,
+      "loss": 0.8596,
+      "step": 36245
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8667866882241149,
+      "learning_rate": 1.0983331481293168e-06,
+      "loss": 0.8638,
+      "step": 36246
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4734185294600206,
+      "learning_rate": 1.0979855058140842e-06,
+      "loss": 0.8068,
+      "step": 36247
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8896706495408846,
+      "learning_rate": 1.097637915329196e-06,
+      "loss": 0.7914,
+      "step": 36248
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5777412979145158,
+      "learning_rate": 1.0972903766766797e-06,
+      "loss": 0.8068,
+      "step": 36249
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.706465087151745,
+      "learning_rate": 1.0969428898585565e-06,
+      "loss": 0.8788,
+      "step": 36250
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.577612065294109,
+      "learning_rate": 1.0965954548768509e-06,
+      "loss": 0.8721,
+      "step": 36251
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2662950679060436,
+      "learning_rate": 1.0962480717335843e-06,
+      "loss": 0.6824,
+      "step": 36252
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5315805052156077,
+      "learning_rate": 1.095900740430783e-06,
+      "loss": 0.8494,
+      "step": 36253
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7534918924011917,
+      "learning_rate": 1.0955534609704644e-06,
+      "loss": 0.7625,
+      "step": 36254
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6298792510203735,
+      "learning_rate": 1.095206233354653e-06,
+      "loss": 0.817,
+      "step": 36255
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5867238030637572,
+      "learning_rate": 1.0948590575853723e-06,
+      "loss": 0.8828,
+      "step": 36256
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8603478178539785,
+      "learning_rate": 1.0945119336646382e-06,
+      "loss": 0.7664,
+      "step": 36257
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0219623467216628,
+      "learning_rate": 1.0941648615944756e-06,
+      "loss": 0.7924,
+      "step": 36258
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6436083170463216,
+      "learning_rate": 1.0938178413769062e-06,
+      "loss": 0.8402,
+      "step": 36259
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7626696026410764,
+      "learning_rate": 1.0934708730139486e-06,
+      "loss": 0.7144,
+      "step": 36260
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.687671056660763,
+      "learning_rate": 1.0931239565076224e-06,
+      "loss": 0.822,
+      "step": 36261
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4870265364534943,
+      "learning_rate": 1.0927770918599478e-06,
+      "loss": 0.9247,
+      "step": 36262
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4731157141276567,
+      "learning_rate": 1.0924302790729468e-06,
+      "loss": 0.7957,
+      "step": 36263
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6010784218989789,
+      "learning_rate": 1.0920835181486378e-06,
+      "loss": 0.9539,
+      "step": 36264
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4949802116983495,
+      "learning_rate": 1.0917368090890357e-06,
+      "loss": 0.86,
+      "step": 36265
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7611752859625812,
+      "learning_rate": 1.0913901518961623e-06,
+      "loss": 0.7141,
+      "step": 36266
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5104201299858706,
+      "learning_rate": 1.0910435465720382e-06,
+      "loss": 0.7132,
+      "step": 36267
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.774223054252603,
+      "learning_rate": 1.0906969931186795e-06,
+      "loss": 0.7408,
+      "step": 36268
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.134261785285547,
+      "learning_rate": 1.0903504915381002e-06,
+      "loss": 0.7974,
+      "step": 36269
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7194112750364046,
+      "learning_rate": 1.0900040418323243e-06,
+      "loss": 0.8814,
+      "step": 36270
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.999607890751226,
+      "learning_rate": 1.0896576440033636e-06,
+      "loss": 0.8647,
+      "step": 36271
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5370595147355626,
+      "learning_rate": 1.0893112980532384e-06,
+      "loss": 0.9391,
+      "step": 36272
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8336367361424502,
+      "learning_rate": 1.0889650039839627e-06,
+      "loss": 0.7941,
+      "step": 36273
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.55088204962636,
+      "learning_rate": 1.0886187617975552e-06,
+      "loss": 0.8417,
+      "step": 36274
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.479394696543991,
+      "learning_rate": 1.0882725714960296e-06,
+      "loss": 0.7978,
+      "step": 36275
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7678901203872792,
+      "learning_rate": 1.0879264330814033e-06,
+      "loss": 0.8585,
+      "step": 36276
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4801253077928542,
+      "learning_rate": 1.0875803465556911e-06,
+      "loss": 0.8816,
+      "step": 36277
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8174245882139597,
+      "learning_rate": 1.087234311920906e-06,
+      "loss": 0.7045,
+      "step": 36278
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5435241337412982,
+      "learning_rate": 1.0868883291790645e-06,
+      "loss": 0.6933,
+      "step": 36279
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.490255352347738,
+      "learning_rate": 1.0865423983321832e-06,
+      "loss": 0.8485,
+      "step": 36280
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.244030485973317,
+      "learning_rate": 1.0861965193822743e-06,
+      "loss": 0.8063,
+      "step": 36281
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.51003280148324,
+      "learning_rate": 1.0858506923313484e-06,
+      "loss": 0.7124,
+      "step": 36282
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.594039357186443,
+      "learning_rate": 1.0855049171814226e-06,
+      "loss": 0.8645,
+      "step": 36283
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.628235503008301,
+      "learning_rate": 1.0851591939345108e-06,
+      "loss": 0.7445,
+      "step": 36284
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.911685378782945,
+      "learning_rate": 1.0848135225926248e-06,
+      "loss": 0.801,
+      "step": 36285
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2618328339901366,
+      "learning_rate": 1.0844679031577754e-06,
+      "loss": 0.7188,
+      "step": 36286
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.517783179085898,
+      "learning_rate": 1.0841223356319763e-06,
+      "loss": 0.7589,
+      "step": 36287
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7356978070954943,
+      "learning_rate": 1.0837768200172417e-06,
+      "loss": 0.7744,
+      "step": 36288
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.652872002374394,
+      "learning_rate": 1.0834313563155807e-06,
+      "loss": 0.8726,
+      "step": 36289
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.003471108731172,
+      "learning_rate": 1.0830859445290044e-06,
+      "loss": 0.8828,
+      "step": 36290
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6146469584497531,
+      "learning_rate": 1.0827405846595251e-06,
+      "loss": 0.8116,
+      "step": 36291
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.931672259333636,
+      "learning_rate": 1.0823952767091527e-06,
+      "loss": 0.893,
+      "step": 36292
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7535182951134093,
+      "learning_rate": 1.0820500206799002e-06,
+      "loss": 0.6932,
+      "step": 36293
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.516247105566327,
+      "learning_rate": 1.0817048165737732e-06,
+      "loss": 0.8591,
+      "step": 36294
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8344603226708922,
+      "learning_rate": 1.0813596643927871e-06,
+      "loss": 0.7857,
+      "step": 36295
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8299820448158537,
+      "learning_rate": 1.0810145641389458e-06,
+      "loss": 0.7159,
+      "step": 36296
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4441188788958086,
+      "learning_rate": 1.0806695158142633e-06,
+      "loss": 0.7992,
+      "step": 36297
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5433684942227222,
+      "learning_rate": 1.0803245194207468e-06,
+      "loss": 0.932,
+      "step": 36298
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6202046009401792,
+      "learning_rate": 1.0799795749604037e-06,
+      "loss": 0.8234,
+      "step": 36299
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6043385640839225,
+      "learning_rate": 1.0796346824352433e-06,
+      "loss": 0.7375,
+      "step": 36300
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4550398515141632,
+      "learning_rate": 1.0792898418472753e-06,
+      "loss": 0.8356,
+      "step": 36301
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4639392604256698,
+      "learning_rate": 1.0789450531985068e-06,
+      "loss": 0.8541,
+      "step": 36302
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6195145180760746,
+      "learning_rate": 1.078600316490942e-06,
+      "loss": 0.7516,
+      "step": 36303
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6563783606924898,
+      "learning_rate": 1.0782556317265902e-06,
+      "loss": 0.8838,
+      "step": 36304
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6062131501986352,
+      "learning_rate": 1.0779109989074621e-06,
+      "loss": 0.8802,
+      "step": 36305
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0938406065795983,
+      "learning_rate": 1.0775664180355593e-06,
+      "loss": 0.7998,
+      "step": 36306
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0821296996164373,
+      "learning_rate": 1.0772218891128882e-06,
+      "loss": 0.9119,
+      "step": 36307
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4869102559787226,
+      "learning_rate": 1.076877412141457e-06,
+      "loss": 0.8274,
+      "step": 36308
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5188231260452905,
+      "learning_rate": 1.076532987123271e-06,
+      "loss": 0.7918,
+      "step": 36309
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5602873741889642,
+      "learning_rate": 1.0761886140603362e-06,
+      "loss": 0.8474,
+      "step": 36310
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5988365106845626,
+      "learning_rate": 1.0758442929546543e-06,
+      "loss": 0.8479,
+      "step": 36311
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4916806936674585,
+      "learning_rate": 1.0755000238082326e-06,
+      "loss": 0.891,
+      "step": 36312
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5622514025172576,
+      "learning_rate": 1.075155806623076e-06,
+      "loss": 0.9629,
+      "step": 36313
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7293913779491901,
+      "learning_rate": 1.074811641401189e-06,
+      "loss": 0.6608,
+      "step": 36314
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4891033138783163,
+      "learning_rate": 1.0744675281445726e-06,
+      "loss": 0.8164,
+      "step": 36315
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7973272505049187,
+      "learning_rate": 1.0741234668552324e-06,
+      "loss": 0.687,
+      "step": 36316
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.476624770600149,
+      "learning_rate": 1.0737794575351712e-06,
+      "loss": 0.9847,
+      "step": 36317
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4468715630527438,
+      "learning_rate": 1.0734355001863904e-06,
+      "loss": 0.8711,
+      "step": 36318
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6966493923327675,
+      "learning_rate": 1.0730915948108966e-06,
+      "loss": 0.8518,
+      "step": 36319
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4893737493397634,
+      "learning_rate": 1.0727477414106903e-06,
+      "loss": 0.9301,
+      "step": 36320
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.722761377998522,
+      "learning_rate": 1.0724039399877705e-06,
+      "loss": 0.8291,
+      "step": 36321
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6195985643803967,
+      "learning_rate": 1.0720601905441419e-06,
+      "loss": 0.8355,
+      "step": 36322
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5252441209283907,
+      "learning_rate": 1.071716493081809e-06,
+      "loss": 0.7395,
+      "step": 36323
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6473091929498993,
+      "learning_rate": 1.0713728476027653e-06,
+      "loss": 0.8405,
+      "step": 36324
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4019956595462555,
+      "learning_rate": 1.0710292541090161e-06,
+      "loss": 0.9635,
+      "step": 36325
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6154955745878423,
+      "learning_rate": 1.0706857126025604e-06,
+      "loss": 0.8184,
+      "step": 36326
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6870949136726303,
+      "learning_rate": 1.0703422230854022e-06,
+      "loss": 0.7023,
+      "step": 36327
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.755598461824686,
+      "learning_rate": 1.0699987855595384e-06,
+      "loss": 0.7641,
+      "step": 36328
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.579973448040378,
+      "learning_rate": 1.0696554000269677e-06,
+      "loss": 0.7489,
+      "step": 36329
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.770783877423544,
+      "learning_rate": 1.0693120664896916e-06,
+      "loss": 0.792,
+      "step": 36330
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.574864221608101,
+      "learning_rate": 1.0689687849497066e-06,
+      "loss": 0.8146,
+      "step": 36331
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6679317905579765,
+      "learning_rate": 1.068625555409014e-06,
+      "loss": 0.8284,
+      "step": 36332
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7528845614309118,
+      "learning_rate": 1.0682823778696105e-06,
+      "loss": 0.6727,
+      "step": 36333
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8363136565106074,
+      "learning_rate": 1.0679392523334954e-06,
+      "loss": 0.7192,
+      "step": 36334
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.801516124476433,
+      "learning_rate": 1.0675961788026646e-06,
+      "loss": 1.0042,
+      "step": 36335
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4527471518813724,
+      "learning_rate": 1.0672531572791178e-06,
+      "loss": 0.8453,
+      "step": 36336
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9236405784061117,
+      "learning_rate": 1.0669101877648524e-06,
+      "loss": 0.9421,
+      "step": 36337
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5122275430572591,
+      "learning_rate": 1.0665672702618612e-06,
+      "loss": 0.8983,
+      "step": 36338
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5206004769065888,
+      "learning_rate": 1.0662244047721437e-06,
+      "loss": 0.6873,
+      "step": 36339
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.617688927574005,
+      "learning_rate": 1.0658815912976983e-06,
+      "loss": 0.7939,
+      "step": 36340
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.88937011692578,
+      "learning_rate": 1.065538829840519e-06,
+      "loss": 0.8207,
+      "step": 36341
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4225979995841036,
+      "learning_rate": 1.0651961204025995e-06,
+      "loss": 0.8334,
+      "step": 36342
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7251312477038407,
+      "learning_rate": 1.0648534629859363e-06,
+      "loss": 0.8666,
+      "step": 36343
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5431719797591485,
+      "learning_rate": 1.0645108575925267e-06,
+      "loss": 0.7778,
+      "step": 36344
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1588277909873343,
+      "learning_rate": 1.0641683042243645e-06,
+      "loss": 0.7476,
+      "step": 36345
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7249235369304923,
+      "learning_rate": 1.0638258028834415e-06,
+      "loss": 0.8426,
+      "step": 36346
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.797124426604256,
+      "learning_rate": 1.0634833535717537e-06,
+      "loss": 0.8413,
+      "step": 36347
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5237798044695756,
+      "learning_rate": 1.0631409562912975e-06,
+      "loss": 0.9012,
+      "step": 36348
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5630790516655069,
+      "learning_rate": 1.0627986110440637e-06,
+      "loss": 0.9091,
+      "step": 36349
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.780588525608447,
+      "learning_rate": 1.0624563178320435e-06,
+      "loss": 0.8241,
+      "step": 36350
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9089645130337225,
+      "learning_rate": 1.0621140766572336e-06,
+      "loss": 0.9099,
+      "step": 36351
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4303842835300995,
+      "learning_rate": 1.0617718875216242e-06,
+      "loss": 0.9718,
+      "step": 36352
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.845433236597433,
+      "learning_rate": 1.0614297504272108e-06,
+      "loss": 0.6761,
+      "step": 36353
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3520016856981836,
+      "learning_rate": 1.0610876653759805e-06,
+      "loss": 0.8847,
+      "step": 36354
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5030184730640168,
+      "learning_rate": 1.0607456323699306e-06,
+      "loss": 0.8615,
+      "step": 36355
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5036560149401441,
+      "learning_rate": 1.0604036514110471e-06,
+      "loss": 0.828,
+      "step": 36356
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.734142246165989,
+      "learning_rate": 1.0600617225013265e-06,
+      "loss": 0.8555,
+      "step": 36357
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3634377841587961,
+      "learning_rate": 1.059719845642756e-06,
+      "loss": 0.7999,
+      "step": 36358
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7813790539715917,
+      "learning_rate": 1.059378020837326e-06,
+      "loss": 0.8881,
+      "step": 36359
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6138078867888825,
+      "learning_rate": 1.0590362480870275e-06,
+      "loss": 0.8214,
+      "step": 36360
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.602264720281043,
+      "learning_rate": 1.058694527393852e-06,
+      "loss": 0.7763,
+      "step": 36361
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0855994822282558,
+      "learning_rate": 1.058352858759788e-06,
+      "loss": 0.7874,
+      "step": 36362
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6915719127908473,
+      "learning_rate": 1.0580112421868226e-06,
+      "loss": 0.8381,
+      "step": 36363
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6087472254428141,
+      "learning_rate": 1.0576696776769469e-06,
+      "loss": 0.8411,
+      "step": 36364
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5736546863604235,
+      "learning_rate": 1.057328165232151e-06,
+      "loss": 0.8107,
+      "step": 36365
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1506274072772595,
+      "learning_rate": 1.0569867048544225e-06,
+      "loss": 0.8758,
+      "step": 36366
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6193115679932004,
+      "learning_rate": 1.0566452965457452e-06,
+      "loss": 0.8441,
+      "step": 36367
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7758569767573464,
+      "learning_rate": 1.056303940308112e-06,
+      "loss": 0.7039,
+      "step": 36368
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4859698565939958,
+      "learning_rate": 1.0559626361435104e-06,
+      "loss": 0.7953,
+      "step": 36369
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7345886317717056,
+      "learning_rate": 1.0556213840539253e-06,
+      "loss": 0.7549,
+      "step": 36370
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5131126310350516,
+      "learning_rate": 1.0552801840413428e-06,
+      "loss": 0.8188,
+      "step": 36371
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4866368793950817,
+      "learning_rate": 1.0549390361077527e-06,
+      "loss": 0.8072,
+      "step": 36372
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6304045835940053,
+      "learning_rate": 1.0545979402551375e-06,
+      "loss": 0.8818,
+      "step": 36373
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.610395961366147,
+      "learning_rate": 1.054256896485487e-06,
+      "loss": 0.8349,
+      "step": 36374
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8038217097181124,
+      "learning_rate": 1.0539159048007842e-06,
+      "loss": 0.842,
+      "step": 36375
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8473714711476121,
+      "learning_rate": 1.0535749652030159e-06,
+      "loss": 0.8482,
+      "step": 36376
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5004751925715332,
+      "learning_rate": 1.0532340776941653e-06,
+      "loss": 0.8621,
+      "step": 36377
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6630288684684416,
+      "learning_rate": 1.0528932422762195e-06,
+      "loss": 0.7518,
+      "step": 36378
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4167992633998163,
+      "learning_rate": 1.0525524589511627e-06,
+      "loss": 0.9584,
+      "step": 36379
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4408502241660266,
+      "learning_rate": 1.0522117277209764e-06,
+      "loss": 0.755,
+      "step": 36380
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9974382881684096,
+      "learning_rate": 1.051871048587646e-06,
+      "loss": 0.7382,
+      "step": 36381
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5025125687138519,
+      "learning_rate": 1.0515304215531563e-06,
+      "loss": 0.793,
+      "step": 36382
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6547372529621518,
+      "learning_rate": 1.0511898466194903e-06,
+      "loss": 0.8283,
+      "step": 36383
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8623604157275657,
+      "learning_rate": 1.0508493237886286e-06,
+      "loss": 0.837,
+      "step": 36384
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7486078126692872,
+      "learning_rate": 1.050508853062555e-06,
+      "loss": 0.908,
+      "step": 36385
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5328348360771342,
+      "learning_rate": 1.0501684344432538e-06,
+      "loss": 0.717,
+      "step": 36386
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7967290383959553,
+      "learning_rate": 1.0498280679327067e-06,
+      "loss": 0.7195,
+      "step": 36387
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6643969594349033,
+      "learning_rate": 1.0494877535328907e-06,
+      "loss": 0.8368,
+      "step": 36388
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7027595460742229,
+      "learning_rate": 1.049147491245791e-06,
+      "loss": 0.9642,
+      "step": 36389
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.046692225628244,
+      "learning_rate": 1.0488072810733906e-06,
+      "loss": 0.8815,
+      "step": 36390
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7660701334330461,
+      "learning_rate": 1.048467123017669e-06,
+      "loss": 0.68,
+      "step": 36391
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0143865467344075,
+      "learning_rate": 1.0481270170806024e-06,
+      "loss": 0.8963,
+      "step": 36392
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6405956233926833,
+      "learning_rate": 1.0477869632641769e-06,
+      "loss": 0.7969,
+      "step": 36393
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7130776658021292,
+      "learning_rate": 1.0474469615703686e-06,
+      "loss": 0.8399,
+      "step": 36394
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6606382530516304,
+      "learning_rate": 1.0471070120011595e-06,
+      "loss": 0.7427,
+      "step": 36395
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5118200745389307,
+      "learning_rate": 1.0467671145585268e-06,
+      "loss": 0.8526,
+      "step": 36396
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8817778030681024,
+      "learning_rate": 1.0464272692444511e-06,
+      "loss": 0.7644,
+      "step": 36397
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5884837886940437,
+      "learning_rate": 1.0460874760609096e-06,
+      "loss": 0.8082,
+      "step": 36398
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7680952175263444,
+      "learning_rate": 1.0457477350098832e-06,
+      "loss": 0.6717,
+      "step": 36399
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5910561227024302,
+      "learning_rate": 1.045408046093348e-06,
+      "loss": 0.8372,
+      "step": 36400
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7670832472707338,
+      "learning_rate": 1.0450684093132813e-06,
+      "loss": 0.8464,
+      "step": 36401
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6006992784286969,
+      "learning_rate": 1.0447288246716602e-06,
+      "loss": 0.8506,
+      "step": 36402
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7335370432778096,
+      "learning_rate": 1.0443892921704635e-06,
+      "loss": 0.764,
+      "step": 36403
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2769577947537774,
+      "learning_rate": 1.0440498118116715e-06,
+      "loss": 0.739,
+      "step": 36404
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8162408920424995,
+      "learning_rate": 1.0437103835972528e-06,
+      "loss": 0.6668,
+      "step": 36405
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.571389250861137,
+      "learning_rate": 1.043371007529188e-06,
+      "loss": 0.8239,
+      "step": 36406
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5661167369746005,
+      "learning_rate": 1.0430316836094555e-06,
+      "loss": 0.941,
+      "step": 36407
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5586059473743274,
+      "learning_rate": 1.0426924118400272e-06,
+      "loss": 0.8879,
+      "step": 36408
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4600556462485041,
+      "learning_rate": 1.042353192222878e-06,
+      "loss": 0.7968,
+      "step": 36409
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8662946906517295,
+      "learning_rate": 1.0420140247599842e-06,
+      "loss": 0.8165,
+      "step": 36410
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6429151049467463,
+      "learning_rate": 1.041674909453323e-06,
+      "loss": 0.8182,
+      "step": 36411
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6477686258898754,
+      "learning_rate": 1.0413358463048674e-06,
+      "loss": 0.8335,
+      "step": 36412
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7377765915137278,
+      "learning_rate": 1.0409968353165878e-06,
+      "loss": 0.8549,
+      "step": 36413
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6042653497080075,
+      "learning_rate": 1.040657876490463e-06,
+      "loss": 0.9271,
+      "step": 36414
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6893270026177047,
+      "learning_rate": 1.040318969828462e-06,
+      "loss": 0.7075,
+      "step": 36415
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6742295897906299,
+      "learning_rate": 1.0399801153325616e-06,
+      "loss": 0.877,
+      "step": 36416
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5039653833215558,
+      "learning_rate": 1.0396413130047356e-06,
+      "loss": 0.8042,
+      "step": 36417
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5517410076502578,
+      "learning_rate": 1.0393025628469545e-06,
+      "loss": 0.8011,
+      "step": 36418
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6661409096369264,
+      "learning_rate": 1.038963864861189e-06,
+      "loss": 0.8623,
+      "step": 36419
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.729020144088196,
+      "learning_rate": 1.0386252190494117e-06,
+      "loss": 0.9204,
+      "step": 36420
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6463461084079005,
+      "learning_rate": 1.0382866254136004e-06,
+      "loss": 0.6843,
+      "step": 36421
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7689599361657067,
+      "learning_rate": 1.0379480839557177e-06,
+      "loss": 0.8505,
+      "step": 36422
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.569094908258485,
+      "learning_rate": 1.0376095946777387e-06,
+      "loss": 0.8556,
+      "step": 36423
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6421520491019692,
+      "learning_rate": 1.0372711575816341e-06,
+      "loss": 0.8677,
+      "step": 36424
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.750005267774813,
+      "learning_rate": 1.0369327726693757e-06,
+      "loss": 0.6895,
+      "step": 36425
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6329709782077384,
+      "learning_rate": 1.0365944399429327e-06,
+      "loss": 0.7615,
+      "step": 36426
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7141930418955773,
+      "learning_rate": 1.036256159404273e-06,
+      "loss": 0.7538,
+      "step": 36427
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6388283628973304,
+      "learning_rate": 1.035917931055369e-06,
+      "loss": 0.7822,
+      "step": 36428
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3320583717818364,
+      "learning_rate": 1.035579754898187e-06,
+      "loss": 0.7626,
+      "step": 36429
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9121358653672291,
+      "learning_rate": 1.0352416309347003e-06,
+      "loss": 0.7101,
+      "step": 36430
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4970383963414577,
+      "learning_rate": 1.0349035591668732e-06,
+      "loss": 0.8307,
+      "step": 36431
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4945847044197094,
+      "learning_rate": 1.034565539596677e-06,
+      "loss": 0.7734,
+      "step": 36432
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.656579648796334,
+      "learning_rate": 1.0342275722260776e-06,
+      "loss": 0.7484,
+      "step": 36433
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.91863363889263,
+      "learning_rate": 1.0338896570570456e-06,
+      "loss": 0.8339,
+      "step": 36434
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.048129424908769,
+      "learning_rate": 1.0335517940915441e-06,
+      "loss": 0.7876,
+      "step": 36435
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.550102668105945,
+      "learning_rate": 1.0332139833315447e-06,
+      "loss": 0.928,
+      "step": 36436
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5886661687434078,
+      "learning_rate": 1.0328762247790103e-06,
+      "loss": 0.8551,
+      "step": 36437
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9896742899202013,
+      "learning_rate": 1.0325385184359126e-06,
+      "loss": 0.8079,
+      "step": 36438
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8585617247835822,
+      "learning_rate": 1.0322008643042136e-06,
+      "loss": 1.0258,
+      "step": 36439
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.373072639882713,
+      "learning_rate": 1.031863262385878e-06,
+      "loss": 0.8411,
+      "step": 36440
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5433491391875847,
+      "learning_rate": 1.0315257126828749e-06,
+      "loss": 0.965,
+      "step": 36441
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3698446601742151,
+      "learning_rate": 1.0311882151971697e-06,
+      "loss": 0.7482,
+      "step": 36442
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4526816094899277,
+      "learning_rate": 1.0308507699307268e-06,
+      "loss": 0.7555,
+      "step": 36443
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6478547320299253,
+      "learning_rate": 1.030513376885508e-06,
+      "loss": 0.7183,
+      "step": 36444
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7326815024998268,
+      "learning_rate": 1.0301760360634794e-06,
+      "loss": 0.6296,
+      "step": 36445
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5513513075302041,
+      "learning_rate": 1.0298387474666083e-06,
+      "loss": 0.8252,
+      "step": 36446
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7176466965423773,
+      "learning_rate": 1.0295015110968564e-06,
+      "loss": 0.7512,
+      "step": 36447
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.511324823288837,
+      "learning_rate": 1.0291643269561835e-06,
+      "loss": 0.8211,
+      "step": 36448
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8122669240858917,
+      "learning_rate": 1.0288271950465566e-06,
+      "loss": 0.7155,
+      "step": 36449
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5911848887284064,
+      "learning_rate": 1.0284901153699412e-06,
+      "loss": 0.8366,
+      "step": 36450
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4525297486765585,
+      "learning_rate": 1.0281530879282953e-06,
+      "loss": 0.844,
+      "step": 36451
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6097647558068249,
+      "learning_rate": 1.0278161127235808e-06,
+      "loss": 0.8137,
+      "step": 36452
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5962197216876786,
+      "learning_rate": 1.027479189757763e-06,
+      "loss": 0.8744,
+      "step": 36453
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8928542493470866,
+      "learning_rate": 1.027142319032801e-06,
+      "loss": 0.8061,
+      "step": 36454
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5976542561213753,
+      "learning_rate": 1.0268055005506583e-06,
+      "loss": 0.8567,
+      "step": 36455
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7547428911227559,
+      "learning_rate": 1.0264687343132928e-06,
+      "loss": 0.7079,
+      "step": 36456
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5546141146168204,
+      "learning_rate": 1.02613202032267e-06,
+      "loss": 0.7266,
+      "step": 36457
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6800191938679003,
+      "learning_rate": 1.0257953585807456e-06,
+      "loss": 0.789,
+      "step": 36458
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5202306336578513,
+      "learning_rate": 1.025458749089484e-06,
+      "loss": 0.6597,
+      "step": 36459
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7663456383505893,
+      "learning_rate": 1.0251221918508425e-06,
+      "loss": 0.7038,
+      "step": 36460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6439804417221187,
+      "learning_rate": 1.0247856868667782e-06,
+      "loss": 0.8036,
+      "step": 36461
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8356572759427574,
+      "learning_rate": 1.0244492341392543e-06,
+      "loss": 0.903,
+      "step": 36462
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6057806612981493,
+      "learning_rate": 1.02411283367023e-06,
+      "loss": 0.6979,
+      "step": 36463
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7897879717859324,
+      "learning_rate": 1.0237764854616628e-06,
+      "loss": 0.8041,
+      "step": 36464
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.563976442030229,
+      "learning_rate": 1.0234401895155076e-06,
+      "loss": 0.8466,
+      "step": 36465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5379372104763223,
+      "learning_rate": 1.0231039458337267e-06,
+      "loss": 0.8275,
+      "step": 36466
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7727497913130394,
+      "learning_rate": 1.022767754418279e-06,
+      "loss": 0.6475,
+      "step": 36467
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.526432631539367,
+      "learning_rate": 1.022431615271119e-06,
+      "loss": 0.9141,
+      "step": 36468
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.530156044521114,
+      "learning_rate": 1.0220955283942024e-06,
+      "loss": 0.8101,
+      "step": 36469
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5835562392081781,
+      "learning_rate": 1.021759493789488e-06,
+      "loss": 0.8807,
+      "step": 36470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7275155586012558,
+      "learning_rate": 1.0214235114589344e-06,
+      "loss": 0.8776,
+      "step": 36471
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6613583466316884,
+      "learning_rate": 1.021087581404495e-06,
+      "loss": 0.9463,
+      "step": 36472
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0323242633347665,
+      "learning_rate": 1.0207517036281256e-06,
+      "loss": 0.793,
+      "step": 36473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8313533593349006,
+      "learning_rate": 1.0204158781317841e-06,
+      "loss": 0.6931,
+      "step": 36474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9519957885538213,
+      "learning_rate": 1.0200801049174214e-06,
+      "loss": 0.8244,
+      "step": 36475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6471473384104234,
+      "learning_rate": 1.019744383986998e-06,
+      "loss": 0.7838,
+      "step": 36476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8194747116463953,
+      "learning_rate": 1.0194087153424648e-06,
+      "loss": 0.7659,
+      "step": 36477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6456865808547003,
+      "learning_rate": 1.0190730989857777e-06,
+      "loss": 0.7767,
+      "step": 36478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6368437529526325,
+      "learning_rate": 1.0187375349188887e-06,
+      "loss": 0.8426,
+      "step": 36479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6522780800878716,
+      "learning_rate": 1.0184020231437564e-06,
+      "loss": 0.839,
+      "step": 36480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.826170310871454,
+      "learning_rate": 1.01806656366233e-06,
+      "loss": 0.9467,
+      "step": 36481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.949500534927905,
+      "learning_rate": 1.0177311564765612e-06,
+      "loss": 0.8093,
+      "step": 36482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6326619492836503,
+      "learning_rate": 1.0173958015884066e-06,
+      "loss": 0.9644,
+      "step": 36483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6630806006532428,
+      "learning_rate": 1.0170604989998189e-06,
+      "loss": 0.8312,
+      "step": 36484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7601439043836773,
+      "learning_rate": 1.0167252487127488e-06,
+      "loss": 0.7422,
+      "step": 36485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5566354166521699,
+      "learning_rate": 1.0163900507291457e-06,
+      "loss": 0.8726,
+      "step": 36486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6020952914220392,
+      "learning_rate": 1.016054905050965e-06,
+      "loss": 0.8279,
+      "step": 36487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.562378443799015,
+      "learning_rate": 1.0157198116801592e-06,
+      "loss": 0.8062,
+      "step": 36488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5082793054992172,
+      "learning_rate": 1.015384770618676e-06,
+      "loss": 0.9375,
+      "step": 36489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7706876798747886,
+      "learning_rate": 1.0150497818684657e-06,
+      "loss": 0.7702,
+      "step": 36490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7742727364242215,
+      "learning_rate": 1.0147148454314792e-06,
+      "loss": 0.8169,
+      "step": 36491
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5277995098172423,
+      "learning_rate": 1.0143799613096706e-06,
+      "loss": 0.7576,
+      "step": 36492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7604194567893257,
+      "learning_rate": 1.014045129504987e-06,
+      "loss": 0.8723,
+      "step": 36493
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.821357456855811,
+      "learning_rate": 1.0137103500193746e-06,
+      "loss": 0.9244,
+      "step": 36494
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.624896447357125,
+      "learning_rate": 1.0133756228547875e-06,
+      "loss": 0.757,
+      "step": 36495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6870967855964165,
+      "learning_rate": 1.013040948013172e-06,
+      "loss": 0.8089,
+      "step": 36496
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5267567418531944,
+      "learning_rate": 1.0127063254964776e-06,
+      "loss": 0.7699,
+      "step": 36497
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9263581126542344,
+      "learning_rate": 1.0123717553066515e-06,
+      "loss": 0.8609,
+      "step": 36498
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8120052997793565,
+      "learning_rate": 1.0120372374456445e-06,
+      "loss": 0.7361,
+      "step": 36499
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9264857001359337,
+      "learning_rate": 1.0117027719153993e-06,
+      "loss": 0.865,
+      "step": 36500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6928019903557034,
+      "learning_rate": 1.011368358717867e-06,
+      "loss": 0.7756,
+      "step": 36501
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6137948316273099,
+      "learning_rate": 1.0110339978549977e-06,
+      "loss": 0.883,
+      "step": 36502
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8813791098206607,
+      "learning_rate": 1.0106996893287303e-06,
+      "loss": 0.7858,
+      "step": 36503
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.518671151733234,
+      "learning_rate": 1.010365433141015e-06,
+      "loss": 0.8758,
+      "step": 36504
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7178486486405079,
+      "learning_rate": 1.0100312292937986e-06,
+      "loss": 0.722,
+      "step": 36505
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1659490813577813,
+      "learning_rate": 1.0096970777890302e-06,
+      "loss": 0.8714,
+      "step": 36506
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7315369397110278,
+      "learning_rate": 1.0093629786286475e-06,
+      "loss": 0.7769,
+      "step": 36507
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.633921448867744,
+      "learning_rate": 1.0090289318146007e-06,
+      "loss": 0.8545,
+      "step": 36508
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.639826242625966,
+      "learning_rate": 1.0086949373488352e-06,
+      "loss": 0.8236,
+      "step": 36509
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6575515708097854,
+      "learning_rate": 1.0083609952332928e-06,
+      "loss": 0.7533,
+      "step": 36510
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6796476520896104,
+      "learning_rate": 1.0080271054699208e-06,
+      "loss": 0.9483,
+      "step": 36511
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7735775380455503,
+      "learning_rate": 1.007693268060661e-06,
+      "loss": 0.8652,
+      "step": 36512
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8117469359896332,
+      "learning_rate": 1.0073594830074596e-06,
+      "loss": 0.8927,
+      "step": 36513
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5464015706806267,
+      "learning_rate": 1.007025750312256e-06,
+      "loss": 0.7515,
+      "step": 36514
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7309479352210335,
+      "learning_rate": 1.0066920699769966e-06,
+      "loss": 0.6796,
+      "step": 36515
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.539795589197382,
+      "learning_rate": 1.0063584420036243e-06,
+      "loss": 0.7839,
+      "step": 36516
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6932646895096455,
+      "learning_rate": 1.0060248663940787e-06,
+      "loss": 0.9505,
+      "step": 36517
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7762565093677576,
+      "learning_rate": 1.0056913431503035e-06,
+      "loss": 0.9024,
+      "step": 36518
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5392567892154982,
+      "learning_rate": 1.005357872274243e-06,
+      "loss": 0.854,
+      "step": 36519
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6094558320134793,
+      "learning_rate": 1.0050244537678356e-06,
+      "loss": 0.9537,
+      "step": 36520
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.001587630554041,
+      "learning_rate": 1.004691087633023e-06,
+      "loss": 0.8141,
+      "step": 36521
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2730113449297904,
+      "learning_rate": 1.0043577738717459e-06,
+      "loss": 0.7706,
+      "step": 36522
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7264397404767855,
+      "learning_rate": 1.0040245124859483e-06,
+      "loss": 0.7852,
+      "step": 36523
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.473379428079174,
+      "learning_rate": 1.0036913034775675e-06,
+      "loss": 0.7514,
+      "step": 36524
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5246760016647813,
+      "learning_rate": 1.0033581468485422e-06,
+      "loss": 0.9064,
+      "step": 36525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8055132625120862,
+      "learning_rate": 1.0030250426008148e-06,
+      "loss": 0.8022,
+      "step": 36526
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6847052489189436,
+      "learning_rate": 1.0026919907363252e-06,
+      "loss": 0.8894,
+      "step": 36527
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7179877866961105,
+      "learning_rate": 1.002358991257012e-06,
+      "loss": 0.6975,
+      "step": 36528
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.735977135956567,
+      "learning_rate": 1.0020260441648111e-06,
+      "loss": 0.7173,
+      "step": 36529
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7406153080303639,
+      "learning_rate": 1.0016931494616644e-06,
+      "loss": 0.7291,
+      "step": 36530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4600197615941022,
+      "learning_rate": 1.001360307149508e-06,
+      "loss": 0.844,
+      "step": 36531
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7021575360238277,
+      "learning_rate": 1.001027517230283e-06,
+      "loss": 0.7385,
+      "step": 36532
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5735075479938079,
+      "learning_rate": 1.000694779705922e-06,
+      "loss": 0.7706,
+      "step": 36533
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.699455590157761,
+      "learning_rate": 1.0003620945783676e-06,
+      "loss": 0.8244,
+      "step": 36534
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.667524938039279,
+      "learning_rate": 1.000029461849552e-06,
+      "loss": 0.8031,
+      "step": 36535
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6385004690075518,
+      "learning_rate": 9.99696881521417e-07,
+      "loss": 0.835,
+      "step": 36536
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4864164331979408,
+      "learning_rate": 9.993643535958952e-07,
+      "loss": 0.758,
+      "step": 36537
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6665881496888817,
+      "learning_rate": 9.990318780749219e-07,
+      "loss": 0.7916,
+      "step": 36538
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6888317263062333,
+      "learning_rate": 9.986994549604346e-07,
+      "loss": 0.7945,
+      "step": 36539
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.736927946260951,
+      "learning_rate": 9.983670842543713e-07,
+      "loss": 0.6732,
+      "step": 36540
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3667761030868006,
+      "learning_rate": 9.980347659586643e-07,
+      "loss": 0.8604,
+      "step": 36541
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6400259810625086,
+      "learning_rate": 9.977025000752472e-07,
+      "loss": 0.8181,
+      "step": 36542
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5828204664606673,
+      "learning_rate": 9.973702866060565e-07,
+      "loss": 0.9511,
+      "step": 36543
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6645532433448815,
+      "learning_rate": 9.97038125553028e-07,
+      "loss": 0.9148,
+      "step": 36544
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5160530464212945,
+      "learning_rate": 9.96706016918093e-07,
+      "loss": 0.8546,
+      "step": 36545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6534155166839972,
+      "learning_rate": 9.963739607031852e-07,
+      "loss": 0.9045,
+      "step": 36546
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4816896005013809,
+      "learning_rate": 9.960419569102375e-07,
+      "loss": 0.7685,
+      "step": 36547
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5039987585950483,
+      "learning_rate": 9.957100055411872e-07,
+      "loss": 0.8962,
+      "step": 36548
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6765587033214802,
+      "learning_rate": 9.95378106597964e-07,
+      "loss": 0.8408,
+      "step": 36549
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8219353025519619,
+      "learning_rate": 9.950462600824973e-07,
+      "loss": 0.7718,
+      "step": 36550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6228268550396154,
+      "learning_rate": 9.947144659967245e-07,
+      "loss": 0.7641,
+      "step": 36551
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1838813590035704,
+      "learning_rate": 9.94382724342573e-07,
+      "loss": 0.764,
+      "step": 36552
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4442941191382686,
+      "learning_rate": 9.94051035121979e-07,
+      "loss": 0.787,
+      "step": 36553
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.618733723458906,
+      "learning_rate": 9.937193983368688e-07,
+      "loss": 0.764,
+      "step": 36554
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2395200436236857,
+      "learning_rate": 9.933878139891772e-07,
+      "loss": 0.93,
+      "step": 36555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4414163941619864,
+      "learning_rate": 9.93056282080832e-07,
+      "loss": 0.9029,
+      "step": 36556
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6661332191719065,
+      "learning_rate": 9.92724802613766e-07,
+      "loss": 0.8157,
+      "step": 36557
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7195147016974754,
+      "learning_rate": 9.923933755899062e-07,
+      "loss": 0.8912,
+      "step": 36558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6501979681221766,
+      "learning_rate": 9.920620010111858e-07,
+      "loss": 0.7318,
+      "step": 36559
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.858711508690711,
+      "learning_rate": 9.9173067887953e-07,
+      "loss": 0.8797,
+      "step": 36560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7336981895509064,
+      "learning_rate": 9.913994091968725e-07,
+      "loss": 0.7252,
+      "step": 36561
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6663995347247345,
+      "learning_rate": 9.910681919651399e-07,
+      "loss": 0.8457,
+      "step": 36562
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5498172448745624,
+      "learning_rate": 9.907370271862582e-07,
+      "loss": 0.8739,
+      "step": 36563
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6606941133372206,
+      "learning_rate": 9.904059148621581e-07,
+      "loss": 0.8381,
+      "step": 36564
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.987453285594018,
+      "learning_rate": 9.900748549947692e-07,
+      "loss": 0.8619,
+      "step": 36565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.582732864496659,
+      "learning_rate": 9.897438475860165e-07,
+      "loss": 0.8158,
+      "step": 36566
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7718423211431834,
+      "learning_rate": 9.894128926378255e-07,
+      "loss": 0.6792,
+      "step": 36567
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5904030811628256,
+      "learning_rate": 9.890819901521266e-07,
+      "loss": 0.7334,
+      "step": 36568
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6137180663182538,
+      "learning_rate": 9.887511401308458e-07,
+      "loss": 0.8283,
+      "step": 36569
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6022286575407594,
+      "learning_rate": 9.884203425759098e-07,
+      "loss": 0.7942,
+      "step": 36570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8281774686077052,
+      "learning_rate": 9.880895974892413e-07,
+      "loss": 0.8125,
+      "step": 36571
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6898141517248244,
+      "learning_rate": 9.877589048727676e-07,
+      "loss": 0.8226,
+      "step": 36572
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5402180556372218,
+      "learning_rate": 9.874282647284173e-07,
+      "loss": 0.8606,
+      "step": 36573
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4900683188671073,
+      "learning_rate": 9.87097677058113e-07,
+      "loss": 0.8722,
+      "step": 36574
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5521401573254663,
+      "learning_rate": 9.86767141863777e-07,
+      "loss": 0.8653,
+      "step": 36575
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2106087428834016,
+      "learning_rate": 9.864366591473385e-07,
+      "loss": 0.9637,
+      "step": 36576
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.544754917889809,
+      "learning_rate": 9.86106228910718e-07,
+      "loss": 0.7942,
+      "step": 36577
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8576849282571994,
+      "learning_rate": 9.857758511558423e-07,
+      "loss": 0.828,
+      "step": 36578
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.599831475725738,
+      "learning_rate": 9.854455258846319e-07,
+      "loss": 0.8046,
+      "step": 36579
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5353770633316794,
+      "learning_rate": 9.851152530990127e-07,
+      "loss": 0.8341,
+      "step": 36580
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6387938280656087,
+      "learning_rate": 9.847850328009056e-07,
+      "loss": 0.8285,
+      "step": 36581
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3984955997525303,
+      "learning_rate": 9.844548649922358e-07,
+      "loss": 0.7619,
+      "step": 36582
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5406554015856087,
+      "learning_rate": 9.84124749674923e-07,
+      "loss": 0.8069,
+      "step": 36583
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6378824063553357,
+      "learning_rate": 9.837946868508897e-07,
+      "loss": 0.7598,
+      "step": 36584
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8253800343227535,
+      "learning_rate": 9.83464676522058e-07,
+      "loss": 0.8845,
+      "step": 36585
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6752703872278658,
+      "learning_rate": 9.831347186903495e-07,
+      "loss": 0.8981,
+      "step": 36586
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.752695195592946,
+      "learning_rate": 9.828048133576884e-07,
+      "loss": 0.6905,
+      "step": 36587
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5146727380726235,
+      "learning_rate": 9.824749605259898e-07,
+      "loss": 0.851,
+      "step": 36588
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7513309047274406,
+      "learning_rate": 9.821451601971755e-07,
+      "loss": 0.6748,
+      "step": 36589
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7983313371291152,
+      "learning_rate": 9.818154123731704e-07,
+      "loss": 0.8553,
+      "step": 36590
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6950625733448907,
+      "learning_rate": 9.8148571705589e-07,
+      "loss": 0.8463,
+      "step": 36591
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9170186403932448,
+      "learning_rate": 9.811560742472537e-07,
+      "loss": 0.7313,
+      "step": 36592
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7665706290394236,
+      "learning_rate": 9.808264839491822e-07,
+      "loss": 0.7344,
+      "step": 36593
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.886625118760147,
+      "learning_rate": 9.80496946163596e-07,
+      "loss": 0.8549,
+      "step": 36594
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.779028175390486,
+      "learning_rate": 9.801674608924118e-07,
+      "loss": 0.8513,
+      "step": 36595
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8029410579167325,
+      "learning_rate": 9.798380281375463e-07,
+      "loss": 0.7533,
+      "step": 36596
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6875179389021093,
+      "learning_rate": 9.795086479009218e-07,
+      "loss": 0.8467,
+      "step": 36597
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.791323119310869,
+      "learning_rate": 9.791793201844525e-07,
+      "loss": 0.7224,
+      "step": 36598
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7312691392529465,
+      "learning_rate": 9.788500449900563e-07,
+      "loss": 0.7815,
+      "step": 36599
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9778301786497776,
+      "learning_rate": 9.78520822319653e-07,
+      "loss": 0.8405,
+      "step": 36600
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.757974152549496,
+      "learning_rate": 9.781916521751577e-07,
+      "loss": 0.7613,
+      "step": 36601
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6108853604947337,
+      "learning_rate": 9.778625345584857e-07,
+      "loss": 0.7335,
+      "step": 36602
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2032877675349973,
+      "learning_rate": 9.775334694715533e-07,
+      "loss": 0.8594,
+      "step": 36603
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.533620362471742,
+      "learning_rate": 9.77204456916282e-07,
+      "loss": 0.943,
+      "step": 36604
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8096826846213232,
+      "learning_rate": 9.768754968945792e-07,
+      "loss": 0.6326,
+      "step": 36605
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7401016722066296,
+      "learning_rate": 9.765465894083637e-07,
+      "loss": 0.7611,
+      "step": 36606
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9174121146492717,
+      "learning_rate": 9.762177344595514e-07,
+      "loss": 0.9959,
+      "step": 36607
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6137347926939831,
+      "learning_rate": 9.758889320500575e-07,
+      "loss": 0.8069,
+      "step": 36608
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6269615969589015,
+      "learning_rate": 9.75560182181795e-07,
+      "loss": 0.7672,
+      "step": 36609
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6237396823257493,
+      "learning_rate": 9.75231484856677e-07,
+      "loss": 0.8203,
+      "step": 36610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.617984729740718,
+      "learning_rate": 9.749028400766191e-07,
+      "loss": 0.8344,
+      "step": 36611
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.256523327233008,
+      "learning_rate": 9.745742478435339e-07,
+      "loss": 0.677,
+      "step": 36612
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.878529825411479,
+      "learning_rate": 9.74245708159336e-07,
+      "loss": 0.8041,
+      "step": 36613
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5082292241282937,
+      "learning_rate": 9.73917221025935e-07,
+      "loss": 0.8367,
+      "step": 36614
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.309584773147033,
+      "learning_rate": 9.735887864452476e-07,
+      "loss": 0.7943,
+      "step": 36615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.60088963161122,
+      "learning_rate": 9.732604044191818e-07,
+      "loss": 0.8962,
+      "step": 36616
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.534868492603769,
+      "learning_rate": 9.72932074949654e-07,
+      "loss": 0.8397,
+      "step": 36617
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6373160298135425,
+      "learning_rate": 9.72603798038574e-07,
+      "loss": 0.9047,
+      "step": 36618
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9155609519511916,
+      "learning_rate": 9.722755736878498e-07,
+      "loss": 0.7934,
+      "step": 36619
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5760903032311635,
+      "learning_rate": 9.719474018993957e-07,
+      "loss": 0.8832,
+      "step": 36620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4914885871695949,
+      "learning_rate": 9.716192826751246e-07,
+      "loss": 0.728,
+      "step": 36621
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.475696649664668,
+      "learning_rate": 9.71291216016944e-07,
+      "loss": 0.8773,
+      "step": 36622
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4667728430595846,
+      "learning_rate": 9.709632019267624e-07,
+      "loss": 0.8025,
+      "step": 36623
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.532064281229545,
+      "learning_rate": 9.706352404064911e-07,
+      "loss": 0.836,
+      "step": 36624
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7172414720145797,
+      "learning_rate": 9.703073314580425e-07,
+      "loss": 0.8567,
+      "step": 36625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6174067253676945,
+      "learning_rate": 9.699794750833235e-07,
+      "loss": 0.8532,
+      "step": 36626
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5996273774548126,
+      "learning_rate": 9.696516712842407e-07,
+      "loss": 0.8323,
+      "step": 36627
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5030604945800765,
+      "learning_rate": 9.693239200627047e-07,
+      "loss": 0.7371,
+      "step": 36628
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5382054657510569,
+      "learning_rate": 9.68996221420626e-07,
+      "loss": 0.7289,
+      "step": 36629
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6362637584681365,
+      "learning_rate": 9.686685753599113e-07,
+      "loss": 0.8588,
+      "step": 36630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.396327679019121,
+      "learning_rate": 9.683409818824641e-07,
+      "loss": 0.764,
+      "step": 36631
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6706726351105983,
+      "learning_rate": 9.680134409901975e-07,
+      "loss": 0.844,
+      "step": 36632
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.661026695725314,
+      "learning_rate": 9.676859526850145e-07,
+      "loss": 0.7607,
+      "step": 36633
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7315246948053555,
+      "learning_rate": 9.67358516968826e-07,
+      "loss": 0.8938,
+      "step": 36634
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.930671843980738,
+      "learning_rate": 9.670311338435324e-07,
+      "loss": 0.762,
+      "step": 36635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5920558862974428,
+      "learning_rate": 9.667038033110465e-07,
+      "loss": 0.7915,
+      "step": 36636
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5436281950795927,
+      "learning_rate": 9.663765253732682e-07,
+      "loss": 0.8988,
+      "step": 36637
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7868563648495763,
+      "learning_rate": 9.660493000321081e-07,
+      "loss": 0.7734,
+      "step": 36638
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.985559091772205,
+      "learning_rate": 9.65722127289469e-07,
+      "loss": 0.7991,
+      "step": 36639
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5931862839708357,
+      "learning_rate": 9.65395007147254e-07,
+      "loss": 0.7113,
+      "step": 36640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5300036655788172,
+      "learning_rate": 9.650679396073691e-07,
+      "loss": 0.7587,
+      "step": 36641
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.508484922968597,
+      "learning_rate": 9.647409246717222e-07,
+      "loss": 0.7941,
+      "step": 36642
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6093302957689475,
+      "learning_rate": 9.644139623422132e-07,
+      "loss": 0.9338,
+      "step": 36643
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.608843412091405,
+      "learning_rate": 9.640870526207445e-07,
+      "loss": 0.7561,
+      "step": 36644
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7390342645846424,
+      "learning_rate": 9.637601955092224e-07,
+      "loss": 0.7709,
+      "step": 36645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.060123987413368,
+      "learning_rate": 9.634333910095516e-07,
+      "loss": 0.8986,
+      "step": 36646
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8010183586367536,
+      "learning_rate": 9.63106639123632e-07,
+      "loss": 0.898,
+      "step": 36647
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6868370973622338,
+      "learning_rate": 9.627799398533644e-07,
+      "loss": 0.8599,
+      "step": 36648
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7668051520475418,
+      "learning_rate": 9.624532932006536e-07,
+      "loss": 0.9103,
+      "step": 36649
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.853154152651222,
+      "learning_rate": 9.621266991674017e-07,
+      "loss": 0.7511,
+      "step": 36650
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1916809377095694,
+      "learning_rate": 9.618001577555113e-07,
+      "loss": 0.7911,
+      "step": 36651
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.607778427086901,
+      "learning_rate": 9.61473668966879e-07,
+      "loss": 0.863,
+      "step": 36652
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6604041024801661,
+      "learning_rate": 9.6114723280341e-07,
+      "loss": 0.7625,
+      "step": 36653
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4895496188444768,
+      "learning_rate": 9.608208492670023e-07,
+      "loss": 0.8036,
+      "step": 36654
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5934368167243051,
+      "learning_rate": 9.604945183595593e-07,
+      "loss": 0.796,
+      "step": 36655
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.621364320406896,
+      "learning_rate": 9.60168240082976e-07,
+      "loss": 0.7729,
+      "step": 36656
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5687737262716663,
+      "learning_rate": 9.598420144391585e-07,
+      "loss": 0.7259,
+      "step": 36657
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4870843148175694,
+      "learning_rate": 9.5951584143e-07,
+      "loss": 0.8448,
+      "step": 36658
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6509849936615975,
+      "learning_rate": 9.591897210574053e-07,
+      "loss": 0.8628,
+      "step": 36659
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.627099676006706,
+      "learning_rate": 9.588636533232697e-07,
+      "loss": 0.8334,
+      "step": 36660
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6707992006054213,
+      "learning_rate": 9.585376382294898e-07,
+      "loss": 0.8614,
+      "step": 36661
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6174892100939071,
+      "learning_rate": 9.582116757779668e-07,
+      "loss": 0.8298,
+      "step": 36662
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6543637983677921,
+      "learning_rate": 9.57885765970601e-07,
+      "loss": 0.7995,
+      "step": 36663
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7629616324443375,
+      "learning_rate": 9.575599088092858e-07,
+      "loss": 0.6797,
+      "step": 36664
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5931602238956024,
+      "learning_rate": 9.572341042959177e-07,
+      "loss": 0.7123,
+      "step": 36665
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6090396546299695,
+      "learning_rate": 9.569083524323952e-07,
+      "loss": 0.7314,
+      "step": 36666
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6055901686808387,
+      "learning_rate": 9.56582653220618e-07,
+      "loss": 0.8779,
+      "step": 36667
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.591870250959759,
+      "learning_rate": 9.562570066624788e-07,
+      "loss": 0.8388,
+      "step": 36668
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.585765386029215,
+      "learning_rate": 9.55931412759874e-07,
+      "loss": 0.8112,
+      "step": 36669
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8431802599635723,
+      "learning_rate": 9.556058715146987e-07,
+      "loss": 0.7627,
+      "step": 36670
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.899760984077001,
+      "learning_rate": 9.552803829288514e-07,
+      "loss": 0.6657,
+      "step": 36671
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7058859705646596,
+      "learning_rate": 9.549549470042251e-07,
+      "loss": 0.9406,
+      "step": 36672
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0874269409849737,
+      "learning_rate": 9.546295637427138e-07,
+      "loss": 0.9253,
+      "step": 36673
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7464344147196486,
+      "learning_rate": 9.543042331462139e-07,
+      "loss": 0.6592,
+      "step": 36674
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.686889816567534,
+      "learning_rate": 9.539789552166167e-07,
+      "loss": 0.8229,
+      "step": 36675
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6442492683179724,
+      "learning_rate": 9.536537299558202e-07,
+      "loss": 0.728,
+      "step": 36676
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9117356211237815,
+      "learning_rate": 9.533285573657137e-07,
+      "loss": 0.84,
+      "step": 36677
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7959479281661299,
+      "learning_rate": 9.530034374481945e-07,
+      "loss": 0.6899,
+      "step": 36678
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8368401966004928,
+      "learning_rate": 9.526783702051512e-07,
+      "loss": 0.7596,
+      "step": 36679
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5524855837810043,
+      "learning_rate": 9.523533556384812e-07,
+      "loss": 0.8877,
+      "step": 36680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4475922530871428,
+      "learning_rate": 9.520283937500718e-07,
+      "loss": 0.7516,
+      "step": 36681
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4764092504174642,
+      "learning_rate": 9.517034845418194e-07,
+      "loss": 0.8676,
+      "step": 36682
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.672094872699308,
+      "learning_rate": 9.513786280156135e-07,
+      "loss": 0.889,
+      "step": 36683
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8347792514544148,
+      "learning_rate": 9.510538241733446e-07,
+      "loss": 0.9184,
+      "step": 36684
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.522991644416518,
+      "learning_rate": 9.507290730169094e-07,
+      "loss": 0.8223,
+      "step": 36685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4525744578298114,
+      "learning_rate": 9.504043745481905e-07,
+      "loss": 0.7488,
+      "step": 36686
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.65297549585124,
+      "learning_rate": 9.500797287690833e-07,
+      "loss": 0.851,
+      "step": 36687
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.040414737156845,
+      "learning_rate": 9.49755135681476e-07,
+      "loss": 0.767,
+      "step": 36688
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7218941802845158,
+      "learning_rate": 9.49430595287264e-07,
+      "loss": 0.8934,
+      "step": 36689
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6710582266550904,
+      "learning_rate": 9.491061075883279e-07,
+      "loss": 0.8129,
+      "step": 36690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6169725082584616,
+      "learning_rate": 9.487816725865618e-07,
+      "loss": 0.9277,
+      "step": 36691
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5912787539658064,
+      "learning_rate": 9.484572902838563e-07,
+      "loss": 0.8731,
+      "step": 36692
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5647245031202686,
+      "learning_rate": 9.481329606820988e-07,
+      "loss": 0.8258,
+      "step": 36693
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.739018024121555,
+      "learning_rate": 9.478086837831746e-07,
+      "loss": 0.7945,
+      "step": 36694
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6298142313778063,
+      "learning_rate": 9.474844595889743e-07,
+      "loss": 0.8444,
+      "step": 36695
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8197797558605797,
+      "learning_rate": 9.471602881013864e-07,
+      "loss": 0.7309,
+      "step": 36696
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7129435040959622,
+      "learning_rate": 9.468361693222972e-07,
+      "loss": 0.6857,
+      "step": 36697
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.559636861009668,
+      "learning_rate": 9.46512103253594e-07,
+      "loss": 0.8641,
+      "step": 36698
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5849036503965388,
+      "learning_rate": 9.461880898971654e-07,
+      "loss": 0.8153,
+      "step": 36699
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5441835798370485,
+      "learning_rate": 9.458641292548931e-07,
+      "loss": 0.826,
+      "step": 36700
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.521748191322866,
+      "learning_rate": 9.455402213286669e-07,
+      "loss": 0.8702,
+      "step": 36701
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6188970106144447,
+      "learning_rate": 9.452163661203728e-07,
+      "loss": 0.7605,
+      "step": 36702
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8114825468062459,
+      "learning_rate": 9.448925636318973e-07,
+      "loss": 0.9356,
+      "step": 36703
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.559378817966965,
+      "learning_rate": 9.44568813865121e-07,
+      "loss": 0.7221,
+      "step": 36704
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7651894949028931,
+      "learning_rate": 9.442451168219325e-07,
+      "loss": 0.747,
+      "step": 36705
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7915795726430148,
+      "learning_rate": 9.439214725042178e-07,
+      "loss": 0.906,
+      "step": 36706
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.215936199339943,
+      "learning_rate": 9.435978809138601e-07,
+      "loss": 0.8722,
+      "step": 36707
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6148175242003948,
+      "learning_rate": 9.432743420527401e-07,
+      "loss": 0.7592,
+      "step": 36708
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7128485083295604,
+      "learning_rate": 9.429508559227441e-07,
+      "loss": 0.9615,
+      "step": 36709
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5833688984315137,
+      "learning_rate": 9.426274225257582e-07,
+      "loss": 0.7086,
+      "step": 36710
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5901896023642221,
+      "learning_rate": 9.42304041863662e-07,
+      "loss": 0.7857,
+      "step": 36711
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1619766077828406,
+      "learning_rate": 9.419807139383374e-07,
+      "loss": 0.9194,
+      "step": 36712
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7458040750335049,
+      "learning_rate": 9.41657438751672e-07,
+      "loss": 0.8405,
+      "step": 36713
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7909252475137064,
+      "learning_rate": 9.413342163055416e-07,
+      "loss": 0.6749,
+      "step": 36714
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4631080482181646,
+      "learning_rate": 9.410110466018341e-07,
+      "loss": 0.8779,
+      "step": 36715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.456268476544034,
+      "learning_rate": 9.406879296424254e-07,
+      "loss": 0.7978,
+      "step": 36716
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5960771053190754,
+      "learning_rate": 9.403648654292019e-07,
+      "loss": 0.7622,
+      "step": 36717
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4841633455608838,
+      "learning_rate": 9.40041853964041e-07,
+      "loss": 0.7872,
+      "step": 36718
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5080621097489302,
+      "learning_rate": 9.397188952488267e-07,
+      "loss": 0.7016,
+      "step": 36719
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6601275092680474,
+      "learning_rate": 9.393959892854366e-07,
+      "loss": 0.7594,
+      "step": 36720
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5842616948841737,
+      "learning_rate": 9.39073136075751e-07,
+      "loss": 0.7853,
+      "step": 36721
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.791944743354026,
+      "learning_rate": 9.387503356216488e-07,
+      "loss": 0.6982,
+      "step": 36722
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5576010569048127,
+      "learning_rate": 9.384275879250138e-07,
+      "loss": 0.6751,
+      "step": 36723
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.653025629198105,
+      "learning_rate": 9.381048929877223e-07,
+      "loss": 0.8362,
+      "step": 36724
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5846482111332179,
+      "learning_rate": 9.377822508116507e-07,
+      "loss": 0.6605,
+      "step": 36725
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7295502868086795,
+      "learning_rate": 9.374596613986797e-07,
+      "loss": 0.6619,
+      "step": 36726
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5454763032970205,
+      "learning_rate": 9.371371247506899e-07,
+      "loss": 0.7276,
+      "step": 36727
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5276032068747656,
+      "learning_rate": 9.368146408695566e-07,
+      "loss": 0.8399,
+      "step": 36728
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.493671108592812,
+      "learning_rate": 9.364922097571572e-07,
+      "loss": 0.8171,
+      "step": 36729
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7824900712882568,
+      "learning_rate": 9.361698314153678e-07,
+      "loss": 0.79,
+      "step": 36730
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5366280112445698,
+      "learning_rate": 9.358475058460704e-07,
+      "loss": 0.9015,
+      "step": 36731
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4090867023842015,
+      "learning_rate": 9.355252330511377e-07,
+      "loss": 0.8065,
+      "step": 36732
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6337628321129565,
+      "learning_rate": 9.352030130324463e-07,
+      "loss": 1.0177,
+      "step": 36733
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7848543618258912,
+      "learning_rate": 9.348808457918734e-07,
+      "loss": 0.7403,
+      "step": 36734
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.764850589086426,
+      "learning_rate": 9.345587313312921e-07,
+      "loss": 0.69,
+      "step": 36735
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7229723930734713,
+      "learning_rate": 9.342366696525829e-07,
+      "loss": 0.6355,
+      "step": 36736
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.668495013995285,
+      "learning_rate": 9.339146607576155e-07,
+      "loss": 0.8847,
+      "step": 36737
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8604816638767439,
+      "learning_rate": 9.335927046482695e-07,
+      "loss": 0.8408,
+      "step": 36738
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5396302355750504,
+      "learning_rate": 9.332708013264147e-07,
+      "loss": 0.7666,
+      "step": 36739
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9308709628178093,
+      "learning_rate": 9.329489507939304e-07,
+      "loss": 0.7882,
+      "step": 36740
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6302431485824673,
+      "learning_rate": 9.326271530526865e-07,
+      "loss": 0.9181,
+      "step": 36741
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.798056494219948,
+      "learning_rate": 9.323054081045579e-07,
+      "loss": 0.8103,
+      "step": 36742
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7539302051953674,
+      "learning_rate": 9.319837159514167e-07,
+      "loss": 0.9499,
+      "step": 36743
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5765768936540918,
+      "learning_rate": 9.31662076595139e-07,
+      "loss": 0.8171,
+      "step": 36744
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6371047524932736,
+      "learning_rate": 9.313404900375966e-07,
+      "loss": 0.7675,
+      "step": 36745
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5837129477638772,
+      "learning_rate": 9.310189562806582e-07,
+      "loss": 0.7679,
+      "step": 36746
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.684359846840804,
+      "learning_rate": 9.306974753261999e-07,
+      "loss": 0.8529,
+      "step": 36747
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6319637414937247,
+      "learning_rate": 9.303760471760925e-07,
+      "loss": 0.8805,
+      "step": 36748
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.505406855667415,
+      "learning_rate": 9.300546718322079e-07,
+      "loss": 0.7764,
+      "step": 36749
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5314871866009634,
+      "learning_rate": 9.297333492964144e-07,
+      "loss": 0.864,
+      "step": 36750
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7808549823058493,
+      "learning_rate": 9.294120795705852e-07,
+      "loss": 0.7993,
+      "step": 36751
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7930959878479964,
+      "learning_rate": 9.290908626565931e-07,
+      "loss": 0.7689,
+      "step": 36752
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7286878584362395,
+      "learning_rate": 9.287696985563044e-07,
+      "loss": 0.7631,
+      "step": 36753
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.016253394919574,
+      "learning_rate": 9.284485872715898e-07,
+      "loss": 0.7488,
+      "step": 36754
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6359570842647375,
+      "learning_rate": 9.281275288043212e-07,
+      "loss": 0.8363,
+      "step": 36755
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7625889161301889,
+      "learning_rate": 9.278065231563637e-07,
+      "loss": 0.7081,
+      "step": 36756
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4388229048022128,
+      "learning_rate": 9.274855703295904e-07,
+      "loss": 0.7627,
+      "step": 36757
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7336691442177774,
+      "learning_rate": 9.271646703258674e-07,
+      "loss": 0.954,
+      "step": 36758
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.874876350293754,
+      "learning_rate": 9.268438231470656e-07,
+      "loss": 0.8483,
+      "step": 36759
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4864910165081566,
+      "learning_rate": 9.265230287950499e-07,
+      "loss": 0.8303,
+      "step": 36760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0185899591559346,
+      "learning_rate": 9.262022872716914e-07,
+      "loss": 0.8135,
+      "step": 36761
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.454232098478635,
+      "learning_rate": 9.25881598578856e-07,
+      "loss": 0.8271,
+      "step": 36762
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.553948540235552,
+      "learning_rate": 9.255609627184092e-07,
+      "loss": 0.8703,
+      "step": 36763
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.828542880252006,
+      "learning_rate": 9.252403796922193e-07,
+      "loss": 0.7476,
+      "step": 36764
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8307393062338292,
+      "learning_rate": 9.249198495021549e-07,
+      "loss": 0.7484,
+      "step": 36765
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4520177544686734,
+      "learning_rate": 9.245993721500801e-07,
+      "loss": 0.6862,
+      "step": 36766
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5538591217604047,
+      "learning_rate": 9.242789476378589e-07,
+      "loss": 0.8316,
+      "step": 36767
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8837913450567894,
+      "learning_rate": 9.239585759673586e-07,
+      "loss": 0.8969,
+      "step": 36768
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6523274885434533,
+      "learning_rate": 9.236382571404478e-07,
+      "loss": 0.7955,
+      "step": 36769
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.595457022938621,
+      "learning_rate": 9.233179911589874e-07,
+      "loss": 0.8895,
+      "step": 36770
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.66580119379759,
+      "learning_rate": 9.229977780248423e-07,
+      "loss": 0.6907,
+      "step": 36771
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0532195909814033,
+      "learning_rate": 9.226776177398778e-07,
+      "loss": 0.7576,
+      "step": 36772
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7416992241640468,
+      "learning_rate": 9.223575103059601e-07,
+      "loss": 0.9023,
+      "step": 36773
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9223276450330393,
+      "learning_rate": 9.2203745572495e-07,
+      "loss": 0.919,
+      "step": 36774
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5444839009845344,
+      "learning_rate": 9.217174539987106e-07,
+      "loss": 0.7753,
+      "step": 36775
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4925417938803398,
+      "learning_rate": 9.21397505129108e-07,
+      "loss": 0.7873,
+      "step": 36776
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4651105866186365,
+      "learning_rate": 9.210776091180018e-07,
+      "loss": 0.8105,
+      "step": 36777
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6136583118050807,
+      "learning_rate": 9.207577659672573e-07,
+      "loss": 0.7168,
+      "step": 36778
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6208915384692544,
+      "learning_rate": 9.204379756787341e-07,
+      "loss": 0.9055,
+      "step": 36779
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4705186910048922,
+      "learning_rate": 9.201182382542983e-07,
+      "loss": 0.791,
+      "step": 36780
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.517910757426711,
+      "learning_rate": 9.197985536958054e-07,
+      "loss": 0.9067,
+      "step": 36781
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7652493975943289,
+      "learning_rate": 9.194789220051214e-07,
+      "loss": 0.8845,
+      "step": 36782
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4479968676050143,
+      "learning_rate": 9.191593431841084e-07,
+      "loss": 0.7557,
+      "step": 36783
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7429711437960975,
+      "learning_rate": 9.188398172346214e-07,
+      "loss": 0.8412,
+      "step": 36784
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4650437214970806,
+      "learning_rate": 9.185203441585244e-07,
+      "loss": 0.7819,
+      "step": 36785
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5560109056370361,
+      "learning_rate": 9.182009239576773e-07,
+      "loss": 0.9685,
+      "step": 36786
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7726299088765103,
+      "learning_rate": 9.178815566339427e-07,
+      "loss": 0.7642,
+      "step": 36787
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.744061166553845,
+      "learning_rate": 9.175622421891738e-07,
+      "loss": 0.7305,
+      "step": 36788
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7192412065981086,
+      "learning_rate": 9.172429806252325e-07,
+      "loss": 0.8662,
+      "step": 36789
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6438481291113227,
+      "learning_rate": 9.169237719439805e-07,
+      "loss": 0.7945,
+      "step": 36790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9708722501461344,
+      "learning_rate": 9.16604616147273e-07,
+      "loss": 0.7172,
+      "step": 36791
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5154732543015628,
+      "learning_rate": 9.162855132369686e-07,
+      "loss": 0.7892,
+      "step": 36792
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6665185185798297,
+      "learning_rate": 9.159664632149245e-07,
+      "loss": 0.9195,
+      "step": 36793
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4867767359032695,
+      "learning_rate": 9.156474660830028e-07,
+      "loss": 0.7485,
+      "step": 36794
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6540915776896559,
+      "learning_rate": 9.153285218430552e-07,
+      "loss": 0.9117,
+      "step": 36795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8482969977681842,
+      "learning_rate": 9.150096304969425e-07,
+      "loss": 0.8514,
+      "step": 36796
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6982822404749465,
+      "learning_rate": 9.146907920465197e-07,
+      "loss": 0.7384,
+      "step": 36797
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8960288910641094,
+      "learning_rate": 9.143720064936423e-07,
+      "loss": 0.7272,
+      "step": 36798
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8385251046303268,
+      "learning_rate": 9.140532738401675e-07,
+      "loss": 0.807,
+      "step": 36799
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6720558397640135,
+      "learning_rate": 9.137345940879538e-07,
+      "loss": 0.9084,
+      "step": 36800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7859729238106994,
+      "learning_rate": 9.134159672388531e-07,
+      "loss": 0.7067,
+      "step": 36801
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5725221809801169,
+      "learning_rate": 9.130973932947196e-07,
+      "loss": 0.9339,
+      "step": 36802
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7919133986594054,
+      "learning_rate": 9.127788722574094e-07,
+      "loss": 0.9705,
+      "step": 36803
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1526457738324636,
+      "learning_rate": 9.124604041287799e-07,
+      "loss": 0.8769,
+      "step": 36804
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5230265332420043,
+      "learning_rate": 9.121419889106831e-07,
+      "loss": 0.777,
+      "step": 36805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7542765002290106,
+      "learning_rate": 9.118236266049707e-07,
+      "loss": 0.7218,
+      "step": 36806
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.1009849041364,
+      "learning_rate": 9.115053172134991e-07,
+      "loss": 0.8766,
+      "step": 36807
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.439391574892687,
+      "learning_rate": 9.111870607381223e-07,
+      "loss": 0.6715,
+      "step": 36808
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5479161884365036,
+      "learning_rate": 9.108688571806912e-07,
+      "loss": 0.9303,
+      "step": 36809
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.443675469329107,
+      "learning_rate": 9.105507065430586e-07,
+      "loss": 0.724,
+      "step": 36810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5662416723539532,
+      "learning_rate": 9.102326088270763e-07,
+      "loss": 0.867,
+      "step": 36811
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6506186877381541,
+      "learning_rate": 9.099145640346007e-07,
+      "loss": 0.8717,
+      "step": 36812
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.65480444283937,
+      "learning_rate": 9.095965721674793e-07,
+      "loss": 0.8391,
+      "step": 36813
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6038411816340254,
+      "learning_rate": 9.092786332275627e-07,
+      "loss": 0.8424,
+      "step": 36814
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.74600837717694,
+      "learning_rate": 9.089607472167061e-07,
+      "loss": 0.9585,
+      "step": 36815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7531687249840794,
+      "learning_rate": 9.08642914136757e-07,
+      "loss": 0.717,
+      "step": 36816
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6072037185848387,
+      "learning_rate": 9.083251339895682e-07,
+      "loss": 0.8102,
+      "step": 36817
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4253845922947985,
+      "learning_rate": 9.080074067769872e-07,
+      "loss": 0.7738,
+      "step": 36818
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.667214983136517,
+      "learning_rate": 9.076897325008671e-07,
+      "loss": 0.921,
+      "step": 36819
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.599239140441366,
+      "learning_rate": 9.073721111630552e-07,
+      "loss": 0.7209,
+      "step": 36820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8457407464376216,
+      "learning_rate": 9.070545427654021e-07,
+      "loss": 0.7721,
+      "step": 36821
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7558206041490533,
+      "learning_rate": 9.067370273097575e-07,
+      "loss": 0.7083,
+      "step": 36822
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6450297702133518,
+      "learning_rate": 9.064195647979656e-07,
+      "loss": 0.8715,
+      "step": 36823
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8878737514921997,
+      "learning_rate": 9.061021552318783e-07,
+      "loss": 0.8665,
+      "step": 36824
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.643798366230988,
+      "learning_rate": 9.057847986133461e-07,
+      "loss": 0.7253,
+      "step": 36825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7649158512725542,
+      "learning_rate": 9.054674949442132e-07,
+      "loss": 0.8447,
+      "step": 36826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6169384110029583,
+      "learning_rate": 9.05150244226326e-07,
+      "loss": 0.8134,
+      "step": 36827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5580441589084744,
+      "learning_rate": 9.04833046461534e-07,
+      "loss": 0.7275,
+      "step": 36828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6978629433691643,
+      "learning_rate": 9.045159016516847e-07,
+      "loss": 0.8176,
+      "step": 36829
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5440299943344518,
+      "learning_rate": 9.041988097986232e-07,
+      "loss": 0.7905,
+      "step": 36830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6918158317299377,
+      "learning_rate": 9.038817709041947e-07,
+      "loss": 0.8611,
+      "step": 36831
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.825994693398904,
+      "learning_rate": 9.035647849702467e-07,
+      "loss": 0.8239,
+      "step": 36832
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.791614617750113,
+      "learning_rate": 9.032478519986254e-07,
+      "loss": 0.7279,
+      "step": 36833
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3077098911911946,
+      "learning_rate": 9.02930971991175e-07,
+      "loss": 0.8802,
+      "step": 36834
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.615086709995255,
+      "learning_rate": 9.026141449497394e-07,
+      "loss": 0.826,
+      "step": 36835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5318569031622697,
+      "learning_rate": 9.022973708761662e-07,
+      "loss": 0.8447,
+      "step": 36836
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5619308758601749,
+      "learning_rate": 9.019806497722961e-07,
+      "loss": 0.804,
+      "step": 36837
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.099625074139387,
+      "learning_rate": 9.016639816399764e-07,
+      "loss": 0.9278,
+      "step": 36838
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5055384299513728,
+      "learning_rate": 9.013473664810479e-07,
+      "loss": 0.7813,
+      "step": 36839
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5168324419593453,
+      "learning_rate": 9.01030804297357e-07,
+      "loss": 0.6527,
+      "step": 36840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0172012874767744,
+      "learning_rate": 9.007142950907444e-07,
+      "loss": 0.816,
+      "step": 36841
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6354855812087128,
+      "learning_rate": 9.003978388630552e-07,
+      "loss": 0.9924,
+      "step": 36842
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7763598409865929,
+      "learning_rate": 9.000814356161314e-07,
+      "loss": 0.7218,
+      "step": 36843
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5219908874748875,
+      "learning_rate": 8.997650853518125e-07,
+      "loss": 0.8613,
+      "step": 36844
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7451645561217655,
+      "learning_rate": 8.994487880719415e-07,
+      "loss": 0.6675,
+      "step": 36845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6940707692669068,
+      "learning_rate": 8.991325437783627e-07,
+      "loss": 0.779,
+      "step": 36846
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6020362912426434,
+      "learning_rate": 8.988163524729165e-07,
+      "loss": 0.7734,
+      "step": 36847
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.866800870604748,
+      "learning_rate": 8.985002141574406e-07,
+      "loss": 0.6858,
+      "step": 36848
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7295590173193662,
+      "learning_rate": 8.981841288337778e-07,
+      "loss": 0.8248,
+      "step": 36849
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8172156906872812,
+      "learning_rate": 8.978680965037712e-07,
+      "loss": 0.8075,
+      "step": 36850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7934505132195302,
+      "learning_rate": 8.975521171692569e-07,
+      "loss": 0.7879,
+      "step": 36851
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.753005542456233,
+      "learning_rate": 8.972361908320748e-07,
+      "loss": 0.72,
+      "step": 36852
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7236277409997117,
+      "learning_rate": 8.969203174940655e-07,
+      "loss": 0.7124,
+      "step": 36853
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7461951648246525,
+      "learning_rate": 8.966044971570698e-07,
+      "loss": 0.8173,
+      "step": 36854
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7543076146850641,
+      "learning_rate": 8.96288729822925e-07,
+      "loss": 0.7423,
+      "step": 36855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.026387200156168,
+      "learning_rate": 8.959730154934676e-07,
+      "loss": 0.9244,
+      "step": 36856
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.779245396353811,
+      "learning_rate": 8.956573541705393e-07,
+      "loss": 0.8068,
+      "step": 36857
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6187518722118777,
+      "learning_rate": 8.953417458559743e-07,
+      "loss": 0.7908,
+      "step": 36858
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.569367321757818,
+      "learning_rate": 8.950261905516145e-07,
+      "loss": 0.8275,
+      "step": 36859
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6187847336113026,
+      "learning_rate": 8.947106882592927e-07,
+      "loss": 0.8617,
+      "step": 36860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8283088243776302,
+      "learning_rate": 8.943952389808497e-07,
+      "loss": 0.8768,
+      "step": 36861
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5350194772012447,
+      "learning_rate": 8.940798427181185e-07,
+      "loss": 0.7459,
+      "step": 36862
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6488689106428474,
+      "learning_rate": 8.937644994729388e-07,
+      "loss": 0.8692,
+      "step": 36863
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5809984471611747,
+      "learning_rate": 8.934492092471459e-07,
+      "loss": 0.8535,
+      "step": 36864
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5303755174146547,
+      "learning_rate": 8.931339720425725e-07,
+      "loss": 0.7688,
+      "step": 36865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6559778215275065,
+      "learning_rate": 8.928187878610561e-07,
+      "loss": 0.8497,
+      "step": 36866
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5478718119644508,
+      "learning_rate": 8.925036567044332e-07,
+      "loss": 0.8278,
+      "step": 36867
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7461712363834183,
+      "learning_rate": 8.921885785745399e-07,
+      "loss": 0.6864,
+      "step": 36868
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6840678998263554,
+      "learning_rate": 8.918735534732048e-07,
+      "loss": 0.7151,
+      "step": 36869
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.905070841676744,
+      "learning_rate": 8.915585814022653e-07,
+      "loss": 0.7293,
+      "step": 36870
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.698036038030639,
+      "learning_rate": 8.912436623635579e-07,
+      "loss": 0.8568,
+      "step": 36871
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6919729274487794,
+      "learning_rate": 8.90928796358913e-07,
+      "loss": 1.0899,
+      "step": 36872
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4655547829351967,
+      "learning_rate": 8.906139833901629e-07,
+      "loss": 0.7044,
+      "step": 36873
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7824461777471547,
+      "learning_rate": 8.902992234591424e-07,
+      "loss": 0.965,
+      "step": 36874
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6044722616305909,
+      "learning_rate": 8.899845165676857e-07,
+      "loss": 0.7064,
+      "step": 36875
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6919338638720882,
+      "learning_rate": 8.896698627176226e-07,
+      "loss": 0.8941,
+      "step": 36876
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.780179130040416,
+      "learning_rate": 8.893552619107848e-07,
+      "loss": 0.8308,
+      "step": 36877
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.28926024070071,
+      "learning_rate": 8.890407141490065e-07,
+      "loss": 0.7009,
+      "step": 36878
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7027386344972402,
+      "learning_rate": 8.887262194341162e-07,
+      "loss": 0.9692,
+      "step": 36879
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.74063457369684,
+      "learning_rate": 8.884117777679458e-07,
+      "loss": 0.6446,
+      "step": 36880
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6802799454956745,
+      "learning_rate": 8.880973891523292e-07,
+      "loss": 0.8882,
+      "step": 36881
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5797522966903512,
+      "learning_rate": 8.877830535890941e-07,
+      "loss": 0.9479,
+      "step": 36882
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6675682634696267,
+      "learning_rate": 8.874687710800689e-07,
+      "loss": 0.9141,
+      "step": 36883
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7197741975271388,
+      "learning_rate": 8.871545416270855e-07,
+      "loss": 0.7707,
+      "step": 36884
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.041672721336215,
+      "learning_rate": 8.868403652319779e-07,
+      "loss": 0.7493,
+      "step": 36885
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5902054639764267,
+      "learning_rate": 8.865262418965659e-07,
+      "loss": 0.7637,
+      "step": 36886
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.989682343432936,
+      "learning_rate": 8.862121716226846e-07,
+      "loss": 0.7509,
+      "step": 36887
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5569344247835097,
+      "learning_rate": 8.858981544121614e-07,
+      "loss": 0.9723,
+      "step": 36888
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.140072133484241,
+      "learning_rate": 8.85584190266825e-07,
+      "loss": 0.9057,
+      "step": 36889
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5928379957511902,
+      "learning_rate": 8.852702791885048e-07,
+      "loss": 0.7532,
+      "step": 36890
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9418902215761145,
+      "learning_rate": 8.84956421179024e-07,
+      "loss": 0.9015,
+      "step": 36891
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.663183591243462,
+      "learning_rate": 8.846426162402144e-07,
+      "loss": 0.7145,
+      "step": 36892
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4490043024477055,
+      "learning_rate": 8.843288643739001e-07,
+      "loss": 0.6924,
+      "step": 36893
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8362944834512256,
+      "learning_rate": 8.840151655819107e-07,
+      "loss": 0.6685,
+      "step": 36894
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4291105904319423,
+      "learning_rate": 8.837015198660681e-07,
+      "loss": 0.7922,
+      "step": 36895
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6335390500363787,
+      "learning_rate": 8.833879272282053e-07,
+      "loss": 0.8318,
+      "step": 36896
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5659344771237331,
+      "learning_rate": 8.830743876701408e-07,
+      "loss": 0.7329,
+      "step": 36897
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8473443758946417,
+      "learning_rate": 8.827609011937066e-07,
+      "loss": 0.7932,
+      "step": 36898
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7062741029430952,
+      "learning_rate": 8.824474678007244e-07,
+      "loss": 0.8425,
+      "step": 36899
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3955386480320924,
+      "learning_rate": 8.821340874930173e-07,
+      "loss": 0.8501,
+      "step": 36900
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5392009967147073,
+      "learning_rate": 8.818207602724127e-07,
+      "loss": 0.843,
+      "step": 36901
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7651998837204481,
+      "learning_rate": 8.815074861407358e-07,
+      "loss": 0.7351,
+      "step": 36902
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5205931984073973,
+      "learning_rate": 8.811942650998096e-07,
+      "loss": 0.7314,
+      "step": 36903
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7352250848348598,
+      "learning_rate": 8.808810971514559e-07,
+      "loss": 0.7083,
+      "step": 36904
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6248430952688961,
+      "learning_rate": 8.805679822974988e-07,
+      "loss": 0.7083,
+      "step": 36905
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3654805005765212,
+      "learning_rate": 8.802549205397637e-07,
+      "loss": 0.6113,
+      "step": 36906
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7304610811209624,
+      "learning_rate": 8.799419118800722e-07,
+      "loss": 0.7059,
+      "step": 36907
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.57474939769855,
+      "learning_rate": 8.796289563202454e-07,
+      "loss": 0.9065,
+      "step": 36908
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6850155971638099,
+      "learning_rate": 8.793160538621048e-07,
+      "loss": 0.8525,
+      "step": 36909
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6150598374582759,
+      "learning_rate": 8.790032045074759e-07,
+      "loss": 0.8535,
+      "step": 36910
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.496231911778042,
+      "learning_rate": 8.786904082581782e-07,
+      "loss": 0.9238,
+      "step": 36911
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5326879252851104,
+      "learning_rate": 8.783776651160303e-07,
+      "loss": 0.9014,
+      "step": 36912
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7101286246482406,
+      "learning_rate": 8.780649750828585e-07,
+      "loss": 0.8447,
+      "step": 36913
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6737547943099853,
+      "learning_rate": 8.77752338160478e-07,
+      "loss": 0.9124,
+      "step": 36914
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.631116664155015,
+      "learning_rate": 8.774397543507129e-07,
+      "loss": 0.8036,
+      "step": 36915
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5555415318335797,
+      "learning_rate": 8.771272236553796e-07,
+      "loss": 0.6848,
+      "step": 36916
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5676989903434317,
+      "learning_rate": 8.768147460763021e-07,
+      "loss": 0.8677,
+      "step": 36917
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.720350656053165,
+      "learning_rate": 8.765023216152969e-07,
+      "loss": 0.8554,
+      "step": 36918
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4878001449402192,
+      "learning_rate": 8.761899502741833e-07,
+      "loss": 0.8356,
+      "step": 36919
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.570189740362734,
+      "learning_rate": 8.758776320547812e-07,
+      "loss": 0.8264,
+      "step": 36920
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7716822413485618,
+      "learning_rate": 8.75565366958907e-07,
+      "loss": 0.901,
+      "step": 36921
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9559753295070472,
+      "learning_rate": 8.752531549883791e-07,
+      "loss": 0.8041,
+      "step": 36922
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7751464466410349,
+      "learning_rate": 8.749409961450183e-07,
+      "loss": 0.7022,
+      "step": 36923
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.940563484575678,
+      "learning_rate": 8.74628890430641e-07,
+      "loss": 0.932,
+      "step": 36924
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6610017851626362,
+      "learning_rate": 8.743168378470601e-07,
+      "loss": 0.7712,
+      "step": 36925
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.567792697017635,
+      "learning_rate": 8.740048383960964e-07,
+      "loss": 0.978,
+      "step": 36926
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.776698033542905,
+      "learning_rate": 8.736928920795674e-07,
+      "loss": 0.7935,
+      "step": 36927
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4308231876366744,
+      "learning_rate": 8.733809988992881e-07,
+      "loss": 0.8941,
+      "step": 36928
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4606370242183773,
+      "learning_rate": 8.730691588570716e-07,
+      "loss": 0.8323,
+      "step": 36929
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8131561063833872,
+      "learning_rate": 8.727573719547367e-07,
+      "loss": 0.7365,
+      "step": 36930
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7141389705430092,
+      "learning_rate": 8.724456381940994e-07,
+      "loss": 0.8557,
+      "step": 36931
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4840266349888618,
+      "learning_rate": 8.721339575769727e-07,
+      "loss": 0.8363,
+      "step": 36932
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6222729402416105,
+      "learning_rate": 8.71822330105171e-07,
+      "loss": 0.8019,
+      "step": 36933
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.008637047460108,
+      "learning_rate": 8.715107557805092e-07,
+      "loss": 0.7965,
+      "step": 36934
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5896234094202857,
+      "learning_rate": 8.711992346048038e-07,
+      "loss": 0.6676,
+      "step": 36935
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.634813437302454,
+      "learning_rate": 8.708877665798666e-07,
+      "loss": 0.7413,
+      "step": 36936
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7214597826748768,
+      "learning_rate": 8.705763517075094e-07,
+      "loss": 0.9289,
+      "step": 36937
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5996380234721648,
+      "learning_rate": 8.702649899895488e-07,
+      "loss": 0.7726,
+      "step": 36938
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.636101536955915,
+      "learning_rate": 8.699536814277931e-07,
+      "loss": 0.8265,
+      "step": 36939
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5260407909540759,
+      "learning_rate": 8.69642426024061e-07,
+      "loss": 0.7942,
+      "step": 36940
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.390550929670926,
+      "learning_rate": 8.693312237801588e-07,
+      "loss": 0.8745,
+      "step": 36941
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6622547678331343,
+      "learning_rate": 8.690200746979016e-07,
+      "loss": 0.8377,
+      "step": 36942
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6401015334260505,
+      "learning_rate": 8.687089787791003e-07,
+      "loss": 0.948,
+      "step": 36943
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6737252476121547,
+      "learning_rate": 8.683979360255668e-07,
+      "loss": 0.8747,
+      "step": 36944
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.450705864530567,
+      "learning_rate": 8.680869464391129e-07,
+      "loss": 0.7926,
+      "step": 36945
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7612272438664123,
+      "learning_rate": 8.677760100215449e-07,
+      "loss": 0.8774,
+      "step": 36946
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8814365626526894,
+      "learning_rate": 8.674651267746769e-07,
+      "loss": 0.8779,
+      "step": 36947
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7426813556545331,
+      "learning_rate": 8.671542967003188e-07,
+      "loss": 0.7459,
+      "step": 36948
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7202231426320358,
+      "learning_rate": 8.668435198002812e-07,
+      "loss": 0.8936,
+      "step": 36949
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5519422784177404,
+      "learning_rate": 8.665327960763692e-07,
+      "loss": 0.8275,
+      "step": 36950
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9222988075641039,
+      "learning_rate": 8.66222125530396e-07,
+      "loss": 0.8355,
+      "step": 36951
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7322033798249528,
+      "learning_rate": 8.659115081641712e-07,
+      "loss": 0.728,
+      "step": 36952
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1119121964179257,
+      "learning_rate": 8.656009439794998e-07,
+      "loss": 0.9112,
+      "step": 36953
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.606299710179284,
+      "learning_rate": 8.652904329781908e-07,
+      "loss": 0.7424,
+      "step": 36954
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.578028365131823,
+      "learning_rate": 8.649799751620536e-07,
+      "loss": 0.7788,
+      "step": 36955
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.756255601763117,
+      "learning_rate": 8.646695705328955e-07,
+      "loss": 0.8257,
+      "step": 36956
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6159218256208647,
+      "learning_rate": 8.643592190925243e-07,
+      "loss": 0.8,
+      "step": 36957
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7247151226923806,
+      "learning_rate": 8.640489208427438e-07,
+      "loss": 0.6755,
+      "step": 36958
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6167815517483148,
+      "learning_rate": 8.637386757853649e-07,
+      "loss": 0.8075,
+      "step": 36959
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5388615241261514,
+      "learning_rate": 8.634284839221907e-07,
+      "loss": 0.763,
+      "step": 36960
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.556017536405375,
+      "learning_rate": 8.631183452550295e-07,
+      "loss": 0.8945,
+      "step": 36961
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4241698471532895,
+      "learning_rate": 8.628082597856835e-07,
+      "loss": 0.8015,
+      "step": 36962
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7153311567429166,
+      "learning_rate": 8.624982275159633e-07,
+      "loss": 0.9038,
+      "step": 36963
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4852123823079404,
+      "learning_rate": 8.621882484476695e-07,
+      "loss": 0.925,
+      "step": 36964
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5252660536087248,
+      "learning_rate": 8.618783225826088e-07,
+      "loss": 0.8585,
+      "step": 36965
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7522339684969037,
+      "learning_rate": 8.615684499225884e-07,
+      "loss": 0.8516,
+      "step": 36966
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4674298319841779,
+      "learning_rate": 8.61258630469407e-07,
+      "loss": 0.9012,
+      "step": 36967
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7160186239101727,
+      "learning_rate": 8.609488642248709e-07,
+      "loss": 0.9091,
+      "step": 36968
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.486644607086232,
+      "learning_rate": 8.606391511907841e-07,
+      "loss": 0.742,
+      "step": 36969
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7737639437728013,
+      "learning_rate": 8.60329491368953e-07,
+      "loss": 0.7355,
+      "step": 36970
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5278989819718567,
+      "learning_rate": 8.60019884761173e-07,
+      "loss": 0.8375,
+      "step": 36971
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6452174337999235,
+      "learning_rate": 8.597103313692512e-07,
+      "loss": 0.8762,
+      "step": 36972
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6218002077295754,
+      "learning_rate": 8.59400831194992e-07,
+      "loss": 0.8812,
+      "step": 36973
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7752746984745618,
+      "learning_rate": 8.590913842401949e-07,
+      "loss": 0.6958,
+      "step": 36974
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.58733070510586,
+      "learning_rate": 8.587819905066597e-07,
+      "loss": 0.8277,
+      "step": 36975
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5710672878608836,
+      "learning_rate": 8.584726499961893e-07,
+      "loss": 0.9092,
+      "step": 36976
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5573946586086496,
+      "learning_rate": 8.581633627105879e-07,
+      "loss": 0.8293,
+      "step": 36977
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5654434133335147,
+      "learning_rate": 8.578541286516506e-07,
+      "loss": 0.81,
+      "step": 36978
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.262597618455371,
+      "learning_rate": 8.575449478211839e-07,
+      "loss": 0.7746,
+      "step": 36979
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7324507894120345,
+      "learning_rate": 8.572358202209851e-07,
+      "loss": 0.6063,
+      "step": 36980
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5683165525606275,
+      "learning_rate": 8.569267458528519e-07,
+      "loss": 0.7315,
+      "step": 36981
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.723016869805832,
+      "learning_rate": 8.566177247185847e-07,
+      "loss": 0.7656,
+      "step": 36982
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6521165609828905,
+      "learning_rate": 8.563087568199868e-07,
+      "loss": 0.8095,
+      "step": 36983
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5867924498395907,
+      "learning_rate": 8.559998421588545e-07,
+      "loss": 0.8995,
+      "step": 36984
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.638952980742679,
+      "learning_rate": 8.55690980736984e-07,
+      "loss": 0.9291,
+      "step": 36985
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3134545921210967,
+      "learning_rate": 8.553821725561751e-07,
+      "loss": 0.8887,
+      "step": 36986
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1771834090662034,
+      "learning_rate": 8.550734176182285e-07,
+      "loss": 0.7913,
+      "step": 36987
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6794050420672977,
+      "learning_rate": 8.547647159249395e-07,
+      "loss": 0.8524,
+      "step": 36988
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5424632244392995,
+      "learning_rate": 8.544560674781033e-07,
+      "loss": 0.8616,
+      "step": 36989
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.709447889021239,
+      "learning_rate": 8.541474722795207e-07,
+      "loss": 0.756,
+      "step": 36990
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7757752198864478,
+      "learning_rate": 8.538389303309879e-07,
+      "loss": 0.7655,
+      "step": 36991
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.687387336561785,
+      "learning_rate": 8.535304416343005e-07,
+      "loss": 0.8361,
+      "step": 36992
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5301193555588264,
+      "learning_rate": 8.532220061912533e-07,
+      "loss": 0.8149,
+      "step": 36993
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6112423676912317,
+      "learning_rate": 8.529136240036439e-07,
+      "loss": 0.7239,
+      "step": 36994
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6547756269365739,
+      "learning_rate": 8.526052950732666e-07,
+      "loss": 0.8378,
+      "step": 36995
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7710978233851533,
+      "learning_rate": 8.522970194019187e-07,
+      "loss": 0.9017,
+      "step": 36996
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4920215055624082,
+      "learning_rate": 8.51988796991392e-07,
+      "loss": 0.7295,
+      "step": 36997
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9115946811736757,
+      "learning_rate": 8.516806278434852e-07,
+      "loss": 0.7333,
+      "step": 36998
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.627713855965269,
+      "learning_rate": 8.51372511959988e-07,
+      "loss": 0.8476,
+      "step": 36999
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.103191089020719,
+      "learning_rate": 8.510644493426979e-07,
+      "loss": 0.8548,
+      "step": 37000
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6430971224947635,
+      "learning_rate": 8.507564399934065e-07,
+      "loss": 0.8267,
+      "step": 37001
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7711383205292782,
+      "learning_rate": 8.504484839139071e-07,
+      "loss": 0.7012,
+      "step": 37002
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6574207497291062,
+      "learning_rate": 8.501405811059926e-07,
+      "loss": 0.7564,
+      "step": 37003
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.571749121502793,
+      "learning_rate": 8.498327315714583e-07,
+      "loss": 0.7506,
+      "step": 37004
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6186231373903686,
+      "learning_rate": 8.495249353120949e-07,
+      "loss": 0.8684,
+      "step": 37005
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7963776049064685,
+      "learning_rate": 8.492171923296922e-07,
+      "loss": 0.7475,
+      "step": 37006
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5365439592936723,
+      "learning_rate": 8.489095026260452e-07,
+      "loss": 0.7217,
+      "step": 37007
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6488129505097133,
+      "learning_rate": 8.48601866202945e-07,
+      "loss": 0.8565,
+      "step": 37008
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7129048064630565,
+      "learning_rate": 8.482942830621821e-07,
+      "loss": 0.8539,
+      "step": 37009
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.450506483776303,
+      "learning_rate": 8.479867532055452e-07,
+      "loss": 0.9415,
+      "step": 37010
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6564113480470015,
+      "learning_rate": 8.476792766348274e-07,
+      "loss": 0.797,
+      "step": 37011
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7943586145474865,
+      "learning_rate": 8.473718533518205e-07,
+      "loss": 0.6894,
+      "step": 37012
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.809223314964033,
+      "learning_rate": 8.470644833583119e-07,
+      "loss": 0.8786,
+      "step": 37013
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7809615837376856,
+      "learning_rate": 8.467571666560903e-07,
+      "loss": 0.7052,
+      "step": 37014
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5072730827740217,
+      "learning_rate": 8.464499032469475e-07,
+      "loss": 0.8408,
+      "step": 37015
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4688965385162975,
+      "learning_rate": 8.461426931326699e-07,
+      "loss": 0.858,
+      "step": 37016
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.640727809043405,
+      "learning_rate": 8.458355363150494e-07,
+      "loss": 0.8077,
+      "step": 37017
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6987912296216412,
+      "learning_rate": 8.455284327958702e-07,
+      "loss": 0.8711,
+      "step": 37018
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5428901398075519,
+      "learning_rate": 8.452213825769251e-07,
+      "loss": 0.6965,
+      "step": 37019
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5735750214614161,
+      "learning_rate": 8.449143856599973e-07,
+      "loss": 0.777,
+      "step": 37020
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5110093451851878,
+      "learning_rate": 8.446074420468786e-07,
+      "loss": 0.8626,
+      "step": 37021
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7898233274823153,
+      "learning_rate": 8.443005517393543e-07,
+      "loss": 0.7515,
+      "step": 37022
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.552960720842615,
+      "learning_rate": 8.439937147392086e-07,
+      "loss": 0.8424,
+      "step": 37023
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7830257467422692,
+      "learning_rate": 8.436869310482298e-07,
+      "loss": 0.6872,
+      "step": 37024
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5409258519353664,
+      "learning_rate": 8.433802006682068e-07,
+      "loss": 0.817,
+      "step": 37025
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7799284069537455,
+      "learning_rate": 8.430735236009225e-07,
+      "loss": 0.7998,
+      "step": 37026
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.806806202852988,
+      "learning_rate": 8.427668998481619e-07,
+      "loss": 0.891,
+      "step": 37027
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7828413704659483,
+      "learning_rate": 8.424603294117118e-07,
+      "loss": 0.7106,
+      "step": 37028
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4866407488740394,
+      "learning_rate": 8.421538122933593e-07,
+      "loss": 0.8345,
+      "step": 37029
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5648775960288435,
+      "learning_rate": 8.418473484948853e-07,
+      "loss": 0.8555,
+      "step": 37030
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.785696807700123,
+      "learning_rate": 8.41540938018074e-07,
+      "loss": 0.8508,
+      "step": 37031
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6549529520361055,
+      "learning_rate": 8.412345808647115e-07,
+      "loss": 0.8029,
+      "step": 37032
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7972685311703729,
+      "learning_rate": 8.409282770365823e-07,
+      "loss": 0.8842,
+      "step": 37033
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.630429262293612,
+      "learning_rate": 8.406220265354681e-07,
+      "loss": 0.673,
+      "step": 37034
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4927638786047426,
+      "learning_rate": 8.403158293631508e-07,
+      "loss": 0.893,
+      "step": 37035
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7291919185346236,
+      "learning_rate": 8.400096855214157e-07,
+      "loss": 0.8934,
+      "step": 37036
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.387491201223209,
+      "learning_rate": 8.397035950120435e-07,
+      "loss": 0.7025,
+      "step": 37037
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.48992863656994,
+      "learning_rate": 8.393975578368186e-07,
+      "loss": 0.8325,
+      "step": 37038
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.849727181166202,
+      "learning_rate": 8.390915739975192e-07,
+      "loss": 0.8803,
+      "step": 37039
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9274387372368988,
+      "learning_rate": 8.387856434959307e-07,
+      "loss": 0.8232,
+      "step": 37040
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5046933198794123,
+      "learning_rate": 8.384797663338307e-07,
+      "loss": 0.886,
+      "step": 37041
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6131371111592525,
+      "learning_rate": 8.381739425130042e-07,
+      "loss": 0.8337,
+      "step": 37042
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7582758288306498,
+      "learning_rate": 8.3786817203523e-07,
+      "loss": 0.6512,
+      "step": 37043
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.572745783342823,
+      "learning_rate": 8.375624549022854e-07,
+      "loss": 0.8849,
+      "step": 37044
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.633927337304644,
+      "learning_rate": 8.372567911159535e-07,
+      "loss": 0.7852,
+      "step": 37045
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7279216670414435,
+      "learning_rate": 8.369511806780151e-07,
+      "loss": 0.9151,
+      "step": 37046
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5442237199592426,
+      "learning_rate": 8.366456235902476e-07,
+      "loss": 0.9021,
+      "step": 37047
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5672947349898285,
+      "learning_rate": 8.363401198544285e-07,
+      "loss": 0.7869,
+      "step": 37048
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9622174880185996,
+      "learning_rate": 8.360346694723387e-07,
+      "loss": 0.7313,
+      "step": 37049
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9689697198926095,
+      "learning_rate": 8.357292724457577e-07,
+      "loss": 0.9052,
+      "step": 37050
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7198227232071803,
+      "learning_rate": 8.35423928776462e-07,
+      "loss": 0.9159,
+      "step": 37051
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6706004443994524,
+      "learning_rate": 8.351186384662291e-07,
+      "loss": 0.9034,
+      "step": 37052
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4870767258420172,
+      "learning_rate": 8.348134015168364e-07,
+      "loss": 0.8691,
+      "step": 37053
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5230907886614662,
+      "learning_rate": 8.345082179300634e-07,
+      "loss": 0.7966,
+      "step": 37054
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8233949965829108,
+      "learning_rate": 8.342030877076856e-07,
+      "loss": 0.8681,
+      "step": 37055
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8154288946686106,
+      "learning_rate": 8.338980108514772e-07,
+      "loss": 0.7448,
+      "step": 37056
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0089384526693768,
+      "learning_rate": 8.335929873632164e-07,
+      "loss": 0.8817,
+      "step": 37057
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6247451013055565,
+      "learning_rate": 8.332880172446811e-07,
+      "loss": 0.8771,
+      "step": 37058
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.485078636560217,
+      "learning_rate": 8.329831004976452e-07,
+      "loss": 0.7781,
+      "step": 37059
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6483412738611123,
+      "learning_rate": 8.326782371238818e-07,
+      "loss": 0.7927,
+      "step": 37060
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.598493059351102,
+      "learning_rate": 8.323734271251704e-07,
+      "loss": 0.772,
+      "step": 37061
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.611360156253855,
+      "learning_rate": 8.32068670503281e-07,
+      "loss": 0.8139,
+      "step": 37062
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.261384968261265,
+      "learning_rate": 8.317639672599909e-07,
+      "loss": 0.8052,
+      "step": 37063
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5260682569320383,
+      "learning_rate": 8.314593173970753e-07,
+      "loss": 0.9291,
+      "step": 37064
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4325599763812782,
+      "learning_rate": 8.311547209163052e-07,
+      "loss": 0.75,
+      "step": 37065
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7749367728739467,
+      "learning_rate": 8.308501778194545e-07,
+      "loss": 0.7317,
+      "step": 37066
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6349354280339423,
+      "learning_rate": 8.305456881082963e-07,
+      "loss": 0.803,
+      "step": 37067
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6403358528241219,
+      "learning_rate": 8.302412517846081e-07,
+      "loss": 0.8062,
+      "step": 37068
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7788713419085073,
+      "learning_rate": 8.299368688501552e-07,
+      "loss": 0.7273,
+      "step": 37069
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5786818242553604,
+      "learning_rate": 8.296325393067128e-07,
+      "loss": 0.8085,
+      "step": 37070
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5305599943284214,
+      "learning_rate": 8.293282631560529e-07,
+      "loss": 0.8417,
+      "step": 37071
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4899028389602118,
+      "learning_rate": 8.290240403999505e-07,
+      "loss": 0.7759,
+      "step": 37072
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5769748461191568,
+      "learning_rate": 8.287198710401701e-07,
+      "loss": 0.8652,
+      "step": 37073
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7808583988413903,
+      "learning_rate": 8.284157550784866e-07,
+      "loss": 0.6993,
+      "step": 37074
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7984818320905194,
+      "learning_rate": 8.281116925166721e-07,
+      "loss": 0.7838,
+      "step": 37075
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5545648219306216,
+      "learning_rate": 8.278076833564929e-07,
+      "loss": 0.8349,
+      "step": 37076
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7799068107647662,
+      "learning_rate": 8.275037275997233e-07,
+      "loss": 0.9416,
+      "step": 37077
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.65662675017873,
+      "learning_rate": 8.271998252481283e-07,
+      "loss": 0.8629,
+      "step": 37078
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.543591236671662,
+      "learning_rate": 8.268959763034834e-07,
+      "loss": 0.8342,
+      "step": 37079
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.799995696958599,
+      "learning_rate": 8.265921807675515e-07,
+      "loss": 0.9141,
+      "step": 37080
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6297081402462958,
+      "learning_rate": 8.262884386421055e-07,
+      "loss": 0.8495,
+      "step": 37081
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.96590522608649,
+      "learning_rate": 8.25984749928913e-07,
+      "loss": 0.8875,
+      "step": 37082
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6925984339407105,
+      "learning_rate": 8.256811146297405e-07,
+      "loss": 0.8565,
+      "step": 37083
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7146616484299246,
+      "learning_rate": 8.253775327463565e-07,
+      "loss": 0.8278,
+      "step": 37084
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5877446613990749,
+      "learning_rate": 8.250740042805317e-07,
+      "loss": 0.7439,
+      "step": 37085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.511137566865447,
+      "learning_rate": 8.247705292340302e-07,
+      "loss": 0.815,
+      "step": 37086
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.64921598696593,
+      "learning_rate": 8.244671076086174e-07,
+      "loss": 0.8679,
+      "step": 37087
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5886759276534184,
+      "learning_rate": 8.241637394060619e-07,
+      "loss": 0.8382,
+      "step": 37088
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7758093083100955,
+      "learning_rate": 8.238604246281323e-07,
+      "loss": 0.6868,
+      "step": 37089
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8507099152073272,
+      "learning_rate": 8.235571632765927e-07,
+      "loss": 0.8496,
+      "step": 37090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.086852279321101,
+      "learning_rate": 8.232539553532071e-07,
+      "loss": 0.8339,
+      "step": 37091
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6379988019822052,
+      "learning_rate": 8.229508008597409e-07,
+      "loss": 0.9216,
+      "step": 37092
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6347952400151904,
+      "learning_rate": 8.226476997979638e-07,
+      "loss": 0.8018,
+      "step": 37093
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6645278262572294,
+      "learning_rate": 8.223446521696366e-07,
+      "loss": 0.8904,
+      "step": 37094
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5882790546491778,
+      "learning_rate": 8.220416579765222e-07,
+      "loss": 0.7798,
+      "step": 37095
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4879131609746554,
+      "learning_rate": 8.217387172203895e-07,
+      "loss": 0.8712,
+      "step": 37096
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6063537782278117,
+      "learning_rate": 8.214358299029968e-07,
+      "loss": 1.0046,
+      "step": 37097
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7550252018925843,
+      "learning_rate": 8.211329960261138e-07,
+      "loss": 0.7272,
+      "step": 37098
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.692616945326824,
+      "learning_rate": 8.208302155914982e-07,
+      "loss": 0.8382,
+      "step": 37099
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8285353169831704,
+      "learning_rate": 8.205274886009162e-07,
+      "loss": 0.8678,
+      "step": 37100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7982324567315346,
+      "learning_rate": 8.202248150561288e-07,
+      "loss": 0.835,
+      "step": 37101
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7322003222590892,
+      "learning_rate": 8.199221949588998e-07,
+      "loss": 0.6891,
+      "step": 37102
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4977802619631737,
+      "learning_rate": 8.196196283109903e-07,
+      "loss": 0.9631,
+      "step": 37103
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8094939799769956,
+      "learning_rate": 8.193171151141599e-07,
+      "loss": 0.7378,
+      "step": 37104
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4592795513866232,
+      "learning_rate": 8.190146553701716e-07,
+      "loss": 0.6581,
+      "step": 37105
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5262676350583742,
+      "learning_rate": 8.187122490807897e-07,
+      "loss": 0.8322,
+      "step": 37106
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7759863291899189,
+      "learning_rate": 8.184098962477705e-07,
+      "loss": 0.6769,
+      "step": 37107
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6576622803280299,
+      "learning_rate": 8.181075968728746e-07,
+      "loss": 0.8757,
+      "step": 37108
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6523838063897438,
+      "learning_rate": 8.178053509578631e-07,
+      "loss": 0.9267,
+      "step": 37109
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8640356640870512,
+      "learning_rate": 8.175031585044968e-07,
+      "loss": 0.7665,
+      "step": 37110
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4965292260871608,
+      "learning_rate": 8.172010195145352e-07,
+      "loss": 0.8815,
+      "step": 37111
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7120888548118335,
+      "learning_rate": 8.168989339897348e-07,
+      "loss": 0.7022,
+      "step": 37112
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7574883746460126,
+      "learning_rate": 8.165969019318554e-07,
+      "loss": 0.8235,
+      "step": 37113
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8455512802418998,
+      "learning_rate": 8.162949233426587e-07,
+      "loss": 0.9311,
+      "step": 37114
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4396887463302541,
+      "learning_rate": 8.159929982239001e-07,
+      "loss": 0.752,
+      "step": 37115
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.61191521840503,
+      "learning_rate": 8.156911265773359e-07,
+      "loss": 0.8138,
+      "step": 37116
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7244467173073459,
+      "learning_rate": 8.153893084047281e-07,
+      "loss": 0.7249,
+      "step": 37117
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7902759903063186,
+      "learning_rate": 8.150875437078298e-07,
+      "loss": 0.8444,
+      "step": 37118
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.699630870338084,
+      "learning_rate": 8.147858324884005e-07,
+      "loss": 0.8423,
+      "step": 37119
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7337883507887026,
+      "learning_rate": 8.144841747481947e-07,
+      "loss": 0.9818,
+      "step": 37120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.035678409925059,
+      "learning_rate": 8.141825704889728e-07,
+      "loss": 0.7272,
+      "step": 37121
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5156447090002405,
+      "learning_rate": 8.138810197124846e-07,
+      "loss": 0.7012,
+      "step": 37122
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.56511271760193,
+      "learning_rate": 8.135795224204923e-07,
+      "loss": 0.9005,
+      "step": 37123
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6071060072086953,
+      "learning_rate": 8.132780786147487e-07,
+      "loss": 0.8261,
+      "step": 37124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7551207889042587,
+      "learning_rate": 8.129766882970069e-07,
+      "loss": 0.662,
+      "step": 37125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6934638003209124,
+      "learning_rate": 8.126753514690222e-07,
+      "loss": 0.8597,
+      "step": 37126
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7528674778389863,
+      "learning_rate": 8.123740681325531e-07,
+      "loss": 0.7227,
+      "step": 37127
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.478947118246367,
+      "learning_rate": 8.120728382893505e-07,
+      "loss": 0.7804,
+      "step": 37128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6604734987353897,
+      "learning_rate": 8.117716619411675e-07,
+      "loss": 0.8249,
+      "step": 37129
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6736991220499964,
+      "learning_rate": 8.114705390897581e-07,
+      "loss": 0.8034,
+      "step": 37130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7314971263819738,
+      "learning_rate": 8.111694697368778e-07,
+      "loss": 0.844,
+      "step": 37131
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5689368016741556,
+      "learning_rate": 8.108684538842781e-07,
+      "loss": 0.7957,
+      "step": 37132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9174557246031583,
+      "learning_rate": 8.105674915337103e-07,
+      "loss": 0.7836,
+      "step": 37133
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0565819589426027,
+      "learning_rate": 8.102665826869283e-07,
+      "loss": 0.7891,
+      "step": 37134
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.069433666978964,
+      "learning_rate": 8.09965727345684e-07,
+      "loss": 0.8105,
+      "step": 37135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6638861307236108,
+      "learning_rate": 8.096649255117294e-07,
+      "loss": 0.9155,
+      "step": 37136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.583731354112546,
+      "learning_rate": 8.093641771868144e-07,
+      "loss": 0.8182,
+      "step": 37137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5842630514785738,
+      "learning_rate": 8.090634823726917e-07,
+      "loss": 0.779,
+      "step": 37138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5123345119552891,
+      "learning_rate": 8.08762841071109e-07,
+      "loss": 0.6746,
+      "step": 37139
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7610534769599346,
+      "learning_rate": 8.084622532838215e-07,
+      "loss": 0.8066,
+      "step": 37140
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.977078552785905,
+      "learning_rate": 8.081617190125745e-07,
+      "loss": 0.8283,
+      "step": 37141
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.786814438648428,
+      "learning_rate": 8.078612382591222e-07,
+      "loss": 0.7247,
+      "step": 37142
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4740525153821886,
+      "learning_rate": 8.075608110252098e-07,
+      "loss": 0.745,
+      "step": 37143
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.590199856987003,
+      "learning_rate": 8.072604373125914e-07,
+      "loss": 0.8077,
+      "step": 37144
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7084838998465157,
+      "learning_rate": 8.069601171230124e-07,
+      "loss": 0.8396,
+      "step": 37145
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7206172882803539,
+      "learning_rate": 8.066598504582201e-07,
+      "loss": 0.8336,
+      "step": 37146
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.741101107726395,
+      "learning_rate": 8.063596373199645e-07,
+      "loss": 0.6911,
+      "step": 37147
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5028385390289751,
+      "learning_rate": 8.060594777099939e-07,
+      "loss": 0.935,
+      "step": 37148
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5032242125960527,
+      "learning_rate": 8.057593716300605e-07,
+      "loss": 0.8627,
+      "step": 37149
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.208124296029028,
+      "learning_rate": 8.054593190819016e-07,
+      "loss": 0.75,
+      "step": 37150
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8022259794682174,
+      "learning_rate": 8.051593200672703e-07,
+      "loss": 0.8803,
+      "step": 37151
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7113459664057258,
+      "learning_rate": 8.048593745879141e-07,
+      "loss": 0.8286,
+      "step": 37152
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4906628536107904,
+      "learning_rate": 8.045594826455783e-07,
+      "loss": 0.901,
+      "step": 37153
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7806372235520211,
+      "learning_rate": 8.042596442420059e-07,
+      "loss": 0.8264,
+      "step": 37154
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.685092048256272,
+      "learning_rate": 8.039598593789455e-07,
+      "loss": 0.8386,
+      "step": 37155
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5868485556800314,
+      "learning_rate": 8.036601280581446e-07,
+      "loss": 0.9268,
+      "step": 37156
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4693611797583328,
+      "learning_rate": 8.033604502813453e-07,
+      "loss": 0.8549,
+      "step": 37157
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1162765285207574,
+      "learning_rate": 8.030608260502915e-07,
+      "loss": 0.7879,
+      "step": 37158
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7050135656661438,
+      "learning_rate": 8.027612553667308e-07,
+      "loss": 0.7199,
+      "step": 37159
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.769874245567462,
+      "learning_rate": 8.02461738232404e-07,
+      "loss": 0.9179,
+      "step": 37160
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5901532023384113,
+      "learning_rate": 8.021622746490565e-07,
+      "loss": 0.7585,
+      "step": 37161
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5697238100964455,
+      "learning_rate": 8.018628646184345e-07,
+      "loss": 0.7429,
+      "step": 37162
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6636453131507951,
+      "learning_rate": 8.01563508142279e-07,
+      "loss": 0.8054,
+      "step": 37163
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7870987791844273,
+      "learning_rate": 8.012642052223319e-07,
+      "loss": 0.7658,
+      "step": 37164
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7591660484700018,
+      "learning_rate": 8.009649558603361e-07,
+      "loss": 0.6956,
+      "step": 37165
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.554391432948271,
+      "learning_rate": 8.006657600580381e-07,
+      "loss": 0.8766,
+      "step": 37166
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5990198204498878,
+      "learning_rate": 8.003666178171732e-07,
+      "loss": 0.8012,
+      "step": 37167
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4215066356568917,
+      "learning_rate": 8.000675291394855e-07,
+      "loss": 0.8451,
+      "step": 37168
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.568396281847158,
+      "learning_rate": 7.997684940267181e-07,
+      "loss": 0.6993,
+      "step": 37169
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.889952382913168,
+      "learning_rate": 7.994695124806152e-07,
+      "loss": 0.8555,
+      "step": 37170
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7232883262389526,
+      "learning_rate": 7.991705845029096e-07,
+      "loss": 0.6952,
+      "step": 37171
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8468159839196423,
+      "learning_rate": 7.988717100953458e-07,
+      "loss": 0.8156,
+      "step": 37172
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8754741979277303,
+      "learning_rate": 7.985728892596656e-07,
+      "loss": 0.9714,
+      "step": 37173
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5429822428858784,
+      "learning_rate": 7.982741219976065e-07,
+      "loss": 0.9036,
+      "step": 37174
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.023956444220997,
+      "learning_rate": 7.979754083109092e-07,
+      "loss": 0.7642,
+      "step": 37175
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8223979627895068,
+      "learning_rate": 7.976767482013115e-07,
+      "loss": 0.8841,
+      "step": 37176
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7264871624680442,
+      "learning_rate": 7.973781416705539e-07,
+      "loss": 0.8272,
+      "step": 37177
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0969594779922,
+      "learning_rate": 7.970795887203731e-07,
+      "loss": 0.8586,
+      "step": 37178
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5779488481898032,
+      "learning_rate": 7.967810893525096e-07,
+      "loss": 0.7996,
+      "step": 37179
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6271352729664332,
+      "learning_rate": 7.964826435687001e-07,
+      "loss": 0.6998,
+      "step": 37180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.340414155090099,
+      "learning_rate": 7.96184251370683e-07,
+      "loss": 0.7487,
+      "step": 37181
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.708615793314534,
+      "learning_rate": 7.958859127601937e-07,
+      "loss": 0.8266,
+      "step": 37182
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6477386908982286,
+      "learning_rate": 7.955876277389718e-07,
+      "loss": 0.8546,
+      "step": 37183
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7416368261508042,
+      "learning_rate": 7.952893963087527e-07,
+      "loss": 0.6851,
+      "step": 37184
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6956411960258946,
+      "learning_rate": 7.949912184712716e-07,
+      "loss": 0.7831,
+      "step": 37185
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.976830765923362,
+      "learning_rate": 7.94693094228266e-07,
+      "loss": 0.7931,
+      "step": 37186
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6465776594066448,
+      "learning_rate": 7.943950235814734e-07,
+      "loss": 0.9147,
+      "step": 37187
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.763080486480156,
+      "learning_rate": 7.94097006532627e-07,
+      "loss": 0.7128,
+      "step": 37188
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.290662063743615,
+      "learning_rate": 7.937990430834607e-07,
+      "loss": 0.8763,
+      "step": 37189
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5527554763066393,
+      "learning_rate": 7.935011332357113e-07,
+      "loss": 0.8439,
+      "step": 37190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1316061536201474,
+      "learning_rate": 7.932032769911146e-07,
+      "loss": 0.8099,
+      "step": 37191
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6785110655582138,
+      "learning_rate": 7.929054743514042e-07,
+      "loss": 0.6515,
+      "step": 37192
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7254814382858705,
+      "learning_rate": 7.926077253183096e-07,
+      "loss": 0.7469,
+      "step": 37193
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8000790489541236,
+      "learning_rate": 7.923100298935693e-07,
+      "loss": 0.884,
+      "step": 37194
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5011396068815344,
+      "learning_rate": 7.920123880789155e-07,
+      "loss": 0.8806,
+      "step": 37195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9244545671556779,
+      "learning_rate": 7.917147998760821e-07,
+      "loss": 0.7466,
+      "step": 37196
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6419595602682469,
+      "learning_rate": 7.91417265286798e-07,
+      "loss": 0.7388,
+      "step": 37197
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.589189614814259,
+      "learning_rate": 7.911197843127993e-07,
+      "loss": 0.7302,
+      "step": 37198
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6403554754951681,
+      "learning_rate": 7.90822356955816e-07,
+      "loss": 0.8761,
+      "step": 37199
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.734866115898975,
+      "learning_rate": 7.905249832175809e-07,
+      "loss": 0.7828,
+      "step": 37200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1051414642927027,
+      "learning_rate": 7.90227663099824e-07,
+      "loss": 0.7927,
+      "step": 37201
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5630798053475192,
+      "learning_rate": 7.899303966042793e-07,
+      "loss": 0.8672,
+      "step": 37202
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7966982490226835,
+      "learning_rate": 7.896331837326743e-07,
+      "loss": 0.85,
+      "step": 37203
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5469441791070957,
+      "learning_rate": 7.893360244867421e-07,
+      "loss": 0.813,
+      "step": 37204
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6014843146790168,
+      "learning_rate": 7.890389188682112e-07,
+      "loss": 0.712,
+      "step": 37205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5300424885900152,
+      "learning_rate": 7.887418668788104e-07,
+      "loss": 0.7032,
+      "step": 37206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4235120620052784,
+      "learning_rate": 7.884448685202717e-07,
+      "loss": 0.7822,
+      "step": 37207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7738278640468819,
+      "learning_rate": 7.881479237943235e-07,
+      "loss": 0.9739,
+      "step": 37208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4346769394413,
+      "learning_rate": 7.878510327026956e-07,
+      "loss": 0.7578,
+      "step": 37209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5485671102767276,
+      "learning_rate": 7.875541952471133e-07,
+      "loss": 0.8304,
+      "step": 37210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7582660523438642,
+      "learning_rate": 7.872574114293075e-07,
+      "loss": 0.7012,
+      "step": 37211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5169134215077726,
+      "learning_rate": 7.869606812510067e-07,
+      "loss": 0.8224,
+      "step": 37212
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5481698490011029,
+      "learning_rate": 7.866640047139384e-07,
+      "loss": 0.9205,
+      "step": 37213
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5161574467654595,
+      "learning_rate": 7.863673818198269e-07,
+      "loss": 0.8685,
+      "step": 37214
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9195287040005866,
+      "learning_rate": 7.860708125704019e-07,
+      "loss": 0.6419,
+      "step": 37215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2643848278257477,
+      "learning_rate": 7.857742969673909e-07,
+      "loss": 0.8287,
+      "step": 37216
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8044242916588753,
+      "learning_rate": 7.854778350125191e-07,
+      "loss": 0.7703,
+      "step": 37217
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4564278854579482,
+      "learning_rate": 7.85181426707512e-07,
+      "loss": 0.9619,
+      "step": 37218
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6835176286877866,
+      "learning_rate": 7.848850720540968e-07,
+      "loss": 0.83,
+      "step": 37219
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6157736211641676,
+      "learning_rate": 7.845887710539957e-07,
+      "loss": 0.7525,
+      "step": 37220
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5229313917544571,
+      "learning_rate": 7.842925237089394e-07,
+      "loss": 0.8047,
+      "step": 37221
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0299721980480405,
+      "learning_rate": 7.839963300206477e-07,
+      "loss": 0.708,
+      "step": 37222
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6358820429483376,
+      "learning_rate": 7.837001899908481e-07,
+      "loss": 0.6315,
+      "step": 37223
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.633124440022307,
+      "learning_rate": 7.834041036212625e-07,
+      "loss": 0.9551,
+      "step": 37224
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.807833276031316,
+      "learning_rate": 7.831080709136185e-07,
+      "loss": 0.6797,
+      "step": 37225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.607811520569836,
+      "learning_rate": 7.828120918696369e-07,
+      "loss": 0.7272,
+      "step": 37226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7769249494546684,
+      "learning_rate": 7.825161664910397e-07,
+      "loss": 0.7149,
+      "step": 37227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.551399677870011,
+      "learning_rate": 7.822202947795509e-07,
+      "loss": 0.7793,
+      "step": 37228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7444783293018051,
+      "learning_rate": 7.81924476736896e-07,
+      "loss": 0.6748,
+      "step": 37229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.411619272253573,
+      "learning_rate": 7.816287123647958e-07,
+      "loss": 0.8618,
+      "step": 37230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.721585991080647,
+      "learning_rate": 7.813330016649689e-07,
+      "loss": 0.9613,
+      "step": 37231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8266505839037885,
+      "learning_rate": 7.810373446391406e-07,
+      "loss": 0.8266,
+      "step": 37232
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.317305432507087,
+      "learning_rate": 7.807417412890339e-07,
+      "loss": 0.8027,
+      "step": 37233
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7640939505792409,
+      "learning_rate": 7.804461916163663e-07,
+      "loss": 0.6987,
+      "step": 37234
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5842063961492439,
+      "learning_rate": 7.801506956228588e-07,
+      "loss": 0.94,
+      "step": 37235
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5208352323222667,
+      "learning_rate": 7.798552533102322e-07,
+      "loss": 0.7639,
+      "step": 37236
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0595320104365147,
+      "learning_rate": 7.795598646802105e-07,
+      "loss": 0.82,
+      "step": 37237
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4026165906577672,
+      "learning_rate": 7.792645297345091e-07,
+      "loss": 0.7848,
+      "step": 37238
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5629926605931916,
+      "learning_rate": 7.789692484748479e-07,
+      "loss": 0.8873,
+      "step": 37239
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7835422717395948,
+      "learning_rate": 7.786740209029475e-07,
+      "loss": 0.8349,
+      "step": 37240
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.47392281791402,
+      "learning_rate": 7.783788470205256e-07,
+      "loss": 0.8494,
+      "step": 37241
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1715406653215314,
+      "learning_rate": 7.78083726829304e-07,
+      "loss": 0.7208,
+      "step": 37242
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5740469256236556,
+      "learning_rate": 7.777886603309958e-07,
+      "loss": 0.7502,
+      "step": 37243
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4834084265024468,
+      "learning_rate": 7.774936475273243e-07,
+      "loss": 0.7107,
+      "step": 37244
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4417067409762419,
+      "learning_rate": 7.771986884200022e-07,
+      "loss": 0.8502,
+      "step": 37245
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9060455333925201,
+      "learning_rate": 7.769037830107484e-07,
+      "loss": 0.8066,
+      "step": 37246
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.041811326014024,
+      "learning_rate": 7.766089313012859e-07,
+      "loss": 0.8796,
+      "step": 37247
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2116776743077233,
+      "learning_rate": 7.763141332933222e-07,
+      "loss": 0.9175,
+      "step": 37248
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4773697751578438,
+      "learning_rate": 7.76019388988577e-07,
+      "loss": 0.7882,
+      "step": 37249
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.552763883254829,
+      "learning_rate": 7.757246983887679e-07,
+      "loss": 0.7643,
+      "step": 37250
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4074391443812038,
+      "learning_rate": 7.754300614956134e-07,
+      "loss": 0.8073,
+      "step": 37251
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6208944856914334,
+      "learning_rate": 7.751354783108223e-07,
+      "loss": 0.8502,
+      "step": 37252
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.777032578573016,
+      "learning_rate": 7.748409488361131e-07,
+      "loss": 0.8807,
+      "step": 37253
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.84653195644715,
+      "learning_rate": 7.745464730732011e-07,
+      "loss": 0.8409,
+      "step": 37254
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5265642312073848,
+      "learning_rate": 7.742520510238016e-07,
+      "loss": 0.8218,
+      "step": 37255
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.944650521846336,
+      "learning_rate": 7.739576826896244e-07,
+      "loss": 0.8267,
+      "step": 37256
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9104572872062076,
+      "learning_rate": 7.736633680723882e-07,
+      "loss": 0.8544,
+      "step": 37257
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8228127877104379,
+      "learning_rate": 7.733691071738047e-07,
+      "loss": 0.9122,
+      "step": 37258
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5847251995741323,
+      "learning_rate": 7.730748999955872e-07,
+      "loss": 0.8045,
+      "step": 37259
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7515255098637084,
+      "learning_rate": 7.7278074653945e-07,
+      "loss": 0.7415,
+      "step": 37260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7523181872761042,
+      "learning_rate": 7.724866468071046e-07,
+      "loss": 0.8163,
+      "step": 37261
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5650807726646445,
+      "learning_rate": 7.721926008002611e-07,
+      "loss": 0.802,
+      "step": 37262
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7957616606019318,
+      "learning_rate": 7.718986085206348e-07,
+      "loss": 0.844,
+      "step": 37263
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6168351097747475,
+      "learning_rate": 7.716046699699376e-07,
+      "loss": 0.8571,
+      "step": 37264
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5534254939884802,
+      "learning_rate": 7.713107851498791e-07,
+      "loss": 0.8668,
+      "step": 37265
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6156515979531598,
+      "learning_rate": 7.710169540621692e-07,
+      "loss": 0.7512,
+      "step": 37266
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8668809422525812,
+      "learning_rate": 7.70723176708521e-07,
+      "loss": 0.8283,
+      "step": 37267
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6583718684665703,
+      "learning_rate": 7.704294530906487e-07,
+      "loss": 0.8392,
+      "step": 37268
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7116370750902594,
+      "learning_rate": 7.70135783210253e-07,
+      "loss": 0.8544,
+      "step": 37269
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.619665860240978,
+      "learning_rate": 7.698421670690503e-07,
+      "loss": 0.832,
+      "step": 37270
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7873751233952893,
+      "learning_rate": 7.695486046687473e-07,
+      "loss": 0.7567,
+      "step": 37271
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.95293337608075,
+      "learning_rate": 7.692550960110578e-07,
+      "loss": 0.9035,
+      "step": 37272
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4652215932271344,
+      "learning_rate": 7.689616410976863e-07,
+      "loss": 0.9205,
+      "step": 37273
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7930847018450489,
+      "learning_rate": 7.686682399303425e-07,
+      "loss": 0.6949,
+      "step": 37274
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.765186868215055,
+      "learning_rate": 7.683748925107348e-07,
+      "loss": 0.8239,
+      "step": 37275
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6064873003798812,
+      "learning_rate": 7.680815988405698e-07,
+      "loss": 0.7387,
+      "step": 37276
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6336374529220485,
+      "learning_rate": 7.677883589215596e-07,
+      "loss": 0.8535,
+      "step": 37277
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5665376314657722,
+      "learning_rate": 7.674951727554058e-07,
+      "loss": 0.8502,
+      "step": 37278
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6206964705809979,
+      "learning_rate": 7.672020403438196e-07,
+      "loss": 0.7304,
+      "step": 37279
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1506302112407054,
+      "learning_rate": 7.66908961688504e-07,
+      "loss": 0.7814,
+      "step": 37280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7244913585749126,
+      "learning_rate": 7.666159367911696e-07,
+      "loss": 0.7177,
+      "step": 37281
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7411998958967652,
+      "learning_rate": 7.663229656535209e-07,
+      "loss": 0.892,
+      "step": 37282
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4026171660874789,
+      "learning_rate": 7.66030048277262e-07,
+      "loss": 0.7951,
+      "step": 37283
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7676277733252914,
+      "learning_rate": 7.657371846640982e-07,
+      "loss": 0.7693,
+      "step": 37284
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7722980424379574,
+      "learning_rate": 7.654443748157391e-07,
+      "loss": 0.6799,
+      "step": 37285
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.175581556019633,
+      "learning_rate": 7.651516187338859e-07,
+      "loss": 0.8827,
+      "step": 37286
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5672651155161057,
+      "learning_rate": 7.648589164202413e-07,
+      "loss": 0.8984,
+      "step": 37287
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7488739826847244,
+      "learning_rate": 7.645662678765131e-07,
+      "loss": 0.7439,
+      "step": 37288
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9032980901745427,
+      "learning_rate": 7.642736731044043e-07,
+      "loss": 0.8807,
+      "step": 37289
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.12051433055579,
+      "learning_rate": 7.639811321056189e-07,
+      "loss": 0.7774,
+      "step": 37290
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5988452078256512,
+      "learning_rate": 7.636886448818582e-07,
+      "loss": 0.9749,
+      "step": 37291
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4350288042399608,
+      "learning_rate": 7.633962114348259e-07,
+      "loss": 0.8808,
+      "step": 37292
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.530926071436605,
+      "learning_rate": 7.631038317662265e-07,
+      "loss": 0.8289,
+      "step": 37293
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5869237153603337,
+      "learning_rate": 7.628115058777619e-07,
+      "loss": 0.7695,
+      "step": 37294
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5690159554349943,
+      "learning_rate": 7.625192337711306e-07,
+      "loss": 0.8795,
+      "step": 37295
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6062710723693183,
+      "learning_rate": 7.622270154480382e-07,
+      "loss": 0.8463,
+      "step": 37296
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4938222052052332,
+      "learning_rate": 7.619348509101843e-07,
+      "loss": 0.8752,
+      "step": 37297
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7213980161996834,
+      "learning_rate": 7.616427401592708e-07,
+      "loss": 0.8048,
+      "step": 37298
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.516264503016187,
+      "learning_rate": 7.613506831969975e-07,
+      "loss": 0.7374,
+      "step": 37299
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6434508645459194,
+      "learning_rate": 7.610586800250663e-07,
+      "loss": 0.7692,
+      "step": 37300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7538732830779957,
+      "learning_rate": 7.607667306451749e-07,
+      "loss": 0.7213,
+      "step": 37301
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7265265934245198,
+      "learning_rate": 7.604748350590274e-07,
+      "loss": 0.873,
+      "step": 37302
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.368417019262874,
+      "learning_rate": 7.60182993268318e-07,
+      "loss": 0.8398,
+      "step": 37303
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.422690832799582,
+      "learning_rate": 7.598912052747498e-07,
+      "loss": 0.926,
+      "step": 37304
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4775016391336298,
+      "learning_rate": 7.595994710800192e-07,
+      "loss": 0.7722,
+      "step": 37305
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6562874451580374,
+      "learning_rate": 7.593077906858281e-07,
+      "loss": 0.8813,
+      "step": 37306
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7755827526091019,
+      "learning_rate": 7.59016164093872e-07,
+      "loss": 0.7016,
+      "step": 37307
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6753822400403728,
+      "learning_rate": 7.587245913058472e-07,
+      "loss": 0.8231,
+      "step": 37308
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.59685851003108,
+      "learning_rate": 7.584330723234545e-07,
+      "loss": 0.9568,
+      "step": 37309
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.467702919664508,
+      "learning_rate": 7.581416071483915e-07,
+      "loss": 0.8771,
+      "step": 37310
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.429569918977498,
+      "learning_rate": 7.578501957823547e-07,
+      "loss": 0.8172,
+      "step": 37311
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5137672963752513,
+      "learning_rate": 7.575588382270372e-07,
+      "loss": 0.8211,
+      "step": 37312
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.566589656002274,
+      "learning_rate": 7.572675344841396e-07,
+      "loss": 0.9075,
+      "step": 37313
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9024284138672038,
+      "learning_rate": 7.569762845553574e-07,
+      "loss": 0.8745,
+      "step": 37314
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7542404158968606,
+      "learning_rate": 7.566850884423859e-07,
+      "loss": 0.7133,
+      "step": 37315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7446014218863954,
+      "learning_rate": 7.563939461469183e-07,
+      "loss": 0.7666,
+      "step": 37316
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9271362367971396,
+      "learning_rate": 7.561028576706519e-07,
+      "loss": 0.8105,
+      "step": 37317
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5759087224377906,
+      "learning_rate": 7.558118230152822e-07,
+      "loss": 0.9113,
+      "step": 37318
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9547200051799327,
+      "learning_rate": 7.555208421825033e-07,
+      "loss": 0.781,
+      "step": 37319
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7709335279830816,
+      "learning_rate": 7.552299151740072e-07,
+      "loss": 0.7195,
+      "step": 37320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.83715125890541,
+      "learning_rate": 7.549390419914915e-07,
+      "loss": 0.8534,
+      "step": 37321
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.462695165168517,
+      "learning_rate": 7.546482226366447e-07,
+      "loss": 0.8769,
+      "step": 37322
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5823999324660984,
+      "learning_rate": 7.543574571111656e-07,
+      "loss": 0.8796,
+      "step": 37323
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7798690309073802,
+      "learning_rate": 7.540667454167427e-07,
+      "loss": 0.6761,
+      "step": 37324
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5798620411770932,
+      "learning_rate": 7.537760875550714e-07,
+      "loss": 0.9093,
+      "step": 37325
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6447186970930383,
+      "learning_rate": 7.534854835278427e-07,
+      "loss": 0.7633,
+      "step": 37326
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7373719913658372,
+      "learning_rate": 7.531949333367495e-07,
+      "loss": 0.8205,
+      "step": 37327
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6930873977199024,
+      "learning_rate": 7.529044369834826e-07,
+      "loss": 0.9023,
+      "step": 37328
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7794928882590642,
+      "learning_rate": 7.526139944697331e-07,
+      "loss": 0.798,
+      "step": 37329
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7087712596096616,
+      "learning_rate": 7.523236057971917e-07,
+      "loss": 0.7974,
+      "step": 37330
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5549172143436143,
+      "learning_rate": 7.520332709675504e-07,
+      "loss": 0.7711,
+      "step": 37331
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6038687085085441,
+      "learning_rate": 7.517429899825035e-07,
+      "loss": 0.7303,
+      "step": 37332
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6743345761641246,
+      "learning_rate": 7.514527628437318e-07,
+      "loss": 0.88,
+      "step": 37333
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5546082758456305,
+      "learning_rate": 7.511625895529317e-07,
+      "loss": 0.8779,
+      "step": 37334
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5765672973609974,
+      "learning_rate": 7.50872470111792e-07,
+      "loss": 0.9611,
+      "step": 37335
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5368532804110588,
+      "learning_rate": 7.505824045220012e-07,
+      "loss": 0.9135,
+      "step": 37336
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5629826784219696,
+      "learning_rate": 7.502923927852457e-07,
+      "loss": 0.635,
+      "step": 37337
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.70538843558322,
+      "learning_rate": 7.500024349032165e-07,
+      "loss": 0.9342,
+      "step": 37338
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4334714098397923,
+      "learning_rate": 7.497125308776043e-07,
+      "loss": 0.8756,
+      "step": 37339
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7093408044821106,
+      "learning_rate": 7.494226807100935e-07,
+      "loss": 0.845,
+      "step": 37340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4895103259469504,
+      "learning_rate": 7.491328844023704e-07,
+      "loss": 0.7792,
+      "step": 37341
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6479991391709459,
+      "learning_rate": 7.48843141956127e-07,
+      "loss": 0.8785,
+      "step": 37342
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6991944591865993,
+      "learning_rate": 7.485534533730454e-07,
+      "loss": 0.9044,
+      "step": 37343
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5704270681801742,
+      "learning_rate": 7.482638186548153e-07,
+      "loss": 0.8741,
+      "step": 37344
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7741152183050164,
+      "learning_rate": 7.47974237803123e-07,
+      "loss": 0.7868,
+      "step": 37345
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7916086421142792,
+      "learning_rate": 7.47684710819655e-07,
+      "loss": 0.8178,
+      "step": 37346
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.604470367408067,
+      "learning_rate": 7.473952377060934e-07,
+      "loss": 0.8981,
+      "step": 37347
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.996680704716585,
+      "learning_rate": 7.471058184641255e-07,
+      "loss": 0.9338,
+      "step": 37348
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5911570151197125,
+      "learning_rate": 7.468164530954414e-07,
+      "loss": 0.8783,
+      "step": 37349
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6710885901957189,
+      "learning_rate": 7.465271416017172e-07,
+      "loss": 0.711,
+      "step": 37350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4230181608606813,
+      "learning_rate": 7.462378839846419e-07,
+      "loss": 0.7352,
+      "step": 37351
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6670351059231978,
+      "learning_rate": 7.459486802458993e-07,
+      "loss": 0.8138,
+      "step": 37352
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7368270092253908,
+      "learning_rate": 7.456595303871772e-07,
+      "loss": 0.7591,
+      "step": 37353
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.518857038720256,
+      "learning_rate": 7.453704344101509e-07,
+      "loss": 0.8264,
+      "step": 37354
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8133121169671094,
+      "learning_rate": 7.45081392316509e-07,
+      "loss": 0.7178,
+      "step": 37355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.594281345854837,
+      "learning_rate": 7.447924041079347e-07,
+      "loss": 0.8233,
+      "step": 37356
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7914722169629522,
+      "learning_rate": 7.445034697861064e-07,
+      "loss": 0.9177,
+      "step": 37357
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5779711497588742,
+      "learning_rate": 7.442145893527119e-07,
+      "loss": 0.794,
+      "step": 37358
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.756399349170966,
+      "learning_rate": 7.439257628094288e-07,
+      "loss": 0.839,
+      "step": 37359
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7985935395564792,
+      "learning_rate": 7.43636990157941e-07,
+      "loss": 0.7609,
+      "step": 37360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7521607292378062,
+      "learning_rate": 7.433482713999285e-07,
+      "loss": 0.7015,
+      "step": 37361
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.81641876520183,
+      "learning_rate": 7.430596065370743e-07,
+      "loss": 0.6817,
+      "step": 37362
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.873737210381998,
+      "learning_rate": 7.427709955710572e-07,
+      "loss": 0.7754,
+      "step": 37363
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6660543933175191,
+      "learning_rate": 7.424824385035568e-07,
+      "loss": 0.8444,
+      "step": 37364
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6168308672687859,
+      "learning_rate": 7.42193935336255e-07,
+      "loss": 0.8773,
+      "step": 37365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5763480099939595,
+      "learning_rate": 7.419054860708319e-07,
+      "loss": 0.8279,
+      "step": 37366
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.477269673048963,
+      "learning_rate": 7.416170907089671e-07,
+      "loss": 0.8379,
+      "step": 37367
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.740786862069943,
+      "learning_rate": 7.413287492523369e-07,
+      "loss": 0.7004,
+      "step": 37368
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4766942467206527,
+      "learning_rate": 7.410404617026212e-07,
+      "loss": 0.9145,
+      "step": 37369
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.530071435209299,
+      "learning_rate": 7.40752228061502e-07,
+      "loss": 0.8159,
+      "step": 37370
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7531030827988368,
+      "learning_rate": 7.404640483306536e-07,
+      "loss": 0.8192,
+      "step": 37371
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6018609557516836,
+      "learning_rate": 7.401759225117544e-07,
+      "loss": 0.795,
+      "step": 37372
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.902549294814886,
+      "learning_rate": 7.398878506064821e-07,
+      "loss": 0.7443,
+      "step": 37373
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8627599066116332,
+      "learning_rate": 7.395998326165165e-07,
+      "loss": 0.7549,
+      "step": 37374
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5862797165083764,
+      "learning_rate": 7.393118685435319e-07,
+      "loss": 0.932,
+      "step": 37375
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8390777324786662,
+      "learning_rate": 7.390239583892045e-07,
+      "loss": 0.808,
+      "step": 37376
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8460880279058582,
+      "learning_rate": 7.38736102155212e-07,
+      "loss": 0.7048,
+      "step": 37377
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6239360001322105,
+      "learning_rate": 7.384482998432296e-07,
+      "loss": 0.7555,
+      "step": 37378
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.575643490026665,
+      "learning_rate": 7.381605514549339e-07,
+      "loss": 0.6765,
+      "step": 37379
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5246283720876197,
+      "learning_rate": 7.378728569919991e-07,
+      "loss": 0.8536,
+      "step": 37380
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7778324136135095,
+      "learning_rate": 7.375852164561026e-07,
+      "loss": 0.8014,
+      "step": 37381
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4471742813375017,
+      "learning_rate": 7.372976298489154e-07,
+      "loss": 0.8036,
+      "step": 37382
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.705723676886154,
+      "learning_rate": 7.370100971721151e-07,
+      "loss": 0.8684,
+      "step": 37383
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6459503365148265,
+      "learning_rate": 7.367226184273746e-07,
+      "loss": 0.8002,
+      "step": 37384
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.535720717514302,
+      "learning_rate": 7.364351936163672e-07,
+      "loss": 0.9589,
+      "step": 37385
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5070186062236244,
+      "learning_rate": 7.361478227407659e-07,
+      "loss": 0.8141,
+      "step": 37386
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.802111319374711,
+      "learning_rate": 7.358605058022461e-07,
+      "loss": 0.7145,
+      "step": 37387
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7765716471318628,
+      "learning_rate": 7.355732428024808e-07,
+      "loss": 0.6914,
+      "step": 37388
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.801193752437951,
+      "learning_rate": 7.352860337431389e-07,
+      "loss": 0.7477,
+      "step": 37389
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8797409424258322,
+      "learning_rate": 7.349988786258944e-07,
+      "loss": 0.8308,
+      "step": 37390
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6628072296722807,
+      "learning_rate": 7.347117774524215e-07,
+      "loss": 0.7905,
+      "step": 37391
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3896605161423752,
+      "learning_rate": 7.344247302243901e-07,
+      "loss": 0.8509,
+      "step": 37392
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.413775499949445,
+      "learning_rate": 7.3413773694347e-07,
+      "loss": 0.7801,
+      "step": 37393
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7593058409816722,
+      "learning_rate": 7.338507976113329e-07,
+      "loss": 0.6743,
+      "step": 37394
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6548667622651971,
+      "learning_rate": 7.335639122296522e-07,
+      "loss": 0.8456,
+      "step": 37395
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7077454197019346,
+      "learning_rate": 7.332770808000955e-07,
+      "loss": 0.8917,
+      "step": 37396
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7031962990380225,
+      "learning_rate": 7.329903033243324e-07,
+      "loss": 0.8331,
+      "step": 37397
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5948024111654715,
+      "learning_rate": 7.32703579804036e-07,
+      "loss": 0.8917,
+      "step": 37398
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.555880948421636,
+      "learning_rate": 7.324169102408707e-07,
+      "loss": 0.8208,
+      "step": 37399
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5571926981928064,
+      "learning_rate": 7.321302946365094e-07,
+      "loss": 0.8208,
+      "step": 37400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.607495511574792,
+      "learning_rate": 7.318437329926187e-07,
+      "loss": 0.7946,
+      "step": 37401
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7732128267371472,
+      "learning_rate": 7.315572253108694e-07,
+      "loss": 0.7255,
+      "step": 37402
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6051663087367327,
+      "learning_rate": 7.312707715929268e-07,
+      "loss": 0.9354,
+      "step": 37403
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.134270527591394,
+      "learning_rate": 7.309843718404619e-07,
+      "loss": 0.8152,
+      "step": 37404
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5826632080596656,
+      "learning_rate": 7.306980260551411e-07,
+      "loss": 0.836,
+      "step": 37405
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4883201104108457,
+      "learning_rate": 7.304117342386286e-07,
+      "loss": 0.8424,
+      "step": 37406
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8930676316423858,
+      "learning_rate": 7.301254963925941e-07,
+      "loss": 0.8484,
+      "step": 37407
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6270986453073712,
+      "learning_rate": 7.298393125187054e-07,
+      "loss": 0.791,
+      "step": 37408
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6091316880520767,
+      "learning_rate": 7.295531826186264e-07,
+      "loss": 0.85,
+      "step": 37409
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4981721521972808,
+      "learning_rate": 7.292671066940226e-07,
+      "loss": 0.7232,
+      "step": 37410
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.700408164343747,
+      "learning_rate": 7.289810847465606e-07,
+      "loss": 0.7709,
+      "step": 37411
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9725870506887744,
+      "learning_rate": 7.286951167779077e-07,
+      "loss": 0.8473,
+      "step": 37412
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.814922125127407,
+      "learning_rate": 7.284092027897272e-07,
+      "loss": 0.7186,
+      "step": 37413
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5762667211024886,
+      "learning_rate": 7.28123342783682e-07,
+      "loss": 1.0056,
+      "step": 37414
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.796699796740834,
+      "learning_rate": 7.278375367614377e-07,
+      "loss": 0.828,
+      "step": 37415
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5042955784815608,
+      "learning_rate": 7.275517847246604e-07,
+      "loss": 0.8644,
+      "step": 37416
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6356843851324354,
+      "learning_rate": 7.272660866750125e-07,
+      "loss": 0.7981,
+      "step": 37417
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6974321061201862,
+      "learning_rate": 7.269804426141558e-07,
+      "loss": 0.9495,
+      "step": 37418
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5193452219398036,
+      "learning_rate": 7.266948525437556e-07,
+      "loss": 0.8209,
+      "step": 37419
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.037256651853334,
+      "learning_rate": 7.264093164654728e-07,
+      "loss": 0.8405,
+      "step": 37420
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5866919245399815,
+      "learning_rate": 7.261238343809728e-07,
+      "loss": 0.7494,
+      "step": 37421
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6988528670272078,
+      "learning_rate": 7.258384062919144e-07,
+      "loss": 0.8174,
+      "step": 37422
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8402944695978227,
+      "learning_rate": 7.255530321999616e-07,
+      "loss": 0.8227,
+      "step": 37423
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5343853845537463,
+      "learning_rate": 7.252677121067742e-07,
+      "loss": 0.7469,
+      "step": 37424
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.761266958535805,
+      "learning_rate": 7.249824460140165e-07,
+      "loss": 0.7929,
+      "step": 37425
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4891010786855439,
+      "learning_rate": 7.246972339233449e-07,
+      "loss": 0.8547,
+      "step": 37426
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6262768625267556,
+      "learning_rate": 7.244120758364248e-07,
+      "loss": 0.8625,
+      "step": 37427
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6309442871547009,
+      "learning_rate": 7.241269717549126e-07,
+      "loss": 0.8736,
+      "step": 37428
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6362507526487418,
+      "learning_rate": 7.238419216804693e-07,
+      "loss": 0.846,
+      "step": 37429
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6212888999468307,
+      "learning_rate": 7.23556925614759e-07,
+      "loss": 0.8448,
+      "step": 37430
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6034278194366063,
+      "learning_rate": 7.232719835594337e-07,
+      "loss": 0.8192,
+      "step": 37431
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.555413761113243,
+      "learning_rate": 7.229870955161555e-07,
+      "loss": 0.8881,
+      "step": 37432
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5118330967628464,
+      "learning_rate": 7.227022614865841e-07,
+      "loss": 0.7984,
+      "step": 37433
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6822059790861064,
+      "learning_rate": 7.224174814723794e-07,
+      "loss": 0.8331,
+      "step": 37434
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7465080064858721,
+      "learning_rate": 7.221327554751956e-07,
+      "loss": 0.7044,
+      "step": 37435
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6223181719841395,
+      "learning_rate": 7.218480834966912e-07,
+      "loss": 0.8812,
+      "step": 37436
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.541817005122465,
+      "learning_rate": 7.215634655385273e-07,
+      "loss": 0.8135,
+      "step": 37437
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5111189353887358,
+      "learning_rate": 7.212789016023569e-07,
+      "loss": 0.7917,
+      "step": 37438
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7772613643399363,
+      "learning_rate": 7.209943916898377e-07,
+      "loss": 0.6503,
+      "step": 37439
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.454380509279572,
+      "learning_rate": 7.20709935802627e-07,
+      "loss": 0.7544,
+      "step": 37440
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7940993341864788,
+      "learning_rate": 7.204255339423815e-07,
+      "loss": 0.8906,
+      "step": 37441
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5893334900307585,
+      "learning_rate": 7.201411861107555e-07,
+      "loss": 0.8368,
+      "step": 37442
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5920886252131323,
+      "learning_rate": 7.198568923094074e-07,
+      "loss": 0.8721,
+      "step": 37443
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7516197787974168,
+      "learning_rate": 7.195726525399905e-07,
+      "loss": 0.8811,
+      "step": 37444
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.502250604586942,
+      "learning_rate": 7.192884668041578e-07,
+      "loss": 0.839,
+      "step": 37445
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4789399238550132,
+      "learning_rate": 7.190043351035658e-07,
+      "loss": 0.8026,
+      "step": 37446
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3586822855700236,
+      "learning_rate": 7.187202574398699e-07,
+      "loss": 0.7509,
+      "step": 37447
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7369073924549308,
+      "learning_rate": 7.184362338147244e-07,
+      "loss": 0.7469,
+      "step": 37448
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6246509314273339,
+      "learning_rate": 7.181522642297789e-07,
+      "loss": 0.7957,
+      "step": 37449
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4065540278954187,
+      "learning_rate": 7.178683486866899e-07,
+      "loss": 0.8333,
+      "step": 37450
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8518734097309497,
+      "learning_rate": 7.175844871871129e-07,
+      "loss": 0.726,
+      "step": 37451
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9032848729884029,
+      "learning_rate": 7.173006797326954e-07,
+      "loss": 0.806,
+      "step": 37452
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6617030027705564,
+      "learning_rate": 7.170169263250914e-07,
+      "loss": 1.0452,
+      "step": 37453
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7212031774457734,
+      "learning_rate": 7.167332269659544e-07,
+      "loss": 0.6456,
+      "step": 37454
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5951862477138106,
+      "learning_rate": 7.164495816569373e-07,
+      "loss": 0.7704,
+      "step": 37455
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4602053584990036,
+      "learning_rate": 7.161659903996887e-07,
+      "loss": 0.8557,
+      "step": 37456
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.675314748273822,
+      "learning_rate": 7.158824531958608e-07,
+      "loss": 0.671,
+      "step": 37457
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6829082903091837,
+      "learning_rate": 7.155989700471056e-07,
+      "loss": 0.768,
+      "step": 37458
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7038550664863863,
+      "learning_rate": 7.153155409550705e-07,
+      "loss": 0.8509,
+      "step": 37459
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6025448591774287,
+      "learning_rate": 7.150321659214099e-07,
+      "loss": 0.8002,
+      "step": 37460
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9337766010893611,
+      "learning_rate": 7.14748844947769e-07,
+      "loss": 0.8511,
+      "step": 37461
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4484705572107197,
+      "learning_rate": 7.144655780358034e-07,
+      "loss": 0.7624,
+      "step": 37462
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.788408430599611,
+      "learning_rate": 7.14182365187156e-07,
+      "loss": 0.8282,
+      "step": 37463
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6429445286705402,
+      "learning_rate": 7.138992064034811e-07,
+      "loss": 0.8551,
+      "step": 37464
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6404102356235226,
+      "learning_rate": 7.136161016864241e-07,
+      "loss": 0.7674,
+      "step": 37465
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5903657196323975,
+      "learning_rate": 7.133330510376335e-07,
+      "loss": 0.8171,
+      "step": 37466
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.548344084903198,
+      "learning_rate": 7.13050054458757e-07,
+      "loss": 0.8556,
+      "step": 37467
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.745820487141187,
+      "learning_rate": 7.127671119514456e-07,
+      "loss": 0.7706,
+      "step": 37468
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.34869962559308,
+      "learning_rate": 7.124842235173445e-07,
+      "loss": 0.7575,
+      "step": 37469
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6187961178644927,
+      "learning_rate": 7.122013891580981e-07,
+      "loss": 0.7695,
+      "step": 37470
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.681767667259237,
+      "learning_rate": 7.11918608875356e-07,
+      "loss": 0.9101,
+      "step": 37471
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8206000701597589,
+      "learning_rate": 7.11635882670767e-07,
+      "loss": 0.7509,
+      "step": 37472
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7270401909534225,
+      "learning_rate": 7.113532105459731e-07,
+      "loss": 0.7693,
+      "step": 37473
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8421316312895568,
+      "learning_rate": 7.110705925026207e-07,
+      "loss": 0.7312,
+      "step": 37474
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7494947400095963,
+      "learning_rate": 7.107880285423563e-07,
+      "loss": 0.7499,
+      "step": 37475
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6095451249437513,
+      "learning_rate": 7.105055186668264e-07,
+      "loss": 0.8865,
+      "step": 37476
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.785619996329725,
+      "learning_rate": 7.10223062877674e-07,
+      "loss": 0.8016,
+      "step": 37477
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7079550716966374,
+      "learning_rate": 7.099406611765425e-07,
+      "loss": 0.8141,
+      "step": 37478
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4348403801100085,
+      "learning_rate": 7.096583135650792e-07,
+      "loss": 0.875,
+      "step": 37479
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4801974297416984,
+      "learning_rate": 7.093760200449251e-07,
+      "loss": 0.8276,
+      "step": 37480
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.288960144546922,
+      "learning_rate": 7.090937806177267e-07,
+      "loss": 0.8472,
+      "step": 37481
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7300999099796521,
+      "learning_rate": 7.088115952851238e-07,
+      "loss": 0.8297,
+      "step": 37482
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6723432225569783,
+      "learning_rate": 7.085294640487639e-07,
+      "loss": 0.7132,
+      "step": 37483
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8873965406017763,
+      "learning_rate": 7.082473869102846e-07,
+      "loss": 0.8377,
+      "step": 37484
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5379176609765508,
+      "learning_rate": 7.079653638713324e-07,
+      "loss": 0.7786,
+      "step": 37485
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.664570855812453,
+      "learning_rate": 7.076833949335482e-07,
+      "loss": 0.8012,
+      "step": 37486
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.422846387952423,
+      "learning_rate": 7.074014800985718e-07,
+      "loss": 0.8887,
+      "step": 37487
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8912897448719026,
+      "learning_rate": 7.071196193680452e-07,
+      "loss": 0.7851,
+      "step": 37488
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.599940845588634,
+      "learning_rate": 7.068378127436115e-07,
+      "loss": 0.7702,
+      "step": 37489
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.836366081402133,
+      "learning_rate": 7.065560602269106e-07,
+      "loss": 0.8865,
+      "step": 37490
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3754925455082554,
+      "learning_rate": 7.062743618195822e-07,
+      "loss": 0.788,
+      "step": 37491
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.788580595101314,
+      "learning_rate": 7.05992717523265e-07,
+      "loss": 0.9205,
+      "step": 37492
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.583797689459761,
+      "learning_rate": 7.057111273396034e-07,
+      "loss": 0.8062,
+      "step": 37493
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8578141008113813,
+      "learning_rate": 7.054295912702336e-07,
+      "loss": 0.8463,
+      "step": 37494
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6418500259583584,
+      "learning_rate": 7.051481093167933e-07,
+      "loss": 0.7944,
+      "step": 37495
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4428931120823902,
+      "learning_rate": 7.048666814809246e-07,
+      "loss": 0.8387,
+      "step": 37496
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7420029454461473,
+      "learning_rate": 7.045853077642661e-07,
+      "loss": 0.7228,
+      "step": 37497
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6241205010680548,
+      "learning_rate": 7.043039881684544e-07,
+      "loss": 0.8666,
+      "step": 37498
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5456172923697362,
+      "learning_rate": 7.040227226951268e-07,
+      "loss": 0.7992,
+      "step": 37499
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4752048752691767,
+      "learning_rate": 7.037415113459223e-07,
+      "loss": 0.7847,
+      "step": 37500
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4438119502880526,
+      "learning_rate": 7.034603541224771e-07,
+      "loss": 0.6588,
+      "step": 37501
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5784433656087353,
+      "learning_rate": 7.031792510264312e-07,
+      "loss": 0.8808,
+      "step": 37502
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5055159430813085,
+      "learning_rate": 7.028982020594166e-07,
+      "loss": 0.9169,
+      "step": 37503
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5702933471649496,
+      "learning_rate": 7.026172072230742e-07,
+      "loss": 0.8298,
+      "step": 37504
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.811943074146777,
+      "learning_rate": 7.023362665190358e-07,
+      "loss": 0.7107,
+      "step": 37505
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7451233481770244,
+      "learning_rate": 7.020553799489415e-07,
+      "loss": 0.6993,
+      "step": 37506
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.561536223539261,
+      "learning_rate": 7.017745475144233e-07,
+      "loss": 0.7001,
+      "step": 37507
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6135950514658837,
+      "learning_rate": 7.014937692171176e-07,
+      "loss": 0.8763,
+      "step": 37508
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.611613573597575,
+      "learning_rate": 7.012130450586574e-07,
+      "loss": 0.7663,
+      "step": 37509
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7038005143145827,
+      "learning_rate": 7.009323750406816e-07,
+      "loss": 0.8224,
+      "step": 37510
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7094356037387157,
+      "learning_rate": 7.00651759164821e-07,
+      "loss": 0.8549,
+      "step": 37511
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.706989080846755,
+      "learning_rate": 7.003711974327076e-07,
+      "loss": 0.9107,
+      "step": 37512
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.687493287565427,
+      "learning_rate": 7.00090689845978e-07,
+      "loss": 0.7846,
+      "step": 37513
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.18998385434998,
+      "learning_rate": 6.998102364062665e-07,
+      "loss": 0.8658,
+      "step": 37514
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7181894747079927,
+      "learning_rate": 6.995298371152037e-07,
+      "loss": 0.7938,
+      "step": 37515
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6850467166493113,
+      "learning_rate": 6.992494919744219e-07,
+      "loss": 0.8367,
+      "step": 37516
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7690454797769046,
+      "learning_rate": 6.989692009855543e-07,
+      "loss": 0.834,
+      "step": 37517
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6577204551974003,
+      "learning_rate": 6.986889641502326e-07,
+      "loss": 0.9176,
+      "step": 37518
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6076434850743266,
+      "learning_rate": 6.984087814700902e-07,
+      "loss": 0.8607,
+      "step": 37519
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6601818576759144,
+      "learning_rate": 6.981286529467546e-07,
+      "loss": 0.8226,
+      "step": 37520
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8016625559690514,
+      "learning_rate": 6.9784857858186e-07,
+      "loss": 0.6922,
+      "step": 37521
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7018667996595689,
+      "learning_rate": 6.975685583770341e-07,
+      "loss": 0.7921,
+      "step": 37522
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.577340947176454,
+      "learning_rate": 6.972885923339123e-07,
+      "loss": 0.7297,
+      "step": 37523
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4371880022530346,
+      "learning_rate": 6.970086804541187e-07,
+      "loss": 0.8648,
+      "step": 37524
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.701683081427313,
+      "learning_rate": 6.967288227392877e-07,
+      "loss": 0.7665,
+      "step": 37525
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.63793733013571,
+      "learning_rate": 6.964490191910456e-07,
+      "loss": 0.8228,
+      "step": 37526
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6306348775345025,
+      "learning_rate": 6.961692698110223e-07,
+      "loss": 0.7837,
+      "step": 37527
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.746604477421463,
+      "learning_rate": 6.958895746008499e-07,
+      "loss": 0.6877,
+      "step": 37528
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.703339530809942,
+      "learning_rate": 6.956099335621513e-07,
+      "loss": 0.7623,
+      "step": 37529
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.793246178516054,
+      "learning_rate": 6.953303466965578e-07,
+      "loss": 0.8613,
+      "step": 37530
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5943646391664978,
+      "learning_rate": 6.950508140056966e-07,
+      "loss": 0.8448,
+      "step": 37531
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.098474820979339,
+      "learning_rate": 6.947713354911978e-07,
+      "loss": 0.9038,
+      "step": 37532
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8897693879024573,
+      "learning_rate": 6.944919111546844e-07,
+      "loss": 0.7088,
+      "step": 37533
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6058450873971626,
+      "learning_rate": 6.94212540997784e-07,
+      "loss": 0.7694,
+      "step": 37534
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.576984145292228,
+      "learning_rate": 6.939332250221265e-07,
+      "loss": 0.8282,
+      "step": 37535
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4622283985935212,
+      "learning_rate": 6.93653963229336e-07,
+      "loss": 0.7756,
+      "step": 37536
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8004634088529643,
+      "learning_rate": 6.933747556210368e-07,
+      "loss": 0.7253,
+      "step": 37537
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6712841986814666,
+      "learning_rate": 6.930956021988567e-07,
+      "loss": 0.8471,
+      "step": 37538
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6076792873567576,
+      "learning_rate": 6.928165029644219e-07,
+      "loss": 0.7039,
+      "step": 37539
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0077125686737176,
+      "learning_rate": 6.925374579193533e-07,
+      "loss": 0.9445,
+      "step": 37540
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6364670968427493,
+      "learning_rate": 6.92258467065281e-07,
+      "loss": 0.9016,
+      "step": 37541
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7725326281244654,
+      "learning_rate": 6.919795304038268e-07,
+      "loss": 0.6757,
+      "step": 37542
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9437457200973056,
+      "learning_rate": 6.917006479366128e-07,
+      "loss": 0.8142,
+      "step": 37543
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6108768386399839,
+      "learning_rate": 6.914218196652645e-07,
+      "loss": 0.8698,
+      "step": 37544
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4978569771754224,
+      "learning_rate": 6.91143045591407e-07,
+      "loss": 0.7772,
+      "step": 37545
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5606317025620466,
+      "learning_rate": 6.908643257166625e-07,
+      "loss": 0.7723,
+      "step": 37546
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6469022881149022,
+      "learning_rate": 6.905856600426508e-07,
+      "loss": 0.8917,
+      "step": 37547
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4279890052090891,
+      "learning_rate": 6.903070485709962e-07,
+      "loss": 0.6114,
+      "step": 37548
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8474192651165435,
+      "learning_rate": 6.90028491303324e-07,
+      "loss": 0.776,
+      "step": 37549
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6855605465268497,
+      "learning_rate": 6.897499882412529e-07,
+      "loss": 0.8229,
+      "step": 37550
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.773453777014319,
+      "learning_rate": 6.894715393864027e-07,
+      "loss": 0.7024,
+      "step": 37551
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4712935680313834,
+      "learning_rate": 6.891931447403977e-07,
+      "loss": 0.8292,
+      "step": 37552
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6636306072522478,
+      "learning_rate": 6.889148043048599e-07,
+      "loss": 0.7739,
+      "step": 37553
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.14041327041807,
+      "learning_rate": 6.886365180814069e-07,
+      "loss": 0.7402,
+      "step": 37554
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6711131696920813,
+      "learning_rate": 6.883582860716598e-07,
+      "loss": 0.8689,
+      "step": 37555
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6521477048756774,
+      "learning_rate": 6.880801082772382e-07,
+      "loss": 0.8633,
+      "step": 37556
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8299230934316202,
+      "learning_rate": 6.87801984699763e-07,
+      "loss": 0.8597,
+      "step": 37557
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.635581839498835,
+      "learning_rate": 6.875239153408541e-07,
+      "loss": 0.8064,
+      "step": 37558
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7452858754846707,
+      "learning_rate": 6.87245900202127e-07,
+      "loss": 0.755,
+      "step": 37559
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.745100095092298,
+      "learning_rate": 6.869679392852035e-07,
+      "loss": 0.8925,
+      "step": 37560
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7674310902217768,
+      "learning_rate": 6.866900325917003e-07,
+      "loss": 0.6616,
+      "step": 37561
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8365819788564053,
+      "learning_rate": 6.864121801232381e-07,
+      "loss": 0.8364,
+      "step": 37562
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6126244317718457,
+      "learning_rate": 6.861343818814315e-07,
+      "loss": 0.9372,
+      "step": 37563
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7414306875129486,
+      "learning_rate": 6.858566378678999e-07,
+      "loss": 0.7712,
+      "step": 37564
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6183562680966364,
+      "learning_rate": 6.855789480842589e-07,
+      "loss": 0.8945,
+      "step": 37565
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6834369315195257,
+      "learning_rate": 6.853013125321273e-07,
+      "loss": 0.8748,
+      "step": 37566
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.56004679086635,
+      "learning_rate": 6.850237312131213e-07,
+      "loss": 0.8128,
+      "step": 37567
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4995205933138385,
+      "learning_rate": 6.847462041288533e-07,
+      "loss": 0.789,
+      "step": 37568
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5666341731919036,
+      "learning_rate": 6.844687312809429e-07,
+      "loss": 0.8147,
+      "step": 37569
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7471307585531386,
+      "learning_rate": 6.841913126710065e-07,
+      "loss": 0.7051,
+      "step": 37570
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5365301050867273,
+      "learning_rate": 6.839139483006574e-07,
+      "loss": 0.7875,
+      "step": 37571
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5160849447259057,
+      "learning_rate": 6.836366381715087e-07,
+      "loss": 0.9164,
+      "step": 37572
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.791405684919964,
+      "learning_rate": 6.83359382285177e-07,
+      "loss": 0.7232,
+      "step": 37573
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.008388399746105,
+      "learning_rate": 6.830821806432786e-07,
+      "loss": 0.7876,
+      "step": 37574
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7704374917879117,
+      "learning_rate": 6.828050332474256e-07,
+      "loss": 0.7132,
+      "step": 37575
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7541948309555505,
+      "learning_rate": 6.825279400992291e-07,
+      "loss": 0.6713,
+      "step": 37576
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7135774677289508,
+      "learning_rate": 6.822509012003054e-07,
+      "loss": 0.8131,
+      "step": 37577
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5296607150682824,
+      "learning_rate": 6.819739165522688e-07,
+      "loss": 0.7947,
+      "step": 37578
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5342300588201754,
+      "learning_rate": 6.816969861567291e-07,
+      "loss": 0.9247,
+      "step": 37579
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7418161182802431,
+      "learning_rate": 6.814201100152995e-07,
+      "loss": 0.7079,
+      "step": 37580
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7351644465055844,
+      "learning_rate": 6.811432881295921e-07,
+      "loss": 0.8002,
+      "step": 37581
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8182090038537742,
+      "learning_rate": 6.808665205012188e-07,
+      "loss": 0.8822,
+      "step": 37582
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2277253333636007,
+      "learning_rate": 6.805898071317918e-07,
+      "loss": 0.8893,
+      "step": 37583
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7808381778872739,
+      "learning_rate": 6.803131480229197e-07,
+      "loss": 0.7323,
+      "step": 37584
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4639073423280606,
+      "learning_rate": 6.800365431762168e-07,
+      "loss": 0.8988,
+      "step": 37585
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0659138216656863,
+      "learning_rate": 6.797599925932907e-07,
+      "loss": 0.9171,
+      "step": 37586
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9153181462932645,
+      "learning_rate": 6.794834962757535e-07,
+      "loss": 0.7289,
+      "step": 37587
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5311372677295434,
+      "learning_rate": 6.792070542252138e-07,
+      "loss": 0.7504,
+      "step": 37588
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7057015699630582,
+      "learning_rate": 6.789306664432815e-07,
+      "loss": 0.8621,
+      "step": 37589
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6232585591954058,
+      "learning_rate": 6.786543329315643e-07,
+      "loss": 0.8693,
+      "step": 37590
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6809089106930177,
+      "learning_rate": 6.783780536916751e-07,
+      "loss": 0.7858,
+      "step": 37591
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5129867876270633,
+      "learning_rate": 6.781018287252206e-07,
+      "loss": 0.7705,
+      "step": 37592
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5836000427500747,
+      "learning_rate": 6.77825658033806e-07,
+      "loss": 0.8599,
+      "step": 37593
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6241491015407767,
+      "learning_rate": 6.775495416190414e-07,
+      "loss": 0.8314,
+      "step": 37594
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6273513083578868,
+      "learning_rate": 6.772734794825375e-07,
+      "loss": 0.8298,
+      "step": 37595
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.172019880886591,
+      "learning_rate": 6.769974716258987e-07,
+      "loss": 0.7984,
+      "step": 37596
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.587766685977451,
+      "learning_rate": 6.767215180507303e-07,
+      "loss": 0.8837,
+      "step": 37597
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6896230222707136,
+      "learning_rate": 6.76445618758641e-07,
+      "loss": 0.8079,
+      "step": 37598
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5869468371968216,
+      "learning_rate": 6.761697737512385e-07,
+      "loss": 0.6944,
+      "step": 37599
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5683227472953987,
+      "learning_rate": 6.75893983030127e-07,
+      "loss": 0.8564,
+      "step": 37600
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5113668620952383,
+      "learning_rate": 6.756182465969119e-07,
+      "loss": 0.8317,
+      "step": 37601
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6046988449168944,
+      "learning_rate": 6.753425644532008e-07,
+      "loss": 0.9366,
+      "step": 37602
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7760561400741544,
+      "learning_rate": 6.750669366005957e-07,
+      "loss": 0.6939,
+      "step": 37603
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6483535921094215,
+      "learning_rate": 6.747913630407054e-07,
+      "loss": 0.8605,
+      "step": 37604
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.772216525874511,
+      "learning_rate": 6.745158437751298e-07,
+      "loss": 0.712,
+      "step": 37605
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5366241416713038,
+      "learning_rate": 6.742403788054775e-07,
+      "loss": 0.947,
+      "step": 37606
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6156865756371739,
+      "learning_rate": 6.739649681333482e-07,
+      "loss": 0.9351,
+      "step": 37607
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6147260366895024,
+      "learning_rate": 6.736896117603487e-07,
+      "loss": 0.8468,
+      "step": 37608
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8181467325199786,
+      "learning_rate": 6.734143096880808e-07,
+      "loss": 0.7292,
+      "step": 37609
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7000533152626034,
+      "learning_rate": 6.731390619181466e-07,
+      "loss": 0.8407,
+      "step": 37610
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5518611560669247,
+      "learning_rate": 6.728638684521493e-07,
+      "loss": 0.8562,
+      "step": 37611
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5770869409895685,
+      "learning_rate": 6.725887292916933e-07,
+      "loss": 0.8239,
+      "step": 37612
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.516301765047708,
+      "learning_rate": 6.723136444383793e-07,
+      "loss": 0.9282,
+      "step": 37613
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4880749218856333,
+      "learning_rate": 6.720386138938062e-07,
+      "loss": 0.8193,
+      "step": 37614
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4240092095451506,
+      "learning_rate": 6.717636376595771e-07,
+      "loss": 0.7436,
+      "step": 37615
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5451941814212216,
+      "learning_rate": 6.714887157372952e-07,
+      "loss": 0.8331,
+      "step": 37616
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7842438131294215,
+      "learning_rate": 6.71213848128559e-07,
+      "loss": 0.7222,
+      "step": 37617
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6870946053653137,
+      "learning_rate": 6.709390348349687e-07,
+      "loss": 0.8707,
+      "step": 37618
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5669368704073987,
+      "learning_rate": 6.706642758581239e-07,
+      "loss": 0.7746,
+      "step": 37619
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9769534569381328,
+      "learning_rate": 6.703895711996278e-07,
+      "loss": 0.6707,
+      "step": 37620
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.745756810103409,
+      "learning_rate": 6.701149208610758e-07,
+      "loss": 0.6843,
+      "step": 37621
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.59650216218478,
+      "learning_rate": 6.698403248440688e-07,
+      "loss": 0.7556,
+      "step": 37622
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5348120391958577,
+      "learning_rate": 6.695657831502056e-07,
+      "loss": 0.9143,
+      "step": 37623
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.525967407578257,
+      "learning_rate": 6.692912957810826e-07,
+      "loss": 0.8821,
+      "step": 37624
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3127249794738796,
+      "learning_rate": 6.69016862738301e-07,
+      "loss": 0.8189,
+      "step": 37625
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6511372190413245,
+      "learning_rate": 6.687424840234579e-07,
+      "loss": 0.7529,
+      "step": 37626
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8124592548537894,
+      "learning_rate": 6.684681596381503e-07,
+      "loss": 0.8213,
+      "step": 37627
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6143441553254785,
+      "learning_rate": 6.681938895839746e-07,
+      "loss": 0.719,
+      "step": 37628
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6008329875962366,
+      "learning_rate": 6.679196738625282e-07,
+      "loss": 0.7237,
+      "step": 37629
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5545277728375106,
+      "learning_rate": 6.676455124754099e-07,
+      "loss": 0.7435,
+      "step": 37630
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.481040326738654,
+      "learning_rate": 6.67371405424212e-07,
+      "loss": 0.8492,
+      "step": 37631
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6944163292898784,
+      "learning_rate": 6.670973527105329e-07,
+      "loss": 0.8459,
+      "step": 37632
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8017751987517823,
+      "learning_rate": 6.66823354335967e-07,
+      "loss": 0.7714,
+      "step": 37633
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.534453058059227,
+      "learning_rate": 6.66549410302113e-07,
+      "loss": 0.8629,
+      "step": 37634
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4977235878472288,
+      "learning_rate": 6.662755206105608e-07,
+      "loss": 0.7285,
+      "step": 37635
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5524914893433055,
+      "learning_rate": 6.66001685262908e-07,
+      "loss": 0.7714,
+      "step": 37636
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4596921546573587,
+      "learning_rate": 6.65727904260749e-07,
+      "loss": 0.8904,
+      "step": 37637
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6242233215795105,
+      "learning_rate": 6.654541776056767e-07,
+      "loss": 0.8143,
+      "step": 37638
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8901159484111822,
+      "learning_rate": 6.651805052992866e-07,
+      "loss": 0.7851,
+      "step": 37639
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8239202166569342,
+      "learning_rate": 6.649068873431697e-07,
+      "loss": 0.7769,
+      "step": 37640
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7522233350957249,
+      "learning_rate": 6.646333237389213e-07,
+      "loss": 0.7044,
+      "step": 37641
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7729491253141598,
+      "learning_rate": 6.643598144881324e-07,
+      "loss": 0.7288,
+      "step": 37642
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8745249865199751,
+      "learning_rate": 6.640863595923975e-07,
+      "loss": 0.8895,
+      "step": 37643
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9104514350104613,
+      "learning_rate": 6.638129590533071e-07,
+      "loss": 0.8616,
+      "step": 37644
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6930700984547835,
+      "learning_rate": 6.635396128724513e-07,
+      "loss": 0.8524,
+      "step": 37645
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.504642852898557,
+      "learning_rate": 6.632663210514257e-07,
+      "loss": 0.7527,
+      "step": 37646
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4458439380323505,
+      "learning_rate": 6.629930835918196e-07,
+      "loss": 1.0352,
+      "step": 37647
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.649194658049436,
+      "learning_rate": 6.627199004952245e-07,
+      "loss": 0.7782,
+      "step": 37648
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7249373089630144,
+      "learning_rate": 6.624467717632277e-07,
+      "loss": 0.8313,
+      "step": 37649
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5648382710082636,
+      "learning_rate": 6.621736973974236e-07,
+      "loss": 0.7576,
+      "step": 37650
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.68439968658252,
+      "learning_rate": 6.619006773994008e-07,
+      "loss": 0.7504,
+      "step": 37651
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5012094953937205,
+      "learning_rate": 6.616277117707493e-07,
+      "loss": 0.8163,
+      "step": 37652
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8094574581527383,
+      "learning_rate": 6.613548005130555e-07,
+      "loss": 0.9108,
+      "step": 37653
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0229775164434516,
+      "learning_rate": 6.610819436279115e-07,
+      "loss": 0.8064,
+      "step": 37654
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5845805223092961,
+      "learning_rate": 6.60809141116906e-07,
+      "loss": 0.7956,
+      "step": 37655
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8083377410671224,
+      "learning_rate": 6.605363929816266e-07,
+      "loss": 0.9334,
+      "step": 37656
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.725453791809892,
+      "learning_rate": 6.6026369922366e-07,
+      "loss": 0.845,
+      "step": 37657
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9860516557688304,
+      "learning_rate": 6.599910598445958e-07,
+      "loss": 0.795,
+      "step": 37658
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7483605459286593,
+      "learning_rate": 6.597184748460195e-07,
+      "loss": 0.7137,
+      "step": 37659
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8147194697226722,
+      "learning_rate": 6.594459442295209e-07,
+      "loss": 0.7172,
+      "step": 37660
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4868250834081849,
+      "learning_rate": 6.591734679966844e-07,
+      "loss": 0.7946,
+      "step": 37661
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.60106809136625,
+      "learning_rate": 6.589010461490974e-07,
+      "loss": 0.8494,
+      "step": 37662
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4558446120652828,
+      "learning_rate": 6.586286786883455e-07,
+      "loss": 0.7482,
+      "step": 37663
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2065305903117722,
+      "learning_rate": 6.583563656160152e-07,
+      "loss": 0.7927,
+      "step": 37664
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0768435702958636,
+      "learning_rate": 6.580841069336929e-07,
+      "loss": 0.7883,
+      "step": 37665
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.534470958183173,
+      "learning_rate": 6.578119026429597e-07,
+      "loss": 0.8149,
+      "step": 37666
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8075664423887465,
+      "learning_rate": 6.575397527454042e-07,
+      "loss": 0.7657,
+      "step": 37667
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.578512116020807,
+      "learning_rate": 6.572676572426118e-07,
+      "loss": 0.7786,
+      "step": 37668
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5522946784061966,
+      "learning_rate": 6.569956161361646e-07,
+      "loss": 0.6784,
+      "step": 37669
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8236932262936043,
+      "learning_rate": 6.567236294276446e-07,
+      "loss": 0.7472,
+      "step": 37670
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7041484097234838,
+      "learning_rate": 6.564516971186385e-07,
+      "loss": 0.8393,
+      "step": 37671
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.552405706407481,
+      "learning_rate": 6.561798192107305e-07,
+      "loss": 0.8847,
+      "step": 37672
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7075313132766992,
+      "learning_rate": 6.559079957055015e-07,
+      "loss": 0.8267,
+      "step": 37673
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.091817301720905,
+      "learning_rate": 6.556362266045324e-07,
+      "loss": 0.8853,
+      "step": 37674
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6795207703255728,
+      "learning_rate": 6.553645119094076e-07,
+      "loss": 0.8168,
+      "step": 37675
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5630268669884713,
+      "learning_rate": 6.550928516217115e-07,
+      "loss": 0.8266,
+      "step": 37676
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4455706185462518,
+      "learning_rate": 6.548212457430225e-07,
+      "loss": 0.8008,
+      "step": 37677
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.461625933463121,
+      "learning_rate": 6.545496942749219e-07,
+      "loss": 0.7813,
+      "step": 37678
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4827503321584894,
+      "learning_rate": 6.542781972189915e-07,
+      "loss": 0.7994,
+      "step": 37679
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.106431704967107,
+      "learning_rate": 6.540067545768136e-07,
+      "loss": 0.8984,
+      "step": 37680
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5744811374336314,
+      "learning_rate": 6.537353663499669e-07,
+      "loss": 0.7152,
+      "step": 37681
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7962378137439563,
+      "learning_rate": 6.53464032540031e-07,
+      "loss": 0.8926,
+      "step": 37682
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4492273489486194,
+      "learning_rate": 6.531927531485882e-07,
+      "loss": 0.8428,
+      "step": 37683
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.531320400286183,
+      "learning_rate": 6.529215281772139e-07,
+      "loss": 0.7996,
+      "step": 37684
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7521256374329,
+      "learning_rate": 6.526503576274923e-07,
+      "loss": 0.8241,
+      "step": 37685
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5125185610495824,
+      "learning_rate": 6.523792415009977e-07,
+      "loss": 0.8527,
+      "step": 37686
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.510120370602564,
+      "learning_rate": 6.521081797993123e-07,
+      "loss": 0.8805,
+      "step": 37687
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6176670088905807,
+      "learning_rate": 6.518371725240103e-07,
+      "loss": 0.9122,
+      "step": 37688
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.591224602864416,
+      "learning_rate": 6.515662196766737e-07,
+      "loss": 0.8338,
+      "step": 37689
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9917306797151313,
+      "learning_rate": 6.512953212588791e-07,
+      "loss": 0.8281,
+      "step": 37690
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.76288520887755,
+      "learning_rate": 6.510244772721996e-07,
+      "loss": 0.822,
+      "step": 37691
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.693310386079436,
+      "learning_rate": 6.507536877182175e-07,
+      "loss": 0.9576,
+      "step": 37692
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7780582101214013,
+      "learning_rate": 6.504829525985079e-07,
+      "loss": 0.7773,
+      "step": 37693
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4921894652941317,
+      "learning_rate": 6.502122719146464e-07,
+      "loss": 0.7987,
+      "step": 37694
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.558134461815748,
+      "learning_rate": 6.499416456682072e-07,
+      "loss": 0.7923,
+      "step": 37695
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.755284143446415,
+      "learning_rate": 6.496710738607681e-07,
+      "loss": 0.656,
+      "step": 37696
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6773833265549964,
+      "learning_rate": 6.494005564939066e-07,
+      "loss": 0.7783,
+      "step": 37697
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.764556845972764,
+      "learning_rate": 6.491300935691947e-07,
+      "loss": 0.6914,
+      "step": 37698
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5108202986583827,
+      "learning_rate": 6.488596850882068e-07,
+      "loss": 0.7793,
+      "step": 37699
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.129855551685225,
+      "learning_rate": 6.485893310525182e-07,
+      "loss": 0.7997,
+      "step": 37700
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7681823174732694,
+      "learning_rate": 6.483190314637045e-07,
+      "loss": 0.6949,
+      "step": 37701
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5979057566004062,
+      "learning_rate": 6.480487863233387e-07,
+      "loss": 0.9003,
+      "step": 37702
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.464041305474171,
+      "learning_rate": 6.477785956329907e-07,
+      "loss": 0.736,
+      "step": 37703
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7405328804659077,
+      "learning_rate": 6.475084593942393e-07,
+      "loss": 0.9105,
+      "step": 37704
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5950221280120207,
+      "learning_rate": 6.472383776086533e-07,
+      "loss": 0.7546,
+      "step": 37705
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5361108412106521,
+      "learning_rate": 6.469683502778068e-07,
+      "loss": 1.0265,
+      "step": 37706
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8244347210912966,
+      "learning_rate": 6.466983774032709e-07,
+      "loss": 0.779,
+      "step": 37707
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5888173416103897,
+      "learning_rate": 6.464284589866199e-07,
+      "loss": 0.8065,
+      "step": 37708
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7038898749498763,
+      "learning_rate": 6.461585950294214e-07,
+      "loss": 0.8057,
+      "step": 37709
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7758594807818922,
+      "learning_rate": 6.458887855332518e-07,
+      "loss": 0.9099,
+      "step": 37710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5104657655523177,
+      "learning_rate": 6.456190304996779e-07,
+      "loss": 0.8342,
+      "step": 37711
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3338893755227534,
+      "learning_rate": 6.453493299302704e-07,
+      "loss": 0.8287,
+      "step": 37712
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.438517306088121,
+      "learning_rate": 6.450796838266015e-07,
+      "loss": 0.8318,
+      "step": 37713
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5715966632988307,
+      "learning_rate": 6.44810092190239e-07,
+      "loss": 0.6911,
+      "step": 37714
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.101566934544886,
+      "learning_rate": 6.44540555022758e-07,
+      "loss": 0.6548,
+      "step": 37715
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6033281082414437,
+      "learning_rate": 6.442710723257206e-07,
+      "loss": 0.792,
+      "step": 37716
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.026373747523548,
+      "learning_rate": 6.440016441006991e-07,
+      "loss": 0.9013,
+      "step": 37717
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6782270799597405,
+      "learning_rate": 6.437322703492633e-07,
+      "loss": 0.7234,
+      "step": 37718
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4659626989563372,
+      "learning_rate": 6.434629510729806e-07,
+      "loss": 0.7083,
+      "step": 37719
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9085724098649262,
+      "learning_rate": 6.431936862734167e-07,
+      "loss": 0.7954,
+      "step": 37720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7705456235992855,
+      "learning_rate": 6.429244759521425e-07,
+      "loss": 0.6786,
+      "step": 37721
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6329384236098872,
+      "learning_rate": 6.426553201107256e-07,
+      "loss": 0.7204,
+      "step": 37722
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1788877318096462,
+      "learning_rate": 6.423862187507291e-07,
+      "loss": 0.8301,
+      "step": 37723
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7330787853142066,
+      "learning_rate": 6.421171718737252e-07,
+      "loss": 0.6807,
+      "step": 37724
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7660088457304093,
+      "learning_rate": 6.418481794812781e-07,
+      "loss": 0.7057,
+      "step": 37725
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.497752767419603,
+      "learning_rate": 6.415792415749522e-07,
+      "loss": 0.8127,
+      "step": 37726
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4604966392043528,
+      "learning_rate": 6.41310358156314e-07,
+      "loss": 0.7745,
+      "step": 37727
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4504030078730932,
+      "learning_rate": 6.410415292269312e-07,
+      "loss": 0.7828,
+      "step": 37728
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5540749816191755,
+      "learning_rate": 6.407727547883691e-07,
+      "loss": 0.881,
+      "step": 37729
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6741649773221217,
+      "learning_rate": 6.405040348421876e-07,
+      "loss": 0.804,
+      "step": 37730
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.611041958370308,
+      "learning_rate": 6.402353693899566e-07,
+      "loss": 0.7582,
+      "step": 37731
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6209964924996023,
+      "learning_rate": 6.399667584332403e-07,
+      "loss": 0.782,
+      "step": 37732
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6639758373645148,
+      "learning_rate": 6.396982019735987e-07,
+      "loss": 0.8991,
+      "step": 37733
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.575616319375446,
+      "learning_rate": 6.394297000125971e-07,
+      "loss": 0.8979,
+      "step": 37734
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5844828118570438,
+      "learning_rate": 6.391612525517988e-07,
+      "loss": 0.7865,
+      "step": 37735
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5444374359552215,
+      "learning_rate": 6.388928595927691e-07,
+      "loss": 0.6557,
+      "step": 37736
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4391176683553641,
+      "learning_rate": 6.38624521137069e-07,
+      "loss": 0.7954,
+      "step": 37737
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7193327966539715,
+      "learning_rate": 6.383562371862595e-07,
+      "loss": 0.9127,
+      "step": 37738
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8085091357193712,
+      "learning_rate": 6.380880077419049e-07,
+      "loss": 0.7695,
+      "step": 37739
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6752841059623487,
+      "learning_rate": 6.378198328055652e-07,
+      "loss": 0.8094,
+      "step": 37740
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7501118722136797,
+      "learning_rate": 6.375517123788033e-07,
+      "loss": 0.7838,
+      "step": 37741
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4899436103708612,
+      "learning_rate": 6.372836464631771e-07,
+      "loss": 0.7896,
+      "step": 37742
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.605263313399785,
+      "learning_rate": 6.37015635060253e-07,
+      "loss": 0.8556,
+      "step": 37743
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6940448332614284,
+      "learning_rate": 6.367476781715853e-07,
+      "loss": 0.7106,
+      "step": 37744
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.238123040574592,
+      "learning_rate": 6.364797757987384e-07,
+      "loss": 0.7996,
+      "step": 37745
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5114219345106796,
+      "learning_rate": 6.36211927943271e-07,
+      "loss": 0.8303,
+      "step": 37746
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0601267753341275,
+      "learning_rate": 6.359441346067419e-07,
+      "loss": 0.8532,
+      "step": 37747
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7421921672646268,
+      "learning_rate": 6.356763957907097e-07,
+      "loss": 0.7886,
+      "step": 37748
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5078307916434759,
+      "learning_rate": 6.354087114967355e-07,
+      "loss": 0.8635,
+      "step": 37749
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6363704030350585,
+      "learning_rate": 6.35141081726377e-07,
+      "loss": 0.6671,
+      "step": 37750
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7690029987694222,
+      "learning_rate": 6.348735064811895e-07,
+      "loss": 0.7834,
+      "step": 37751
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5497936688079832,
+      "learning_rate": 6.34605985762734e-07,
+      "loss": 0.7935,
+      "step": 37752
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8023623173157355,
+      "learning_rate": 6.343385195725693e-07,
+      "loss": 0.7929,
+      "step": 37753
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6801079757962276,
+      "learning_rate": 6.340711079122508e-07,
+      "loss": 0.8228,
+      "step": 37754
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7557408044459661,
+      "learning_rate": 6.338037507833339e-07,
+      "loss": 0.7703,
+      "step": 37755
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4735078310722445,
+      "learning_rate": 6.335364481873762e-07,
+      "loss": 0.7751,
+      "step": 37756
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4632607980523185,
+      "learning_rate": 6.332692001259377e-07,
+      "loss": 0.8849,
+      "step": 37757
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2316102323142464,
+      "learning_rate": 6.330020066005704e-07,
+      "loss": 0.8573,
+      "step": 37758
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5962182290446787,
+      "learning_rate": 6.327348676128298e-07,
+      "loss": 0.7729,
+      "step": 37759
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5518882420870659,
+      "learning_rate": 6.324677831642734e-07,
+      "loss": 0.7808,
+      "step": 37760
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7018446250042412,
+      "learning_rate": 6.322007532564533e-07,
+      "loss": 0.9225,
+      "step": 37761
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4199206253310914,
+      "learning_rate": 6.319337778909295e-07,
+      "loss": 0.8829,
+      "step": 37762
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6837936675806384,
+      "learning_rate": 6.316668570692497e-07,
+      "loss": 0.7209,
+      "step": 37763
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7045330408917776,
+      "learning_rate": 6.313999907929746e-07,
+      "loss": 0.8418,
+      "step": 37764
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7594453280878932,
+      "learning_rate": 6.311331790636521e-07,
+      "loss": 0.899,
+      "step": 37765
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4348252000141737,
+      "learning_rate": 6.308664218828397e-07,
+      "loss": 0.8707,
+      "step": 37766
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.565650888171138,
+      "learning_rate": 6.305997192520908e-07,
+      "loss": 0.7159,
+      "step": 37767
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5732955239473092,
+      "learning_rate": 6.303330711729539e-07,
+      "loss": 0.7707,
+      "step": 37768
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6826865427513809,
+      "learning_rate": 6.300664776469845e-07,
+      "loss": 0.758,
+      "step": 37769
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4447351291165158,
+      "learning_rate": 6.297999386757359e-07,
+      "loss": 0.8179,
+      "step": 37770
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7879868773704324,
+      "learning_rate": 6.29533454260759e-07,
+      "loss": 0.6855,
+      "step": 37771
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7975605762394802,
+      "learning_rate": 6.292670244036026e-07,
+      "loss": 0.8371,
+      "step": 37772
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4710699219433838,
+      "learning_rate": 6.290006491058209e-07,
+      "loss": 0.773,
+      "step": 37773
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5598649985001793,
+      "learning_rate": 6.287343283689662e-07,
+      "loss": 0.8387,
+      "step": 37774
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.230401761176202,
+      "learning_rate": 6.28468062194586e-07,
+      "loss": 0.7244,
+      "step": 37775
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.650430927369987,
+      "learning_rate": 6.282018505842314e-07,
+      "loss": 0.8242,
+      "step": 37776
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5037165575176996,
+      "learning_rate": 6.27935693539452e-07,
+      "loss": 0.7367,
+      "step": 37777
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4975491253658,
+      "learning_rate": 6.276695910618002e-07,
+      "loss": 0.8976,
+      "step": 37778
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.013565962522234,
+      "learning_rate": 6.274035431528225e-07,
+      "loss": 0.9026,
+      "step": 37779
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6432358181836315,
+      "learning_rate": 6.271375498140675e-07,
+      "loss": 0.9608,
+      "step": 37780
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6695684003284892,
+      "learning_rate": 6.268716110470863e-07,
+      "loss": 0.8612,
+      "step": 37781
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4541634921484454,
+      "learning_rate": 6.266057268534254e-07,
+      "loss": 0.7973,
+      "step": 37782
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4975224555056368,
+      "learning_rate": 6.263398972346346e-07,
+      "loss": 0.7319,
+      "step": 37783
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.098656793538778,
+      "learning_rate": 6.260741221922583e-07,
+      "loss": 0.875,
+      "step": 37784
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.429933257896465,
+      "learning_rate": 6.258084017278487e-07,
+      "loss": 0.805,
+      "step": 37785
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.557759677190238,
+      "learning_rate": 6.255427358429489e-07,
+      "loss": 0.8083,
+      "step": 37786
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0411847393002494,
+      "learning_rate": 6.252771245391087e-07,
+      "loss": 0.9802,
+      "step": 37787
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4779452216467346,
+      "learning_rate": 6.250115678178725e-07,
+      "loss": 0.7037,
+      "step": 37788
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.74136149479736,
+      "learning_rate": 6.247460656807857e-07,
+      "loss": 0.7662,
+      "step": 37789
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5386339875038653,
+      "learning_rate": 6.24480618129395e-07,
+      "loss": 0.7919,
+      "step": 37790
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.556247492054005,
+      "learning_rate": 6.242152251652489e-07,
+      "loss": 0.744,
+      "step": 37791
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7946770338925802,
+      "learning_rate": 6.239498867898897e-07,
+      "loss": 0.9148,
+      "step": 37792
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.594659768712965,
+      "learning_rate": 6.236846030048605e-07,
+      "loss": 0.9259,
+      "step": 37793
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.498442751344131,
+      "learning_rate": 6.234193738117089e-07,
+      "loss": 0.8672,
+      "step": 37794
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7106846784600345,
+      "learning_rate": 6.231541992119794e-07,
+      "loss": 0.9089,
+      "step": 37795
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7164578367660694,
+      "learning_rate": 6.22889079207215e-07,
+      "loss": 0.7799,
+      "step": 37796
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5564829699478673,
+      "learning_rate": 6.226240137989581e-07,
+      "loss": 0.7693,
+      "step": 37797
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.568842331930387,
+      "learning_rate": 6.223590029887516e-07,
+      "loss": 0.8434,
+      "step": 37798
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4756965678179688,
+      "learning_rate": 6.220940467781422e-07,
+      "loss": 0.7926,
+      "step": 37799
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8481865646918894,
+      "learning_rate": 6.218291451686709e-07,
+      "loss": 0.7751,
+      "step": 37800
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5659506774376137,
+      "learning_rate": 6.215642981618775e-07,
+      "loss": 0.8108,
+      "step": 37801
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5739478825621969,
+      "learning_rate": 6.212995057593052e-07,
+      "loss": 0.8736,
+      "step": 37802
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.784959927232507,
+      "learning_rate": 6.210347679624984e-07,
+      "loss": 0.7397,
+      "step": 37803
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5661331815567887,
+      "learning_rate": 6.207700847729969e-07,
+      "loss": 0.9575,
+      "step": 37804
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5331910871683123,
+      "learning_rate": 6.205054561923396e-07,
+      "loss": 0.8857,
+      "step": 37805
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.471201366858658,
+      "learning_rate": 6.202408822220696e-07,
+      "loss": 0.7696,
+      "step": 37806
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.575556807944303,
+      "learning_rate": 6.199763628637267e-07,
+      "loss": 0.9313,
+      "step": 37807
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6921636734953587,
+      "learning_rate": 6.197118981188499e-07,
+      "loss": 0.7172,
+      "step": 37808
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.993316312953679,
+      "learning_rate": 6.194474879889823e-07,
+      "loss": 0.8135,
+      "step": 37809
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6274403976382368,
+      "learning_rate": 6.191831324756615e-07,
+      "loss": 0.7693,
+      "step": 37810
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6593784207958553,
+      "learning_rate": 6.189188315804251e-07,
+      "loss": 0.6872,
+      "step": 37811
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.327849607127844,
+      "learning_rate": 6.18654585304812e-07,
+      "loss": 0.9014,
+      "step": 37812
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5534905125793461,
+      "learning_rate": 6.183903936503666e-07,
+      "loss": 0.7357,
+      "step": 37813
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7852299521877818,
+      "learning_rate": 6.181262566186186e-07,
+      "loss": 0.7434,
+      "step": 37814
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8358920042888927,
+      "learning_rate": 6.178621742111102e-07,
+      "loss": 0.9227,
+      "step": 37815
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5376605729746577,
+      "learning_rate": 6.175981464293779e-07,
+      "loss": 0.8449,
+      "step": 37816
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4273590831464464,
+      "learning_rate": 6.173341732749627e-07,
+      "loss": 0.9316,
+      "step": 37817
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6654318116622315,
+      "learning_rate": 6.170702547493956e-07,
+      "loss": 0.8408,
+      "step": 37818
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.768662900865135,
+      "learning_rate": 6.168063908542166e-07,
+      "loss": 0.8534,
+      "step": 37819
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4741441554339099,
+      "learning_rate": 6.16542581590962e-07,
+      "loss": 0.8492,
+      "step": 37820
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.837512120536664,
+      "learning_rate": 6.162788269611664e-07,
+      "loss": 0.7929,
+      "step": 37821
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8341105866504446,
+      "learning_rate": 6.160151269663684e-07,
+      "loss": 0.7611,
+      "step": 37822
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4913706610212547,
+      "learning_rate": 6.157514816080989e-07,
+      "loss": 0.7844,
+      "step": 37823
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.464051073417403,
+      "learning_rate": 6.15487890887897e-07,
+      "loss": 0.8396,
+      "step": 37824
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.749160640413357,
+      "learning_rate": 6.152243548072933e-07,
+      "loss": 0.8483,
+      "step": 37825
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.615351065021567,
+      "learning_rate": 6.149608733678269e-07,
+      "loss": 0.8178,
+      "step": 37826
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9227726205818365,
+      "learning_rate": 6.146974465710298e-07,
+      "loss": 0.8908,
+      "step": 37827
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6823712791651693,
+      "learning_rate": 6.144340744184341e-07,
+      "loss": 0.6831,
+      "step": 37828
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7413449842560607,
+      "learning_rate": 6.14170756911574e-07,
+      "loss": 0.8182,
+      "step": 37829
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5677724394460062,
+      "learning_rate": 6.139074940519851e-07,
+      "loss": 0.8684,
+      "step": 37830
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.765272722670033,
+      "learning_rate": 6.136442858411984e-07,
+      "loss": 1.0015,
+      "step": 37831
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7566251513506828,
+      "learning_rate": 6.133811322807437e-07,
+      "loss": 0.8555,
+      "step": 37832
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7150074736995597,
+      "learning_rate": 6.131180333721576e-07,
+      "loss": 0.7411,
+      "step": 37833
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6712564794248672,
+      "learning_rate": 6.128549891169711e-07,
+      "loss": 0.7982,
+      "step": 37834
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8979329092574078,
+      "learning_rate": 6.12591999516714e-07,
+      "loss": 0.9587,
+      "step": 37835
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.740994692580552,
+      "learning_rate": 6.123290645729174e-07,
+      "loss": 0.7343,
+      "step": 37836
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5618173961541455,
+      "learning_rate": 6.120661842871134e-07,
+      "loss": 0.8066,
+      "step": 37837
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8837342672519994,
+      "learning_rate": 6.11803358660833e-07,
+      "loss": 0.763,
+      "step": 37838
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.003189092461196,
+      "learning_rate": 6.11540587695606e-07,
+      "loss": 0.782,
+      "step": 37839
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4994747770402224,
+      "learning_rate": 6.112778713929601e-07,
+      "loss": 0.7234,
+      "step": 37840
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6786606445387664,
+      "learning_rate": 6.110152097544286e-07,
+      "loss": 0.7132,
+      "step": 37841
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5973752442921367,
+      "learning_rate": 6.10752602781538e-07,
+      "loss": 0.7543,
+      "step": 37842
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8983306715355033,
+      "learning_rate": 6.104900504758193e-07,
+      "loss": 0.9244,
+      "step": 37843
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9001167249511863,
+      "learning_rate": 6.10227552838799e-07,
+      "loss": 0.7539,
+      "step": 37844
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.490653096123779,
+      "learning_rate": 6.099651098720083e-07,
+      "loss": 0.8676,
+      "step": 37845
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7254149287988374,
+      "learning_rate": 6.097027215769725e-07,
+      "loss": 0.8197,
+      "step": 37846
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9144866817888995,
+      "learning_rate": 6.094403879552213e-07,
+      "loss": 0.8315,
+      "step": 37847
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5312665182092142,
+      "learning_rate": 6.091781090082804e-07,
+      "loss": 0.7624,
+      "step": 37848
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7065973099490912,
+      "learning_rate": 6.089158847376775e-07,
+      "loss": 0.6679,
+      "step": 37849
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6422836388108102,
+      "learning_rate": 6.08653715144939e-07,
+      "loss": 0.9273,
+      "step": 37850
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.06971726510454,
+      "learning_rate": 6.083916002315926e-07,
+      "loss": 0.8264,
+      "step": 37851
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4650493565348268,
+      "learning_rate": 6.081295399991649e-07,
+      "loss": 0.7955,
+      "step": 37852
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8926511618245265,
+      "learning_rate": 6.07867534449178e-07,
+      "loss": 0.8959,
+      "step": 37853
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6672391348389386,
+      "learning_rate": 6.076055835831596e-07,
+      "loss": 0.8462,
+      "step": 37854
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9042480102865909,
+      "learning_rate": 6.073436874026373e-07,
+      "loss": 0.8273,
+      "step": 37855
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6539333651990298,
+      "learning_rate": 6.070818459091333e-07,
+      "loss": 0.7603,
+      "step": 37856
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5192364426856317,
+      "learning_rate": 6.068200591041717e-07,
+      "loss": 0.8115,
+      "step": 37857
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6921731370606286,
+      "learning_rate": 6.065583269892761e-07,
+      "loss": 0.8459,
+      "step": 37858
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8022875253452384,
+      "learning_rate": 6.06296649565975e-07,
+      "loss": 0.7276,
+      "step": 37859
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.528189701411751,
+      "learning_rate": 6.060350268357873e-07,
+      "loss": 0.7853,
+      "step": 37860
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6808645811356024,
+      "learning_rate": 6.057734588002373e-07,
+      "loss": 0.8455,
+      "step": 37861
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.549997196987651,
+      "learning_rate": 6.055119454608493e-07,
+      "loss": 0.6613,
+      "step": 37862
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9111469618044972,
+      "learning_rate": 6.052504868191444e-07,
+      "loss": 0.8995,
+      "step": 37863
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.248449180957299,
+      "learning_rate": 6.049890828766458e-07,
+      "loss": 0.7644,
+      "step": 37864
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.871773298333265,
+      "learning_rate": 6.047277336348745e-07,
+      "loss": 0.983,
+      "step": 37865
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4998468358141999,
+      "learning_rate": 6.044664390953547e-07,
+      "loss": 0.8008,
+      "step": 37866
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4215061827120181,
+      "learning_rate": 6.042051992596032e-07,
+      "loss": 0.8078,
+      "step": 37867
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3829664603458567,
+      "learning_rate": 6.039440141291464e-07,
+      "loss": 0.6473,
+      "step": 37868
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7762642323182568,
+      "learning_rate": 6.036828837055008e-07,
+      "loss": 0.7031,
+      "step": 37869
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7320562227576832,
+      "learning_rate": 6.034218079901877e-07,
+      "loss": 0.7017,
+      "step": 37870
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7719710943692131,
+      "learning_rate": 6.031607869847278e-07,
+      "loss": 0.6899,
+      "step": 37871
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5435834334777647,
+      "learning_rate": 6.028998206906423e-07,
+      "loss": 0.7829,
+      "step": 37872
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5686155378682496,
+      "learning_rate": 6.026389091094487e-07,
+      "loss": 1.0581,
+      "step": 37873
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7901768973056921,
+      "learning_rate": 6.023780522426659e-07,
+      "loss": 0.6825,
+      "step": 37874
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4330898697944814,
+      "learning_rate": 6.021172500918127e-07,
+      "loss": 0.8568,
+      "step": 37875
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9164981774226293,
+      "learning_rate": 6.018565026584089e-07,
+      "loss": 0.6863,
+      "step": 37876
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8615788104770863,
+      "learning_rate": 6.015958099439734e-07,
+      "loss": 0.8564,
+      "step": 37877
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5039337570597802,
+      "learning_rate": 6.013351719500205e-07,
+      "loss": 0.817,
+      "step": 37878
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7227065176580847,
+      "learning_rate": 6.010745886780689e-07,
+      "loss": 0.8604,
+      "step": 37879
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.791986668391994,
+      "learning_rate": 6.008140601296398e-07,
+      "loss": 0.7477,
+      "step": 37880
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5202339318241014,
+      "learning_rate": 6.005535863062461e-07,
+      "loss": 0.8741,
+      "step": 37881
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7908809252024454,
+      "learning_rate": 6.002931672094037e-07,
+      "loss": 0.9292,
+      "step": 37882
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5366561648212522,
+      "learning_rate": 6.000328028406322e-07,
+      "loss": 0.8102,
+      "step": 37883
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.189460326749329,
+      "learning_rate": 5.997724932014437e-07,
+      "loss": 0.8833,
+      "step": 37884
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3825427938086787,
+      "learning_rate": 5.995122382933571e-07,
+      "loss": 0.6707,
+      "step": 37885
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8067701407351893,
+      "learning_rate": 5.992520381178857e-07,
+      "loss": 0.8945,
+      "step": 37886
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.515326012372003,
+      "learning_rate": 5.98991892676547e-07,
+      "loss": 0.8741,
+      "step": 37887
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5459039632134595,
+      "learning_rate": 5.98731801970851e-07,
+      "loss": 0.7744,
+      "step": 37888
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5876444104839706,
+      "learning_rate": 5.984717660023176e-07,
+      "loss": 0.7947,
+      "step": 37889
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6157622826035152,
+      "learning_rate": 5.982117847724567e-07,
+      "loss": 0.7033,
+      "step": 37890
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8722615300898946,
+      "learning_rate": 5.979518582827826e-07,
+      "loss": 1.0109,
+      "step": 37891
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5661422918898913,
+      "learning_rate": 5.976919865348097e-07,
+      "loss": 0.694,
+      "step": 37892
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7025193085109915,
+      "learning_rate": 5.974321695300523e-07,
+      "loss": 0.7019,
+      "step": 37893
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.681691279635103,
+      "learning_rate": 5.971724072700213e-07,
+      "loss": 0.912,
+      "step": 37894
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6410067086927032,
+      "learning_rate": 5.96912699756228e-07,
+      "loss": 0.9114,
+      "step": 37895
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5907695174633176,
+      "learning_rate": 5.966530469901854e-07,
+      "loss": 0.846,
+      "step": 37896
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7336059783883113,
+      "learning_rate": 5.96393448973408e-07,
+      "loss": 0.7041,
+      "step": 37897
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6976354574204775,
+      "learning_rate": 5.961339057074045e-07,
+      "loss": 0.7477,
+      "step": 37898
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4153339169368686,
+      "learning_rate": 5.958744171936859e-07,
+      "loss": 0.8227,
+      "step": 37899
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4903099394310042,
+      "learning_rate": 5.956149834337632e-07,
+      "loss": 0.7719,
+      "step": 37900
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3878883507370332,
+      "learning_rate": 5.953556044291487e-07,
+      "loss": 0.7128,
+      "step": 37901
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.787243566692772,
+      "learning_rate": 5.950962801813521e-07,
+      "loss": 0.7598,
+      "step": 37902
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9490223931826212,
+      "learning_rate": 5.948370106918799e-07,
+      "loss": 0.7735,
+      "step": 37903
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.574580813384026,
+      "learning_rate": 5.945777959622467e-07,
+      "loss": 0.8499,
+      "step": 37904
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.73529001162753,
+      "learning_rate": 5.943186359939579e-07,
+      "loss": 0.793,
+      "step": 37905
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7276748541503143,
+      "learning_rate": 5.940595307885233e-07,
+      "loss": 0.7111,
+      "step": 37906
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7742150441708311,
+      "learning_rate": 5.93800480347454e-07,
+      "loss": 0.7149,
+      "step": 37907
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8568211875545162,
+      "learning_rate": 5.935414846722554e-07,
+      "loss": 0.8911,
+      "step": 37908
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.712856803380086,
+      "learning_rate": 5.932825437644352e-07,
+      "loss": 0.8457,
+      "step": 37909
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.538285150910181,
+      "learning_rate": 5.930236576255022e-07,
+      "loss": 0.9046,
+      "step": 37910
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8931069287343252,
+      "learning_rate": 5.927648262569663e-07,
+      "loss": 0.7238,
+      "step": 37911
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6610396838167116,
+      "learning_rate": 5.925060496603297e-07,
+      "loss": 0.8631,
+      "step": 37912
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6284189004945449,
+      "learning_rate": 5.92247327837101e-07,
+      "loss": 0.7962,
+      "step": 37913
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5887152370225253,
+      "learning_rate": 5.919886607887859e-07,
+      "loss": 0.7891,
+      "step": 37914
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5215864939171564,
+      "learning_rate": 5.917300485168953e-07,
+      "loss": 0.7362,
+      "step": 37915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7780535800570525,
+      "learning_rate": 5.91471491022928e-07,
+      "loss": 0.7904,
+      "step": 37916
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7502255117601496,
+      "learning_rate": 5.912129883083917e-07,
+      "loss": 0.7348,
+      "step": 37917
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7257862843192737,
+      "learning_rate": 5.909545403747941e-07,
+      "loss": 0.8495,
+      "step": 37918
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5762091022403226,
+      "learning_rate": 5.906961472236361e-07,
+      "loss": 0.9089,
+      "step": 37919
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5101338695188347,
+      "learning_rate": 5.904378088564255e-07,
+      "loss": 0.7647,
+      "step": 37920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.617867735426475,
+      "learning_rate": 5.901795252746644e-07,
+      "loss": 0.8372,
+      "step": 37921
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.404546742487272,
+      "learning_rate": 5.899212964798584e-07,
+      "loss": 0.9239,
+      "step": 37922
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6326375090676815,
+      "learning_rate": 5.896631224735072e-07,
+      "loss": 0.9035,
+      "step": 37923
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5244210738872357,
+      "learning_rate": 5.894050032571197e-07,
+      "loss": 0.9032,
+      "step": 37924
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6699496549950323,
+      "learning_rate": 5.891469388321936e-07,
+      "loss": 0.7749,
+      "step": 37925
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7938462547300613,
+      "learning_rate": 5.888889292002342e-07,
+      "loss": 0.8575,
+      "step": 37926
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9021037846639257,
+      "learning_rate": 5.886309743627416e-07,
+      "loss": 0.9978,
+      "step": 37927
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6259978161244701,
+      "learning_rate": 5.883730743212213e-07,
+      "loss": 0.9029,
+      "step": 37928
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.642023323796328,
+      "learning_rate": 5.881152290771719e-07,
+      "loss": 0.8198,
+      "step": 37929
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5173046013502862,
+      "learning_rate": 5.878574386320945e-07,
+      "loss": 0.853,
+      "step": 37930
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8867597263908578,
+      "learning_rate": 5.875997029874903e-07,
+      "loss": 0.8242,
+      "step": 37931
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6280613392901688,
+      "learning_rate": 5.873420221448623e-07,
+      "loss": 0.8176,
+      "step": 37932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5501805574107332,
+      "learning_rate": 5.870843961057093e-07,
+      "loss": 0.6919,
+      "step": 37933
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4267219411089371,
+      "learning_rate": 5.868268248715292e-07,
+      "loss": 0.8485,
+      "step": 37934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5113860849025684,
+      "learning_rate": 5.86569308443824e-07,
+      "loss": 0.8781,
+      "step": 37935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9410922659280774,
+      "learning_rate": 5.863118468240936e-07,
+      "loss": 0.8457,
+      "step": 37936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.547831480154693,
+      "learning_rate": 5.860544400138357e-07,
+      "loss": 0.8569,
+      "step": 37937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.671526558769813,
+      "learning_rate": 5.857970880145492e-07,
+      "loss": 0.8512,
+      "step": 37938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8012875969249182,
+      "learning_rate": 5.855397908277305e-07,
+      "loss": 0.8294,
+      "step": 37939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7460660232525642,
+      "learning_rate": 5.85282548454883e-07,
+      "loss": 0.9129,
+      "step": 37940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5249590867932379,
+      "learning_rate": 5.850253608974998e-07,
+      "loss": 0.6234,
+      "step": 37941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4726961662706701,
+      "learning_rate": 5.847682281570788e-07,
+      "loss": 0.8038,
+      "step": 37942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6936149324579732,
+      "learning_rate": 5.845111502351197e-07,
+      "loss": 0.7993,
+      "step": 37943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4720856330196004,
+      "learning_rate": 5.842541271331148e-07,
+      "loss": 0.8146,
+      "step": 37944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9456311015198904,
+      "learning_rate": 5.839971588525661e-07,
+      "loss": 0.929,
+      "step": 37945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7710092034530436,
+      "learning_rate": 5.837402453949648e-07,
+      "loss": 0.8567,
+      "step": 37946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.559791241117992,
+      "learning_rate": 5.834833867618117e-07,
+      "loss": 0.961,
+      "step": 37947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7399367957113445,
+      "learning_rate": 5.832265829545969e-07,
+      "loss": 0.7158,
+      "step": 37948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7654851538571726,
+      "learning_rate": 5.829698339748202e-07,
+      "loss": 0.6825,
+      "step": 37949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5734164967330448,
+      "learning_rate": 5.827131398239738e-07,
+      "loss": 0.8572,
+      "step": 37950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4957063566485802,
+      "learning_rate": 5.824565005035521e-07,
+      "loss": 0.8714,
+      "step": 37951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7519599885596348,
+      "learning_rate": 5.821999160150504e-07,
+      "loss": 0.6542,
+      "step": 37952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6890048256727037,
+      "learning_rate": 5.819433863599633e-07,
+      "loss": 0.8251,
+      "step": 37953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7388357626130742,
+      "learning_rate": 5.816869115397838e-07,
+      "loss": 0.6731,
+      "step": 37954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4668267739644953,
+      "learning_rate": 5.814304915560031e-07,
+      "loss": 0.7597,
+      "step": 37955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6110023120910342,
+      "learning_rate": 5.811741264101167e-07,
+      "loss": 0.7909,
+      "step": 37956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7271595247735774,
+      "learning_rate": 5.809178161036166e-07,
+      "loss": 0.795,
+      "step": 37957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5373189786948391,
+      "learning_rate": 5.806615606379961e-07,
+      "loss": 0.9258,
+      "step": 37958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7421250573904234,
+      "learning_rate": 5.804053600147452e-07,
+      "loss": 0.6872,
+      "step": 37959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.65070842454935,
+      "learning_rate": 5.801492142353549e-07,
+      "loss": 0.8733,
+      "step": 37960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6219883655265883,
+      "learning_rate": 5.798931233013205e-07,
+      "loss": 0.9384,
+      "step": 37961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.355907899709926,
+      "learning_rate": 5.796370872141311e-07,
+      "loss": 0.6891,
+      "step": 37962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6257468794503365,
+      "learning_rate": 5.793811059752752e-07,
+      "loss": 0.8217,
+      "step": 37963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6814915969716238,
+      "learning_rate": 5.791251795862463e-07,
+      "loss": 0.8264,
+      "step": 37964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5772823017659,
+      "learning_rate": 5.788693080485319e-07,
+      "loss": 0.7879,
+      "step": 37965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4935003784689105,
+      "learning_rate": 5.786134913636243e-07,
+      "loss": 0.7245,
+      "step": 37966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.472898468040509,
+      "learning_rate": 5.7835772953301e-07,
+      "loss": 0.8029,
+      "step": 37967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.779764709515461,
+      "learning_rate": 5.781020225581813e-07,
+      "loss": 0.8594,
+      "step": 37968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9211669799858593,
+      "learning_rate": 5.778463704406245e-07,
+      "loss": 0.8071,
+      "step": 37969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5719180194517652,
+      "learning_rate": 5.775907731818308e-07,
+      "loss": 0.9025,
+      "step": 37970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7770891907390828,
+      "learning_rate": 5.773352307832869e-07,
+      "loss": 0.6562,
+      "step": 37971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5791456705743945,
+      "learning_rate": 5.770797432464781e-07,
+      "loss": 0.8183,
+      "step": 37972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.339895753473523,
+      "learning_rate": 5.768243105728954e-07,
+      "loss": 0.8425,
+      "step": 37973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.973999199077869,
+      "learning_rate": 5.765689327640256e-07,
+      "loss": 0.8008,
+      "step": 37974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4927906621595626,
+      "learning_rate": 5.763136098213551e-07,
+      "loss": 0.8421,
+      "step": 37975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.64957994471899,
+      "learning_rate": 5.760583417463683e-07,
+      "loss": 0.8996,
+      "step": 37976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6198108186102562,
+      "learning_rate": 5.758031285405541e-07,
+      "loss": 0.8535,
+      "step": 37977
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5299276550917849,
+      "learning_rate": 5.755479702053979e-07,
+      "loss": 0.8218,
+      "step": 37978
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5025786631236093,
+      "learning_rate": 5.752928667423862e-07,
+      "loss": 0.7563,
+      "step": 37979
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4696905262385098,
+      "learning_rate": 5.750378181530025e-07,
+      "loss": 0.8578,
+      "step": 37980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7634727279882234,
+      "learning_rate": 5.74782824438731e-07,
+      "loss": 0.8972,
+      "step": 37981
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5837491572396123,
+      "learning_rate": 5.745278856010605e-07,
+      "loss": 0.8246,
+      "step": 37982
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.840866072045499,
+      "learning_rate": 5.742730016414722e-07,
+      "loss": 0.7718,
+      "step": 37983
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.580866837290975,
+      "learning_rate": 5.740181725614491e-07,
+      "loss": 0.7201,
+      "step": 37984
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5454486596175785,
+      "learning_rate": 5.737633983624779e-07,
+      "loss": 0.7204,
+      "step": 37985
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6313786629123883,
+      "learning_rate": 5.735086790460387e-07,
+      "loss": 0.7899,
+      "step": 37986
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5236267596412667,
+      "learning_rate": 5.73254014613619e-07,
+      "loss": 0.6675,
+      "step": 37987
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7508992749002352,
+      "learning_rate": 5.729994050666965e-07,
+      "loss": 0.6846,
+      "step": 37988
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6139893391494664,
+      "learning_rate": 5.727448504067578e-07,
+      "loss": 0.8297,
+      "step": 37989
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6054183603344296,
+      "learning_rate": 5.724903506352808e-07,
+      "loss": 1.003,
+      "step": 37990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7386465336441537,
+      "learning_rate": 5.722359057537519e-07,
+      "loss": 0.8949,
+      "step": 37991
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7364244188016434,
+      "learning_rate": 5.71981515763651e-07,
+      "loss": 0.7433,
+      "step": 37992
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6475244465618124,
+      "learning_rate": 5.71727180666456e-07,
+      "loss": 0.9197,
+      "step": 37993
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5105318501626694,
+      "learning_rate": 5.71472900463651e-07,
+      "loss": 0.7274,
+      "step": 37994
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6939483395404504,
+      "learning_rate": 5.71218675156715e-07,
+      "loss": 0.7083,
+      "step": 37995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5281835271253477,
+      "learning_rate": 5.709645047471324e-07,
+      "loss": 0.8148,
+      "step": 37996
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6905773041469914,
+      "learning_rate": 5.707103892363774e-07,
+      "loss": 0.7515,
+      "step": 37997
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9940446349844652,
+      "learning_rate": 5.704563286259313e-07,
+      "loss": 0.8581,
+      "step": 37998
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5874534098451387,
+      "learning_rate": 5.702023229172749e-07,
+      "loss": 0.6715,
+      "step": 37999
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5358533720088732,
+      "learning_rate": 5.699483721118859e-07,
+      "loss": 0.8646,
+      "step": 38000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5704483753639833,
+      "learning_rate": 5.696944762112422e-07,
+      "loss": 0.8046,
+      "step": 38001
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6929770966472293,
+      "learning_rate": 5.694406352168214e-07,
+      "loss": 0.8609,
+      "step": 38002
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.603121023065403,
+      "learning_rate": 5.691868491301056e-07,
+      "loss": 0.7689,
+      "step": 38003
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6642148857990278,
+      "learning_rate": 5.689331179525681e-07,
+      "loss": 0.9371,
+      "step": 38004
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5703167056247098,
+      "learning_rate": 5.686794416856889e-07,
+      "loss": 0.7537,
+      "step": 38005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7069362005869357,
+      "learning_rate": 5.684258203309446e-07,
+      "loss": 0.9645,
+      "step": 38006
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8770350670536302,
+      "learning_rate": 5.681722538898093e-07,
+      "loss": 0.7664,
+      "step": 38007
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6663873660831572,
+      "learning_rate": 5.679187423637611e-07,
+      "loss": 0.9072,
+      "step": 38008
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5765486041215488,
+      "learning_rate": 5.676652857542775e-07,
+      "loss": 0.7547,
+      "step": 38009
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7207213158576411,
+      "learning_rate": 5.674118840628329e-07,
+      "loss": 0.6279,
+      "step": 38010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5080546388841312,
+      "learning_rate": 5.671585372909016e-07,
+      "loss": 0.8454,
+      "step": 38011
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6774066835849517,
+      "learning_rate": 5.669052454399593e-07,
+      "loss": 0.8128,
+      "step": 38012
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5390608229755085,
+      "learning_rate": 5.666520085114846e-07,
+      "loss": 0.8323,
+      "step": 38013
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.226975154962209,
+      "learning_rate": 5.663988265069454e-07,
+      "loss": 0.8055,
+      "step": 38014
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8046008960016788,
+      "learning_rate": 5.661456994278181e-07,
+      "loss": 0.7153,
+      "step": 38015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4891410657195114,
+      "learning_rate": 5.658926272755772e-07,
+      "loss": 0.8931,
+      "step": 38016
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6377936416443994,
+      "learning_rate": 5.656396100516981e-07,
+      "loss": 0.8888,
+      "step": 38017
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.619517087131213,
+      "learning_rate": 5.65386647757652e-07,
+      "loss": 0.9181,
+      "step": 38018
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7696691428622007,
+      "learning_rate": 5.651337403949109e-07,
+      "loss": 0.6816,
+      "step": 38019
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.963254656117752,
+      "learning_rate": 5.648808879649493e-07,
+      "loss": 0.7699,
+      "step": 38020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5250654384148896,
+      "learning_rate": 5.64628090469237e-07,
+      "loss": 0.7673,
+      "step": 38021
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4874400006778152,
+      "learning_rate": 5.643753479092473e-07,
+      "loss": 0.7956,
+      "step": 38022
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0681343211422547,
+      "learning_rate": 5.641226602864514e-07,
+      "loss": 0.6941,
+      "step": 38023
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5936614691215742,
+      "learning_rate": 5.638700276023223e-07,
+      "loss": 0.9993,
+      "step": 38024
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.562169950279378,
+      "learning_rate": 5.636174498583269e-07,
+      "loss": 0.7713,
+      "step": 38025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6570347893984745,
+      "learning_rate": 5.633649270559405e-07,
+      "loss": 0.8566,
+      "step": 38026
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4595585498584054,
+      "learning_rate": 5.631124591966308e-07,
+      "loss": 0.7784,
+      "step": 38027
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7988894007241849,
+      "learning_rate": 5.628600462818668e-07,
+      "loss": 0.7207,
+      "step": 38028
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5242633159880734,
+      "learning_rate": 5.626076883131193e-07,
+      "loss": 0.7193,
+      "step": 38029
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6104537160184436,
+      "learning_rate": 5.623553852918584e-07,
+      "loss": 0.6698,
+      "step": 38030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.700909403473865,
+      "learning_rate": 5.621031372195529e-07,
+      "loss": 0.9711,
+      "step": 38031
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5183858300549653,
+      "learning_rate": 5.618509440976694e-07,
+      "loss": 0.8746,
+      "step": 38032
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6997826729123295,
+      "learning_rate": 5.615988059276777e-07,
+      "loss": 0.8931,
+      "step": 38033
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8036721448826445,
+      "learning_rate": 5.613467227110481e-07,
+      "loss": 0.7996,
+      "step": 38034
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5022870316646788,
+      "learning_rate": 5.610946944492457e-07,
+      "loss": 0.6959,
+      "step": 38035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.651042296691089,
+      "learning_rate": 5.608427211437362e-07,
+      "loss": 0.7789,
+      "step": 38036
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6670901004427547,
+      "learning_rate": 5.605908027959894e-07,
+      "loss": 0.749,
+      "step": 38037
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4447155702835328,
+      "learning_rate": 5.603389394074732e-07,
+      "loss": 0.7412,
+      "step": 38038
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8052233483543119,
+      "learning_rate": 5.600871309796519e-07,
+      "loss": 0.805,
+      "step": 38039
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7624003341677035,
+      "learning_rate": 5.598353775139898e-07,
+      "loss": 0.8548,
+      "step": 38040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4297373736016086,
+      "learning_rate": 5.595836790119569e-07,
+      "loss": 0.8314,
+      "step": 38041
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.590811305001554,
+      "learning_rate": 5.593320354750154e-07,
+      "loss": 0.8223,
+      "step": 38042
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5768135606948614,
+      "learning_rate": 5.59080446904633e-07,
+      "loss": 0.8251,
+      "step": 38043
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5485505788767264,
+      "learning_rate": 5.588289133022706e-07,
+      "loss": 0.8344,
+      "step": 38044
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.992640251920453,
+      "learning_rate": 5.585774346693984e-07,
+      "loss": 0.8199,
+      "step": 38045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7447442112531604,
+      "learning_rate": 5.583260110074751e-07,
+      "loss": 0.6626,
+      "step": 38046
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7362003328836166,
+      "learning_rate": 5.580746423179684e-07,
+      "loss": 0.7785,
+      "step": 38047
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8662439331980913,
+      "learning_rate": 5.578233286023393e-07,
+      "loss": 0.7815,
+      "step": 38048
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5777300235479792,
+      "learning_rate": 5.575720698620535e-07,
+      "loss": 0.713,
+      "step": 38049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6066889698753564,
+      "learning_rate": 5.573208660985707e-07,
+      "loss": 0.7337,
+      "step": 38050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4362225875539691,
+      "learning_rate": 5.570697173133565e-07,
+      "loss": 0.6946,
+      "step": 38051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5469442539670293,
+      "learning_rate": 5.568186235078732e-07,
+      "loss": 0.9065,
+      "step": 38052
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8435540206565766,
+      "learning_rate": 5.565675846835794e-07,
+      "loss": 0.6711,
+      "step": 38053
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5947604434568843,
+      "learning_rate": 5.563166008419385e-07,
+      "loss": 0.8284,
+      "step": 38054
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0813671581177053,
+      "learning_rate": 5.560656719844149e-07,
+      "loss": 0.8285,
+      "step": 38055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7356481811317055,
+      "learning_rate": 5.55814798112465e-07,
+      "loss": 0.8516,
+      "step": 38056
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6063529079682926,
+      "learning_rate": 5.555639792275514e-07,
+      "loss": 0.7275,
+      "step": 38057
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5496774034836858,
+      "learning_rate": 5.553132153311335e-07,
+      "loss": 0.7691,
+      "step": 38058
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8321488456180903,
+      "learning_rate": 5.55062506424674e-07,
+      "loss": 0.7836,
+      "step": 38059
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5350058079177473,
+      "learning_rate": 5.548118525096313e-07,
+      "loss": 0.8691,
+      "step": 38060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7269617379737279,
+      "learning_rate": 5.545612535874622e-07,
+      "loss": 0.8384,
+      "step": 38061
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1158027889061644,
+      "learning_rate": 5.543107096596279e-07,
+      "loss": 0.8239,
+      "step": 38062
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0417343611094907,
+      "learning_rate": 5.54060220727588e-07,
+      "loss": 0.9333,
+      "step": 38063
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5012038389773354,
+      "learning_rate": 5.538097867928016e-07,
+      "loss": 0.8909,
+      "step": 38064
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8081492624869206,
+      "learning_rate": 5.53559407856722e-07,
+      "loss": 0.8502,
+      "step": 38065
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8122903431644,
+      "learning_rate": 5.533090839208133e-07,
+      "loss": 0.7366,
+      "step": 38066
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.69633612650846,
+      "learning_rate": 5.530588149865268e-07,
+      "loss": 0.8996,
+      "step": 38067
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.653566690132535,
+      "learning_rate": 5.528086010553257e-07,
+      "loss": 0.7829,
+      "step": 38068
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6739745568266626,
+      "learning_rate": 5.525584421286612e-07,
+      "loss": 0.6715,
+      "step": 38069
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7804395947442992,
+      "learning_rate": 5.523083382079952e-07,
+      "loss": 0.7517,
+      "step": 38070
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3542887225927238,
+      "learning_rate": 5.520582892947779e-07,
+      "loss": 0.6769,
+      "step": 38071
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5050837668761265,
+      "learning_rate": 5.518082953904714e-07,
+      "loss": 0.9025,
+      "step": 38072
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6230209978783017,
+      "learning_rate": 5.515583564965277e-07,
+      "loss": 0.8197,
+      "step": 38073
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.583741479337011,
+      "learning_rate": 5.513084726144013e-07,
+      "loss": 0.8907,
+      "step": 38074
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1634608728802602,
+      "learning_rate": 5.510586437455478e-07,
+      "loss": 0.8218,
+      "step": 38075
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4630371941133353,
+      "learning_rate": 5.508088698914249e-07,
+      "loss": 0.8087,
+      "step": 38076
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5805317159844634,
+      "learning_rate": 5.505591510534836e-07,
+      "loss": 0.9129,
+      "step": 38077
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6688650355813086,
+      "learning_rate": 5.503094872331771e-07,
+      "loss": 0.7968,
+      "step": 38078
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.576423623492915,
+      "learning_rate": 5.5005987843196e-07,
+      "loss": 0.8237,
+      "step": 38079
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6761468431257582,
+      "learning_rate": 5.498103246512887e-07,
+      "loss": 0.8578,
+      "step": 38080
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5458398617423532,
+      "learning_rate": 5.495608258926133e-07,
+      "loss": 0.9184,
+      "step": 38081
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2866532271607984,
+      "learning_rate": 5.493113821573847e-07,
+      "loss": 0.7891,
+      "step": 38082
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0545542078066994,
+      "learning_rate": 5.490619934470576e-07,
+      "loss": 0.8729,
+      "step": 38083
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.602821712310915,
+      "learning_rate": 5.48812659763085e-07,
+      "loss": 0.879,
+      "step": 38084
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9789379070866702,
+      "learning_rate": 5.485633811069169e-07,
+      "loss": 0.7065,
+      "step": 38085
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8102085457500703,
+      "learning_rate": 5.483141574800044e-07,
+      "loss": 0.6541,
+      "step": 38086
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2984517068021892,
+      "learning_rate": 5.480649888838008e-07,
+      "loss": 0.9231,
+      "step": 38087
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5462382982450729,
+      "learning_rate": 5.478158753197527e-07,
+      "loss": 0.8958,
+      "step": 38088
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.499840319058618,
+      "learning_rate": 5.475668167893156e-07,
+      "loss": 0.7816,
+      "step": 38089
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7725502719710602,
+      "learning_rate": 5.47317813293935e-07,
+      "loss": 0.9577,
+      "step": 38090
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7466650828962995,
+      "learning_rate": 5.470688648350642e-07,
+      "loss": 0.7543,
+      "step": 38091
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8195689840297182,
+      "learning_rate": 5.468199714141509e-07,
+      "loss": 0.9371,
+      "step": 38092
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6286638498189756,
+      "learning_rate": 5.465711330326428e-07,
+      "loss": 0.8446,
+      "step": 38093
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7139877736939833,
+      "learning_rate": 5.463223496919945e-07,
+      "loss": 0.9085,
+      "step": 38094
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6369257438599538,
+      "learning_rate": 5.460736213936479e-07,
+      "loss": 0.8656,
+      "step": 38095
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7356999046242201,
+      "learning_rate": 5.458249481390532e-07,
+      "loss": 0.6901,
+      "step": 38096
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5333202579653555,
+      "learning_rate": 5.45576329929658e-07,
+      "loss": 0.8497,
+      "step": 38097
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7476697697721596,
+      "learning_rate": 5.453277667669143e-07,
+      "loss": 0.6959,
+      "step": 38098
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6986460029107566,
+      "learning_rate": 5.450792586522624e-07,
+      "loss": 0.7372,
+      "step": 38099
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5959630933561464,
+      "learning_rate": 5.448308055871532e-07,
+      "loss": 0.765,
+      "step": 38100
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6353473650327344,
+      "learning_rate": 5.445824075730333e-07,
+      "loss": 0.999,
+      "step": 38101
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7576073128421706,
+      "learning_rate": 5.443340646113471e-07,
+      "loss": 0.7183,
+      "step": 38102
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6471471476533706,
+      "learning_rate": 5.440857767035434e-07,
+      "loss": 0.8114,
+      "step": 38103
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4515791075089943,
+      "learning_rate": 5.438375438510635e-07,
+      "loss": 0.8094,
+      "step": 38104
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6601263884355126,
+      "learning_rate": 5.43589366055357e-07,
+      "loss": 0.7552,
+      "step": 38105
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7105229129373873,
+      "learning_rate": 5.433412433178664e-07,
+      "loss": 0.9113,
+      "step": 38106
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4678632661486584,
+      "learning_rate": 5.43093175640037e-07,
+      "loss": 0.825,
+      "step": 38107
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7274021838215712,
+      "learning_rate": 5.428451630233144e-07,
+      "loss": 0.8271,
+      "step": 38108
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8356571041549126,
+      "learning_rate": 5.425972054691397e-07,
+      "loss": 0.8919,
+      "step": 38109
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7465026014973395,
+      "learning_rate": 5.423493029789573e-07,
+      "loss": 0.8968,
+      "step": 38110
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.78010854955357,
+      "learning_rate": 5.421014555542137e-07,
+      "loss": 0.8515,
+      "step": 38111
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7440335835119526,
+      "learning_rate": 5.418536631963501e-07,
+      "loss": 0.6449,
+      "step": 38112
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5331048251989894,
+      "learning_rate": 5.416059259068063e-07,
+      "loss": 0.871,
+      "step": 38113
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6656167357367895,
+      "learning_rate": 5.41358243687028e-07,
+      "loss": 0.869,
+      "step": 38114
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4831770320805975,
+      "learning_rate": 5.411106165384583e-07,
+      "loss": 0.8207,
+      "step": 38115
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6554719613876943,
+      "learning_rate": 5.408630444625373e-07,
+      "loss": 0.828,
+      "step": 38116
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5698392975959545,
+      "learning_rate": 5.406155274607039e-07,
+      "loss": 0.8742,
+      "step": 38117
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7026793912436249,
+      "learning_rate": 5.403680655344023e-07,
+      "loss": 0.7922,
+      "step": 38118
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5767812621454862,
+      "learning_rate": 5.40120658685075e-07,
+      "loss": 0.9699,
+      "step": 38119
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7691051908413056,
+      "learning_rate": 5.398733069141593e-07,
+      "loss": 0.8216,
+      "step": 38120
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7554071734253724,
+      "learning_rate": 5.396260102230944e-07,
+      "loss": 0.7191,
+      "step": 38121
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.746642010175921,
+      "learning_rate": 5.393787686133234e-07,
+      "loss": 0.6853,
+      "step": 38122
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7033778514067974,
+      "learning_rate": 5.391315820862841e-07,
+      "loss": 0.9449,
+      "step": 38123
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.502651830274816,
+      "learning_rate": 5.388844506434166e-07,
+      "loss": 0.7559,
+      "step": 38124
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6698039818615344,
+      "learning_rate": 5.386373742861572e-07,
+      "loss": 0.8593,
+      "step": 38125
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5500278487440309,
+      "learning_rate": 5.383903530159484e-07,
+      "loss": 0.6648,
+      "step": 38126
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5714203189666398,
+      "learning_rate": 5.381433868342256e-07,
+      "loss": 0.7813,
+      "step": 38127
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7496874104376803,
+      "learning_rate": 5.378964757424287e-07,
+      "loss": 0.684,
+      "step": 38128
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6431412150363043,
+      "learning_rate": 5.376496197419934e-07,
+      "loss": 0.9852,
+      "step": 38129
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6131591367847842,
+      "learning_rate": 5.374028188343572e-07,
+      "loss": 0.8906,
+      "step": 38130
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7008608432849253,
+      "learning_rate": 5.371560730209579e-07,
+      "loss": 0.8042,
+      "step": 38131
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.671279467241929,
+      "learning_rate": 5.369093823032323e-07,
+      "loss": 0.7534,
+      "step": 38132
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6027383811672773,
+      "learning_rate": 5.366627466826169e-07,
+      "loss": 0.7193,
+      "step": 38133
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.608605536880982,
+      "learning_rate": 5.364161661605449e-07,
+      "loss": 0.9021,
+      "step": 38134
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7085429479797951,
+      "learning_rate": 5.361696407384543e-07,
+      "loss": 0.9491,
+      "step": 38135
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7577786626307437,
+      "learning_rate": 5.359231704177825e-07,
+      "loss": 0.7919,
+      "step": 38136
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.258683312339643,
+      "learning_rate": 5.356767551999608e-07,
+      "loss": 0.7571,
+      "step": 38137
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6328200980059877,
+      "learning_rate": 5.354303950864248e-07,
+      "loss": 0.878,
+      "step": 38138
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4500603908292395,
+      "learning_rate": 5.351840900786098e-07,
+      "loss": 0.7528,
+      "step": 38139
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9898833664394762,
+      "learning_rate": 5.349378401779504e-07,
+      "loss": 0.7487,
+      "step": 38140
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5958106605034486,
+      "learning_rate": 5.346916453858797e-07,
+      "loss": 0.8565,
+      "step": 38141
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5721450423639336,
+      "learning_rate": 5.3444550570383e-07,
+      "loss": 0.7084,
+      "step": 38142
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5476305384879545,
+      "learning_rate": 5.341994211332358e-07,
+      "loss": 0.9776,
+      "step": 38143
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7296817936801149,
+      "learning_rate": 5.339533916755291e-07,
+      "loss": 0.6852,
+      "step": 38144
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.596821188507741,
+      "learning_rate": 5.337074173321433e-07,
+      "loss": 0.8247,
+      "step": 38145
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8053259109924449,
+      "learning_rate": 5.334614981045083e-07,
+      "loss": 0.6821,
+      "step": 38146
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4669835941697924,
+      "learning_rate": 5.332156339940598e-07,
+      "loss": 0.7838,
+      "step": 38147
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.610918530384194,
+      "learning_rate": 5.329698250022241e-07,
+      "loss": 0.8718,
+      "step": 38148
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6581258136566728,
+      "learning_rate": 5.327240711304383e-07,
+      "loss": 0.809,
+      "step": 38149
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5358402918865108,
+      "learning_rate": 5.324783723801296e-07,
+      "loss": 0.9095,
+      "step": 38150
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3831144497566417,
+      "learning_rate": 5.322327287527274e-07,
+      "loss": 0.9055,
+      "step": 38151
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5271353004669659,
+      "learning_rate": 5.319871402496645e-07,
+      "loss": 0.8358,
+      "step": 38152
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.142906606608793,
+      "learning_rate": 5.317416068723702e-07,
+      "loss": 0.855,
+      "step": 38153
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4362610213444507,
+      "learning_rate": 5.314961286222752e-07,
+      "loss": 0.8053,
+      "step": 38154
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.855076643740522,
+      "learning_rate": 5.312507055008054e-07,
+      "loss": 0.8304,
+      "step": 38155
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7261063645254802,
+      "learning_rate": 5.310053375093915e-07,
+      "loss": 0.7935,
+      "step": 38156
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6421857224039322,
+      "learning_rate": 5.307600246494638e-07,
+      "loss": 0.8734,
+      "step": 38157
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4245067702730854,
+      "learning_rate": 5.305147669224497e-07,
+      "loss": 1.0079,
+      "step": 38158
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5421493033978508,
+      "learning_rate": 5.302695643297751e-07,
+      "loss": 0.8705,
+      "step": 38159
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8383777038980633,
+      "learning_rate": 5.300244168728686e-07,
+      "loss": 0.7946,
+      "step": 38160
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6096375631553663,
+      "learning_rate": 5.297793245531602e-07,
+      "loss": 0.865,
+      "step": 38161
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.577330149857462,
+      "learning_rate": 5.295342873720743e-07,
+      "loss": 0.7905,
+      "step": 38162
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7742332044913507,
+      "learning_rate": 5.292893053310366e-07,
+      "loss": 0.7616,
+      "step": 38163
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8454089172293464,
+      "learning_rate": 5.29044378431477e-07,
+      "loss": 0.8478,
+      "step": 38164
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7644652772697782,
+      "learning_rate": 5.287995066748175e-07,
+      "loss": 0.7796,
+      "step": 38165
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6554671063117958,
+      "learning_rate": 5.285546900624872e-07,
+      "loss": 0.7215,
+      "step": 38166
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5188429933653116,
+      "learning_rate": 5.283099285959081e-07,
+      "loss": 0.9453,
+      "step": 38167
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.556747975321337,
+      "learning_rate": 5.280652222765103e-07,
+      "loss": 0.7796,
+      "step": 38168
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.96197462940789,
+      "learning_rate": 5.278205711057127e-07,
+      "loss": 0.846,
+      "step": 38169
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.485800320643869,
+      "learning_rate": 5.275759750849441e-07,
+      "loss": 0.7719,
+      "step": 38170
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9094607356258617,
+      "learning_rate": 5.273314342156266e-07,
+      "loss": 0.8756,
+      "step": 38171
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2885739482082275,
+      "learning_rate": 5.270869484991858e-07,
+      "loss": 0.8777,
+      "step": 38172
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8995754059264895,
+      "learning_rate": 5.268425179370418e-07,
+      "loss": 0.842,
+      "step": 38173
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.485539306049715,
+      "learning_rate": 5.26598142530621e-07,
+      "loss": 0.8169,
+      "step": 38174
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7130076378083098,
+      "learning_rate": 5.263538222813457e-07,
+      "loss": 0.7647,
+      "step": 38175
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7112388027271284,
+      "learning_rate": 5.261095571906361e-07,
+      "loss": 0.7956,
+      "step": 38176
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7835967670210078,
+      "learning_rate": 5.258653472599162e-07,
+      "loss": 0.6611,
+      "step": 38177
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5777465766898946,
+      "learning_rate": 5.256211924906074e-07,
+      "loss": 0.9182,
+      "step": 38178
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.562568282483125,
+      "learning_rate": 5.253770928841351e-07,
+      "loss": 0.7047,
+      "step": 38179
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5383710527880976,
+      "learning_rate": 5.251330484419137e-07,
+      "loss": 0.9692,
+      "step": 38180
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6977057472767925,
+      "learning_rate": 5.248890591653666e-07,
+      "loss": 0.8259,
+      "step": 38181
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5267084118439422,
+      "learning_rate": 5.246451250559182e-07,
+      "loss": 0.7819,
+      "step": 38182
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9479933314762972,
+      "learning_rate": 5.24401246114985e-07,
+      "loss": 1.0241,
+      "step": 38183
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.685502217717592,
+      "learning_rate": 5.24157422343986e-07,
+      "loss": 0.8623,
+      "step": 38184
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8318445915960744,
+      "learning_rate": 5.239136537443423e-07,
+      "loss": 0.7978,
+      "step": 38185
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5511492390283583,
+      "learning_rate": 5.23669940317475e-07,
+      "loss": 0.8321,
+      "step": 38186
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5448464616793143,
+      "learning_rate": 5.234262820648006e-07,
+      "loss": 0.9315,
+      "step": 38187
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3838011973130973,
+      "learning_rate": 5.231826789877392e-07,
+      "loss": 0.8093,
+      "step": 38188
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.419138224976671,
+      "learning_rate": 5.229391310877085e-07,
+      "loss": 0.7379,
+      "step": 38189
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4385146013940477,
+      "learning_rate": 5.226956383661263e-07,
+      "loss": 0.858,
+      "step": 38190
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7519969642657756,
+      "learning_rate": 5.224522008244092e-07,
+      "loss": 0.6241,
+      "step": 38191
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5128010871332436,
+      "learning_rate": 5.222088184639785e-07,
+      "loss": 0.7175,
+      "step": 38192
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5052165285543784,
+      "learning_rate": 5.219654912862482e-07,
+      "loss": 0.8574,
+      "step": 38193
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.773930328948956,
+      "learning_rate": 5.217222192926342e-07,
+      "loss": 0.77,
+      "step": 38194
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6655027523913684,
+      "learning_rate": 5.214790024845539e-07,
+      "loss": 0.8239,
+      "step": 38195
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5805734632069075,
+      "learning_rate": 5.212358408634266e-07,
+      "loss": 0.7953,
+      "step": 38196
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7047891400262756,
+      "learning_rate": 5.20992734430662e-07,
+      "loss": 0.8236,
+      "step": 38197
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7214117859884586,
+      "learning_rate": 5.207496831876791e-07,
+      "loss": 0.8448,
+      "step": 38198
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5988034112975948,
+      "learning_rate": 5.205066871358932e-07,
+      "loss": 0.678,
+      "step": 38199
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9866796338435369,
+      "learning_rate": 5.20263746276719e-07,
+      "loss": 0.8391,
+      "step": 38200
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5435485226674464,
+      "learning_rate": 5.200208606115709e-07,
+      "loss": 0.8665,
+      "step": 38201
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8347879884236407,
+      "learning_rate": 5.197780301418609e-07,
+      "loss": 0.888,
+      "step": 38202
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4621645297081254,
+      "learning_rate": 5.195352548690069e-07,
+      "loss": 0.8201,
+      "step": 38203
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7128851829921372,
+      "learning_rate": 5.192925347944178e-07,
+      "loss": 0.8352,
+      "step": 38204
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.879770116400542,
+      "learning_rate": 5.190498699195112e-07,
+      "loss": 0.8721,
+      "step": 38205
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.156972636247597,
+      "learning_rate": 5.188072602456962e-07,
+      "loss": 0.6956,
+      "step": 38206
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8402533358274704,
+      "learning_rate": 5.185647057743881e-07,
+      "loss": 0.9774,
+      "step": 38207
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7743841743077748,
+      "learning_rate": 5.18322206506997e-07,
+      "loss": 0.7133,
+      "step": 38208
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5914396788451581,
+      "learning_rate": 5.180797624449363e-07,
+      "loss": 0.9119,
+      "step": 38209
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5132604067814626,
+      "learning_rate": 5.17837373589618e-07,
+      "loss": 0.7884,
+      "step": 38210
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6721088881247228,
+      "learning_rate": 5.175950399424512e-07,
+      "loss": 0.8895,
+      "step": 38211
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6956136843572325,
+      "learning_rate": 5.173527615048491e-07,
+      "loss": 0.7968,
+      "step": 38212
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7840771431401737,
+      "learning_rate": 5.171105382782215e-07,
+      "loss": 0.7315,
+      "step": 38213
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6077908875298559,
+      "learning_rate": 5.168683702639788e-07,
+      "loss": 0.7536,
+      "step": 38214
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1019488858792723,
+      "learning_rate": 5.166262574635295e-07,
+      "loss": 0.8379,
+      "step": 38215
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5028157003709144,
+      "learning_rate": 5.163841998782837e-07,
+      "loss": 0.8741,
+      "step": 38216
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6673942730046052,
+      "learning_rate": 5.161421975096537e-07,
+      "loss": 0.8129,
+      "step": 38217
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8033212519355788,
+      "learning_rate": 5.159002503590472e-07,
+      "loss": 0.82,
+      "step": 38218
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6715451522623808,
+      "learning_rate": 5.156583584278696e-07,
+      "loss": 0.8941,
+      "step": 38219
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5120107038238038,
+      "learning_rate": 5.154165217175311e-07,
+      "loss": 0.8042,
+      "step": 38220
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.638424252930701,
+      "learning_rate": 5.151747402294427e-07,
+      "loss": 0.7043,
+      "step": 38221
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6364314151693569,
+      "learning_rate": 5.1493301396501e-07,
+      "loss": 0.7428,
+      "step": 38222
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6189749161111877,
+      "learning_rate": 5.146913429256384e-07,
+      "loss": 0.8624,
+      "step": 38223
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9382220614353758,
+      "learning_rate": 5.14449727112738e-07,
+      "loss": 0.8884,
+      "step": 38224
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.966663766339503,
+      "learning_rate": 5.142081665277132e-07,
+      "loss": 0.6683,
+      "step": 38225
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6079536371871088,
+      "learning_rate": 5.13966661171974e-07,
+      "loss": 0.8278,
+      "step": 38226
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.407208599650816,
+      "learning_rate": 5.137252110469215e-07,
+      "loss": 0.8897,
+      "step": 38227
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6193831287564395,
+      "learning_rate": 5.134838161539657e-07,
+      "loss": 0.8666,
+      "step": 38228
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5474519024303455,
+      "learning_rate": 5.132424764945088e-07,
+      "loss": 0.738,
+      "step": 38229
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7173823014299194,
+      "learning_rate": 5.130011920699596e-07,
+      "loss": 0.7139,
+      "step": 38230
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.46686771013908,
+      "learning_rate": 5.127599628817215e-07,
+      "loss": 0.8644,
+      "step": 38231
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5475052752682579,
+      "learning_rate": 5.125187889311966e-07,
+      "loss": 0.7698,
+      "step": 38232
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6371388369609463,
+      "learning_rate": 5.122776702197907e-07,
+      "loss": 0.8976,
+      "step": 38233
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.72500032753802,
+      "learning_rate": 5.120366067489091e-07,
+      "loss": 0.7162,
+      "step": 38234
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5651839212506744,
+      "learning_rate": 5.117955985199552e-07,
+      "loss": 0.887,
+      "step": 38235
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6267570067607917,
+      "learning_rate": 5.115546455343289e-07,
+      "loss": 0.7878,
+      "step": 38236
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6491282887889736,
+      "learning_rate": 5.11313747793436e-07,
+      "loss": 0.7933,
+      "step": 38237
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7478870357727665,
+      "learning_rate": 5.110729052986796e-07,
+      "loss": 0.8906,
+      "step": 38238
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5993697144798347,
+      "learning_rate": 5.108321180514608e-07,
+      "loss": 0.7623,
+      "step": 38239
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5311620853768533,
+      "learning_rate": 5.105913860531808e-07,
+      "loss": 0.8264,
+      "step": 38240
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6901576998859165,
+      "learning_rate": 5.103507093052418e-07,
+      "loss": 0.7772,
+      "step": 38241
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7497832800578312,
+      "learning_rate": 5.101100878090471e-07,
+      "loss": 0.77,
+      "step": 38242
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6959358676087464,
+      "learning_rate": 5.098695215659954e-07,
+      "loss": 0.7759,
+      "step": 38243
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4975447372870418,
+      "learning_rate": 5.096290105774859e-07,
+      "loss": 0.8617,
+      "step": 38244
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7068913321485102,
+      "learning_rate": 5.093885548449229e-07,
+      "loss": 0.921,
+      "step": 38245
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4688888674027412,
+      "learning_rate": 5.091481543697042e-07,
+      "loss": 0.8889,
+      "step": 38246
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6344831928343297,
+      "learning_rate": 5.089078091532296e-07,
+      "loss": 0.9264,
+      "step": 38247
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6627614379482218,
+      "learning_rate": 5.086675191968971e-07,
+      "loss": 0.8793,
+      "step": 38248
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7571538926826271,
+      "learning_rate": 5.084272845021088e-07,
+      "loss": 0.8064,
+      "step": 38249
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8159481511437456,
+      "learning_rate": 5.081871050702614e-07,
+      "loss": 0.7989,
+      "step": 38250
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7437791820436903,
+      "learning_rate": 5.079469809027537e-07,
+      "loss": 0.8357,
+      "step": 38251
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6693490805925286,
+      "learning_rate": 5.077069120009848e-07,
+      "loss": 0.8404,
+      "step": 38252
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6814991960086183,
+      "learning_rate": 5.0746689836635e-07,
+      "loss": 0.854,
+      "step": 38253
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.107659219248448,
+      "learning_rate": 5.072269400002483e-07,
+      "loss": 0.7775,
+      "step": 38254
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0335363181811488,
+      "learning_rate": 5.069870369040775e-07,
+      "loss": 0.9124,
+      "step": 38255
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6107946136506928,
+      "learning_rate": 5.067471890792341e-07,
+      "loss": 0.7516,
+      "step": 38256
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9658032548824547,
+      "learning_rate": 5.065073965271128e-07,
+      "loss": 0.8592,
+      "step": 38257
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5174474716710962,
+      "learning_rate": 5.062676592491101e-07,
+      "loss": 0.8651,
+      "step": 38258
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6375913999676164,
+      "learning_rate": 5.060279772466248e-07,
+      "loss": 0.8894,
+      "step": 38259
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5086103413339074,
+      "learning_rate": 5.057883505210504e-07,
+      "loss": 0.8298,
+      "step": 38260
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.642182042244986,
+      "learning_rate": 5.055487790737812e-07,
+      "loss": 0.7794,
+      "step": 38261
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6862386201623774,
+      "learning_rate": 5.053092629062117e-07,
+      "loss": 0.9142,
+      "step": 38262
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6726068840845894,
+      "learning_rate": 5.050698020197398e-07,
+      "loss": 0.9365,
+      "step": 38263
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5971694900434894,
+      "learning_rate": 5.048303964157575e-07,
+      "loss": 0.859,
+      "step": 38264
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7140489622456696,
+      "learning_rate": 5.045910460956572e-07,
+      "loss": 0.7522,
+      "step": 38265
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.542424905314493,
+      "learning_rate": 5.043517510608354e-07,
+      "loss": 0.6932,
+      "step": 38266
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5611383728130022,
+      "learning_rate": 5.041125113126821e-07,
+      "loss": 0.705,
+      "step": 38267
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8206747373380899,
+      "learning_rate": 5.038733268525941e-07,
+      "loss": 0.6865,
+      "step": 38268
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7453437087896815,
+      "learning_rate": 5.036341976819615e-07,
+      "loss": 0.9797,
+      "step": 38269
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5010574884376071,
+      "learning_rate": 5.033951238021773e-07,
+      "loss": 0.836,
+      "step": 38270
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6366723266765872,
+      "learning_rate": 5.031561052146316e-07,
+      "loss": 0.7481,
+      "step": 38271
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5099408663455431,
+      "learning_rate": 5.029171419207202e-07,
+      "loss": 0.8564,
+      "step": 38272
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5421717372658086,
+      "learning_rate": 5.026782339218306e-07,
+      "loss": 0.9723,
+      "step": 38273
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6072910904439877,
+      "learning_rate": 5.024393812193551e-07,
+      "loss": 0.8743,
+      "step": 38274
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5464476983690325,
+      "learning_rate": 5.022005838146837e-07,
+      "loss": 0.7218,
+      "step": 38275
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7492866848489776,
+      "learning_rate": 5.019618417092075e-07,
+      "loss": 0.9171,
+      "step": 38276
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6085520220359752,
+      "learning_rate": 5.017231549043189e-07,
+      "loss": 0.855,
+      "step": 38277
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8036116074094575,
+      "learning_rate": 5.014845234014032e-07,
+      "loss": 0.7183,
+      "step": 38278
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6959116271122625,
+      "learning_rate": 5.012459472018516e-07,
+      "loss": 0.8705,
+      "step": 38279
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5234059329766618,
+      "learning_rate": 5.010074263070541e-07,
+      "loss": 0.7825,
+      "step": 38280
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.080435436282269,
+      "learning_rate": 5.007689607183997e-07,
+      "loss": 0.9326,
+      "step": 38281
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6757742861414646,
+      "learning_rate": 5.005305504372738e-07,
+      "loss": 0.7842,
+      "step": 38282
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6996622699328109,
+      "learning_rate": 5.002921954650664e-07,
+      "loss": 0.7085,
+      "step": 38283
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7312460532727306,
+      "learning_rate": 5.000538958031664e-07,
+      "loss": 0.7235,
+      "step": 38284
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6005296959142203,
+      "learning_rate": 4.998156514529595e-07,
+      "loss": 0.7107,
+      "step": 38285
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6964954553211153,
+      "learning_rate": 4.995774624158356e-07,
+      "loss": 0.7514,
+      "step": 38286
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.67745368546005,
+      "learning_rate": 4.993393286931781e-07,
+      "loss": 0.8493,
+      "step": 38287
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0014407445411098,
+      "learning_rate": 4.991012502863745e-07,
+      "loss": 0.7849,
+      "step": 38288
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.643027531020821,
+      "learning_rate": 4.988632271968108e-07,
+      "loss": 0.7805,
+      "step": 38289
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.597672693146378,
+      "learning_rate": 4.986252594258756e-07,
+      "loss": 0.7362,
+      "step": 38290
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6167687438859442,
+      "learning_rate": 4.983873469749511e-07,
+      "loss": 0.7759,
+      "step": 38291
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7184487734183402,
+      "learning_rate": 4.981494898454232e-07,
+      "loss": 0.7791,
+      "step": 38292
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.726154597459881,
+      "learning_rate": 4.97911688038677e-07,
+      "loss": 0.9019,
+      "step": 38293
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7594656887956337,
+      "learning_rate": 4.976739415560982e-07,
+      "loss": 0.7675,
+      "step": 38294
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7507425457574172,
+      "learning_rate": 4.974362503990704e-07,
+      "loss": 0.6706,
+      "step": 38295
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8361978104594174,
+      "learning_rate": 4.971986145689756e-07,
+      "loss": 0.8846,
+      "step": 38296
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.169531495664704,
+      "learning_rate": 4.969610340671993e-07,
+      "loss": 0.8217,
+      "step": 38297
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4985132539757584,
+      "learning_rate": 4.967235088951261e-07,
+      "loss": 0.9071,
+      "step": 38298
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7388283066294858,
+      "learning_rate": 4.96486039054136e-07,
+      "loss": 0.8633,
+      "step": 38299
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4914177674533018,
+      "learning_rate": 4.962486245456133e-07,
+      "loss": 0.8246,
+      "step": 38300
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.321547613325857,
+      "learning_rate": 4.960112653709392e-07,
+      "loss": 0.6781,
+      "step": 38301
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7473789338842043,
+      "learning_rate": 4.957739615314972e-07,
+      "loss": 0.6999,
+      "step": 38302
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7460401092961053,
+      "learning_rate": 4.955367130286692e-07,
+      "loss": 0.6708,
+      "step": 38303
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.777427321983606,
+      "learning_rate": 4.952995198638333e-07,
+      "loss": 0.7092,
+      "step": 38304
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5515872051752604,
+      "learning_rate": 4.950623820383749e-07,
+      "loss": 0.8077,
+      "step": 38305
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.603266064944895,
+      "learning_rate": 4.948252995536707e-07,
+      "loss": 0.7732,
+      "step": 38306
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.396675436706313,
+      "learning_rate": 4.94588272411104e-07,
+      "loss": 0.7753,
+      "step": 38307
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8542872484550985,
+      "learning_rate": 4.943513006120526e-07,
+      "loss": 0.6705,
+      "step": 38308
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.761109807266483,
+      "learning_rate": 4.941143841578988e-07,
+      "loss": 0.805,
+      "step": 38309
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4215116543616069,
+      "learning_rate": 4.938775230500192e-07,
+      "loss": 0.6581,
+      "step": 38310
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5302266715592825,
+      "learning_rate": 4.936407172897961e-07,
+      "loss": 0.8662,
+      "step": 38311
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.801333824193659,
+      "learning_rate": 4.934039668786061e-07,
+      "loss": 0.7262,
+      "step": 38312
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6011442100510862,
+      "learning_rate": 4.931672718178259e-07,
+      "loss": 0.7911,
+      "step": 38313
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.775354975385622,
+      "learning_rate": 4.929306321088368e-07,
+      "loss": 0.7184,
+      "step": 38314
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.583966112520306,
+      "learning_rate": 4.926940477530151e-07,
+      "loss": 0.7121,
+      "step": 38315
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6388420221888276,
+      "learning_rate": 4.924575187517399e-07,
+      "loss": 0.7902,
+      "step": 38316
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5470378477991844,
+      "learning_rate": 4.922210451063858e-07,
+      "loss": 0.7262,
+      "step": 38317
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6216785947913614,
+      "learning_rate": 4.919846268183303e-07,
+      "loss": 0.8784,
+      "step": 38318
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7819748354676898,
+      "learning_rate": 4.917482638889525e-07,
+      "loss": 0.6894,
+      "step": 38319
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.701717064328486,
+      "learning_rate": 4.915119563196258e-07,
+      "loss": 0.6663,
+      "step": 38320
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6258882597767588,
+      "learning_rate": 4.912757041117266e-07,
+      "loss": 0.8472,
+      "step": 38321
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.710206327873552,
+      "learning_rate": 4.910395072666297e-07,
+      "loss": 0.7739,
+      "step": 38322
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8739161702032905,
+      "learning_rate": 4.908033657857125e-07,
+      "loss": 0.8302,
+      "step": 38323
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4674313894449404,
+      "learning_rate": 4.905672796703498e-07,
+      "loss": 0.838,
+      "step": 38324
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.048867500918769,
+      "learning_rate": 4.903312489219125e-07,
+      "loss": 0.7125,
+      "step": 38325
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5002049753324,
+      "learning_rate": 4.900952735417786e-07,
+      "loss": 0.7465,
+      "step": 38326
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5079124766441814,
+      "learning_rate": 4.898593535313201e-07,
+      "loss": 0.7917,
+      "step": 38327
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7869824943553305,
+      "learning_rate": 4.896234888919127e-07,
+      "loss": 0.6245,
+      "step": 38328
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6959111020318136,
+      "learning_rate": 4.893876796249264e-07,
+      "loss": 0.9033,
+      "step": 38329
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.67068168374358,
+      "learning_rate": 4.891519257317379e-07,
+      "loss": 0.9589,
+      "step": 38330
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.844688035504338,
+      "learning_rate": 4.889162272137171e-07,
+      "loss": 0.773,
+      "step": 38331
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4741815605631556,
+      "learning_rate": 4.886805840722375e-07,
+      "loss": 0.7374,
+      "step": 38332
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8163214715501579,
+      "learning_rate": 4.884449963086712e-07,
+      "loss": 0.88,
+      "step": 38333
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5420130966223273,
+      "learning_rate": 4.882094639243884e-07,
+      "loss": 0.834,
+      "step": 38334
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.720965653606184,
+      "learning_rate": 4.879739869207611e-07,
+      "loss": 0.8947,
+      "step": 38335
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.712431385939123,
+      "learning_rate": 4.877385652991618e-07,
+      "loss": 0.857,
+      "step": 38336
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5431802507159942,
+      "learning_rate": 4.875031990609602e-07,
+      "loss": 0.7667,
+      "step": 38337
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6478117727429449,
+      "learning_rate": 4.872678882075255e-07,
+      "loss": 0.9206,
+      "step": 38338
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7618750463052043,
+      "learning_rate": 4.870326327402297e-07,
+      "loss": 0.7214,
+      "step": 38339
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.800101728690433,
+      "learning_rate": 4.867974326604419e-07,
+      "loss": 0.6951,
+      "step": 38340
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5382879895375154,
+      "learning_rate": 4.865622879695309e-07,
+      "loss": 0.8223,
+      "step": 38341
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.62122623212128,
+      "learning_rate": 4.863271986688656e-07,
+      "loss": 0.8411,
+      "step": 38342
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8586844785670738,
+      "learning_rate": 4.860921647598149e-07,
+      "loss": 0.8138,
+      "step": 38343
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5865095440932362,
+      "learning_rate": 4.8585718624375e-07,
+      "loss": 0.8804,
+      "step": 38344
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.613715659835187,
+      "learning_rate": 4.856222631220354e-07,
+      "loss": 0.7642,
+      "step": 38345
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5645179851174043,
+      "learning_rate": 4.853873953960397e-07,
+      "loss": 0.8106,
+      "step": 38346
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.587127050686854,
+      "learning_rate": 4.851525830671322e-07,
+      "loss": 0.8258,
+      "step": 38347
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.861688146343257,
+      "learning_rate": 4.849178261366771e-07,
+      "loss": 0.9279,
+      "step": 38348
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7636504006458005,
+      "learning_rate": 4.846831246060435e-07,
+      "loss": 0.715,
+      "step": 38349
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8473042640246735,
+      "learning_rate": 4.844484784765968e-07,
+      "loss": 0.897,
+      "step": 38350
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7572992277356864,
+      "learning_rate": 4.842138877497049e-07,
+      "loss": 0.6593,
+      "step": 38351
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8878340973531913,
+      "learning_rate": 4.839793524267311e-07,
+      "loss": 0.8139,
+      "step": 38352
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8377916379902284,
+      "learning_rate": 4.837448725090433e-07,
+      "loss": 0.8717,
+      "step": 38353
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5573081923233043,
+      "learning_rate": 4.835104479980058e-07,
+      "loss": 0.8547,
+      "step": 38354
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7367139178570596,
+      "learning_rate": 4.832760788949809e-07,
+      "loss": 0.8593,
+      "step": 38355
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.818835887214808,
+      "learning_rate": 4.830417652013364e-07,
+      "loss": 0.9193,
+      "step": 38356
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8472353659938197,
+      "learning_rate": 4.828075069184379e-07,
+      "loss": 0.6627,
+      "step": 38357
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5497757779072436,
+      "learning_rate": 4.825733040476465e-07,
+      "loss": 0.76,
+      "step": 38358
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.588089514893298,
+      "learning_rate": 4.823391565903257e-07,
+      "loss": 0.9196,
+      "step": 38359
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.623843761975287,
+      "learning_rate": 4.821050645478386e-07,
+      "loss": 0.8235,
+      "step": 38360
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.079359416375061,
+      "learning_rate": 4.81871027921551e-07,
+      "loss": 0.7874,
+      "step": 38361
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7559425332621172,
+      "learning_rate": 4.81637046712824e-07,
+      "loss": 0.9352,
+      "step": 38362
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7565256205056423,
+      "learning_rate": 4.814031209230175e-07,
+      "loss": 0.7868,
+      "step": 38363
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.937972608168557,
+      "learning_rate": 4.811692505534959e-07,
+      "loss": 0.8705,
+      "step": 38364
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5321843537194906,
+      "learning_rate": 4.809354356056217e-07,
+      "loss": 0.8702,
+      "step": 38365
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6099387314054168,
+      "learning_rate": 4.807016760807559e-07,
+      "loss": 0.8325,
+      "step": 38366
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.126267040181865,
+      "learning_rate": 4.804679719802563e-07,
+      "loss": 0.8194,
+      "step": 38367
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.629124111883684,
+      "learning_rate": 4.802343233054884e-07,
+      "loss": 0.7472,
+      "step": 38368
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5150682758970373,
+      "learning_rate": 4.80000730057808e-07,
+      "loss": 0.9824,
+      "step": 38369
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7248039242493528,
+      "learning_rate": 4.797671922385783e-07,
+      "loss": 0.6863,
+      "step": 38370
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4456255849751145,
+      "learning_rate": 4.795337098491582e-07,
+      "loss": 0.7988,
+      "step": 38371
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9626692968729202,
+      "learning_rate": 4.793002828909066e-07,
+      "loss": 0.8668,
+      "step": 38372
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.688196674707681,
+      "learning_rate": 4.790669113651825e-07,
+      "loss": 0.8346,
+      "step": 38373
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5158682135992154,
+      "learning_rate": 4.788335952733459e-07,
+      "loss": 0.7738,
+      "step": 38374
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.571365469907187,
+      "learning_rate": 4.786003346167556e-07,
+      "loss": 0.6472,
+      "step": 38375
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6822540987167245,
+      "learning_rate": 4.783671293967673e-07,
+      "loss": 0.9114,
+      "step": 38376
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8093795003483986,
+      "learning_rate": 4.781339796147389e-07,
+      "loss": 0.7266,
+      "step": 38377
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6139735067344179,
+      "learning_rate": 4.77900885272029e-07,
+      "loss": 0.7124,
+      "step": 38378
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6157765445249284,
+      "learning_rate": 4.77667846369998e-07,
+      "loss": 0.8106,
+      "step": 38379
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4566123658257313,
+      "learning_rate": 4.774348629099967e-07,
+      "loss": 0.7594,
+      "step": 38380
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6771620809900363,
+      "learning_rate": 4.772019348933854e-07,
+      "loss": 0.8941,
+      "step": 38381
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6028128899035214,
+      "learning_rate": 4.769690623215195e-07,
+      "loss": 0.7676,
+      "step": 38382
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8655420571504915,
+      "learning_rate": 4.7673624519575355e-07,
+      "loss": 0.728,
+      "step": 38383
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4410859033418497,
+      "learning_rate": 4.7650348351744533e-07,
+      "loss": 0.8377,
+      "step": 38384
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8110686915471574,
+      "learning_rate": 4.7627077728794824e-07,
+      "loss": 0.7306,
+      "step": 38385
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5768041791552043,
+      "learning_rate": 4.7603812650861893e-07,
+      "loss": 0.845,
+      "step": 38386
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5909867009034544,
+      "learning_rate": 4.758055311808096e-07,
+      "loss": 0.9179,
+      "step": 38387
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5890269094790292,
+      "learning_rate": 4.755729913058771e-07,
+      "loss": 0.7054,
+      "step": 38388
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8788898241241843,
+      "learning_rate": 4.753405068851735e-07,
+      "loss": 1.0402,
+      "step": 38389
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6047546695110548,
+      "learning_rate": 4.751080779200512e-07,
+      "loss": 0.8824,
+      "step": 38390
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5859590447384058,
+      "learning_rate": 4.748757044118668e-07,
+      "loss": 0.8768,
+      "step": 38391
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6213347532689741,
+      "learning_rate": 4.746433863619715e-07,
+      "loss": 0.932,
+      "step": 38392
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5175193546028025,
+      "learning_rate": 4.744111237717186e-07,
+      "loss": 0.7534,
+      "step": 38393
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4690443340899781,
+      "learning_rate": 4.741789166424593e-07,
+      "loss": 0.8241,
+      "step": 38394
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.70457867406591,
+      "learning_rate": 4.739467649755447e-07,
+      "loss": 0.7772,
+      "step": 38395
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5504537596446626,
+      "learning_rate": 4.7371466877233043e-07,
+      "loss": 0.8306,
+      "step": 38396
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4930226482082953,
+      "learning_rate": 4.734826280341653e-07,
+      "loss": 0.9071,
+      "step": 38397
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6116940811764926,
+      "learning_rate": 4.732506427623984e-07,
+      "loss": 0.807,
+      "step": 38398
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5968114592375076,
+      "learning_rate": 4.7301871295838297e-07,
+      "loss": 0.7844,
+      "step": 38399
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.856707089589291,
+      "learning_rate": 4.727868386234702e-07,
+      "loss": 0.8334,
+      "step": 38400
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.726876841587419,
+      "learning_rate": 4.725550197590079e-07,
+      "loss": 0.9563,
+      "step": 38401
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6276400451242223,
+      "learning_rate": 4.7232325636634603e-07,
+      "loss": 0.7632,
+      "step": 38402
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6367610624205344,
+      "learning_rate": 4.720915484468358e-07,
+      "loss": 0.8333,
+      "step": 38403
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7829360209167321,
+      "learning_rate": 4.718598960018228e-07,
+      "loss": 0.6854,
+      "step": 38404
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.795525431841775,
+      "learning_rate": 4.716282990326604e-07,
+      "loss": 0.753,
+      "step": 38405
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.512186861188532,
+      "learning_rate": 4.7139675754069304e-07,
+      "loss": 0.8248,
+      "step": 38406
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.740339954000017,
+      "learning_rate": 4.7116527152727187e-07,
+      "loss": 0.8958,
+      "step": 38407
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6687198170106528,
+      "learning_rate": 4.709338409937414e-07,
+      "loss": 0.8609,
+      "step": 38408
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.788883133783825,
+      "learning_rate": 4.7070246594145276e-07,
+      "loss": 0.8417,
+      "step": 38409
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6792215764264717,
+      "learning_rate": 4.704711463717515e-07,
+      "loss": 0.8051,
+      "step": 38410
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6227854727708944,
+      "learning_rate": 4.7023988228598326e-07,
+      "loss": 0.9459,
+      "step": 38411
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6128471428139661,
+      "learning_rate": 4.7000867368549473e-07,
+      "loss": 0.8457,
+      "step": 38412
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.615144974081402,
+      "learning_rate": 4.697775205716348e-07,
+      "loss": 0.7689,
+      "step": 38413
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7262615130421758,
+      "learning_rate": 4.695464229457469e-07,
+      "loss": 0.6375,
+      "step": 38414
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.701613282871379,
+      "learning_rate": 4.693153808091755e-07,
+      "loss": 0.776,
+      "step": 38415
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.396953017933607,
+      "learning_rate": 4.690843941632672e-07,
+      "loss": 0.7079,
+      "step": 38416
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.839265256339836,
+      "learning_rate": 4.6885346300936885e-07,
+      "loss": 0.6916,
+      "step": 38417
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.432760460881862,
+      "learning_rate": 4.686225873488226e-07,
+      "loss": 0.8426,
+      "step": 38418
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5712654698487292,
+      "learning_rate": 4.6839176718297297e-07,
+      "loss": 0.7974,
+      "step": 38419
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7295807861995615,
+      "learning_rate": 4.6816100251316333e-07,
+      "loss": 0.7138,
+      "step": 38420
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8251980517168402,
+      "learning_rate": 4.6793029334073926e-07,
+      "loss": 0.7902,
+      "step": 38421
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6684138273706568,
+      "learning_rate": 4.67699639667043e-07,
+      "loss": 0.8325,
+      "step": 38422
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.796954281498799,
+      "learning_rate": 4.6746904149341576e-07,
+      "loss": 0.8704,
+      "step": 38423
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7696705261518634,
+      "learning_rate": 4.67238498821202e-07,
+      "loss": 0.7324,
+      "step": 38424
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4654303341455086,
+      "learning_rate": 4.6700801165174504e-07,
+      "loss": 0.8171,
+      "step": 38425
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.758568313865541,
+      "learning_rate": 4.667775799863861e-07,
+      "loss": 0.9164,
+      "step": 38426
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6022786778150155,
+      "learning_rate": 4.66547203826464e-07,
+      "loss": 0.8073,
+      "step": 38427
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7593685666990644,
+      "learning_rate": 4.6631688317332447e-07,
+      "loss": 0.8576,
+      "step": 38428
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7139774529872618,
+      "learning_rate": 4.660866180283041e-07,
+      "loss": 0.8449,
+      "step": 38429
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5717761277518199,
+      "learning_rate": 4.658564083927475e-07,
+      "loss": 0.8774,
+      "step": 38430
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7287703296805677,
+      "learning_rate": 4.6562625426799237e-07,
+      "loss": 0.9586,
+      "step": 38431
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6905265398972813,
+      "learning_rate": 4.65396155655381e-07,
+      "loss": 0.869,
+      "step": 38432
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.781923029220228,
+      "learning_rate": 4.651661125562501e-07,
+      "loss": 0.7048,
+      "step": 38433
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.125887320536363,
+      "learning_rate": 4.64936124971942e-07,
+      "loss": 0.6751,
+      "step": 38434
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6762107806323354,
+      "learning_rate": 4.6470619290379547e-07,
+      "loss": 0.8511,
+      "step": 38435
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5621953215116682,
+      "learning_rate": 4.6447631635314626e-07,
+      "loss": 0.9273,
+      "step": 38436
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7365134849799927,
+      "learning_rate": 4.6424649532133437e-07,
+      "loss": 0.6481,
+      "step": 38437
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3659349335541304,
+      "learning_rate": 4.640167298097009e-07,
+      "loss": 0.8003,
+      "step": 38438
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5360358211995613,
+      "learning_rate": 4.637870198195804e-07,
+      "loss": 0.8595,
+      "step": 38439
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5492210589147961,
+      "learning_rate": 4.6355736535230956e-07,
+      "loss": 0.9761,
+      "step": 38440
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7941346300293022,
+      "learning_rate": 4.633277664092262e-07,
+      "loss": 0.8016,
+      "step": 38441
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.657602843470964,
+      "learning_rate": 4.630982229916703e-07,
+      "loss": 0.7456,
+      "step": 38442
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0213821230357465,
+      "learning_rate": 4.628687351009753e-07,
+      "loss": 0.7755,
+      "step": 38443
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5523440494210692,
+      "learning_rate": 4.626393027384768e-07,
+      "loss": 0.8,
+      "step": 38444
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6342485515679903,
+      "learning_rate": 4.624099259055126e-07,
+      "loss": 0.774,
+      "step": 38445
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7076445411865708,
+      "learning_rate": 4.6218060460341716e-07,
+      "loss": 0.8309,
+      "step": 38446
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5195686638269394,
+      "learning_rate": 4.619513388335273e-07,
+      "loss": 0.8096,
+      "step": 38447
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.494379178611496,
+      "learning_rate": 4.6172212859717403e-07,
+      "loss": 0.7506,
+      "step": 38448
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7841645502050759,
+      "learning_rate": 4.6149297389569634e-07,
+      "loss": 0.7021,
+      "step": 38449
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6650819734283537,
+      "learning_rate": 4.612638747304243e-07,
+      "loss": 0.8868,
+      "step": 38450
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8193270778062756,
+      "learning_rate": 4.6103483110269575e-07,
+      "loss": 0.7734,
+      "step": 38451
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.31057535254325,
+      "learning_rate": 4.608058430138407e-07,
+      "loss": 0.8819,
+      "step": 38452
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4416444250810279,
+      "learning_rate": 4.605769104651958e-07,
+      "loss": 0.8562,
+      "step": 38453
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6481744542554129,
+      "learning_rate": 4.603480334580912e-07,
+      "loss": 0.7211,
+      "step": 38454
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7201386494265167,
+      "learning_rate": 4.6011921199386136e-07,
+      "loss": 0.78,
+      "step": 38455
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4939264335628906,
+      "learning_rate": 4.5989044607383846e-07,
+      "loss": 0.7998,
+      "step": 38456
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.579804020022573,
+      "learning_rate": 4.5966173569935156e-07,
+      "loss": 0.8846,
+      "step": 38457
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1574275525998896,
+      "learning_rate": 4.594330808717351e-07,
+      "loss": 0.9069,
+      "step": 38458
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.629620890895407,
+      "learning_rate": 4.592044815923191e-07,
+      "loss": 0.8797,
+      "step": 38459
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8015269744129632,
+      "learning_rate": 4.589759378624381e-07,
+      "loss": 0.8011,
+      "step": 38460
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6789695635079611,
+      "learning_rate": 4.587474496834166e-07,
+      "loss": 0.9586,
+      "step": 38461
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8865140796870234,
+      "learning_rate": 4.58519017056589e-07,
+      "loss": 0.8823,
+      "step": 38462
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7534998875531274,
+      "learning_rate": 4.582906399832854e-07,
+      "loss": 0.799,
+      "step": 38463
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.550295086787872,
+      "learning_rate": 4.5806231846483475e-07,
+      "loss": 0.9638,
+      "step": 38464
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6602663875608468,
+      "learning_rate": 4.5783405250256487e-07,
+      "loss": 0.817,
+      "step": 38465
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.54987124338903,
+      "learning_rate": 4.5760584209780577e-07,
+      "loss": 0.821,
+      "step": 38466
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6744315593677002,
+      "learning_rate": 4.573776872518887e-07,
+      "loss": 0.8668,
+      "step": 38467
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.593493966181244,
+      "learning_rate": 4.571495879661392e-07,
+      "loss": 0.7005,
+      "step": 38468
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5593416385083663,
+      "learning_rate": 4.5692154424188506e-07,
+      "loss": 0.8141,
+      "step": 38469
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5192756861205614,
+      "learning_rate": 4.566935560804564e-07,
+      "loss": 0.9331,
+      "step": 38470
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.99585931845587,
+      "learning_rate": 4.5646562348317657e-07,
+      "loss": 0.6473,
+      "step": 38471
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6862563975418121,
+      "learning_rate": 4.5623774645137675e-07,
+      "loss": 0.854,
+      "step": 38472
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7859578440484123,
+      "learning_rate": 4.5600992498638364e-07,
+      "loss": 0.6796,
+      "step": 38473
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6039704098861889,
+      "learning_rate": 4.557821590895217e-07,
+      "loss": 0.7407,
+      "step": 38474
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5896134638838015,
+      "learning_rate": 4.5555444876211664e-07,
+      "loss": 0.8001,
+      "step": 38475
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7696587770534811,
+      "learning_rate": 4.553267940054951e-07,
+      "loss": 0.7276,
+      "step": 38476
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4029234293501764,
+      "learning_rate": 4.5509919482098486e-07,
+      "loss": 0.7254,
+      "step": 38477
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7839040541684947,
+      "learning_rate": 4.5487165120990716e-07,
+      "loss": 0.8496,
+      "step": 38478
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6568761560760876,
+      "learning_rate": 4.546441631735887e-07,
+      "loss": 0.8811,
+      "step": 38479
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5454521720946464,
+      "learning_rate": 4.544167307133529e-07,
+      "loss": 0.9464,
+      "step": 38480
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.446063185476188,
+      "learning_rate": 4.5418935383052756e-07,
+      "loss": 0.9182,
+      "step": 38481
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6791149837310657,
+      "learning_rate": 4.5396203252643377e-07,
+      "loss": 0.904,
+      "step": 38482
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5403713416731977,
+      "learning_rate": 4.5373476680239283e-07,
+      "loss": 0.9012,
+      "step": 38483
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4453242421680057,
+      "learning_rate": 4.5350755665973243e-07,
+      "loss": 0.8338,
+      "step": 38484
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.444547525019434,
+      "learning_rate": 4.532804020997728e-07,
+      "loss": 0.8331,
+      "step": 38485
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7586091749175365,
+      "learning_rate": 4.5305330312383713e-07,
+      "loss": 0.7066,
+      "step": 38486
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8353154073275588,
+      "learning_rate": 4.5282625973324666e-07,
+      "loss": 0.7859,
+      "step": 38487
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5651811683494172,
+      "learning_rate": 4.52599271929326e-07,
+      "loss": 0.842,
+      "step": 38488
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4636565846871166,
+      "learning_rate": 4.523723397133939e-07,
+      "loss": 0.8539,
+      "step": 38489
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5520384869386794,
+      "learning_rate": 4.5214546308677386e-07,
+      "loss": 0.7204,
+      "step": 38490
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6570136504099127,
+      "learning_rate": 4.5191864205078596e-07,
+      "loss": 0.7801,
+      "step": 38491
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6629542800833061,
+      "learning_rate": 4.516918766067491e-07,
+      "loss": 0.846,
+      "step": 38492
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.524554309004315,
+      "learning_rate": 4.514651667559844e-07,
+      "loss": 0.8821,
+      "step": 38493
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.638390375843148,
+      "learning_rate": 4.512385124998142e-07,
+      "loss": 0.8612,
+      "step": 38494
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8216659712282758,
+      "learning_rate": 4.510119138395563e-07,
+      "loss": 0.6735,
+      "step": 38495
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.928891184735028,
+      "learning_rate": 4.5078537077652975e-07,
+      "loss": 0.8737,
+      "step": 38496
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3546769498533826,
+      "learning_rate": 4.5055888331205335e-07,
+      "loss": 0.9176,
+      "step": 38497
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7599475904973488,
+      "learning_rate": 4.503324514474483e-07,
+      "loss": 0.7248,
+      "step": 38498
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4217111098354693,
+      "learning_rate": 4.5010607518403024e-07,
+      "loss": 0.8356,
+      "step": 38499
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4726321201385533,
+      "learning_rate": 4.498797545231171e-07,
+      "loss": 0.6778,
+      "step": 38500
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.688115835629164,
+      "learning_rate": 4.496534894660276e-07,
+      "loss": 0.8404,
+      "step": 38501
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6808882216074152,
+      "learning_rate": 4.4942728001408087e-07,
+      "loss": 0.8715,
+      "step": 38502
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7858930638677332,
+      "learning_rate": 4.4920112616859135e-07,
+      "loss": 0.7027,
+      "step": 38503
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6433827914472867,
+      "learning_rate": 4.4897502793087576e-07,
+      "loss": 0.8424,
+      "step": 38504
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5325188749044618,
+      "learning_rate": 4.487489853022531e-07,
+      "loss": 0.8199,
+      "step": 38505
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6338384476398495,
+      "learning_rate": 4.485229982840356e-07,
+      "loss": 0.8007,
+      "step": 38506
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6130136792658094,
+      "learning_rate": 4.4829706687754213e-07,
+      "loss": 0.7964,
+      "step": 38507
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6914793468582536,
+      "learning_rate": 4.480711910840874e-07,
+      "loss": 0.6749,
+      "step": 38508
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3267279436812967,
+      "learning_rate": 4.478453709049857e-07,
+      "loss": 0.8221,
+      "step": 38509
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4797041890223206,
+      "learning_rate": 4.476196063415528e-07,
+      "loss": 0.9545,
+      "step": 38510
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4752388131434315,
+      "learning_rate": 4.4739389739510306e-07,
+      "loss": 0.7879,
+      "step": 38511
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5363401871042996,
+      "learning_rate": 4.4716824406695005e-07,
+      "loss": 0.8562,
+      "step": 38512
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7706824266276728,
+      "learning_rate": 4.4694264635840814e-07,
+      "loss": 0.7789,
+      "step": 38513
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9150193934285231,
+      "learning_rate": 4.467171042707896e-07,
+      "loss": 0.8685,
+      "step": 38514
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7939104604850193,
+      "learning_rate": 4.4649161780540907e-07,
+      "loss": 0.722,
+      "step": 38515
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1917946357848184,
+      "learning_rate": 4.4626618696357983e-07,
+      "loss": 0.8472,
+      "step": 38516
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5576011264505116,
+      "learning_rate": 4.460408117466131e-07,
+      "loss": 0.8183,
+      "step": 38517
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5876864778638262,
+      "learning_rate": 4.4581549215582e-07,
+      "loss": 0.786,
+      "step": 38518
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6106531690005563,
+      "learning_rate": 4.455902281925162e-07,
+      "loss": 0.7163,
+      "step": 38519
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.052576405440692,
+      "learning_rate": 4.453650198580106e-07,
+      "loss": 0.6685,
+      "step": 38520
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7468740182464484,
+      "learning_rate": 4.4513986715361336e-07,
+      "loss": 0.827,
+      "step": 38521
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.459911791666215,
+      "learning_rate": 4.4491477008063666e-07,
+      "loss": 0.8595,
+      "step": 38522
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.726835135168576,
+      "learning_rate": 4.446897286403928e-07,
+      "loss": 0.8624,
+      "step": 38523
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6659686306167658,
+      "learning_rate": 4.444647428341908e-07,
+      "loss": 0.9054,
+      "step": 38524
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7455920701413026,
+      "learning_rate": 4.4423981266333847e-07,
+      "loss": 0.8044,
+      "step": 38525
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.648208830552741,
+      "learning_rate": 4.440149381291492e-07,
+      "loss": 0.8514,
+      "step": 38526
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5902372105182423,
+      "learning_rate": 4.437901192329286e-07,
+      "loss": 0.854,
+      "step": 38527
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6046411083625929,
+      "learning_rate": 4.435653559759878e-07,
+      "loss": 0.775,
+      "step": 38528
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7764125940100348,
+      "learning_rate": 4.4334064835963475e-07,
+      "loss": 0.7919,
+      "step": 38529
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5450465176955668,
+      "learning_rate": 4.4311599638517944e-07,
+      "loss": 0.7907,
+      "step": 38530
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7512779251238637,
+      "learning_rate": 4.4289140005392637e-07,
+      "loss": 0.7139,
+      "step": 38531
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5553014621748098,
+      "learning_rate": 4.426668593671879e-07,
+      "loss": 0.8807,
+      "step": 38532
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5658400163125112,
+      "learning_rate": 4.4244237432626737e-07,
+      "loss": 0.8696,
+      "step": 38533
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8290429815887617,
+      "learning_rate": 4.422179449324726e-07,
+      "loss": 0.7344,
+      "step": 38534
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7881967350045999,
+      "learning_rate": 4.419935711871115e-07,
+      "loss": 0.7672,
+      "step": 38535
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8067337840940905,
+      "learning_rate": 4.417692530914908e-07,
+      "loss": 0.8608,
+      "step": 38536
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5709512453116556,
+      "learning_rate": 4.415449906469149e-07,
+      "loss": 0.8577,
+      "step": 38537
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5220767507394906,
+      "learning_rate": 4.4132078385468956e-07,
+      "loss": 0.7866,
+      "step": 38538
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6261966905172585,
+      "learning_rate": 4.4109663271612036e-07,
+      "loss": 0.7351,
+      "step": 38539
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.600592612933546,
+      "learning_rate": 4.40872537232514e-07,
+      "loss": 0.9077,
+      "step": 38540
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5004787476963253,
+      "learning_rate": 4.40648497405175e-07,
+      "loss": 0.819,
+      "step": 38541
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4744271748642608,
+      "learning_rate": 4.4042451323540457e-07,
+      "loss": 0.7043,
+      "step": 38542
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6718752019664274,
+      "learning_rate": 4.4020058472450946e-07,
+      "loss": 0.8256,
+      "step": 38543
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4755260452534333,
+      "learning_rate": 4.399767118737952e-07,
+      "loss": 0.7736,
+      "step": 38544
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8456504584359594,
+      "learning_rate": 4.3975289468456194e-07,
+      "loss": 0.9092,
+      "step": 38545
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.704765035284419,
+      "learning_rate": 4.395291331581142e-07,
+      "loss": 0.8585,
+      "step": 38546
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4600282779524123,
+      "learning_rate": 4.393054272957531e-07,
+      "loss": 0.8539,
+      "step": 38547
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6489802934297535,
+      "learning_rate": 4.390817770987854e-07,
+      "loss": 0.7682,
+      "step": 38548
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7269873894599383,
+      "learning_rate": 4.38858182568509e-07,
+      "loss": 0.7961,
+      "step": 38549
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6661656975305763,
+      "learning_rate": 4.3863464370622723e-07,
+      "loss": 0.7633,
+      "step": 38550
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9544186531428736,
+      "learning_rate": 4.384111605132424e-07,
+      "loss": 0.8438,
+      "step": 38551
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4887332396526956,
+      "learning_rate": 4.381877329908535e-07,
+      "loss": 0.8817,
+      "step": 38552
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6730198601880875,
+      "learning_rate": 4.37964361140365e-07,
+      "loss": 0.714,
+      "step": 38553
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6029823040103335,
+      "learning_rate": 4.377410449630737e-07,
+      "loss": 0.7965,
+      "step": 38554
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9448607262447901,
+      "learning_rate": 4.375177844602818e-07,
+      "loss": 0.8951,
+      "step": 38555
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5760681998726314,
+      "learning_rate": 4.3729457963328837e-07,
+      "loss": 0.8436,
+      "step": 38556
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.565578779355469,
+      "learning_rate": 4.370714304833923e-07,
+      "loss": 0.8839,
+      "step": 38557
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7367960982746437,
+      "learning_rate": 4.368483370118981e-07,
+      "loss": 0.6685,
+      "step": 38558
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5061454017163265,
+      "learning_rate": 4.36625299220097e-07,
+      "loss": 0.9366,
+      "step": 38559
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8384382241837787,
+      "learning_rate": 4.3640231710929124e-07,
+      "loss": 0.8649,
+      "step": 38560
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.672653681747749,
+      "learning_rate": 4.3617939068077874e-07,
+      "loss": 0.8899,
+      "step": 38561
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8823742817493732,
+      "learning_rate": 4.359565199358606e-07,
+      "loss": 0.8554,
+      "step": 38562
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4638711892764213,
+      "learning_rate": 4.357337048758292e-07,
+      "loss": 0.8201,
+      "step": 38563
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6646547666011944,
+      "learning_rate": 4.355109455019835e-07,
+      "loss": 0.8255,
+      "step": 38564
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6687350100523668,
+      "learning_rate": 4.3528824181562237e-07,
+      "loss": 0.8729,
+      "step": 38565
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.095475171470099,
+      "learning_rate": 4.350655938180426e-07,
+      "loss": 0.8586,
+      "step": 38566
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5779687471531831,
+      "learning_rate": 4.348430015105365e-07,
+      "loss": 0.895,
+      "step": 38567
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.634022192623683,
+      "learning_rate": 4.34620464894403e-07,
+      "loss": 0.986,
+      "step": 38568
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.074440639545319,
+      "learning_rate": 4.343979839709389e-07,
+      "loss": 0.7647,
+      "step": 38569
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5447991512075645,
+      "learning_rate": 4.3417555874143644e-07,
+      "loss": 0.8656,
+      "step": 38570
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6349538624902573,
+      "learning_rate": 4.3395318920719465e-07,
+      "loss": 0.9464,
+      "step": 38571
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5642967465754274,
+      "learning_rate": 4.3373087536950574e-07,
+      "loss": 0.781,
+      "step": 38572
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8314209315392729,
+      "learning_rate": 4.335086172296621e-07,
+      "loss": 0.694,
+      "step": 38573
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5279289663237603,
+      "learning_rate": 4.3328641478896037e-07,
+      "loss": 0.8595,
+      "step": 38574
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.625014528979144,
+      "learning_rate": 4.330642680486952e-07,
+      "loss": 0.8069,
+      "step": 38575
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.43537529675004,
+      "learning_rate": 4.3284217701015985e-07,
+      "loss": 0.7318,
+      "step": 38576
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8001407116390554,
+      "learning_rate": 4.3262014167464337e-07,
+      "loss": 0.7377,
+      "step": 38577
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5481189018572363,
+      "learning_rate": 4.3239816204344256e-07,
+      "loss": 0.7943,
+      "step": 38578
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5700057166929615,
+      "learning_rate": 4.321762381178496e-07,
+      "loss": 0.8279,
+      "step": 38579
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4435173358054674,
+      "learning_rate": 4.319543698991546e-07,
+      "loss": 0.8508,
+      "step": 38580
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7936052465290878,
+      "learning_rate": 4.3173255738865104e-07,
+      "loss": 0.7084,
+      "step": 38581
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5243310857077668,
+      "learning_rate": 4.3151080058762897e-07,
+      "loss": 0.7418,
+      "step": 38582
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.884956536605586,
+      "learning_rate": 4.312890994973806e-07,
+      "loss": 0.8071,
+      "step": 38583
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5373086866370507,
+      "learning_rate": 4.310674541191984e-07,
+      "loss": 0.9246,
+      "step": 38584
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.980156961040861,
+      "learning_rate": 4.308458644543678e-07,
+      "loss": 0.8031,
+      "step": 38585
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7334245977183806,
+      "learning_rate": 4.306243305041846e-07,
+      "loss": 0.6675,
+      "step": 38586
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.667905285876451,
+      "learning_rate": 4.304028522699344e-07,
+      "loss": 0.8285,
+      "step": 38587
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.503590816407067,
+      "learning_rate": 4.3018142975291054e-07,
+      "loss": 0.937,
+      "step": 38588
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6318297090702598,
+      "learning_rate": 4.2996006295439764e-07,
+      "loss": 0.8643,
+      "step": 38589
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6287828203990553,
+      "learning_rate": 4.2973875187568903e-07,
+      "loss": 0.849,
+      "step": 38590
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.577574232696068,
+      "learning_rate": 4.295174965180704e-07,
+      "loss": 0.7763,
+      "step": 38591
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6588823633812737,
+      "learning_rate": 4.292962968828318e-07,
+      "loss": 0.8778,
+      "step": 38592
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6798013938919378,
+      "learning_rate": 4.2907515297126003e-07,
+      "loss": 0.7623,
+      "step": 38593
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.381299079154539,
+      "learning_rate": 4.288540647846418e-07,
+      "loss": 0.7821,
+      "step": 38594
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6161375611120474,
+      "learning_rate": 4.286330323242649e-07,
+      "loss": 0.7279,
+      "step": 38595
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3915996938729687,
+      "learning_rate": 4.284120555914184e-07,
+      "loss": 0.7837,
+      "step": 38596
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6749903701253923,
+      "learning_rate": 4.281911345873868e-07,
+      "loss": 0.8395,
+      "step": 38597
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6046381829612202,
+      "learning_rate": 4.279702693134569e-07,
+      "loss": 0.7549,
+      "step": 38598
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9856639536895155,
+      "learning_rate": 4.277494597709131e-07,
+      "loss": 0.8405,
+      "step": 38599
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9716428068180172,
+      "learning_rate": 4.2752870596104446e-07,
+      "loss": 0.9135,
+      "step": 38600
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.601733906462651,
+      "learning_rate": 4.273080078851333e-07,
+      "loss": 0.7491,
+      "step": 38601
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7380879855586513,
+      "learning_rate": 4.270873655444652e-07,
+      "loss": 0.6624,
+      "step": 38602
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.642372904320335,
+      "learning_rate": 4.268667789403247e-07,
+      "loss": 0.8587,
+      "step": 38603
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7829914509670605,
+      "learning_rate": 4.2664624807399857e-07,
+      "loss": 0.7054,
+      "step": 38604
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6005557114773996,
+      "learning_rate": 4.264257729467691e-07,
+      "loss": 0.964,
+      "step": 38605
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8543685666264198,
+      "learning_rate": 4.2620535355991756e-07,
+      "loss": 0.8103,
+      "step": 38606
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5390746103605413,
+      "learning_rate": 4.2598498991473057e-07,
+      "loss": 0.7876,
+      "step": 38607
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5007364610611744,
+      "learning_rate": 4.2576468201248946e-07,
+      "loss": 0.9688,
+      "step": 38608
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7169208026540985,
+      "learning_rate": 4.2554442985447864e-07,
+      "loss": 0.6686,
+      "step": 38609
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5810810848109431,
+      "learning_rate": 4.2532423344197825e-07,
+      "loss": 0.8266,
+      "step": 38610
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.067695538684126,
+      "learning_rate": 4.2510409277627174e-07,
+      "loss": 0.8719,
+      "step": 38611
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0745846793664713,
+      "learning_rate": 4.2488400785864027e-07,
+      "loss": 0.7893,
+      "step": 38612
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9468910716665773,
+      "learning_rate": 4.2466397869036724e-07,
+      "loss": 0.6891,
+      "step": 38613
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6941895996304173,
+      "learning_rate": 4.2444400527273056e-07,
+      "loss": 0.9016,
+      "step": 38614
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.645130757293772,
+      "learning_rate": 4.2422408760701253e-07,
+      "loss": 0.9454,
+      "step": 38615
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7357305179414297,
+      "learning_rate": 4.240042256944932e-07,
+      "loss": 0.6991,
+      "step": 38616
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7701260479085308,
+      "learning_rate": 4.2378441953645375e-07,
+      "loss": 0.7733,
+      "step": 38617
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7916275177339891,
+      "learning_rate": 4.2356466913417326e-07,
+      "loss": 0.7262,
+      "step": 38618
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.806371690283769,
+      "learning_rate": 4.233449744889295e-07,
+      "loss": 0.7581,
+      "step": 38619
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.735855307205015,
+      "learning_rate": 4.2312533560200376e-07,
+      "loss": 0.777,
+      "step": 38620
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5613827870542667,
+      "learning_rate": 4.22905752474676e-07,
+      "loss": 0.8162,
+      "step": 38621
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.85872888070316,
+      "learning_rate": 4.226862251082231e-07,
+      "loss": 0.7651,
+      "step": 38622
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5975714810565906,
+      "learning_rate": 4.224667535039206e-07,
+      "loss": 0.895,
+      "step": 38623
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0060946302842737,
+      "learning_rate": 4.222473376630498e-07,
+      "loss": 0.7631,
+      "step": 38624
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7231989950497189,
+      "learning_rate": 4.2202797758688853e-07,
+      "loss": 0.7958,
+      "step": 38625
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5046968846702171,
+      "learning_rate": 4.2180867327671237e-07,
+      "loss": 0.7916,
+      "step": 38626
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5540261653715726,
+      "learning_rate": 4.2158942473379706e-07,
+      "loss": 0.8816,
+      "step": 38627
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4681575912537799,
+      "learning_rate": 4.2137023195942264e-07,
+      "loss": 0.8639,
+      "step": 38628
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5851977590653097,
+      "learning_rate": 4.211510949548614e-07,
+      "loss": 0.7674,
+      "step": 38629
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7593527271052916,
+      "learning_rate": 4.209320137213924e-07,
+      "loss": 0.7882,
+      "step": 38630
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6297667658036437,
+      "learning_rate": 4.2071298826028894e-07,
+      "loss": 0.7836,
+      "step": 38631
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8301494107596227,
+      "learning_rate": 4.204940185728279e-07,
+      "loss": 0.6628,
+      "step": 38632
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6313461338777984,
+      "learning_rate": 4.2027510466028265e-07,
+      "loss": 0.9099,
+      "step": 38633
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7267870834996741,
+      "learning_rate": 4.2005624652392996e-07,
+      "loss": 0.8341,
+      "step": 38634
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7804231124322902,
+      "learning_rate": 4.1983744416504323e-07,
+      "loss": 0.7505,
+      "step": 38635
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8189669849307761,
+      "learning_rate": 4.196186975848937e-07,
+      "loss": 0.9311,
+      "step": 38636
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5547544553161265,
+      "learning_rate": 4.19400006784757e-07,
+      "loss": 0.8533,
+      "step": 38637
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5346680185880541,
+      "learning_rate": 4.1918137176590876e-07,
+      "loss": 0.8579,
+      "step": 38638
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5584217971733172,
+      "learning_rate": 4.189627925296202e-07,
+      "loss": 0.8309,
+      "step": 38639
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6360098239097824,
+      "learning_rate": 4.187442690771615e-07,
+      "loss": 0.8546,
+      "step": 38640
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.603642617976196,
+      "learning_rate": 4.18525801409807e-07,
+      "loss": 0.8884,
+      "step": 38641
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.67958269144009,
+      "learning_rate": 4.183073895288303e-07,
+      "loss": 0.8151,
+      "step": 38642
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.593077017252534,
+      "learning_rate": 4.1808903343550145e-07,
+      "loss": 0.9081,
+      "step": 38643
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7717917141807038,
+      "learning_rate": 4.178707331310905e-07,
+      "loss": 0.7903,
+      "step": 38644
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7519418622608294,
+      "learning_rate": 4.176524886168698e-07,
+      "loss": 0.7537,
+      "step": 38645
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9714153721606764,
+      "learning_rate": 4.1743429989411054e-07,
+      "loss": 0.8088,
+      "step": 38646
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6187485796094727,
+      "learning_rate": 4.1721616696408394e-07,
+      "loss": 0.859,
+      "step": 38647
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.389640769548633,
+      "learning_rate": 4.169980898280568e-07,
+      "loss": 0.798,
+      "step": 38648
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5093032647354927,
+      "learning_rate": 4.167800684873013e-07,
+      "loss": 0.7448,
+      "step": 38649
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4115828247401794,
+      "learning_rate": 4.165621029430855e-07,
+      "loss": 0.7586,
+      "step": 38650
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.388239880186425,
+      "learning_rate": 4.1634419319668163e-07,
+      "loss": 0.7763,
+      "step": 38651
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.522112947319442,
+      "learning_rate": 4.1612633924935307e-07,
+      "loss": 0.7791,
+      "step": 38652
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4823017018054796,
+      "learning_rate": 4.1590854110237333e-07,
+      "loss": 0.8701,
+      "step": 38653
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7636537825047436,
+      "learning_rate": 4.156907987570069e-07,
+      "loss": 0.8586,
+      "step": 38654
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7240501319582152,
+      "learning_rate": 4.154731122145228e-07,
+      "loss": 0.6954,
+      "step": 38655
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.253574627287512,
+      "learning_rate": 4.152554814761911e-07,
+      "loss": 0.858,
+      "step": 38656
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7517551284288173,
+      "learning_rate": 4.1503790654327414e-07,
+      "loss": 0.7466,
+      "step": 38657
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.863613791095082,
+      "learning_rate": 4.148203874170409e-07,
+      "loss": 0.7973,
+      "step": 38658
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.43863040543446,
+      "learning_rate": 4.1460292409875815e-07,
+      "loss": 0.7459,
+      "step": 38659
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6241626911157483,
+      "learning_rate": 4.1438551658969484e-07,
+      "loss": 0.7299,
+      "step": 38660
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.650615775234376,
+      "learning_rate": 4.1416816489111e-07,
+      "loss": 0.8246,
+      "step": 38661
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7463783662344623,
+      "learning_rate": 4.139508690042737e-07,
+      "loss": 0.7186,
+      "step": 38662
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6801665841294042,
+      "learning_rate": 4.137336289304517e-07,
+      "loss": 0.8305,
+      "step": 38663
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6236261712641924,
+      "learning_rate": 4.1351644467090614e-07,
+      "loss": 0.8022,
+      "step": 38664
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5374439942709728,
+      "learning_rate": 4.132993162269039e-07,
+      "loss": 0.6815,
+      "step": 38665
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.719757540338926,
+      "learning_rate": 4.130822435997073e-07,
+      "loss": 0.7694,
+      "step": 38666
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5456696316025091,
+      "learning_rate": 4.128652267905819e-07,
+      "loss": 0.866,
+      "step": 38667
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7134558474092847,
+      "learning_rate": 4.126482658007891e-07,
+      "loss": 0.8471,
+      "step": 38668
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5841069852036282,
+      "learning_rate": 4.124313606315955e-07,
+      "loss": 0.8481,
+      "step": 38669
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7415345222875169,
+      "learning_rate": 4.1221451128426014e-07,
+      "loss": 0.6856,
+      "step": 38670
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5792754599814183,
+      "learning_rate": 4.119977177600476e-07,
+      "loss": 0.7896,
+      "step": 38671
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6628799646051036,
+      "learning_rate": 4.117809800602202e-07,
+      "loss": 0.7349,
+      "step": 38672
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.729824606349566,
+      "learning_rate": 4.11564298186039e-07,
+      "loss": 0.8716,
+      "step": 38673
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6313340276561616,
+      "learning_rate": 4.113476721387677e-07,
+      "loss": 0.799,
+      "step": 38674
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.413645252277906,
+      "learning_rate": 4.11131101919664e-07,
+      "loss": 0.8722,
+      "step": 38675
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5620818197666742,
+      "learning_rate": 4.1091458752999136e-07,
+      "loss": 0.845,
+      "step": 38676
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5922812127425976,
+      "learning_rate": 4.106981289710099e-07,
+      "loss": 0.7398,
+      "step": 38677
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.729834944054773,
+      "learning_rate": 4.104817262439809e-07,
+      "loss": 0.8292,
+      "step": 38678
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6445194411445523,
+      "learning_rate": 4.102653793501621e-07,
+      "loss": 0.8679,
+      "step": 38679
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5126124932537162,
+      "learning_rate": 4.100490882908137e-07,
+      "loss": 0.7944,
+      "step": 38680
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7518682636890417,
+      "learning_rate": 4.098328530671969e-07,
+      "loss": 0.9007,
+      "step": 38681
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9675771985161647,
+      "learning_rate": 4.0961667368056957e-07,
+      "loss": 0.767,
+      "step": 38682
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9452655199463629,
+      "learning_rate": 4.094005501321896e-07,
+      "loss": 0.8649,
+      "step": 38683
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7477842626616085,
+      "learning_rate": 4.09184482423316e-07,
+      "loss": 0.7154,
+      "step": 38684
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.779070843030663,
+      "learning_rate": 4.089684705552066e-07,
+      "loss": 0.734,
+      "step": 38685
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.449627151156367,
+      "learning_rate": 4.087525145291205e-07,
+      "loss": 0.8042,
+      "step": 38686
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5102704261082243,
+      "learning_rate": 4.085366143463121e-07,
+      "loss": 0.6185,
+      "step": 38687
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5186968246824004,
+      "learning_rate": 4.083207700080416e-07,
+      "loss": 0.8066,
+      "step": 38688
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8530460963286672,
+      "learning_rate": 4.0810498151556246e-07,
+      "loss": 0.7902,
+      "step": 38689
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7778195305028306,
+      "learning_rate": 4.078892488701347e-07,
+      "loss": 0.7459,
+      "step": 38690
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.582136611461743,
+      "learning_rate": 4.0767357207301073e-07,
+      "loss": 0.8085,
+      "step": 38691
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7845444332202638,
+      "learning_rate": 4.074579511254495e-07,
+      "loss": 0.8834,
+      "step": 38692
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.483974605873255,
+      "learning_rate": 4.072423860287045e-07,
+      "loss": 0.7075,
+      "step": 38693
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5434985295539487,
+      "learning_rate": 4.070268767840324e-07,
+      "loss": 0.7527,
+      "step": 38694
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6838409580015,
+      "learning_rate": 4.0681142339268564e-07,
+      "loss": 0.7479,
+      "step": 38695
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6011176375546707,
+      "learning_rate": 4.0659602585591984e-07,
+      "loss": 0.9035,
+      "step": 38696
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6694720736398188,
+      "learning_rate": 4.0638068417498955e-07,
+      "loss": 0.7719,
+      "step": 38697
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5960501894705077,
+      "learning_rate": 4.0616539835114823e-07,
+      "loss": 0.8459,
+      "step": 38698
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6040177525941686,
+      "learning_rate": 4.059501683856504e-07,
+      "loss": 0.9166,
+      "step": 38699
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7302946195726574,
+      "learning_rate": 4.0573499427974616e-07,
+      "loss": 0.8854,
+      "step": 38700
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5147661269488326,
+      "learning_rate": 4.055198760346901e-07,
+      "loss": 0.7821,
+      "step": 38701
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5932989418999135,
+      "learning_rate": 4.053048136517368e-07,
+      "loss": 0.8738,
+      "step": 38702
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4789575832504667,
+      "learning_rate": 4.050898071321363e-07,
+      "loss": 0.7122,
+      "step": 38703
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.960440406298516,
+      "learning_rate": 4.048748564771399e-07,
+      "loss": 0.9189,
+      "step": 38704
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.526878316543331,
+      "learning_rate": 4.046599616879987e-07,
+      "loss": 0.8191,
+      "step": 38705
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.693803971684796,
+      "learning_rate": 4.044451227659674e-07,
+      "loss": 0.8011,
+      "step": 38706
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5892485823265972,
+      "learning_rate": 4.0423033971229375e-07,
+      "loss": 0.8489,
+      "step": 38707
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6449339789687725,
+      "learning_rate": 4.0401561252822796e-07,
+      "loss": 0.7858,
+      "step": 38708
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5416550595790908,
+      "learning_rate": 4.0380094121502236e-07,
+      "loss": 0.777,
+      "step": 38709
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.664833244571994,
+      "learning_rate": 4.035863257739248e-07,
+      "loss": 0.8082,
+      "step": 38710
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7241456353278302,
+      "learning_rate": 4.0337176620618756e-07,
+      "loss": 0.8402,
+      "step": 38711
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8460388140238468,
+      "learning_rate": 4.031572625130564e-07,
+      "loss": 0.8036,
+      "step": 38712
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.901942109030208,
+      "learning_rate": 4.0294281469578364e-07,
+      "loss": 0.8042,
+      "step": 38713
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7517194528445472,
+      "learning_rate": 4.0272842275561385e-07,
+      "loss": 0.6955,
+      "step": 38714
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5685153897834347,
+      "learning_rate": 4.025140866937993e-07,
+      "loss": 0.9619,
+      "step": 38715
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.590972932568192,
+      "learning_rate": 4.0229980651158686e-07,
+      "loss": 0.8622,
+      "step": 38716
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.327425772214256,
+      "learning_rate": 4.020855822102221e-07,
+      "loss": 0.8548,
+      "step": 38717
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6822247505219752,
+      "learning_rate": 4.0187141379095294e-07,
+      "loss": 0.8061,
+      "step": 38718
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5203416028501093,
+      "learning_rate": 4.0165730125502956e-07,
+      "loss": 0.8161,
+      "step": 38719
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7245338276981717,
+      "learning_rate": 4.0144324460369535e-07,
+      "loss": 0.7341,
+      "step": 38720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7277040145394142,
+      "learning_rate": 4.012292438381971e-07,
+      "loss": 0.8526,
+      "step": 38721
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5442656905876886,
+      "learning_rate": 4.010152989597804e-07,
+      "loss": 0.6824,
+      "step": 38722
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.533934119095017,
+      "learning_rate": 4.008014099696922e-07,
+      "loss": 0.8487,
+      "step": 38723
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5065169538747547,
+      "learning_rate": 4.0058757686917805e-07,
+      "loss": 0.7885,
+      "step": 38724
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5341131884460706,
+      "learning_rate": 4.003737996594803e-07,
+      "loss": 0.763,
+      "step": 38725
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.618361530189871,
+      "learning_rate": 4.001600783418469e-07,
+      "loss": 0.8938,
+      "step": 38726
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6115780095304473,
+      "learning_rate": 3.999464129175201e-07,
+      "loss": 0.7899,
+      "step": 38727
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7387196536490694,
+      "learning_rate": 3.997328033877457e-07,
+      "loss": 0.7432,
+      "step": 38728
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7456619605909318,
+      "learning_rate": 3.995192497537659e-07,
+      "loss": 0.6704,
+      "step": 38729
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.693583178418717,
+      "learning_rate": 3.993057520168242e-07,
+      "loss": 0.9209,
+      "step": 38730
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6971356812656455,
+      "learning_rate": 3.99092310178163e-07,
+      "loss": 0.8033,
+      "step": 38731
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4102524069692977,
+      "learning_rate": 3.988789242390279e-07,
+      "loss": 0.7851,
+      "step": 38732
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.278770801307642,
+      "learning_rate": 3.986655942006579e-07,
+      "loss": 0.8098,
+      "step": 38733
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.611010735475056,
+      "learning_rate": 3.9845232006429755e-07,
+      "loss": 0.8559,
+      "step": 38734
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.944093671183058,
+      "learning_rate": 3.98239101831186e-07,
+      "loss": 0.887,
+      "step": 38735
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6884262103377219,
+      "learning_rate": 3.9802593950256875e-07,
+      "loss": 0.9658,
+      "step": 38736
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3576125711135205,
+      "learning_rate": 3.9781283307968266e-07,
+      "loss": 0.8167,
+      "step": 38737
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.484651420957775,
+      "learning_rate": 3.9759978256377006e-07,
+      "loss": 0.8114,
+      "step": 38738
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4964714369779974,
+      "learning_rate": 3.973867879560722e-07,
+      "loss": 0.7846,
+      "step": 38739
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.766220926191087,
+      "learning_rate": 3.9717384925782697e-07,
+      "loss": 0.6922,
+      "step": 38740
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5295828173086388,
+      "learning_rate": 3.9696096647028004e-07,
+      "loss": 0.8343,
+      "step": 38741
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6018845522393619,
+      "learning_rate": 3.9674813959466263e-07,
+      "loss": 0.9155,
+      "step": 38742
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.593995825950233,
+      "learning_rate": 3.965353686322193e-07,
+      "loss": 0.865,
+      "step": 38743
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.996456011538213,
+      "learning_rate": 3.9632265358418796e-07,
+      "loss": 0.7628,
+      "step": 38744
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4957045258205761,
+      "learning_rate": 3.961099944518065e-07,
+      "loss": 0.714,
+      "step": 38745
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7600013066582962,
+      "learning_rate": 3.9589739123631177e-07,
+      "loss": 0.6905,
+      "step": 38746
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.783862866278725,
+      "learning_rate": 3.956848439389427e-07,
+      "loss": 0.7397,
+      "step": 38747
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.722157249835854,
+      "learning_rate": 3.954723525609394e-07,
+      "loss": 0.8495,
+      "step": 38748
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.470265632523063,
+      "learning_rate": 3.952599171035365e-07,
+      "loss": 0.8162,
+      "step": 38749
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.896697333661784,
+      "learning_rate": 3.950475375679697e-07,
+      "loss": 0.8584,
+      "step": 38750
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6121592983567714,
+      "learning_rate": 3.9483521395547895e-07,
+      "loss": 0.7164,
+      "step": 38751
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6459372706060145,
+      "learning_rate": 3.946229462672957e-07,
+      "loss": 0.8186,
+      "step": 38752
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.818742074462186,
+      "learning_rate": 3.9441073450465997e-07,
+      "loss": 0.9254,
+      "step": 38753
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8203868060822705,
+      "learning_rate": 3.941985786688074e-07,
+      "loss": 0.7062,
+      "step": 38754
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.699101973413667,
+      "learning_rate": 3.939864787609715e-07,
+      "loss": 0.8201,
+      "step": 38755
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3190839186023102,
+      "learning_rate": 3.9377443478238575e-07,
+      "loss": 0.7103,
+      "step": 38756
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7204498016658558,
+      "learning_rate": 3.935624467342869e-07,
+      "loss": 0.7411,
+      "step": 38757
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5823825600647967,
+      "learning_rate": 3.933505146179117e-07,
+      "loss": 0.7482,
+      "step": 38758
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5666871131020477,
+      "learning_rate": 3.9313863843448817e-07,
+      "loss": 0.7421,
+      "step": 38759
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.074320266657498,
+      "learning_rate": 3.92926818185253e-07,
+      "loss": 0.9493,
+      "step": 38760
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5471284958829108,
+      "learning_rate": 3.927150538714397e-07,
+      "loss": 0.7517,
+      "step": 38761
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7083406953126326,
+      "learning_rate": 3.925033454942828e-07,
+      "loss": 0.704,
+      "step": 38762
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.827569248438599,
+      "learning_rate": 3.922916930550125e-07,
+      "loss": 0.7834,
+      "step": 38763
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5832179248582432,
+      "learning_rate": 3.920800965548599e-07,
+      "loss": 0.8082,
+      "step": 38764
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7059052944895572,
+      "learning_rate": 3.9186855599506077e-07,
+      "loss": 0.792,
+      "step": 38765
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.101648858611632,
+      "learning_rate": 3.916570713768442e-07,
+      "loss": 0.7481,
+      "step": 38766
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6553512012028415,
+      "learning_rate": 3.914456427014424e-07,
+      "loss": 0.896,
+      "step": 38767
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5273226259144577,
+      "learning_rate": 3.912342699700844e-07,
+      "loss": 0.7103,
+      "step": 38768
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5051005778434257,
+      "learning_rate": 3.9102295318400487e-07,
+      "loss": 0.8606,
+      "step": 38769
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8212866893067634,
+      "learning_rate": 3.908116923444294e-07,
+      "loss": 0.8048,
+      "step": 38770
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8450575122940425,
+      "learning_rate": 3.906004874525926e-07,
+      "loss": 0.6905,
+      "step": 38771
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.015420176049182,
+      "learning_rate": 3.903893385097224e-07,
+      "loss": 0.8629,
+      "step": 38772
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5902565945214957,
+      "learning_rate": 3.9017824551704554e-07,
+      "loss": 0.7657,
+      "step": 38773
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7145934901236248,
+      "learning_rate": 3.8996720847579437e-07,
+      "loss": 0.874,
+      "step": 38774
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.649441733715403,
+      "learning_rate": 3.8975622738719685e-07,
+      "loss": 0.8579,
+      "step": 38775
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6036518477892472,
+      "learning_rate": 3.89545302252482e-07,
+      "loss": 0.8493,
+      "step": 38776
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.066291542606889,
+      "learning_rate": 3.893344330728754e-07,
+      "loss": 0.8098,
+      "step": 38777
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7719653237959267,
+      "learning_rate": 3.891236198496062e-07,
+      "loss": 0.7219,
+      "step": 38778
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.516622812059461,
+      "learning_rate": 3.889128625839034e-07,
+      "loss": 0.813,
+      "step": 38779
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5158386690880454,
+      "learning_rate": 3.887021612769937e-07,
+      "loss": 0.7387,
+      "step": 38780
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6556849085353402,
+      "learning_rate": 3.884915159301006e-07,
+      "loss": 0.9215,
+      "step": 38781
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7647280079966106,
+      "learning_rate": 3.8828092654445314e-07,
+      "loss": 0.7156,
+      "step": 38782
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4591758507748471,
+      "learning_rate": 3.8807039312127813e-07,
+      "loss": 0.6949,
+      "step": 38783
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6447089451002148,
+      "learning_rate": 3.8785991566180017e-07,
+      "loss": 0.9486,
+      "step": 38784
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6917433138731957,
+      "learning_rate": 3.876494941672437e-07,
+      "loss": 0.8162,
+      "step": 38785
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6163235036684205,
+      "learning_rate": 3.874391286388368e-07,
+      "loss": 0.866,
+      "step": 38786
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2928800779484515,
+      "learning_rate": 3.8722881907780173e-07,
+      "loss": 0.7916,
+      "step": 38787
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7462474141226055,
+      "learning_rate": 3.870185654853642e-07,
+      "loss": 0.6473,
+      "step": 38788
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4254321403732026,
+      "learning_rate": 3.8680836786274764e-07,
+      "loss": 0.8504,
+      "step": 38789
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5626683305242204,
+      "learning_rate": 3.865982262111767e-07,
+      "loss": 0.8022,
+      "step": 38790
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5394168797647028,
+      "learning_rate": 3.863881405318737e-07,
+      "loss": 0.8053,
+      "step": 38791
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2770526885651137,
+      "learning_rate": 3.8617811082606317e-07,
+      "loss": 0.7272,
+      "step": 38792
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6441311289670422,
+      "learning_rate": 3.859681370949664e-07,
+      "loss": 0.7761,
+      "step": 38793
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9424476367586851,
+      "learning_rate": 3.8575821933980906e-07,
+      "loss": 0.828,
+      "step": 38794
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.861365224673824,
+      "learning_rate": 3.855483575618091e-07,
+      "loss": 0.8325,
+      "step": 38795
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.503616141152231,
+      "learning_rate": 3.85338551762191e-07,
+      "loss": 0.8521,
+      "step": 38796
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7926373197255621,
+      "learning_rate": 3.851288019421773e-07,
+      "loss": 0.6727,
+      "step": 38797
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.659479516905461,
+      "learning_rate": 3.8491910810298574e-07,
+      "loss": 0.679,
+      "step": 38798
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8669394065235922,
+      "learning_rate": 3.847094702458387e-07,
+      "loss": 0.8639,
+      "step": 38799
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7572799606429468,
+      "learning_rate": 3.8449988837195975e-07,
+      "loss": 0.6877,
+      "step": 38800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9153116802422852,
+      "learning_rate": 3.842903624825656e-07,
+      "loss": 0.8385,
+      "step": 38801
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5980208869115076,
+      "learning_rate": 3.840808925788764e-07,
+      "loss": 0.7661,
+      "step": 38802
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7212101703324407,
+      "learning_rate": 3.838714786621123e-07,
+      "loss": 0.8281,
+      "step": 38803
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7462603104945107,
+      "learning_rate": 3.8366212073349453e-07,
+      "loss": 0.702,
+      "step": 38804
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6455292525578469,
+      "learning_rate": 3.834528187942399e-07,
+      "loss": 0.8773,
+      "step": 38805
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.57656768701996,
+      "learning_rate": 3.8324357284556744e-07,
+      "loss": 0.875,
+      "step": 38806
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7549907627111389,
+      "learning_rate": 3.8303438288869397e-07,
+      "loss": 0.7023,
+      "step": 38807
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2585971125705435,
+      "learning_rate": 3.828252489248407e-07,
+      "loss": 0.7906,
+      "step": 38808
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.708584723523837,
+      "learning_rate": 3.826161709552245e-07,
+      "loss": 0.7515,
+      "step": 38809
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7260721868627154,
+      "learning_rate": 3.824071489810599e-07,
+      "loss": 0.7448,
+      "step": 38810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6337273760147384,
+      "learning_rate": 3.8219818300356706e-07,
+      "loss": 0.7972,
+      "step": 38811
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.936063135475273,
+      "learning_rate": 3.819892730239594e-07,
+      "loss": 0.782,
+      "step": 38812
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3689218476544533,
+      "learning_rate": 3.817804190434571e-07,
+      "loss": 0.6765,
+      "step": 38813
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5293701699050162,
+      "learning_rate": 3.815716210632736e-07,
+      "loss": 0.7678,
+      "step": 38814
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8429844506253101,
+      "learning_rate": 3.813628790846258e-07,
+      "loss": 0.7914,
+      "step": 38815
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7328023671599349,
+      "learning_rate": 3.81154193108727e-07,
+      "loss": 0.887,
+      "step": 38816
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4943653733358981,
+      "learning_rate": 3.809455631367953e-07,
+      "loss": 0.7401,
+      "step": 38817
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8343816368419714,
+      "learning_rate": 3.807369891700441e-07,
+      "loss": 0.7545,
+      "step": 38818
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4174675546457642,
+      "learning_rate": 3.805284712096857e-07,
+      "loss": 0.7841,
+      "step": 38819
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.790473683095816,
+      "learning_rate": 3.80320009256937e-07,
+      "loss": 0.7082,
+      "step": 38820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7501631534240671,
+      "learning_rate": 3.8011160331301147e-07,
+      "loss": 0.8414,
+      "step": 38821
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4618075449104166,
+      "learning_rate": 3.7990325337912136e-07,
+      "loss": 0.8641,
+      "step": 38822
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5399089359431049,
+      "learning_rate": 3.7969495945647915e-07,
+      "loss": 0.7619,
+      "step": 38823
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9868393615374838,
+      "learning_rate": 3.794867215462994e-07,
+      "loss": 0.8293,
+      "step": 38824
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4843025836094914,
+      "learning_rate": 3.7927853964979443e-07,
+      "loss": 0.8872,
+      "step": 38825
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7899712907359542,
+      "learning_rate": 3.7907041376817554e-07,
+      "loss": 0.8643,
+      "step": 38826
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5178861589637627,
+      "learning_rate": 3.7886234390265286e-07,
+      "loss": 0.7288,
+      "step": 38827
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6047887565768795,
+      "learning_rate": 3.78654330054441e-07,
+      "loss": 0.7586,
+      "step": 38828
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5481194319836844,
+      "learning_rate": 3.7844637222475e-07,
+      "loss": 0.8596,
+      "step": 38829
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.535046031993362,
+      "learning_rate": 3.782384704147901e-07,
+      "loss": 0.8242,
+      "step": 38830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5603050641678313,
+      "learning_rate": 3.7803062462577147e-07,
+      "loss": 0.7374,
+      "step": 38831
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6104632074551677,
+      "learning_rate": 3.778228348589064e-07,
+      "loss": 0.7499,
+      "step": 38832
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5949701269679515,
+      "learning_rate": 3.7761510111540054e-07,
+      "loss": 0.9377,
+      "step": 38833
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6846790614477989,
+      "learning_rate": 3.774074233964686e-07,
+      "loss": 0.744,
+      "step": 38834
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5233406803620304,
+      "learning_rate": 3.7719980170331516e-07,
+      "loss": 0.7013,
+      "step": 38835
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7587491206217623,
+      "learning_rate": 3.769922360371525e-07,
+      "loss": 0.7361,
+      "step": 38836
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5515750854096766,
+      "learning_rate": 3.767847263991853e-07,
+      "loss": 0.8108,
+      "step": 38837
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.525245987434025,
+      "learning_rate": 3.7657727279062585e-07,
+      "loss": 0.8809,
+      "step": 38838
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.099867670450357,
+      "learning_rate": 3.7636987521268097e-07,
+      "loss": 0.717,
+      "step": 38839
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7694184755539654,
+      "learning_rate": 3.761625336665564e-07,
+      "loss": 0.7237,
+      "step": 38840
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3136393089712524,
+      "learning_rate": 3.759552481534601e-07,
+      "loss": 0.7311,
+      "step": 38841
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5860373468037765,
+      "learning_rate": 3.7574801867459877e-07,
+      "loss": 0.8753,
+      "step": 38842
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5926170021053196,
+      "learning_rate": 3.755408452311826e-07,
+      "loss": 0.8864,
+      "step": 38843
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6303146029556792,
+      "learning_rate": 3.753337278244118e-07,
+      "loss": 0.8298,
+      "step": 38844
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7043751553688873,
+      "learning_rate": 3.7512666645549533e-07,
+      "loss": 0.822,
+      "step": 38845
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.460222443353159,
+      "learning_rate": 3.7491966112563893e-07,
+      "loss": 0.8564,
+      "step": 38846
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7614750332004069,
+      "learning_rate": 3.747127118360483e-07,
+      "loss": 0.7193,
+      "step": 38847
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.771478786464291,
+      "learning_rate": 3.745058185879269e-07,
+      "loss": 0.9477,
+      "step": 38848
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1922933037981713,
+      "learning_rate": 3.7429898138247933e-07,
+      "loss": 0.7834,
+      "step": 38849
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5910739015809685,
+      "learning_rate": 3.740922002209113e-07,
+      "loss": 0.8827,
+      "step": 38850
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5724795962794575,
+      "learning_rate": 3.738854751044252e-07,
+      "loss": 0.8709,
+      "step": 38851
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9061382853165318,
+      "learning_rate": 3.736788060342267e-07,
+      "loss": 0.8423,
+      "step": 38852
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.73995386292749,
+      "learning_rate": 3.7347219301151815e-07,
+      "loss": 0.6847,
+      "step": 38853
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.504043349470462,
+      "learning_rate": 3.7326563603750085e-07,
+      "loss": 0.7864,
+      "step": 38854
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.722507303362917,
+      "learning_rate": 3.7305913511337834e-07,
+      "loss": 0.8248,
+      "step": 38855
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.752614511104108,
+      "learning_rate": 3.728526902403551e-07,
+      "loss": 0.792,
+      "step": 38856
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5261224828095676,
+      "learning_rate": 3.726463014196313e-07,
+      "loss": 0.7586,
+      "step": 38857
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5800759909275612,
+      "learning_rate": 3.724399686524072e-07,
+      "loss": 0.9994,
+      "step": 38858
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4435175225223453,
+      "learning_rate": 3.722336919398861e-07,
+      "loss": 0.8699,
+      "step": 38859
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.808866219093435,
+      "learning_rate": 3.7202747128326943e-07,
+      "loss": 0.8345,
+      "step": 38860
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7086966643996795,
+      "learning_rate": 3.7182130668375726e-07,
+      "loss": 0.9089,
+      "step": 38861
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.475394429944467,
+      "learning_rate": 3.7161519814254864e-07,
+      "loss": 0.8306,
+      "step": 38862
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4743578083617537,
+      "learning_rate": 3.7140914566084376e-07,
+      "loss": 0.7754,
+      "step": 38863
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6709557918601818,
+      "learning_rate": 3.712031492398449e-07,
+      "loss": 0.9045,
+      "step": 38864
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7671358745694036,
+      "learning_rate": 3.7099720888075006e-07,
+      "loss": 0.766,
+      "step": 38865
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.402898043676994,
+      "learning_rate": 3.7079132458475606e-07,
+      "loss": 0.6632,
+      "step": 38866
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7173635800544866,
+      "learning_rate": 3.705854963530653e-07,
+      "loss": 0.8515,
+      "step": 38867
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.661522625885908,
+      "learning_rate": 3.703797241868734e-07,
+      "loss": 0.8515,
+      "step": 38868
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4586168704281364,
+      "learning_rate": 3.7017400808737946e-07,
+      "loss": 0.8266,
+      "step": 38869
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6794511161057497,
+      "learning_rate": 3.699683480557814e-07,
+      "loss": 0.8266,
+      "step": 38870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7663678634211132,
+      "learning_rate": 3.6976274409327717e-07,
+      "loss": 0.6714,
+      "step": 38871
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4964991037230015,
+      "learning_rate": 3.6955719620106135e-07,
+      "loss": 0.8777,
+      "step": 38872
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.147398394684661,
+      "learning_rate": 3.693517043803341e-07,
+      "loss": 0.9231,
+      "step": 38873
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7728308609866656,
+      "learning_rate": 3.6914626863229e-07,
+      "loss": 0.8344,
+      "step": 38874
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7715799761340311,
+      "learning_rate": 3.689408889581247e-07,
+      "loss": 0.7134,
+      "step": 38875
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7514160590201189,
+      "learning_rate": 3.687355653590341e-07,
+      "loss": 0.8254,
+      "step": 38876
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.712980284843317,
+      "learning_rate": 3.6853029783621596e-07,
+      "loss": 0.8433,
+      "step": 38877
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5986332977115918,
+      "learning_rate": 3.6832508639086386e-07,
+      "loss": 0.7819,
+      "step": 38878
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7493122550332876,
+      "learning_rate": 3.681199310241701e-07,
+      "loss": 0.9063,
+      "step": 38879
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6330528477727226,
+      "learning_rate": 3.6791483173733267e-07,
+      "loss": 0.9357,
+      "step": 38880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5944633707806095,
+      "learning_rate": 3.67709788531545e-07,
+      "loss": 0.8721,
+      "step": 38881
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.695665721597778,
+      "learning_rate": 3.675048014080007e-07,
+      "loss": 0.9213,
+      "step": 38882
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5913263680703809,
+      "learning_rate": 3.6729987036789205e-07,
+      "loss": 0.9454,
+      "step": 38883
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7842777664819713,
+      "learning_rate": 3.6709499541241367e-07,
+      "loss": 0.8704,
+      "step": 38884
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8318507928293203,
+      "learning_rate": 3.6689017654275905e-07,
+      "loss": 0.6718,
+      "step": 38885
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5100547438825127,
+      "learning_rate": 3.6668541376011836e-07,
+      "loss": 0.8106,
+      "step": 38886
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9831812551468968,
+      "learning_rate": 3.664807070656851e-07,
+      "loss": 0.8019,
+      "step": 38887
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5388106083496942,
+      "learning_rate": 3.6627605646065157e-07,
+      "loss": 0.8497,
+      "step": 38888
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.752996216510332,
+      "learning_rate": 3.66071461946208e-07,
+      "loss": 0.7889,
+      "step": 38889
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7431902811672548,
+      "learning_rate": 3.6586692352354793e-07,
+      "loss": 0.8058,
+      "step": 38890
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6961948264316928,
+      "learning_rate": 3.656624411938592e-07,
+      "loss": 0.8238,
+      "step": 38891
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7588575544985878,
+      "learning_rate": 3.6545801495833534e-07,
+      "loss": 0.7804,
+      "step": 38892
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8088791656714305,
+      "learning_rate": 3.6525364481816315e-07,
+      "loss": 0.7444,
+      "step": 38893
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5603746283429583,
+      "learning_rate": 3.650493307745362e-07,
+      "loss": 0.9396,
+      "step": 38894
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6488666401502203,
+      "learning_rate": 3.6484507282864233e-07,
+      "loss": 0.7977,
+      "step": 38895
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7596363004792275,
+      "learning_rate": 3.646408709816707e-07,
+      "loss": 0.8536,
+      "step": 38896
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7346215507482527,
+      "learning_rate": 3.644367252348091e-07,
+      "loss": 0.8667,
+      "step": 38897
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7871570665989687,
+      "learning_rate": 3.6423263558924895e-07,
+      "loss": 0.7704,
+      "step": 38898
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.734388084445497,
+      "learning_rate": 3.64028602046177e-07,
+      "loss": 0.8612,
+      "step": 38899
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6546256792665963,
+      "learning_rate": 3.6382462460678113e-07,
+      "loss": 0.7314,
+      "step": 38900
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7739781953860398,
+      "learning_rate": 3.636207032722483e-07,
+      "loss": 0.8589,
+      "step": 38901
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5840315976009367,
+      "learning_rate": 3.634168380437675e-07,
+      "loss": 0.8561,
+      "step": 38902
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6992385035495712,
+      "learning_rate": 3.632130289225244e-07,
+      "loss": 0.8445,
+      "step": 38903
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8970641638803336,
+      "learning_rate": 3.630092759097059e-07,
+      "loss": 0.8367,
+      "step": 38904
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5662702059663172,
+      "learning_rate": 3.628055790064977e-07,
+      "loss": 0.9449,
+      "step": 38905
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.678260218236293,
+      "learning_rate": 3.6260193821408776e-07,
+      "loss": 0.8278,
+      "step": 38906
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8533429779782773,
+      "learning_rate": 3.623983535336606e-07,
+      "loss": 0.8886,
+      "step": 38907
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8098842354955045,
+      "learning_rate": 3.621948249663998e-07,
+      "loss": 0.7353,
+      "step": 38908
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5696894884136656,
+      "learning_rate": 3.6199135251349324e-07,
+      "loss": 0.7099,
+      "step": 38909
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.634315979034748,
+      "learning_rate": 3.617879361761245e-07,
+      "loss": 0.8131,
+      "step": 38910
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5503579138684005,
+      "learning_rate": 3.615845759554781e-07,
+      "loss": 0.9831,
+      "step": 38911
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.577363620943049,
+      "learning_rate": 3.613812718527365e-07,
+      "loss": 0.8099,
+      "step": 38912
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5428430094065224,
+      "learning_rate": 3.611780238690854e-07,
+      "loss": 0.8711,
+      "step": 38913
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6194900422833203,
+      "learning_rate": 3.6097483200570714e-07,
+      "loss": 0.9075,
+      "step": 38914
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9094747497951647,
+      "learning_rate": 3.6077169626378526e-07,
+      "loss": 0.8347,
+      "step": 38915
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.013363770904225,
+      "learning_rate": 3.6056861664450214e-07,
+      "loss": 0.8386,
+      "step": 38916
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5366397714441886,
+      "learning_rate": 3.603655931490413e-07,
+      "loss": 0.9435,
+      "step": 38917
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5533919181662528,
+      "learning_rate": 3.6016262577858174e-07,
+      "loss": 0.9277,
+      "step": 38918
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5461819777143715,
+      "learning_rate": 3.599597145343092e-07,
+      "loss": 0.9164,
+      "step": 38919
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4501242882141165,
+      "learning_rate": 3.597568594174028e-07,
+      "loss": 0.8552,
+      "step": 38920
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7404077496013795,
+      "learning_rate": 3.595540604290437e-07,
+      "loss": 0.634,
+      "step": 38921
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4834506543306831,
+      "learning_rate": 3.593513175704122e-07,
+      "loss": 0.8539,
+      "step": 38922
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.741451540874306,
+      "learning_rate": 3.5914863084268946e-07,
+      "loss": 0.727,
+      "step": 38923
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.515813762069786,
+      "learning_rate": 3.5894600024705904e-07,
+      "loss": 0.8391,
+      "step": 38924
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.619753478221728,
+      "learning_rate": 3.5874342578469337e-07,
+      "loss": 0.8377,
+      "step": 38925
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9623370914628078,
+      "learning_rate": 3.5854090745677696e-07,
+      "loss": 0.7723,
+      "step": 38926
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6583134716933687,
+      "learning_rate": 3.583384452644889e-07,
+      "loss": 0.8536,
+      "step": 38927
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.739862128893046,
+      "learning_rate": 3.581360392090072e-07,
+      "loss": 0.7121,
+      "step": 38928
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5695991722627096,
+      "learning_rate": 3.579336892915075e-07,
+      "loss": 0.7355,
+      "step": 38929
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5199365446908009,
+      "learning_rate": 3.5773139551317226e-07,
+      "loss": 0.8798,
+      "step": 38930
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0290314038611297,
+      "learning_rate": 3.5752915787517716e-07,
+      "loss": 0.8251,
+      "step": 38931
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7837928167624413,
+      "learning_rate": 3.573269763787013e-07,
+      "loss": 0.9038,
+      "step": 38932
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.606520133609534,
+      "learning_rate": 3.571248510249181e-07,
+      "loss": 0.8349,
+      "step": 38933
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.848812536925586,
+      "learning_rate": 3.5692278181500893e-07,
+      "loss": 0.8824,
+      "step": 38934
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3617375546303312,
+      "learning_rate": 3.5672076875014727e-07,
+      "loss": 0.7904,
+      "step": 38935
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6011072620316302,
+      "learning_rate": 3.5651881183150995e-07,
+      "loss": 1.0171,
+      "step": 38936
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.592528575972055,
+      "learning_rate": 3.563169110602749e-07,
+      "loss": 0.8831,
+      "step": 38937
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.796497908105861,
+      "learning_rate": 3.5611506643761453e-07,
+      "loss": 0.8114,
+      "step": 38938
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7973426003600723,
+      "learning_rate": 3.5591327796470567e-07,
+      "loss": 0.683,
+      "step": 38939
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9980337579876695,
+      "learning_rate": 3.5571154564272293e-07,
+      "loss": 0.8656,
+      "step": 38940
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.783074913672439,
+      "learning_rate": 3.5550986947284205e-07,
+      "loss": 0.8457,
+      "step": 38941
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0198348210786925,
+      "learning_rate": 3.553082494562354e-07,
+      "loss": 0.8998,
+      "step": 38942
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.58366283513118,
+      "learning_rate": 3.5510668559407656e-07,
+      "loss": 0.7914,
+      "step": 38943
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8601664022510744,
+      "learning_rate": 3.549051778875401e-07,
+      "loss": 0.8399,
+      "step": 38944
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.464392056084396,
+      "learning_rate": 3.547037263378017e-07,
+      "loss": 0.8084,
+      "step": 38945
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5616828352026575,
+      "learning_rate": 3.545023309460294e-07,
+      "loss": 0.8341,
+      "step": 38946
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7428281251290964,
+      "learning_rate": 3.543009917133988e-07,
+      "loss": 0.6932,
+      "step": 38947
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8164233898436029,
+      "learning_rate": 3.5409970864108247e-07,
+      "loss": 0.7694,
+      "step": 38948
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5691131321879705,
+      "learning_rate": 3.538984817302504e-07,
+      "loss": 0.7946,
+      "step": 38949
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5356860115425777,
+      "learning_rate": 3.536973109820763e-07,
+      "loss": 0.7803,
+      "step": 38950
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.756691475347056,
+      "learning_rate": 3.5349619639772903e-07,
+      "loss": 0.7603,
+      "step": 38951
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9230038733675354,
+      "learning_rate": 3.532951379783822e-07,
+      "loss": 0.8368,
+      "step": 38952
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7849203354981584,
+      "learning_rate": 3.530941357252038e-07,
+      "loss": 0.8172,
+      "step": 38953
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.826075951572668,
+      "learning_rate": 3.5289318963936726e-07,
+      "loss": 0.6462,
+      "step": 38954
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4081873352194616,
+      "learning_rate": 3.526922997220406e-07,
+      "loss": 0.8303,
+      "step": 38955
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5832292100718464,
+      "learning_rate": 3.524914659743928e-07,
+      "loss": 0.8406,
+      "step": 38956
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.453649907352256,
+      "learning_rate": 3.5229068839759297e-07,
+      "loss": 0.7962,
+      "step": 38957
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9825122211259703,
+      "learning_rate": 3.520899669928135e-07,
+      "loss": 0.8267,
+      "step": 38958
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4727822058682145,
+      "learning_rate": 3.5188930176122016e-07,
+      "loss": 0.8032,
+      "step": 38959
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6622597080653765,
+      "learning_rate": 3.5168869270398085e-07,
+      "loss": 0.8257,
+      "step": 38960
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.620800226331632,
+      "learning_rate": 3.514881398222647e-07,
+      "loss": 0.6929,
+      "step": 38961
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6695872491265802,
+      "learning_rate": 3.512876431172396e-07,
+      "loss": 0.8862,
+      "step": 38962
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4646592804071448,
+      "learning_rate": 3.5108720259007356e-07,
+      "loss": 0.8872,
+      "step": 38963
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6548723459630323,
+      "learning_rate": 3.508868182419312e-07,
+      "loss": 0.786,
+      "step": 38964
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8113528252085068,
+      "learning_rate": 3.506864900739804e-07,
+      "loss": 0.6715,
+      "step": 38965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.017227794524111,
+      "learning_rate": 3.5048621808738913e-07,
+      "loss": 0.8291,
+      "step": 38966
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8441401279993352,
+      "learning_rate": 3.502860022833221e-07,
+      "loss": 0.8286,
+      "step": 38967
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6121454920248715,
+      "learning_rate": 3.500858426629439e-07,
+      "loss": 0.879,
+      "step": 38968
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7191168119111219,
+      "learning_rate": 3.498857392274213e-07,
+      "loss": 0.7313,
+      "step": 38969
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.499878349612278,
+      "learning_rate": 3.49685691977919e-07,
+      "loss": 0.9668,
+      "step": 38970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.530037070162967,
+      "learning_rate": 3.4948570091560276e-07,
+      "loss": 0.87,
+      "step": 38971
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9270135514456306,
+      "learning_rate": 3.4928576604163487e-07,
+      "loss": 0.7673,
+      "step": 38972
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.50116404222354,
+      "learning_rate": 3.4908588735718116e-07,
+      "loss": 0.7806,
+      "step": 38973
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5052997832331745,
+      "learning_rate": 3.4888606486340515e-07,
+      "loss": 0.8146,
+      "step": 38974
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5950463612021586,
+      "learning_rate": 3.486862985614703e-07,
+      "loss": 0.8746,
+      "step": 38975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4904045524998784,
+      "learning_rate": 3.4848658845253903e-07,
+      "loss": 0.7811,
+      "step": 38976
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.779776639457406,
+      "learning_rate": 3.4828693453777375e-07,
+      "loss": 0.6466,
+      "step": 38977
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5989599727676314,
+      "learning_rate": 3.4808733681833686e-07,
+      "loss": 0.8161,
+      "step": 38978
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.226322790089279,
+      "learning_rate": 3.4788779529539297e-07,
+      "loss": 0.7251,
+      "step": 38979
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1232346329534013,
+      "learning_rate": 3.476883099701023e-07,
+      "loss": 0.7036,
+      "step": 38980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.593174633910965,
+      "learning_rate": 3.4748888084362497e-07,
+      "loss": 0.7893,
+      "step": 38981
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.588067341307136,
+      "learning_rate": 3.4728950791712343e-07,
+      "loss": 0.8046,
+      "step": 38982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4976185354471403,
+      "learning_rate": 3.47090191191759e-07,
+      "loss": 0.8065,
+      "step": 38983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.709642397135235,
+      "learning_rate": 3.4689093066869184e-07,
+      "loss": 0.9143,
+      "step": 38984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4730594904516061,
+      "learning_rate": 3.4669172634908097e-07,
+      "loss": 0.8411,
+      "step": 38985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5978453659844514,
+      "learning_rate": 3.4649257823408667e-07,
+      "loss": 0.8072,
+      "step": 38986
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6923968657413915,
+      "learning_rate": 3.462934863248701e-07,
+      "loss": 0.8365,
+      "step": 38987
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.858422065658324,
+      "learning_rate": 3.460944506225894e-07,
+      "loss": 0.803,
+      "step": 38988
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.671199087440741,
+      "learning_rate": 3.458954711284013e-07,
+      "loss": 0.7884,
+      "step": 38989
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9194421451703687,
+      "learning_rate": 3.4569654784346816e-07,
+      "loss": 0.8566,
+      "step": 38990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7281727275689471,
+      "learning_rate": 3.4549768076894473e-07,
+      "loss": 0.6829,
+      "step": 38991
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8314717661457505,
+      "learning_rate": 3.4529886990599113e-07,
+      "loss": 0.8322,
+      "step": 38992
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5724150448378067,
+      "learning_rate": 3.4510011525576317e-07,
+      "loss": 0.7817,
+      "step": 38993
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7363221773474488,
+      "learning_rate": 3.449014168194209e-07,
+      "loss": 0.6855,
+      "step": 38994
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6797205403489661,
+      "learning_rate": 3.44702774598118e-07,
+      "loss": 0.8014,
+      "step": 38995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0693632455858384,
+      "learning_rate": 3.4450418859301226e-07,
+      "loss": 0.9097,
+      "step": 38996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6972113057629667,
+      "learning_rate": 3.443056588052618e-07,
+      "loss": 0.9955,
+      "step": 38997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5341153528205715,
+      "learning_rate": 3.441071852360178e-07,
+      "loss": 0.8162,
+      "step": 38998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.200911350620671,
+      "learning_rate": 3.4390876788644056e-07,
+      "loss": 0.8436,
+      "step": 38999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9450856835333674,
+      "learning_rate": 3.437104067576835e-07,
+      "loss": 0.8677,
+      "step": 39000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6525279136506388,
+      "learning_rate": 3.4351210185090133e-07,
+      "loss": 0.7889,
+      "step": 39001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.839757982849075,
+      "learning_rate": 3.4331385316724864e-07,
+      "loss": 0.7823,
+      "step": 39002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5382969835977858,
+      "learning_rate": 3.4311566070788004e-07,
+      "loss": 0.7644,
+      "step": 39003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.542414481921269,
+      "learning_rate": 3.429175244739502e-07,
+      "loss": 0.9691,
+      "step": 39004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8328760558653522,
+      "learning_rate": 3.4271944446661153e-07,
+      "loss": 0.7107,
+      "step": 39005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5147532690270662,
+      "learning_rate": 3.4252142068701645e-07,
+      "loss": 0.8878,
+      "step": 39006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9926719795771264,
+      "learning_rate": 3.423234531363195e-07,
+      "loss": 0.7253,
+      "step": 39007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2646873051265657,
+      "learning_rate": 3.421255418156744e-07,
+      "loss": 0.8597,
+      "step": 39008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5769641302394823,
+      "learning_rate": 3.4192768672623225e-07,
+      "loss": 0.8132,
+      "step": 39009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7428062140778564,
+      "learning_rate": 3.417298878691433e-07,
+      "loss": 0.9446,
+      "step": 39010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7796738236585127,
+      "learning_rate": 3.415321452455611e-07,
+      "loss": 0.6645,
+      "step": 39011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7491127872656098,
+      "learning_rate": 3.4133445885663584e-07,
+      "loss": 0.7422,
+      "step": 39012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.788844427622589,
+      "learning_rate": 3.4113682870352103e-07,
+      "loss": 0.7718,
+      "step": 39013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0582617854479714,
+      "learning_rate": 3.409392547873636e-07,
+      "loss": 0.9522,
+      "step": 39014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6116980775500627,
+      "learning_rate": 3.4074173710931804e-07,
+      "loss": 0.8091,
+      "step": 39015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0930339214420397,
+      "learning_rate": 3.405442756705302e-07,
+      "loss": 0.7155,
+      "step": 39016
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7355917204670186,
+      "learning_rate": 3.4034687047215245e-07,
+      "loss": 0.8179,
+      "step": 39017
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6311212884538195,
+      "learning_rate": 3.40149521515335e-07,
+      "loss": 0.8911,
+      "step": 39018
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4170051851553898,
+      "learning_rate": 3.3995222880122245e-07,
+      "loss": 0.7736,
+      "step": 39019
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8229702182940566,
+      "learning_rate": 3.3975499233096734e-07,
+      "loss": 0.8506,
+      "step": 39020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6682279789670897,
+      "learning_rate": 3.395578121057164e-07,
+      "loss": 0.8215,
+      "step": 39021
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7027460682069924,
+      "learning_rate": 3.3936068812662094e-07,
+      "loss": 0.7391,
+      "step": 39022
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.448205719568731,
+      "learning_rate": 3.391636203948234e-07,
+      "loss": 0.7682,
+      "step": 39023
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.535144781617802,
+      "learning_rate": 3.3896660891147516e-07,
+      "loss": 0.8077,
+      "step": 39024
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7794988742260851,
+      "learning_rate": 3.3876965367772183e-07,
+      "loss": 0.6719,
+      "step": 39025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0347520810947044,
+      "learning_rate": 3.385727546947104e-07,
+      "loss": 0.9107,
+      "step": 39026
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.621354623602517,
+      "learning_rate": 3.3837591196358657e-07,
+      "loss": 0.6553,
+      "step": 39027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5130269905310503,
+      "learning_rate": 3.38179125485496e-07,
+      "loss": 0.8865,
+      "step": 39028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1471832682245613,
+      "learning_rate": 3.3798239526158793e-07,
+      "loss": 0.8358,
+      "step": 39029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9575134840082329,
+      "learning_rate": 3.377857212930047e-07,
+      "loss": 0.8611,
+      "step": 39030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6722035068740786,
+      "learning_rate": 3.375891035808898e-07,
+      "loss": 0.9076,
+      "step": 39031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.62253921369432,
+      "learning_rate": 3.373925421263924e-07,
+      "loss": 0.7484,
+      "step": 39032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.676652656194903,
+      "learning_rate": 3.3719603693065373e-07,
+      "loss": 0.6408,
+      "step": 39033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.521984037156949,
+      "learning_rate": 3.369995879948185e-07,
+      "loss": 0.7112,
+      "step": 39034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1137589366320846,
+      "learning_rate": 3.3680319532003234e-07,
+      "loss": 0.7478,
+      "step": 39035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.823803817101952,
+      "learning_rate": 3.366068589074378e-07,
+      "loss": 0.8724,
+      "step": 39036
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.563417535363955,
+      "learning_rate": 3.3641057875817507e-07,
+      "loss": 0.7431,
+      "step": 39037
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5755590432505815,
+      "learning_rate": 3.3621435487339095e-07,
+      "loss": 0.7284,
+      "step": 39038
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5045368274079756,
+      "learning_rate": 3.360181872542256e-07,
+      "loss": 0.8001,
+      "step": 39039
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4327609748239356,
+      "learning_rate": 3.358220759018238e-07,
+      "loss": 0.8626,
+      "step": 39040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4801063591032775,
+      "learning_rate": 3.3562602081732344e-07,
+      "loss": 0.8433,
+      "step": 39041
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1849899878733434,
+      "learning_rate": 3.3543002200186803e-07,
+      "loss": 0.8282,
+      "step": 39042
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5612101594367276,
+      "learning_rate": 3.3523407945660004e-07,
+      "loss": 0.8633,
+      "step": 39043
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8344118678972214,
+      "learning_rate": 3.350381931826585e-07,
+      "loss": 0.9355,
+      "step": 39044
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.579810338645835,
+      "learning_rate": 3.348423631811848e-07,
+      "loss": 0.8588,
+      "step": 39045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4155871473899218,
+      "learning_rate": 3.3464658945331796e-07,
+      "loss": 0.9132,
+      "step": 39046
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9001264556077682,
+      "learning_rate": 3.3445087200019934e-07,
+      "loss": 0.8837,
+      "step": 39047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.592774552391517,
+      "learning_rate": 3.3425521082296797e-07,
+      "loss": 0.7466,
+      "step": 39048
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.702226237121117,
+      "learning_rate": 3.340596059227619e-07,
+      "loss": 0.8417,
+      "step": 39049
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.758170085479274,
+      "learning_rate": 3.3386405730072237e-07,
+      "loss": 0.7517,
+      "step": 39050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.660337623933525,
+      "learning_rate": 3.336685649579852e-07,
+      "loss": 0.8074,
+      "step": 39051
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5987417642209967,
+      "learning_rate": 3.334731288956916e-07,
+      "loss": 0.8471,
+      "step": 39052
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8623570684316233,
+      "learning_rate": 3.332777491149763e-07,
+      "loss": 0.913,
+      "step": 39053
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7536995213418807,
+      "learning_rate": 3.3308242561697956e-07,
+      "loss": 0.8001,
+      "step": 39054
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7620916218682687,
+      "learning_rate": 3.328871584028359e-07,
+      "loss": 0.6867,
+      "step": 39055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7782883277062239,
+      "learning_rate": 3.326919474736856e-07,
+      "loss": 0.7119,
+      "step": 39056
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9415422830016635,
+      "learning_rate": 3.324967928306622e-07,
+      "loss": 0.7934,
+      "step": 39057
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.627373167104356,
+      "learning_rate": 3.3230169447490356e-07,
+      "loss": 0.7182,
+      "step": 39058
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.461427903436687,
+      "learning_rate": 3.3210665240754446e-07,
+      "loss": 0.8342,
+      "step": 39059
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7488008471712759,
+      "learning_rate": 3.3191166662972174e-07,
+      "loss": 0.715,
+      "step": 39060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.665755533106787,
+      "learning_rate": 3.3171673714257113e-07,
+      "loss": 0.8359,
+      "step": 39061
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7583904329845367,
+      "learning_rate": 3.3152186394722506e-07,
+      "loss": 0.6971,
+      "step": 39062
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6769790336678476,
+      "learning_rate": 3.3132704704481934e-07,
+      "loss": 0.7849,
+      "step": 39063
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4763681425389452,
+      "learning_rate": 3.3113228643649075e-07,
+      "loss": 0.8539,
+      "step": 39064
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.733248953518866,
+      "learning_rate": 3.309375821233696e-07,
+      "loss": 0.6745,
+      "step": 39065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9161668063233945,
+      "learning_rate": 3.3074293410659154e-07,
+      "loss": 0.745,
+      "step": 39066
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7937646970669513,
+      "learning_rate": 3.3054834238728795e-07,
+      "loss": 0.6535,
+      "step": 39067
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5531988497363383,
+      "learning_rate": 3.303538069665957e-07,
+      "loss": 0.7688,
+      "step": 39068
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7349215858243372,
+      "learning_rate": 3.3015932784564385e-07,
+      "loss": 0.6967,
+      "step": 39069
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6202170595129832,
+      "learning_rate": 3.2996490502556486e-07,
+      "loss": 0.8225,
+      "step": 39070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4988708525219505,
+      "learning_rate": 3.297705385074934e-07,
+      "loss": 0.962,
+      "step": 39071
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7864151433417712,
+      "learning_rate": 3.2957622829255744e-07,
+      "loss": 0.703,
+      "step": 39072
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7381198133528959,
+      "learning_rate": 3.2938197438189154e-07,
+      "loss": 0.6853,
+      "step": 39073
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.772629523863883,
+      "learning_rate": 3.2918777677662495e-07,
+      "loss": 0.8648,
+      "step": 39074
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6558032438414068,
+      "learning_rate": 3.2899363547789e-07,
+      "loss": 0.6871,
+      "step": 39075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6289246048753652,
+      "learning_rate": 3.2879955048681465e-07,
+      "loss": 0.9491,
+      "step": 39076
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7388289659570237,
+      "learning_rate": 3.286055218045303e-07,
+      "loss": 0.7081,
+      "step": 39077
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7832783163102787,
+      "learning_rate": 3.2841154943216714e-07,
+      "loss": 0.7143,
+      "step": 39078
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.628207814215404,
+      "learning_rate": 3.282176333708531e-07,
+      "loss": 0.8554,
+      "step": 39079
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1429543235815514,
+      "learning_rate": 3.2802377362171845e-07,
+      "loss": 0.8045,
+      "step": 39080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6786027049094243,
+      "learning_rate": 3.278299701858922e-07,
+      "loss": 0.9066,
+      "step": 39081
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7682584462169474,
+      "learning_rate": 3.276362230645014e-07,
+      "loss": 0.7183,
+      "step": 39082
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4835230184151023,
+      "learning_rate": 3.2744253225867386e-07,
+      "loss": 0.8436,
+      "step": 39083
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6970963323696753,
+      "learning_rate": 3.2724889776953874e-07,
+      "loss": 0.8763,
+      "step": 39084
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5277117997789766,
+      "learning_rate": 3.270553195982229e-07,
+      "loss": 0.8161,
+      "step": 39085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7664684013968583,
+      "learning_rate": 3.2686179774585434e-07,
+      "loss": 0.8286,
+      "step": 39086
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7023184107567682,
+      "learning_rate": 3.2666833221355774e-07,
+      "loss": 0.8274,
+      "step": 39087
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5967522030150063,
+      "learning_rate": 3.2647492300246e-07,
+      "loss": 0.8498,
+      "step": 39088
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.056127080907105,
+      "learning_rate": 3.262815701136901e-07,
+      "loss": 0.8482,
+      "step": 39089
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6454766732532384,
+      "learning_rate": 3.2608827354837057e-07,
+      "loss": 0.807,
+      "step": 39090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5929548035782461,
+      "learning_rate": 3.258950333076272e-07,
+      "loss": 0.7861,
+      "step": 39091
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.35039413033392,
+      "learning_rate": 3.2570184939258674e-07,
+      "loss": 0.7781,
+      "step": 39092
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.89598072046322,
+      "learning_rate": 3.255087218043718e-07,
+      "loss": 0.8594,
+      "step": 39093
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.68023919083849,
+      "learning_rate": 3.2531565054410797e-07,
+      "loss": 0.9298,
+      "step": 39094
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.446114350854323,
+      "learning_rate": 3.2512263561292e-07,
+      "loss": 0.8203,
+      "step": 39095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8119666005720863,
+      "learning_rate": 3.2492967701193144e-07,
+      "loss": 0.8413,
+      "step": 39096
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8241870284066561,
+      "learning_rate": 3.2473677474226363e-07,
+      "loss": 0.782,
+      "step": 39097
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9639044782540038,
+      "learning_rate": 3.245439288050434e-07,
+      "loss": 0.7786,
+      "step": 39098
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5864930063066096,
+      "learning_rate": 3.2435113920139094e-07,
+      "loss": 0.8545,
+      "step": 39099
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.526711072519847,
+      "learning_rate": 3.2415840593242874e-07,
+      "loss": 0.8684,
+      "step": 39100
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5077352100968229,
+      "learning_rate": 3.2396572899928037e-07,
+      "loss": 0.7513,
+      "step": 39101
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5222563466193597,
+      "learning_rate": 3.237731084030671e-07,
+      "loss": 0.8614,
+      "step": 39102
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5677733755815622,
+      "learning_rate": 3.2358054414491024e-07,
+      "loss": 0.8946,
+      "step": 39103
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6291579154595541,
+      "learning_rate": 3.2338803622593006e-07,
+      "loss": 0.7878,
+      "step": 39104
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8133347127426964,
+      "learning_rate": 3.231955846472479e-07,
+      "loss": 0.8391,
+      "step": 39105
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4695924513364995,
+      "learning_rate": 3.2300318940998607e-07,
+      "loss": 0.8717,
+      "step": 39106
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.670190662682564,
+      "learning_rate": 3.2281085051526383e-07,
+      "loss": 1.014,
+      "step": 39107
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6216074852308262,
+      "learning_rate": 3.226185679641991e-07,
+      "loss": 0.8707,
+      "step": 39108
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6914345129821369,
+      "learning_rate": 3.224263417579132e-07,
+      "loss": 0.7646,
+      "step": 39109
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6075719842624259,
+      "learning_rate": 3.2223417189752525e-07,
+      "loss": 0.6769,
+      "step": 39110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6450591572816415,
+      "learning_rate": 3.2204205838415437e-07,
+      "loss": 0.8306,
+      "step": 39111
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.494581367898153,
+      "learning_rate": 3.218500012189174e-07,
+      "loss": 0.7478,
+      "step": 39112
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.571027196280881,
+      "learning_rate": 3.216580004029357e-07,
+      "loss": 0.8851,
+      "step": 39113
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6031408587466889,
+      "learning_rate": 3.214660559373228e-07,
+      "loss": 0.7991,
+      "step": 39114
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4949613418853536,
+      "learning_rate": 3.212741678232012e-07,
+      "loss": 0.8668,
+      "step": 39115
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0446401769974956,
+      "learning_rate": 3.210823360616833e-07,
+      "loss": 0.7263,
+      "step": 39116
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5076895771637282,
+      "learning_rate": 3.2089056065388924e-07,
+      "loss": 0.8419,
+      "step": 39117
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4297307756771445,
+      "learning_rate": 3.206988416009349e-07,
+      "loss": 0.9394,
+      "step": 39118
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6666375441105419,
+      "learning_rate": 3.20507178903936e-07,
+      "loss": 0.7071,
+      "step": 39119
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7439744000990856,
+      "learning_rate": 3.203155725640106e-07,
+      "loss": 0.8195,
+      "step": 39120
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0121139854238272,
+      "learning_rate": 3.2012402258226995e-07,
+      "loss": 0.8434,
+      "step": 39121
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.597467324039293,
+      "learning_rate": 3.1993252895983205e-07,
+      "loss": 0.8238,
+      "step": 39122
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4986381439826701,
+      "learning_rate": 3.1974109169781165e-07,
+      "loss": 0.8066,
+      "step": 39123
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3537831671844578,
+      "learning_rate": 3.1954971079732554e-07,
+      "loss": 0.8803,
+      "step": 39124
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.689270608337258,
+      "learning_rate": 3.1935838625948403e-07,
+      "loss": 0.74,
+      "step": 39125
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6585491866638864,
+      "learning_rate": 3.191671180854028e-07,
+      "loss": 0.765,
+      "step": 39126
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6207292812213354,
+      "learning_rate": 3.1897590627619546e-07,
+      "loss": 0.932,
+      "step": 39127
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8344031509358711,
+      "learning_rate": 3.1878475083297667e-07,
+      "loss": 0.8016,
+      "step": 39128
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.95459704179794,
+      "learning_rate": 3.1859365175685666e-07,
+      "loss": 0.8273,
+      "step": 39129
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9035052410707425,
+      "learning_rate": 3.1840260904894896e-07,
+      "loss": 0.8406,
+      "step": 39130
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6426064071250266,
+      "learning_rate": 3.182116227103682e-07,
+      "loss": 0.8963,
+      "step": 39131
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6621380970464639,
+      "learning_rate": 3.180206927422236e-07,
+      "loss": 0.8504,
+      "step": 39132
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6499563736379346,
+      "learning_rate": 3.178298191456275e-07,
+      "loss": 0.7408,
+      "step": 39133
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.704186877838055,
+      "learning_rate": 3.176390019216924e-07,
+      "loss": 0.8146,
+      "step": 39134
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5471975157848532,
+      "learning_rate": 3.174482410715274e-07,
+      "loss": 0.6951,
+      "step": 39135
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9648854956484199,
+      "learning_rate": 3.172575365962449e-07,
+      "loss": 0.8053,
+      "step": 39136
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5287963487062548,
+      "learning_rate": 3.170668884969552e-07,
+      "loss": 0.7641,
+      "step": 39137
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6935875467633907,
+      "learning_rate": 3.168762967747674e-07,
+      "loss": 0.8266,
+      "step": 39138
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8672271910810936,
+      "learning_rate": 3.1668576143078946e-07,
+      "loss": 0.8322,
+      "step": 39139
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6899704448805695,
+      "learning_rate": 3.164952824661338e-07,
+      "loss": 0.7948,
+      "step": 39140
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5292991939081926,
+      "learning_rate": 3.163048598819096e-07,
+      "loss": 0.7789,
+      "step": 39141
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5082092579820596,
+      "learning_rate": 3.1611449367922374e-07,
+      "loss": 0.8152,
+      "step": 39142
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7664595703647201,
+      "learning_rate": 3.1592418385918424e-07,
+      "loss": 0.7078,
+      "step": 39143
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7173972523247262,
+      "learning_rate": 3.1573393042290015e-07,
+      "loss": 0.6384,
+      "step": 39144
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6643289136746222,
+      "learning_rate": 3.155437333714795e-07,
+      "loss": 0.8557,
+      "step": 39145
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.492395556962361,
+      "learning_rate": 3.1535359270603027e-07,
+      "loss": 0.8359,
+      "step": 39146
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5098009000701738,
+      "learning_rate": 3.151635084276572e-07,
+      "loss": 0.7386,
+      "step": 39147
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7547387311234774,
+      "learning_rate": 3.1497348053746934e-07,
+      "loss": 0.6961,
+      "step": 39148
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5805808052450645,
+      "learning_rate": 3.1478350903657026e-07,
+      "loss": 0.8142,
+      "step": 39149
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7799186641892143,
+      "learning_rate": 3.1459359392607025e-07,
+      "loss": 0.7061,
+      "step": 39150
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6660916931148277,
+      "learning_rate": 3.1440373520707055e-07,
+      "loss": 0.94,
+      "step": 39151
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5937611150770339,
+      "learning_rate": 3.1421393288068035e-07,
+      "loss": 0.8767,
+      "step": 39152
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.830031205237916,
+      "learning_rate": 3.1402418694800207e-07,
+      "loss": 0.7734,
+      "step": 39153
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4592181973533707,
+      "learning_rate": 3.138344974101426e-07,
+      "loss": 0.7351,
+      "step": 39154
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4809287923714012,
+      "learning_rate": 3.136448642682044e-07,
+      "loss": 0.802,
+      "step": 39155
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7826914817642708,
+      "learning_rate": 3.1345528752329214e-07,
+      "loss": 0.6556,
+      "step": 39156
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5385014574826295,
+      "learning_rate": 3.132657671765105e-07,
+      "loss": 0.8361,
+      "step": 39157
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8554815971406584,
+      "learning_rate": 3.130763032289619e-07,
+      "loss": 0.7247,
+      "step": 39158
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.816764369820725,
+      "learning_rate": 3.12886895681751e-07,
+      "loss": 0.7856,
+      "step": 39159
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4706727593654128,
+      "learning_rate": 3.126975445359781e-07,
+      "loss": 0.8221,
+      "step": 39160
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6777029923752953,
+      "learning_rate": 3.1250824979274675e-07,
+      "loss": 0.814,
+      "step": 39161
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.681071820949613,
+      "learning_rate": 3.1231901145316156e-07,
+      "loss": 0.9753,
+      "step": 39162
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6382957379604843,
+      "learning_rate": 3.1212982951832063e-07,
+      "loss": 0.8655,
+      "step": 39163
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5292144872483677,
+      "learning_rate": 3.119407039893274e-07,
+      "loss": 0.7889,
+      "step": 39164
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7155718499348125,
+      "learning_rate": 3.1175163486728113e-07,
+      "loss": 0.921,
+      "step": 39165
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5753298688206763,
+      "learning_rate": 3.115626221532864e-07,
+      "loss": 0.9042,
+      "step": 39166
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9998393916880706,
+      "learning_rate": 3.113736658484412e-07,
+      "loss": 0.8247,
+      "step": 39167
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4311588765546432,
+      "learning_rate": 3.1118476595384363e-07,
+      "loss": 0.8615,
+      "step": 39168
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0720222229471394,
+      "learning_rate": 3.1099592247059715e-07,
+      "loss": 0.7534,
+      "step": 39169
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5821296742038418,
+      "learning_rate": 3.108071353997999e-07,
+      "loss": 0.8698,
+      "step": 39170
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6150420187449528,
+      "learning_rate": 3.10618404742552e-07,
+      "loss": 0.774,
+      "step": 39171
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8077452405221149,
+      "learning_rate": 3.1042973049994927e-07,
+      "loss": 0.6584,
+      "step": 39172
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.418768876704808,
+      "learning_rate": 3.102411126730942e-07,
+      "loss": 0.7096,
+      "step": 39173
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8646579104714125,
+      "learning_rate": 3.1005255126308255e-07,
+      "loss": 0.7913,
+      "step": 39174
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5818986159742512,
+      "learning_rate": 3.098640462710123e-07,
+      "loss": 0.7772,
+      "step": 39175
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6029170698936421,
+      "learning_rate": 3.096755976979815e-07,
+      "loss": 0.9414,
+      "step": 39176
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5495436103860476,
+      "learning_rate": 3.094872055450882e-07,
+      "loss": 0.8161,
+      "step": 39177
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4041778726004521,
+      "learning_rate": 3.0929886981342694e-07,
+      "loss": 0.836,
+      "step": 39178
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6327665629952093,
+      "learning_rate": 3.0911059050409696e-07,
+      "loss": 0.8931,
+      "step": 39179
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7657325195335984,
+      "learning_rate": 3.0892236761819297e-07,
+      "loss": 0.687,
+      "step": 39180
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7761204658714679,
+      "learning_rate": 3.0873420115680955e-07,
+      "loss": 0.6764,
+      "step": 39181
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7086356373279312,
+      "learning_rate": 3.085460911210447e-07,
+      "loss": 0.8358,
+      "step": 39182
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.530424392910183,
+      "learning_rate": 3.083580375119932e-07,
+      "loss": 0.77,
+      "step": 39183
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5249158121072142,
+      "learning_rate": 3.0817004033075083e-07,
+      "loss": 0.7706,
+      "step": 39184
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.74625768991306,
+      "learning_rate": 3.0798209957840886e-07,
+      "loss": 0.7249,
+      "step": 39185
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.022831793731361,
+      "learning_rate": 3.077942152560631e-07,
+      "loss": 0.8311,
+      "step": 39186
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.728539511289655,
+      "learning_rate": 3.076063873648094e-07,
+      "loss": 0.6497,
+      "step": 39187
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7752588291238873,
+      "learning_rate": 3.074186159057402e-07,
+      "loss": 0.8516,
+      "step": 39188
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6919020088038816,
+      "learning_rate": 3.0723090087994676e-07,
+      "loss": 0.7851,
+      "step": 39189
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.66721594071558,
+      "learning_rate": 3.070432422885239e-07,
+      "loss": 0.8945,
+      "step": 39190
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5495510212590595,
+      "learning_rate": 3.068556401325651e-07,
+      "loss": 0.8295,
+      "step": 39191
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7107157217653048,
+      "learning_rate": 3.066680944131617e-07,
+      "loss": 0.9107,
+      "step": 39192
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6561199479618254,
+      "learning_rate": 3.06480605131404e-07,
+      "loss": 0.8162,
+      "step": 39193
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3618455519468948,
+      "learning_rate": 3.062931722883866e-07,
+      "loss": 0.7999,
+      "step": 39194
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5687082151063145,
+      "learning_rate": 3.0610579588519874e-07,
+      "loss": 0.8376,
+      "step": 39195
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6981881299396684,
+      "learning_rate": 3.0591847592293277e-07,
+      "loss": 0.8908,
+      "step": 39196
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.526833058617219,
+      "learning_rate": 3.057312124026768e-07,
+      "loss": 0.793,
+      "step": 39197
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7094109011472967,
+      "learning_rate": 3.055440053255243e-07,
+      "loss": 0.7241,
+      "step": 39198
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6385340914547961,
+      "learning_rate": 3.0535685469256336e-07,
+      "loss": 0.8207,
+      "step": 39199
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.642533814674413,
+      "learning_rate": 3.051697605048842e-07,
+      "loss": 0.8134,
+      "step": 39200
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.75756573743104,
+      "learning_rate": 3.049827227635771e-07,
+      "loss": 0.8775,
+      "step": 39201
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7374217930771716,
+      "learning_rate": 3.0479574146972777e-07,
+      "loss": 0.6993,
+      "step": 39202
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4563854483946215,
+      "learning_rate": 3.046088166244276e-07,
+      "loss": 0.8254,
+      "step": 39203
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8337149799146866,
+      "learning_rate": 3.044219482287647e-07,
+      "loss": 0.6848,
+      "step": 39204
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.61486558820623,
+      "learning_rate": 3.0423513628382805e-07,
+      "loss": 0.8399,
+      "step": 39205
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4427420409576088,
+      "learning_rate": 3.040483807907024e-07,
+      "loss": 0.681,
+      "step": 39206
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5412747280460193,
+      "learning_rate": 3.0386168175047694e-07,
+      "loss": 0.9074,
+      "step": 39207
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4783147590171661,
+      "learning_rate": 3.0367503916423956e-07,
+      "loss": 0.781,
+      "step": 39208
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6009645258133811,
+      "learning_rate": 3.034884530330762e-07,
+      "loss": 0.7482,
+      "step": 39209
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.609103074960357,
+      "learning_rate": 3.033019233580714e-07,
+      "loss": 0.854,
+      "step": 39210
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5951292187320905,
+      "learning_rate": 3.0311545014031327e-07,
+      "loss": 0.8143,
+      "step": 39211
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7617750993762469,
+      "learning_rate": 3.0292903338088654e-07,
+      "loss": 0.7642,
+      "step": 39212
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8552720526537378,
+      "learning_rate": 3.0274267308087913e-07,
+      "loss": 0.8669,
+      "step": 39213
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9248721496429255,
+      "learning_rate": 3.0255636924137134e-07,
+      "loss": 0.7747,
+      "step": 39214
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1407460864566694,
+      "learning_rate": 3.0237012186345225e-07,
+      "loss": 0.7707,
+      "step": 39215
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8090154206772009,
+      "learning_rate": 3.021839309482033e-07,
+      "loss": 0.7378,
+      "step": 39216
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.606459213873536,
+      "learning_rate": 3.0199779649670915e-07,
+      "loss": 0.953,
+      "step": 39217
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.619517376063851,
+      "learning_rate": 3.018117185100566e-07,
+      "loss": 0.8787,
+      "step": 39218
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.558345052447104,
+      "learning_rate": 3.016256969893261e-07,
+      "loss": 0.9367,
+      "step": 39219
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4763626740486684,
+      "learning_rate": 3.0143973193559994e-07,
+      "loss": 0.8553,
+      "step": 39220
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7671075351037536,
+      "learning_rate": 3.012538233499618e-07,
+      "loss": 0.6749,
+      "step": 39221
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.669772417926058,
+      "learning_rate": 3.010679712334974e-07,
+      "loss": 0.7612,
+      "step": 39222
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7338960497472662,
+      "learning_rate": 3.008821755872826e-07,
+      "loss": 0.8376,
+      "step": 39223
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6621061833117914,
+      "learning_rate": 3.006964364124032e-07,
+      "loss": 0.8553,
+      "step": 39224
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2163840448076106,
+      "learning_rate": 3.005107537099405e-07,
+      "loss": 0.7448,
+      "step": 39225
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7290402629391226,
+      "learning_rate": 3.0032512748097595e-07,
+      "loss": 0.7313,
+      "step": 39226
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.523599943724605,
+      "learning_rate": 3.0013955772658864e-07,
+      "loss": 0.8216,
+      "step": 39227
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8252270884194517,
+      "learning_rate": 2.999540444478588e-07,
+      "loss": 0.8161,
+      "step": 39228
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8096909554686553,
+      "learning_rate": 2.997685876458689e-07,
+      "loss": 0.9245,
+      "step": 39229
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.624435676313325,
+      "learning_rate": 2.9958318732169587e-07,
+      "loss": 0.7862,
+      "step": 39230
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7066156144021938,
+      "learning_rate": 2.993978434764222e-07,
+      "loss": 0.9015,
+      "step": 39231
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7663813155216145,
+      "learning_rate": 2.9921255611112363e-07,
+      "loss": 0.7403,
+      "step": 39232
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.514399071028993,
+      "learning_rate": 2.9902732522688273e-07,
+      "loss": 0.8245,
+      "step": 39233
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6887336666072312,
+      "learning_rate": 2.988421508247741e-07,
+      "loss": 0.904,
+      "step": 39234
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6433426244516818,
+      "learning_rate": 2.986570329058791e-07,
+      "loss": 0.9548,
+      "step": 39235
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7186891289316737,
+      "learning_rate": 2.9847197147127473e-07,
+      "loss": 0.8063,
+      "step": 39236
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7760782623389682,
+      "learning_rate": 2.982869665220367e-07,
+      "loss": 0.6787,
+      "step": 39237
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7501998784160324,
+      "learning_rate": 2.981020180592431e-07,
+      "loss": 0.8075,
+      "step": 39238
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7242271574817152,
+      "learning_rate": 2.9791712608397306e-07,
+      "loss": 0.6809,
+      "step": 39239
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7078761782857746,
+      "learning_rate": 2.977322905973001e-07,
+      "loss": 0.6824,
+      "step": 39240
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.825352121199382,
+      "learning_rate": 2.975475116003013e-07,
+      "loss": 0.9713,
+      "step": 39241
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6213157655458268,
+      "learning_rate": 2.973627890940522e-07,
+      "loss": 0.7092,
+      "step": 39242
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6552318248469413,
+      "learning_rate": 2.971781230796289e-07,
+      "loss": 0.7608,
+      "step": 39243
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8435508132167517,
+      "learning_rate": 2.969935135581081e-07,
+      "loss": 0.8358,
+      "step": 39244
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6834730984252602,
+      "learning_rate": 2.968089605305613e-07,
+      "loss": 0.8314,
+      "step": 39245
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.426634409107347,
+      "learning_rate": 2.966244639980642e-07,
+      "loss": 0.8904,
+      "step": 39246
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7926161720074396,
+      "learning_rate": 2.964400239616927e-07,
+      "loss": 0.8745,
+      "step": 39247
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5909056142474298,
+      "learning_rate": 2.962556404225192e-07,
+      "loss": 0.8188,
+      "step": 39248
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7457988353148022,
+      "learning_rate": 2.9607131338161733e-07,
+      "loss": 0.7882,
+      "step": 39249
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7066473330944545,
+      "learning_rate": 2.9588704284006176e-07,
+      "loss": 0.8157,
+      "step": 39250
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.602904111902288,
+      "learning_rate": 2.9570282879892166e-07,
+      "loss": 0.7334,
+      "step": 39251
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7632671129495965,
+      "learning_rate": 2.955186712592739e-07,
+      "loss": 0.8164,
+      "step": 39252
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.037241364781264,
+      "learning_rate": 2.953345702221877e-07,
+      "loss": 0.8536,
+      "step": 39253
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4872833303527733,
+      "learning_rate": 2.951505256887377e-07,
+      "loss": 0.8966,
+      "step": 39254
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.650243974217331,
+      "learning_rate": 2.9496653765999196e-07,
+      "loss": 0.8779,
+      "step": 39255
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6306163654974577,
+      "learning_rate": 2.9478260613702513e-07,
+      "loss": 0.8861,
+      "step": 39256
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6297716952441514,
+      "learning_rate": 2.9459873112090644e-07,
+      "loss": 0.8944,
+      "step": 39257
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6405277130397442,
+      "learning_rate": 2.944149126127049e-07,
+      "loss": 0.8341,
+      "step": 39258
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7730791002124482,
+      "learning_rate": 2.9423115061349315e-07,
+      "loss": 0.8081,
+      "step": 39259
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.562976496728614,
+      "learning_rate": 2.940474451243414e-07,
+      "loss": 0.7953,
+      "step": 39260
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5249528492901256,
+      "learning_rate": 2.9386379614631866e-07,
+      "loss": 0.7048,
+      "step": 39261
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5930666559157254,
+      "learning_rate": 2.93680203680492e-07,
+      "loss": 0.904,
+      "step": 39262
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.371855101260835,
+      "learning_rate": 2.934966677279327e-07,
+      "loss": 0.7541,
+      "step": 39263
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6380635779153498,
+      "learning_rate": 2.9331318828971e-07,
+      "loss": 0.6706,
+      "step": 39264
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1010029667174563,
+      "learning_rate": 2.9312976536689076e-07,
+      "loss": 0.7195,
+      "step": 39265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7958966044050488,
+      "learning_rate": 2.9294639896054187e-07,
+      "loss": 0.7778,
+      "step": 39266
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.50052749900326,
+      "learning_rate": 2.927630890717326e-07,
+      "loss": 0.8433,
+      "step": 39267
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4867146693071,
+      "learning_rate": 2.925798357015308e-07,
+      "loss": 0.7138,
+      "step": 39268
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0538062674183095,
+      "learning_rate": 2.9239663885100244e-07,
+      "loss": 0.8317,
+      "step": 39269
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7473893048118851,
+      "learning_rate": 2.922134985212133e-07,
+      "loss": 0.8766,
+      "step": 39270
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5738631785642712,
+      "learning_rate": 2.9203041471323136e-07,
+      "loss": 0.8255,
+      "step": 39271
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5915903750996727,
+      "learning_rate": 2.918473874281214e-07,
+      "loss": 0.8219,
+      "step": 39272
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7891955978046468,
+      "learning_rate": 2.9166441666695023e-07,
+      "loss": 0.7264,
+      "step": 39273
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5848100614328213,
+      "learning_rate": 2.9148150243078154e-07,
+      "loss": 0.8294,
+      "step": 39274
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6147589445107446,
+      "learning_rate": 2.912986447206811e-07,
+      "loss": 0.7753,
+      "step": 39275
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6248889903303299,
+      "learning_rate": 2.911158435377137e-07,
+      "loss": 0.8051,
+      "step": 39276
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6367638152032216,
+      "learning_rate": 2.9093309888294507e-07,
+      "loss": 0.8053,
+      "step": 39277
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7892540926132596,
+      "learning_rate": 2.907504107574377e-07,
+      "loss": 0.8577,
+      "step": 39278
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6720178492581235,
+      "learning_rate": 2.90567779162253e-07,
+      "loss": 0.8113,
+      "step": 39279
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6420282830146504,
+      "learning_rate": 2.9038520409845785e-07,
+      "loss": 0.8411,
+      "step": 39280
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9056663604473523,
+      "learning_rate": 2.9020268556711475e-07,
+      "loss": 0.7755,
+      "step": 39281
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4067333379246174,
+      "learning_rate": 2.9002022356928615e-07,
+      "loss": 0.7677,
+      "step": 39282
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0334981336063804,
+      "learning_rate": 2.898378181060324e-07,
+      "loss": 0.8527,
+      "step": 39283
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.785925540719505,
+      "learning_rate": 2.8965546917841705e-07,
+      "loss": 0.7972,
+      "step": 39284
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6851427576208455,
+      "learning_rate": 2.894731767875025e-07,
+      "loss": 0.8531,
+      "step": 39285
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7079675651386355,
+      "learning_rate": 2.892909409343503e-07,
+      "loss": 0.6983,
+      "step": 39286
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6715087139831941,
+      "learning_rate": 2.8910876162001835e-07,
+      "loss": 0.9126,
+      "step": 39287
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.761642721776016,
+      "learning_rate": 2.8892663884557024e-07,
+      "loss": 0.8855,
+      "step": 39288
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7842656715714995,
+      "learning_rate": 2.8874457261206743e-07,
+      "loss": 0.822,
+      "step": 39289
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1077684265791117,
+      "learning_rate": 2.8856256292056797e-07,
+      "loss": 0.8486,
+      "step": 39290
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1428323743506126,
+      "learning_rate": 2.883806097721298e-07,
+      "loss": 0.8013,
+      "step": 39291
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.015079931332382,
+      "learning_rate": 2.8819871316781543e-07,
+      "loss": 0.7539,
+      "step": 39292
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6556110924554115,
+      "learning_rate": 2.88016873108683e-07,
+      "loss": 0.7635,
+      "step": 39293
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.548319478019079,
+      "learning_rate": 2.8783508959579153e-07,
+      "loss": 0.9131,
+      "step": 39294
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5958473226026753,
+      "learning_rate": 2.8765336263019805e-07,
+      "loss": 0.7704,
+      "step": 39295
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8138144136154293,
+      "learning_rate": 2.874716922129628e-07,
+      "loss": 0.7921,
+      "step": 39296
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7897003969600273,
+      "learning_rate": 2.8729007834514157e-07,
+      "loss": 0.7297,
+      "step": 39297
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0564605598629897,
+      "learning_rate": 2.871085210277924e-07,
+      "loss": 0.9722,
+      "step": 39298
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.562869800396502,
+      "learning_rate": 2.869270202619723e-07,
+      "loss": 0.9893,
+      "step": 39299
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6557589323931026,
+      "learning_rate": 2.867455760487392e-07,
+      "loss": 1.0035,
+      "step": 39300
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7679654176902826,
+      "learning_rate": 2.865641883891468e-07,
+      "loss": 0.7859,
+      "step": 39301
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5697767438502717,
+      "learning_rate": 2.863828572842542e-07,
+      "loss": 0.7379,
+      "step": 39302
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6195367428074134,
+      "learning_rate": 2.862015827351172e-07,
+      "loss": 0.99,
+      "step": 39303
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7174906713260818,
+      "learning_rate": 2.860203647427873e-07,
+      "loss": 0.9407,
+      "step": 39304
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.590398555285282,
+      "learning_rate": 2.8583920330832237e-07,
+      "loss": 0.7901,
+      "step": 39305
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.044124023960372,
+      "learning_rate": 2.8565809843277726e-07,
+      "loss": 0.7998,
+      "step": 39306
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7436379378370547,
+      "learning_rate": 2.8547705011720885e-07,
+      "loss": 0.7144,
+      "step": 39307
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7510793920199832,
+      "learning_rate": 2.8529605836266625e-07,
+      "loss": 0.8719,
+      "step": 39308
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6458726588826873,
+      "learning_rate": 2.851151231702054e-07,
+      "loss": 0.7455,
+      "step": 39309
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5466154765957576,
+      "learning_rate": 2.849342445408809e-07,
+      "loss": 0.9215,
+      "step": 39310
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5951516424725953,
+      "learning_rate": 2.8475342247574535e-07,
+      "loss": 0.8445,
+      "step": 39311
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.757998410963584,
+      "learning_rate": 2.8457265697585e-07,
+      "loss": 0.6977,
+      "step": 39312
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5748876826585014,
+      "learning_rate": 2.8439194804224857e-07,
+      "loss": 0.7965,
+      "step": 39313
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7384894821260461,
+      "learning_rate": 2.842112956759946e-07,
+      "loss": 0.8171,
+      "step": 39314
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4499903915219938,
+      "learning_rate": 2.8403069987813613e-07,
+      "loss": 0.9017,
+      "step": 39315
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8687008664716835,
+      "learning_rate": 2.8385016064972905e-07,
+      "loss": 0.9154,
+      "step": 39316
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.064298659846387,
+      "learning_rate": 2.836696779918224e-07,
+      "loss": 0.7317,
+      "step": 39317
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1473331185691626,
+      "learning_rate": 2.834892519054655e-07,
+      "loss": 0.8521,
+      "step": 39318
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7485147194318262,
+      "learning_rate": 2.8330888239171074e-07,
+      "loss": 0.7208,
+      "step": 39319
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6215426323703916,
+      "learning_rate": 2.831285694516084e-07,
+      "loss": 0.7636,
+      "step": 39320
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5549297135099023,
+      "learning_rate": 2.829483130862076e-07,
+      "loss": 0.7181,
+      "step": 39321
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7951467255564842,
+      "learning_rate": 2.8276811329655764e-07,
+      "loss": 0.6882,
+      "step": 39322
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6761114930834766,
+      "learning_rate": 2.8258797008370865e-07,
+      "loss": 1.0063,
+      "step": 39323
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7673643304093025,
+      "learning_rate": 2.8240788344871094e-07,
+      "loss": 0.706,
+      "step": 39324
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.528392884154665,
+      "learning_rate": 2.822278533926082e-07,
+      "loss": 0.8221,
+      "step": 39325
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.463707600616808,
+      "learning_rate": 2.8204787991645163e-07,
+      "loss": 0.8743,
+      "step": 39326
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4335061802659619,
+      "learning_rate": 2.818679630212895e-07,
+      "loss": 0.8359,
+      "step": 39327
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7417513824099542,
+      "learning_rate": 2.816881027081697e-07,
+      "loss": 0.7181,
+      "step": 39328
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.791279704498468,
+      "learning_rate": 2.8150829897813814e-07,
+      "loss": 0.7782,
+      "step": 39329
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7646417826847692,
+      "learning_rate": 2.813285518322417e-07,
+      "loss": 0.6987,
+      "step": 39330
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5394549954398045,
+      "learning_rate": 2.811488612715274e-07,
+      "loss": 0.825,
+      "step": 39331
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7424807666140198,
+      "learning_rate": 2.80969227297041e-07,
+      "loss": 0.6678,
+      "step": 39332
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7933429297414982,
+      "learning_rate": 2.807896499098295e-07,
+      "loss": 0.8648,
+      "step": 39333
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7225536750902408,
+      "learning_rate": 2.8061012911093756e-07,
+      "loss": 0.7714,
+      "step": 39334
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.535629653574312,
+      "learning_rate": 2.804306649014099e-07,
+      "loss": 0.7636,
+      "step": 39335
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7894886143959667,
+      "learning_rate": 2.802512572822924e-07,
+      "loss": 0.8884,
+      "step": 39336
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7518986053255721,
+      "learning_rate": 2.800719062546298e-07,
+      "loss": 0.6883,
+      "step": 39337
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5936195274878107,
+      "learning_rate": 2.798926118194667e-07,
+      "loss": 0.9532,
+      "step": 39338
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6238673425839343,
+      "learning_rate": 2.797133739778435e-07,
+      "loss": 0.7352,
+      "step": 39339
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7597170651414415,
+      "learning_rate": 2.795341927308082e-07,
+      "loss": 0.6841,
+      "step": 39340
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6794967489341976,
+      "learning_rate": 2.793550680794022e-07,
+      "loss": 0.8716,
+      "step": 39341
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.517813904499785,
+      "learning_rate": 2.79176000024669e-07,
+      "loss": 0.8884,
+      "step": 39342
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8588358777038636,
+      "learning_rate": 2.789969885676491e-07,
+      "loss": 0.8973,
+      "step": 39343
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.895308028805095,
+      "learning_rate": 2.78818033709386e-07,
+      "loss": 0.7242,
+      "step": 39344
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1913462083085027,
+      "learning_rate": 2.7863913545092434e-07,
+      "loss": 0.8818,
+      "step": 39345
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5571772490532017,
+      "learning_rate": 2.784602937933023e-07,
+      "loss": 0.8285,
+      "step": 39346
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7037893202729257,
+      "learning_rate": 2.7828150873756124e-07,
+      "loss": 0.7335,
+      "step": 39347
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8255428393921398,
+      "learning_rate": 2.781027802847436e-07,
+      "loss": 0.7924,
+      "step": 39348
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6602444614262972,
+      "learning_rate": 2.779241084358908e-07,
+      "loss": 0.7704,
+      "step": 39349
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6725533827054972,
+      "learning_rate": 2.777454931920409e-07,
+      "loss": 0.8948,
+      "step": 39350
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8904513053794048,
+      "learning_rate": 2.775669345542342e-07,
+      "loss": 0.7344,
+      "step": 39351
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4485594475257009,
+      "learning_rate": 2.773884325235121e-07,
+      "loss": 0.8665,
+      "step": 39352
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7221028399042784,
+      "learning_rate": 2.772099871009115e-07,
+      "loss": 0.9541,
+      "step": 39353
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8267967761984384,
+      "learning_rate": 2.770315982874727e-07,
+      "loss": 0.7018,
+      "step": 39354
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.568171131752318,
+      "learning_rate": 2.768532660842349e-07,
+      "loss": 0.8333,
+      "step": 39355
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8839261429888903,
+      "learning_rate": 2.76674990492235e-07,
+      "loss": 0.795,
+      "step": 39356
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5370151891451969,
+      "learning_rate": 2.764967715125111e-07,
+      "loss": 0.7858,
+      "step": 39357
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6687464952069304,
+      "learning_rate": 2.763186091461034e-07,
+      "loss": 0.7604,
+      "step": 39358
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5939181789442072,
+      "learning_rate": 2.761405033940467e-07,
+      "loss": 0.878,
+      "step": 39359
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8910406414302978,
+      "learning_rate": 2.759624542573769e-07,
+      "loss": 0.9517,
+      "step": 39360
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5244721439295685,
+      "learning_rate": 2.7578446173713305e-07,
+      "loss": 0.7608,
+      "step": 39361
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6253952679130022,
+      "learning_rate": 2.7560652583435097e-07,
+      "loss": 0.8131,
+      "step": 39362
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.169574566993104,
+      "learning_rate": 2.754286465500666e-07,
+      "loss": 0.934,
+      "step": 39363
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6308368914733287,
+      "learning_rate": 2.7525082388531465e-07,
+      "loss": 0.7814,
+      "step": 39364
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7010544556142504,
+      "learning_rate": 2.7507305784113094e-07,
+      "loss": 0.7907,
+      "step": 39365
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.608826994854698,
+      "learning_rate": 2.7489534841855235e-07,
+      "loss": 0.6486,
+      "step": 39366
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4095775276200677,
+      "learning_rate": 2.747176956186115e-07,
+      "loss": 0.7405,
+      "step": 39367
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4526876160776896,
+      "learning_rate": 2.7454009944234195e-07,
+      "loss": 0.8906,
+      "step": 39368
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4734000318937874,
+      "learning_rate": 2.7436255989077844e-07,
+      "loss": 0.8243,
+      "step": 39369
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9463329952778992,
+      "learning_rate": 2.741850769649579e-07,
+      "loss": 0.815,
+      "step": 39370
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7845723387310934,
+      "learning_rate": 2.740076506659095e-07,
+      "loss": 0.7559,
+      "step": 39371
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5143549440399238,
+      "learning_rate": 2.738302809946669e-07,
+      "loss": 0.844,
+      "step": 39372
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5495478019277376,
+      "learning_rate": 2.736529679522659e-07,
+      "loss": 0.8081,
+      "step": 39373
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.648123801185013,
+      "learning_rate": 2.734757115397346e-07,
+      "loss": 0.8278,
+      "step": 39374
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.755324152466724,
+      "learning_rate": 2.732985117581077e-07,
+      "loss": 0.8686,
+      "step": 39375
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8040915645974793,
+      "learning_rate": 2.7312136860841666e-07,
+      "loss": 0.9475,
+      "step": 39376
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7928811420338253,
+      "learning_rate": 2.729442820916928e-07,
+      "loss": 0.8607,
+      "step": 39377
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9297517751486812,
+      "learning_rate": 2.7276725220896526e-07,
+      "loss": 0.7428,
+      "step": 39378
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7761975568040869,
+      "learning_rate": 2.7259027896126776e-07,
+      "loss": 0.689,
+      "step": 39379
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5454166214666076,
+      "learning_rate": 2.7241336234962943e-07,
+      "loss": 0.7629,
+      "step": 39380
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8191822077947604,
+      "learning_rate": 2.722365023750795e-07,
+      "loss": 0.7883,
+      "step": 39381
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.560852971563544,
+      "learning_rate": 2.7205969903864814e-07,
+      "loss": 0.8714,
+      "step": 39382
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5520628868907667,
+      "learning_rate": 2.7188295234136575e-07,
+      "loss": 0.8758,
+      "step": 39383
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7406701261431048,
+      "learning_rate": 2.717062622842603e-07,
+      "loss": 0.8843,
+      "step": 39384
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6833279580279208,
+      "learning_rate": 2.715296288683611e-07,
+      "loss": 0.6785,
+      "step": 39385
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6376090488898296,
+      "learning_rate": 2.7135305209469607e-07,
+      "loss": 0.8465,
+      "step": 39386
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4817206962363212,
+      "learning_rate": 2.7117653196429337e-07,
+      "loss": 0.8161,
+      "step": 39387
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6484659621727316,
+      "learning_rate": 2.7100006847818214e-07,
+      "loss": 0.8897,
+      "step": 39388
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4683474868926862,
+      "learning_rate": 2.7082366163738825e-07,
+      "loss": 0.6877,
+      "step": 39389
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6930467383226742,
+      "learning_rate": 2.7064731144293755e-07,
+      "loss": 0.8478,
+      "step": 39390
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4769261704049497,
+      "learning_rate": 2.7047101789586026e-07,
+      "loss": 0.7666,
+      "step": 39391
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5521542404434623,
+      "learning_rate": 2.702947809971812e-07,
+      "loss": 0.8973,
+      "step": 39392
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7235038349144669,
+      "learning_rate": 2.7011860074792616e-07,
+      "loss": 0.7157,
+      "step": 39393
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9126279514060567,
+      "learning_rate": 2.6994247714912103e-07,
+      "loss": 0.844,
+      "step": 39394
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.528394873955311,
+      "learning_rate": 2.697664102017905e-07,
+      "loss": 0.7397,
+      "step": 39395
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8156874923887424,
+      "learning_rate": 2.6959039990696046e-07,
+      "loss": 0.7113,
+      "step": 39396
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7633855000850475,
+      "learning_rate": 2.694144462656556e-07,
+      "loss": 0.6641,
+      "step": 39397
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6549663597502586,
+      "learning_rate": 2.692385492789018e-07,
+      "loss": 0.8039,
+      "step": 39398
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.623089208503412,
+      "learning_rate": 2.6906270894772047e-07,
+      "loss": 0.8462,
+      "step": 39399
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7608623039856224,
+      "learning_rate": 2.688869252731374e-07,
+      "loss": 0.8112,
+      "step": 39400
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.613259936291781,
+      "learning_rate": 2.687111982561763e-07,
+      "loss": 0.7956,
+      "step": 39401
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7026262904179565,
+      "learning_rate": 2.6853552789785855e-07,
+      "loss": 0.7906,
+      "step": 39402
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6263803339065868,
+      "learning_rate": 2.683599141992077e-07,
+      "loss": 0.7759,
+      "step": 39403
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7578191456583225,
+      "learning_rate": 2.6818435716124525e-07,
+      "loss": 0.7,
+      "step": 39404
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5450579305134757,
+      "learning_rate": 2.680088567849981e-07,
+      "loss": 0.7989,
+      "step": 39405
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6145314744687231,
+      "learning_rate": 2.6783341307148103e-07,
+      "loss": 0.7614,
+      "step": 39406
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.684698797384698,
+      "learning_rate": 2.6765802602171987e-07,
+      "loss": 0.8143,
+      "step": 39407
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4136186998950466,
+      "learning_rate": 2.674826956367349e-07,
+      "loss": 0.7062,
+      "step": 39408
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6995265724694442,
+      "learning_rate": 2.6730742191754864e-07,
+      "loss": 0.9069,
+      "step": 39409
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6937038687954702,
+      "learning_rate": 2.671322048651781e-07,
+      "loss": 0.7662,
+      "step": 39410
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7717804689557979,
+      "learning_rate": 2.6695704448064575e-07,
+      "loss": 0.8661,
+      "step": 39411
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6590206346522183,
+      "learning_rate": 2.6678194076497187e-07,
+      "loss": 0.8139,
+      "step": 39412
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6107103849561732,
+      "learning_rate": 2.6660689371917346e-07,
+      "loss": 0.7077,
+      "step": 39413
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0166796003267824,
+      "learning_rate": 2.6643190334427194e-07,
+      "loss": 0.7395,
+      "step": 39414
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7180695362590535,
+      "learning_rate": 2.662569696412842e-07,
+      "loss": 0.8429,
+      "step": 39415
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6273929585946247,
+      "learning_rate": 2.6608209261123176e-07,
+      "loss": 0.778,
+      "step": 39416
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8500224024632845,
+      "learning_rate": 2.659072722551304e-07,
+      "loss": 0.7454,
+      "step": 39417
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7907850192428914,
+      "learning_rate": 2.657325085739992e-07,
+      "loss": 0.7074,
+      "step": 39418
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8674814330302266,
+      "learning_rate": 2.6555780156885533e-07,
+      "loss": 0.7776,
+      "step": 39419
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7940892383666985,
+      "learning_rate": 2.6538315124071454e-07,
+      "loss": 0.6802,
+      "step": 39420
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8099677700395076,
+      "learning_rate": 2.652085575905949e-07,
+      "loss": 0.8632,
+      "step": 39421
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4840019416070602,
+      "learning_rate": 2.6503402061951456e-07,
+      "loss": 0.8006,
+      "step": 39422
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8245246265768493,
+      "learning_rate": 2.6485954032848814e-07,
+      "loss": 0.8201,
+      "step": 39423
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.842298218404409,
+      "learning_rate": 2.6468511671853047e-07,
+      "loss": 0.7579,
+      "step": 39424
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6772547000313,
+      "learning_rate": 2.645107497906574e-07,
+      "loss": 0.8157,
+      "step": 39425
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6146666329456734,
+      "learning_rate": 2.6433643954588695e-07,
+      "loss": 0.8943,
+      "step": 39426
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8505840606132644,
+      "learning_rate": 2.641621859852317e-07,
+      "loss": 0.8697,
+      "step": 39427
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.158523631274032,
+      "learning_rate": 2.6398798910970525e-07,
+      "loss": 0.8178,
+      "step": 39428
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5684652144694986,
+      "learning_rate": 2.638138489203235e-07,
+      "loss": 0.8389,
+      "step": 39429
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8419363659190313,
+      "learning_rate": 2.636397654181011e-07,
+      "loss": 0.9851,
+      "step": 39430
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.685736561495827,
+      "learning_rate": 2.6346573860404954e-07,
+      "loss": 0.7904,
+      "step": 39431
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5957644449636088,
+      "learning_rate": 2.6329176847918245e-07,
+      "loss": 0.8043,
+      "step": 39432
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6543875264660737,
+      "learning_rate": 2.631178550445146e-07,
+      "loss": 0.8246,
+      "step": 39433
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7329779632051291,
+      "learning_rate": 2.629439983010562e-07,
+      "loss": 0.6458,
+      "step": 39434
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7697243451255718,
+      "learning_rate": 2.627701982498221e-07,
+      "loss": 0.7727,
+      "step": 39435
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7291278747874033,
+      "learning_rate": 2.625964548918203e-07,
+      "loss": 0.7861,
+      "step": 39436
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7552517098131952,
+      "learning_rate": 2.6242276822806666e-07,
+      "loss": 0.7373,
+      "step": 39437
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.428421279451459,
+      "learning_rate": 2.6224913825956933e-07,
+      "loss": 0.8546,
+      "step": 39438
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6177254207765985,
+      "learning_rate": 2.6207556498734076e-07,
+      "loss": 0.9415,
+      "step": 39439
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7551884511287822,
+      "learning_rate": 2.6190204841239244e-07,
+      "loss": 0.7581,
+      "step": 39440
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7953332089141016,
+      "learning_rate": 2.6172858853573125e-07,
+      "loss": 0.6493,
+      "step": 39441
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.570679393785061,
+      "learning_rate": 2.6155518535836976e-07,
+      "loss": 0.828,
+      "step": 39442
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6912087521241275,
+      "learning_rate": 2.6138183888131827e-07,
+      "loss": 0.9045,
+      "step": 39443
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6886022921278283,
+      "learning_rate": 2.612085491055849e-07,
+      "loss": 0.8844,
+      "step": 39444
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5992212403402484,
+      "learning_rate": 2.610353160321777e-07,
+      "loss": 0.7225,
+      "step": 39445
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5486884122135125,
+      "learning_rate": 2.6086213966210697e-07,
+      "loss": 0.9172,
+      "step": 39446
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6950496272906803,
+      "learning_rate": 2.6068901999637965e-07,
+      "loss": 0.7046,
+      "step": 39447
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8630053762511025,
+      "learning_rate": 2.6051595703600494e-07,
+      "loss": 0.7975,
+      "step": 39448
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5779130084261102,
+      "learning_rate": 2.603429507819888e-07,
+      "loss": 0.7596,
+      "step": 39449
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6490425643762718,
+      "learning_rate": 2.601700012353392e-07,
+      "loss": 0.7757,
+      "step": 39450
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4951233692013364,
+      "learning_rate": 2.5999710839706537e-07,
+      "loss": 0.8059,
+      "step": 39451
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6744688386334259,
+      "learning_rate": 2.59824272268171e-07,
+      "loss": 0.8504,
+      "step": 39452
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5954394293894993,
+      "learning_rate": 2.5965149284966297e-07,
+      "loss": 0.7784,
+      "step": 39453
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5247839903272733,
+      "learning_rate": 2.594787701425483e-07,
+      "loss": 0.7743,
+      "step": 39454
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5537345895182268,
+      "learning_rate": 2.5930610414783177e-07,
+      "loss": 0.7264,
+      "step": 39455
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7267582303061937,
+      "learning_rate": 2.591334948665192e-07,
+      "loss": 0.8793,
+      "step": 39456
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.510586268633382,
+      "learning_rate": 2.589609422996142e-07,
+      "loss": 0.8669,
+      "step": 39457
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7280483513543043,
+      "learning_rate": 2.587884464481238e-07,
+      "loss": 0.9049,
+      "step": 39458
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.419365132258493,
+      "learning_rate": 2.586160073130506e-07,
+      "loss": 0.8716,
+      "step": 39459
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7350485953957941,
+      "learning_rate": 2.584436248954003e-07,
+      "loss": 0.7645,
+      "step": 39460
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7213684472598025,
+      "learning_rate": 2.582712991961744e-07,
+      "loss": 0.6294,
+      "step": 39461
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.576542622363645,
+      "learning_rate": 2.580990302163766e-07,
+      "loss": 0.7408,
+      "step": 39462
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.736410116996907,
+      "learning_rate": 2.5792681795701046e-07,
+      "loss": 0.8141,
+      "step": 39463
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.667215774808566,
+      "learning_rate": 2.577546624190797e-07,
+      "loss": 0.8965,
+      "step": 39464
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.466265593833893,
+      "learning_rate": 2.5758256360358444e-07,
+      "loss": 0.7699,
+      "step": 39465
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5410520305808504,
+      "learning_rate": 2.5741052151152855e-07,
+      "loss": 0.8916,
+      "step": 39466
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6229436088913,
+      "learning_rate": 2.5723853614391225e-07,
+      "loss": 0.7909,
+      "step": 39467
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.797852250850956,
+      "learning_rate": 2.5706660750173916e-07,
+      "loss": 0.8515,
+      "step": 39468
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6174911659551507,
+      "learning_rate": 2.568947355860085e-07,
+      "loss": 0.8297,
+      "step": 39469
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7986597078486,
+      "learning_rate": 2.567229203977206e-07,
+      "loss": 0.7404,
+      "step": 39470
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8192901720750032,
+      "learning_rate": 2.565511619378758e-07,
+      "loss": 0.8751,
+      "step": 39471
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6265463788204597,
+      "learning_rate": 2.563794602074765e-07,
+      "loss": 0.9054,
+      "step": 39472
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6808242930312942,
+      "learning_rate": 2.562078152075209e-07,
+      "loss": 0.7605,
+      "step": 39473
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.392385366330525,
+      "learning_rate": 2.560362269390071e-07,
+      "loss": 0.729,
+      "step": 39474
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.572391597109795,
+      "learning_rate": 2.5586469540293643e-07,
+      "loss": 0.7502,
+      "step": 39475
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3955850097668008,
+      "learning_rate": 2.5569322060030487e-07,
+      "loss": 0.8165,
+      "step": 39476
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7370336673932842,
+      "learning_rate": 2.555218025321138e-07,
+      "loss": 0.7736,
+      "step": 39477
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6425199116520337,
+      "learning_rate": 2.5535044119935903e-07,
+      "loss": 0.812,
+      "step": 39478
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4359999069964435,
+      "learning_rate": 2.5517913660303984e-07,
+      "loss": 0.7934,
+      "step": 39479
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.690333987881096,
+      "learning_rate": 2.5500788874415097e-07,
+      "loss": 0.738,
+      "step": 39480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.604771666588923,
+      "learning_rate": 2.5483669762369377e-07,
+      "loss": 0.8924,
+      "step": 39481
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6694396907936895,
+      "learning_rate": 2.54665563242662e-07,
+      "loss": 0.7351,
+      "step": 39482
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4589812586392334,
+      "learning_rate": 2.544944856020515e-07,
+      "loss": 0.8818,
+      "step": 39483
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.613494702698277,
+      "learning_rate": 2.543234647028603e-07,
+      "loss": 0.8104,
+      "step": 39484
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4933309600283486,
+      "learning_rate": 2.541525005460821e-07,
+      "loss": 0.6753,
+      "step": 39485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4166144466917074,
+      "learning_rate": 2.539815931327161e-07,
+      "loss": 0.7494,
+      "step": 39486
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6598488169249896,
+      "learning_rate": 2.538107424637526e-07,
+      "loss": 0.9088,
+      "step": 39487
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5858486349677121,
+      "learning_rate": 2.536399485401897e-07,
+      "loss": 0.9699,
+      "step": 39488
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7400869924241033,
+      "learning_rate": 2.5346921136301995e-07,
+      "loss": 0.7254,
+      "step": 39489
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9277545180489437,
+      "learning_rate": 2.532985309332392e-07,
+      "loss": 0.6773,
+      "step": 39490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8903337691484523,
+      "learning_rate": 2.531279072518389e-07,
+      "loss": 0.7343,
+      "step": 39491
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.735513822236833,
+      "learning_rate": 2.529573403198138e-07,
+      "loss": 0.9101,
+      "step": 39492
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5313980156216067,
+      "learning_rate": 2.5278683013815755e-07,
+      "loss": 0.8183,
+      "step": 39493
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5779977387124313,
+      "learning_rate": 2.5261637670786266e-07,
+      "loss": 0.8238,
+      "step": 39494
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4249294626406668,
+      "learning_rate": 2.524459800299206e-07,
+      "loss": 0.707,
+      "step": 39495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7198076250387409,
+      "learning_rate": 2.522756401053239e-07,
+      "loss": 0.8023,
+      "step": 39496
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6494902517450123,
+      "learning_rate": 2.521053569350651e-07,
+      "loss": 0.7398,
+      "step": 39497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8101699029682864,
+      "learning_rate": 2.5193513052013343e-07,
+      "loss": 0.7442,
+      "step": 39498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7437953192659792,
+      "learning_rate": 2.517649608615236e-07,
+      "loss": 0.9853,
+      "step": 39499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5667852089515226,
+      "learning_rate": 2.515948479602237e-07,
+      "loss": 0.7382,
+      "step": 39500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8210695803344075,
+      "learning_rate": 2.5142479181722525e-07,
+      "loss": 0.7479,
+      "step": 39501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4921471755086522,
+      "learning_rate": 2.512547924335162e-07,
+      "loss": 0.8687,
+      "step": 39502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.531638323830304,
+      "learning_rate": 2.510848498100915e-07,
+      "loss": 0.8581,
+      "step": 39503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7519985989994364,
+      "learning_rate": 2.509149639479347e-07,
+      "loss": 0.7329,
+      "step": 39504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3662621618322934,
+      "learning_rate": 2.507451348480372e-07,
+      "loss": 0.7409,
+      "step": 39505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.062534799068314,
+      "learning_rate": 2.505753625113894e-07,
+      "loss": 0.8824,
+      "step": 39506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7551755995856885,
+      "learning_rate": 2.504056469389793e-07,
+      "loss": 0.6812,
+      "step": 39507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7609996822417513,
+      "learning_rate": 2.502359881317917e-07,
+      "loss": 0.7836,
+      "step": 39508
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9911391370321105,
+      "learning_rate": 2.500663860908181e-07,
+      "loss": 0.8375,
+      "step": 39509
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6137588900094064,
+      "learning_rate": 2.4989684081704545e-07,
+      "loss": 0.8333,
+      "step": 39510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1716478516512674,
+      "learning_rate": 2.497273523114585e-07,
+      "loss": 0.9339,
+      "step": 39511
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6492539397480266,
+      "learning_rate": 2.495579205750476e-07,
+      "loss": 0.8783,
+      "step": 39512
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.789990859854414,
+      "learning_rate": 2.493885456087963e-07,
+      "loss": 0.798,
+      "step": 39513
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4796123964705388,
+      "learning_rate": 2.4921922741369286e-07,
+      "loss": 0.8062,
+      "step": 39514
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6395732705924793,
+      "learning_rate": 2.4904996599072087e-07,
+      "loss": 0.6901,
+      "step": 39515
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5454948110015796,
+      "learning_rate": 2.4888076134086834e-07,
+      "loss": 0.8617,
+      "step": 39516
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.943128920718523,
+      "learning_rate": 2.4871161346511906e-07,
+      "loss": 0.7121,
+      "step": 39517
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7393751044284376,
+      "learning_rate": 2.4854252236445664e-07,
+      "loss": 0.877,
+      "step": 39518
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3487880330962336,
+      "learning_rate": 2.4837348803986803e-07,
+      "loss": 0.647,
+      "step": 39519
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4855475825412892,
+      "learning_rate": 2.4820451049233585e-07,
+      "loss": 0.8215,
+      "step": 39520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6043245409321762,
+      "learning_rate": 2.480355897228448e-07,
+      "loss": 0.8043,
+      "step": 39521
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4757661595397822,
+      "learning_rate": 2.4786672573237744e-07,
+      "loss": 0.9227,
+      "step": 39522
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7901808388897074,
+      "learning_rate": 2.476979185219175e-07,
+      "loss": 0.8193,
+      "step": 39523
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.589497265665126,
+      "learning_rate": 2.4752916809244856e-07,
+      "loss": 0.9027,
+      "step": 39524
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7670985953768441,
+      "learning_rate": 2.4736047444495206e-07,
+      "loss": 0.6645,
+      "step": 39525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.609202814399713,
+      "learning_rate": 2.4719183758041056e-07,
+      "loss": 0.8195,
+      "step": 39526
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.421125225070306,
+      "learning_rate": 2.470232574998044e-07,
+      "loss": 0.7836,
+      "step": 39527
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5572536076869388,
+      "learning_rate": 2.468547342041183e-07,
+      "loss": 0.8835,
+      "step": 39528
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6287721457140825,
+      "learning_rate": 2.4668626769433155e-07,
+      "loss": 0.8738,
+      "step": 39529
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5480060638295583,
+      "learning_rate": 2.4651785797142447e-07,
+      "loss": 0.7406,
+      "step": 39530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.52407977581007,
+      "learning_rate": 2.4634950503637957e-07,
+      "loss": 0.8701,
+      "step": 39531
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7394464677360528,
+      "learning_rate": 2.461812088901749e-07,
+      "loss": 0.9414,
+      "step": 39532
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7046812643402354,
+      "learning_rate": 2.4601296953379094e-07,
+      "loss": 0.8043,
+      "step": 39533
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6258943302106421,
+      "learning_rate": 2.458447869682079e-07,
+      "loss": 0.7975,
+      "step": 39534
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6255828336863154,
+      "learning_rate": 2.4567666119440504e-07,
+      "loss": 0.8845,
+      "step": 39535
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6038747493791445,
+      "learning_rate": 2.455085922133604e-07,
+      "loss": 0.6948,
+      "step": 39536
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1759318525495446,
+      "learning_rate": 2.4534058002605334e-07,
+      "loss": 0.9315,
+      "step": 39537
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8355628726152604,
+      "learning_rate": 2.451726246334607e-07,
+      "loss": 0.81,
+      "step": 39538
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5228735755387042,
+      "learning_rate": 2.450047260365618e-07,
+      "loss": 0.7653,
+      "step": 39539
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7026663287877948,
+      "learning_rate": 2.4483688423633356e-07,
+      "loss": 0.9174,
+      "step": 39540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.752849499868994,
+      "learning_rate": 2.4466909923375414e-07,
+      "loss": 0.7157,
+      "step": 39541
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6191761004115737,
+      "learning_rate": 2.445013710298005e-07,
+      "loss": 0.8809,
+      "step": 39542
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4164174009347472,
+      "learning_rate": 2.443336996254464e-07,
+      "loss": 0.8361,
+      "step": 39543
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.761044863256903,
+      "learning_rate": 2.4416608502167095e-07,
+      "loss": 0.6629,
+      "step": 39544
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5918229570292441,
+      "learning_rate": 2.4399852721944894e-07,
+      "loss": 0.8064,
+      "step": 39545
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4816930180068477,
+      "learning_rate": 2.438310262197574e-07,
+      "loss": 0.9017,
+      "step": 39546
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5750412697523037,
+      "learning_rate": 2.4366358202356887e-07,
+      "loss": 0.7971,
+      "step": 39547
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.436022403152342,
+      "learning_rate": 2.4349619463186036e-07,
+      "loss": 0.8224,
+      "step": 39548
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6353550628998577,
+      "learning_rate": 2.4332886404560553e-07,
+      "loss": 0.7609,
+      "step": 39549
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7628403217953902,
+      "learning_rate": 2.431615902657802e-07,
+      "loss": 0.7054,
+      "step": 39550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7220998087209507,
+      "learning_rate": 2.4299437329335595e-07,
+      "loss": 0.918,
+      "step": 39551
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5986238928056948,
+      "learning_rate": 2.428272131293075e-07,
+      "loss": 0.8868,
+      "step": 39552
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6018415154399266,
+      "learning_rate": 2.426601097746084e-07,
+      "loss": 0.8868,
+      "step": 39553
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7475184213591418,
+      "learning_rate": 2.424930632302314e-07,
+      "loss": 0.7328,
+      "step": 39554
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.856019520767637,
+      "learning_rate": 2.4232607349714775e-07,
+      "loss": 0.9005,
+      "step": 39555
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6515750875042399,
+      "learning_rate": 2.421591405763324e-07,
+      "loss": 0.821,
+      "step": 39556
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6547017645246356,
+      "learning_rate": 2.4199226446875557e-07,
+      "loss": 0.7995,
+      "step": 39557
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5940015927127562,
+      "learning_rate": 2.4182544517538874e-07,
+      "loss": 0.9334,
+      "step": 39558
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5623934842289222,
+      "learning_rate": 2.416586826972034e-07,
+      "loss": 0.8554,
+      "step": 39559
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6748645798175443,
+      "learning_rate": 2.414919770351709e-07,
+      "loss": 0.7907,
+      "step": 39560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7898854776040197,
+      "learning_rate": 2.4132532819026165e-07,
+      "loss": 0.7103,
+      "step": 39561
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6660217474618666,
+      "learning_rate": 2.41158736163446e-07,
+      "loss": 0.9352,
+      "step": 39562
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.580927687671809,
+      "learning_rate": 2.409922009556942e-07,
+      "loss": 0.9124,
+      "step": 39563
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.639670453209559,
+      "learning_rate": 2.408257225679744e-07,
+      "loss": 0.9665,
+      "step": 39564
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4367854336210042,
+      "learning_rate": 2.40659301001257e-07,
+      "loss": 0.7805,
+      "step": 39565
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4629247247063315,
+      "learning_rate": 2.4049293625651224e-07,
+      "loss": 0.7765,
+      "step": 39566
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4821518951877228,
+      "learning_rate": 2.403266283347061e-07,
+      "loss": 0.9104,
+      "step": 39567
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7202167491011688,
+      "learning_rate": 2.4016037723680886e-07,
+      "loss": 0.9843,
+      "step": 39568
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7629125271051184,
+      "learning_rate": 2.3999418296378643e-07,
+      "loss": 0.7551,
+      "step": 39569
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5200875995513872,
+      "learning_rate": 2.398280455166091e-07,
+      "loss": 0.8167,
+      "step": 39570
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5173386422190906,
+      "learning_rate": 2.396619648962439e-07,
+      "loss": 0.7844,
+      "step": 39571
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.71284853726605,
+      "learning_rate": 2.394959411036546e-07,
+      "loss": 0.9233,
+      "step": 39572
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4210915001325928,
+      "learning_rate": 2.393299741398114e-07,
+      "loss": 0.771,
+      "step": 39573
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.68539122672156,
+      "learning_rate": 2.3916406400567915e-07,
+      "loss": 0.7822,
+      "step": 39574
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7500815379425859,
+      "learning_rate": 2.3899821070222375e-07,
+      "loss": 0.694,
+      "step": 39575
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6626773421094043,
+      "learning_rate": 2.38832414230411e-07,
+      "loss": 0.7896,
+      "step": 39576
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2468105065435817,
+      "learning_rate": 2.38666674591207e-07,
+      "loss": 0.896,
+      "step": 39577
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4097545775636389,
+      "learning_rate": 2.3850099178557517e-07,
+      "loss": 0.8218,
+      "step": 39578
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8377443500459083,
+      "learning_rate": 2.3833536581448157e-07,
+      "loss": 0.8221,
+      "step": 39579
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5678704857490926,
+      "learning_rate": 2.3816979667888872e-07,
+      "loss": 0.8743,
+      "step": 39580
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6570967866216122,
+      "learning_rate": 2.3800428437976364e-07,
+      "loss": 0.8094,
+      "step": 39581
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.760301219037152,
+      "learning_rate": 2.3783882891806664e-07,
+      "loss": 0.8746,
+      "step": 39582
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9386683230475155,
+      "learning_rate": 2.3767343029476253e-07,
+      "loss": 0.8691,
+      "step": 39583
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6351761335984545,
+      "learning_rate": 2.3750808851081607e-07,
+      "loss": 0.8529,
+      "step": 39584
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4467892115232053,
+      "learning_rate": 2.3734280356718764e-07,
+      "loss": 0.8064,
+      "step": 39585
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.604659174028939,
+      "learning_rate": 2.3717757546483866e-07,
+      "loss": 0.723,
+      "step": 39586
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5194514062165183,
+      "learning_rate": 2.370124042047339e-07,
+      "loss": 0.9399,
+      "step": 39587
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.510238309196787,
+      "learning_rate": 2.368472897878349e-07,
+      "loss": 0.794,
+      "step": 39588
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6309340338728442,
+      "learning_rate": 2.3668223221510078e-07,
+      "loss": 0.7299,
+      "step": 39589
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2359899502559664,
+      "learning_rate": 2.3651723148749305e-07,
+      "loss": 0.81,
+      "step": 39590
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5617712742794692,
+      "learning_rate": 2.3635228760597429e-07,
+      "loss": 0.8249,
+      "step": 39591
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7770379613481893,
+      "learning_rate": 2.3618740057150368e-07,
+      "loss": 0.8048,
+      "step": 39592
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7286784283294663,
+      "learning_rate": 2.360225703850405e-07,
+      "loss": 0.7417,
+      "step": 39593
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7000218771099769,
+      "learning_rate": 2.3585779704754507e-07,
+      "loss": 0.8048,
+      "step": 39594
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6250181182961108,
+      "learning_rate": 2.3569308055997776e-07,
+      "loss": 0.7481,
+      "step": 39595
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.308954120966912,
+      "learning_rate": 2.3552842092329554e-07,
+      "loss": 0.8118,
+      "step": 39596
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4534194532032605,
+      "learning_rate": 2.3536381813845878e-07,
+      "loss": 0.8692,
+      "step": 39597
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.536351809893756,
+      "learning_rate": 2.3519927220642558e-07,
+      "loss": 0.8368,
+      "step": 39598
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7369257109592022,
+      "learning_rate": 2.3503478312815298e-07,
+      "loss": 0.7902,
+      "step": 39599
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7640904053871165,
+      "learning_rate": 2.348703509045991e-07,
+      "loss": 0.6659,
+      "step": 39600
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5848964475797596,
+      "learning_rate": 2.3470597553672314e-07,
+      "loss": 0.8403,
+      "step": 39601
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6361216747923866,
+      "learning_rate": 2.3454165702547993e-07,
+      "loss": 0.8147,
+      "step": 39602
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7692237672511644,
+      "learning_rate": 2.3437739537182646e-07,
+      "loss": 0.744,
+      "step": 39603
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6268855309984773,
+      "learning_rate": 2.3421319057671975e-07,
+      "loss": 0.8281,
+      "step": 39604
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8404042743250216,
+      "learning_rate": 2.3404904264111683e-07,
+      "loss": 0.7182,
+      "step": 39605
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8840952227568672,
+      "learning_rate": 2.3388495156597136e-07,
+      "loss": 0.7091,
+      "step": 39606
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.599233022805728,
+      "learning_rate": 2.3372091735223812e-07,
+      "loss": 0.6793,
+      "step": 39607
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8043808893280233,
+      "learning_rate": 2.3355694000087525e-07,
+      "loss": 0.7235,
+      "step": 39608
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6291991213782373,
+      "learning_rate": 2.3339301951283532e-07,
+      "loss": 0.9212,
+      "step": 39609
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6641139452539195,
+      "learning_rate": 2.3322915588907425e-07,
+      "loss": 0.8266,
+      "step": 39610
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.312444482250993,
+      "learning_rate": 2.3306534913054347e-07,
+      "loss": 0.8641,
+      "step": 39611
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4866889157663512,
+      "learning_rate": 2.329015992382e-07,
+      "loss": 0.7356,
+      "step": 39612
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7199971563081989,
+      "learning_rate": 2.327379062129942e-07,
+      "loss": 0.6833,
+      "step": 39613
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7435252731502329,
+      "learning_rate": 2.3257427005588197e-07,
+      "loss": 0.7508,
+      "step": 39614
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8833997566202356,
+      "learning_rate": 2.3241069076781365e-07,
+      "loss": 0.8833,
+      "step": 39615
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7419351468725359,
+      "learning_rate": 2.3224716834974404e-07,
+      "loss": 0.7527,
+      "step": 39616
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7335331001829026,
+      "learning_rate": 2.3208370280262127e-07,
+      "loss": 0.9172,
+      "step": 39617
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5309377541690763,
+      "learning_rate": 2.3192029412740237e-07,
+      "loss": 0.8961,
+      "step": 39618
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.020159210719817,
+      "learning_rate": 2.3175694232503542e-07,
+      "loss": 0.7917,
+      "step": 39619
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7744431428384498,
+      "learning_rate": 2.3159364739647084e-07,
+      "loss": 0.7093,
+      "step": 39620
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.762533792610644,
+      "learning_rate": 2.3143040934266115e-07,
+      "loss": 0.812,
+      "step": 39621
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.792086512161345,
+      "learning_rate": 2.312672281645567e-07,
+      "loss": 0.742,
+      "step": 39622
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7722910879000363,
+      "learning_rate": 2.311041038631079e-07,
+      "loss": 0.7512,
+      "step": 39623
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.905305345484149,
+      "learning_rate": 2.3094103643926168e-07,
+      "loss": 0.7013,
+      "step": 39624
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6395252429576863,
+      "learning_rate": 2.3077802589397068e-07,
+      "loss": 0.7479,
+      "step": 39625
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0372067342605,
+      "learning_rate": 2.3061507222818303e-07,
+      "loss": 0.8237,
+      "step": 39626
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.570575130835014,
+      "learning_rate": 2.3045217544284791e-07,
+      "loss": 0.7444,
+      "step": 39627
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5435189556321545,
+      "learning_rate": 2.302893355389124e-07,
+      "loss": 0.7546,
+      "step": 39628
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5483437912228109,
+      "learning_rate": 2.3012655251732574e-07,
+      "loss": 0.8346,
+      "step": 39629
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6785128181043771,
+      "learning_rate": 2.2996382637903492e-07,
+      "loss": 0.9353,
+      "step": 39630
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6220632681558729,
+      "learning_rate": 2.2980115712498918e-07,
+      "loss": 0.7925,
+      "step": 39631
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4881187868155366,
+      "learning_rate": 2.2963854475613223e-07,
+      "loss": 0.822,
+      "step": 39632
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5538467088319008,
+      "learning_rate": 2.2947598927341442e-07,
+      "loss": 0.7648,
+      "step": 39633
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6673458971160526,
+      "learning_rate": 2.2931349067778053e-07,
+      "loss": 0.8267,
+      "step": 39634
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.643120602185788,
+      "learning_rate": 2.2915104897017648e-07,
+      "loss": 0.8531,
+      "step": 39635
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6169559293274487,
+      "learning_rate": 2.2898866415154818e-07,
+      "loss": 0.8181,
+      "step": 39636
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5671605619202311,
+      "learning_rate": 2.2882633622284268e-07,
+      "loss": 0.8081,
+      "step": 39637
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.579894003408163,
+      "learning_rate": 2.2866406518500252e-07,
+      "loss": 0.8568,
+      "step": 39638
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6991294667475985,
+      "learning_rate": 2.285018510389736e-07,
+      "loss": 0.8563,
+      "step": 39639
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.625465101180348,
+      "learning_rate": 2.2833969378570188e-07,
+      "loss": 0.8875,
+      "step": 39640
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.631347398224844,
+      "learning_rate": 2.2817759342612878e-07,
+      "loss": 0.6676,
+      "step": 39641
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5140385390538351,
+      "learning_rate": 2.2801554996119912e-07,
+      "loss": 0.8488,
+      "step": 39642
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8268447891710216,
+      "learning_rate": 2.2785356339185772e-07,
+      "loss": 0.8719,
+      "step": 39643
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7840518197322044,
+      "learning_rate": 2.2769163371904713e-07,
+      "loss": 0.8788,
+      "step": 39644
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.216192532024049,
+      "learning_rate": 2.275297609437077e-07,
+      "loss": 0.8798,
+      "step": 39645
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5278208750827695,
+      "learning_rate": 2.2736794506678539e-07,
+      "loss": 0.7311,
+      "step": 39646
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4602314031495682,
+      "learning_rate": 2.272061860892205e-07,
+      "loss": 0.6668,
+      "step": 39647
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.139673175433716,
+      "learning_rate": 2.2704448401195567e-07,
+      "loss": 0.879,
+      "step": 39648
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6640451852503213,
+      "learning_rate": 2.268828388359312e-07,
+      "loss": 0.8139,
+      "step": 39649
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5897139409966625,
+      "learning_rate": 2.267212505620886e-07,
+      "loss": 0.88,
+      "step": 39650
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7728710837107948,
+      "learning_rate": 2.2655971919137044e-07,
+      "loss": 0.709,
+      "step": 39651
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6150022484641726,
+      "learning_rate": 2.2639824472471594e-07,
+      "loss": 0.9044,
+      "step": 39652
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7319596452033514,
+      "learning_rate": 2.262368271630633e-07,
+      "loss": 0.7358,
+      "step": 39653
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4390125478309925,
+      "learning_rate": 2.2607546650735613e-07,
+      "loss": 0.9614,
+      "step": 39654
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6320187571053926,
+      "learning_rate": 2.259141627585304e-07,
+      "loss": 0.8676,
+      "step": 39655
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5611578723748942,
+      "learning_rate": 2.2575291591752756e-07,
+      "loss": 0.724,
+      "step": 39656
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7181186601023475,
+      "learning_rate": 2.2559172598528578e-07,
+      "loss": 0.7377,
+      "step": 39657
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.645449018276999,
+      "learning_rate": 2.2543059296274316e-07,
+      "loss": 0.7859,
+      "step": 39658
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5044060686801004,
+      "learning_rate": 2.2526951685083787e-07,
+      "loss": 0.7969,
+      "step": 39659
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4674521901777289,
+      "learning_rate": 2.2510849765050914e-07,
+      "loss": 0.827,
+      "step": 39660
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6052885973980688,
+      "learning_rate": 2.2494753536269288e-07,
+      "loss": 0.7525,
+      "step": 39661
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.79924612027705,
+      "learning_rate": 2.2478662998832725e-07,
+      "loss": 0.9724,
+      "step": 39662
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6168460674302472,
+      "learning_rate": 2.2462578152834812e-07,
+      "loss": 0.9528,
+      "step": 39663
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7586539570621953,
+      "learning_rate": 2.2446498998369259e-07,
+      "loss": 0.8654,
+      "step": 39664
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9320487661958121,
+      "learning_rate": 2.2430425535529765e-07,
+      "loss": 0.8256,
+      "step": 39665
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7960661596989627,
+      "learning_rate": 2.24143577644097e-07,
+      "loss": 0.6681,
+      "step": 39666
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7227955096775303,
+      "learning_rate": 2.2398295685102655e-07,
+      "loss": 0.9222,
+      "step": 39667
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7085212673026342,
+      "learning_rate": 2.2382239297702447e-07,
+      "loss": 0.9705,
+      "step": 39668
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5048115830316933,
+      "learning_rate": 2.236618860230233e-07,
+      "loss": 0.8527,
+      "step": 39669
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.834997156924784,
+      "learning_rate": 2.2350143598995676e-07,
+      "loss": 0.8122,
+      "step": 39670
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8186048197091864,
+      "learning_rate": 2.2334104287875969e-07,
+      "loss": 0.8561,
+      "step": 39671
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4334716563187027,
+      "learning_rate": 2.2318070669036686e-07,
+      "loss": 0.7217,
+      "step": 39672
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6868883350730757,
+      "learning_rate": 2.2302042742571194e-07,
+      "loss": 0.7527,
+      "step": 39673
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.788918198672213,
+      "learning_rate": 2.2286020508572648e-07,
+      "loss": 0.8427,
+      "step": 39674
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7903569102899535,
+      "learning_rate": 2.2270003967134414e-07,
+      "loss": 0.8669,
+      "step": 39675
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9045812259410255,
+      "learning_rate": 2.225399311834986e-07,
+      "loss": 1.0067,
+      "step": 39676
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5990945407613932,
+      "learning_rate": 2.2237987962312134e-07,
+      "loss": 0.9371,
+      "step": 39677
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8910368105612356,
+      "learning_rate": 2.2221988499114278e-07,
+      "loss": 0.8295,
+      "step": 39678
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5812788027884883,
+      "learning_rate": 2.2205994728849655e-07,
+      "loss": 0.741,
+      "step": 39679
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4338539400879886,
+      "learning_rate": 2.2190006651611197e-07,
+      "loss": 0.7784,
+      "step": 39680
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9291936747859437,
+      "learning_rate": 2.2174024267492155e-07,
+      "loss": 0.8546,
+      "step": 39681
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6447814734371784,
+      "learning_rate": 2.2158047576585572e-07,
+      "loss": 0.7486,
+      "step": 39682
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4829910311102061,
+      "learning_rate": 2.2142076578984372e-07,
+      "loss": 0.8663,
+      "step": 39683
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8408768423846218,
+      "learning_rate": 2.212611127478148e-07,
+      "loss": 0.7083,
+      "step": 39684
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7843485934242351,
+      "learning_rate": 2.2110151664070044e-07,
+      "loss": 0.8155,
+      "step": 39685
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5068305863608664,
+      "learning_rate": 2.209419774694299e-07,
+      "loss": 0.7907,
+      "step": 39686
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.419807392828278,
+      "learning_rate": 2.2078249523493022e-07,
+      "loss": 0.7507,
+      "step": 39687
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.80525568309523,
+      "learning_rate": 2.2062306993813064e-07,
+      "loss": 0.7985,
+      "step": 39688
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7478405342844952,
+      "learning_rate": 2.2046370157995934e-07,
+      "loss": 0.6799,
+      "step": 39689
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5453567479458152,
+      "learning_rate": 2.2030439016134664e-07,
+      "loss": 0.7433,
+      "step": 39690
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6034323681452503,
+      "learning_rate": 2.2014513568321628e-07,
+      "loss": 0.7796,
+      "step": 39691
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5874210359022276,
+      "learning_rate": 2.1998593814649637e-07,
+      "loss": 0.7807,
+      "step": 39692
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.500615277953701,
+      "learning_rate": 2.1982679755211622e-07,
+      "loss": 0.7381,
+      "step": 39693
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.659322971405397,
+      "learning_rate": 2.196677139009995e-07,
+      "loss": 0.836,
+      "step": 39694
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7584672022229448,
+      "learning_rate": 2.1950868719407548e-07,
+      "loss": 0.68,
+      "step": 39695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6944034570692394,
+      "learning_rate": 2.1934971743226674e-07,
+      "loss": 0.8994,
+      "step": 39696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5827500616203651,
+      "learning_rate": 2.1919080461650144e-07,
+      "loss": 0.8156,
+      "step": 39697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7150730690366591,
+      "learning_rate": 2.190319487477033e-07,
+      "loss": 0.7716,
+      "step": 39698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.759099224871437,
+      "learning_rate": 2.188731498267982e-07,
+      "loss": 0.812,
+      "step": 39699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5393372373277838,
+      "learning_rate": 2.1871440785470987e-07,
+      "loss": 0.8102,
+      "step": 39700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.563457945375378,
+      "learning_rate": 2.1855572283236314e-07,
+      "loss": 0.7517,
+      "step": 39701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6714432535917116,
+      "learning_rate": 2.183970947606806e-07,
+      "loss": 0.8536,
+      "step": 39702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0430805926184603,
+      "learning_rate": 2.1823852364058928e-07,
+      "loss": 0.6649,
+      "step": 39703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6120223734709442,
+      "learning_rate": 2.1808000947300843e-07,
+      "loss": 0.8529,
+      "step": 39704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8089340062825296,
+      "learning_rate": 2.179215522588629e-07,
+      "loss": 0.6661,
+      "step": 39705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6232467952430394,
+      "learning_rate": 2.1776315199907526e-07,
+      "loss": 0.7667,
+      "step": 39706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.648340714420326,
+      "learning_rate": 2.1760480869456812e-07,
+      "loss": 0.7641,
+      "step": 39707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.599443293719061,
+      "learning_rate": 2.1744652234626295e-07,
+      "loss": 0.8633,
+      "step": 39708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7994693412886308,
+      "learning_rate": 2.1728829295508013e-07,
+      "loss": 0.7257,
+      "step": 39709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5669565364620825,
+      "learning_rate": 2.1713012052194114e-07,
+      "loss": 0.7856,
+      "step": 39710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6148909303261536,
+      "learning_rate": 2.1697200504776973e-07,
+      "loss": 0.808,
+      "step": 39711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7251018587465875,
+      "learning_rate": 2.1681394653348508e-07,
+      "loss": 0.7255,
+      "step": 39712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9315185150138525,
+      "learning_rate": 2.166559449800043e-07,
+      "loss": 0.9455,
+      "step": 39713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.516049089950255,
+      "learning_rate": 2.164980003882522e-07,
+      "loss": 0.9039,
+      "step": 39714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7254714315521693,
+      "learning_rate": 2.1634011275914468e-07,
+      "loss": 0.6767,
+      "step": 39715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6932765408907056,
+      "learning_rate": 2.1618228209360438e-07,
+      "loss": 0.9231,
+      "step": 39716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6536390502150249,
+      "learning_rate": 2.1602450839254608e-07,
+      "loss": 0.8123,
+      "step": 39717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.161900377149472,
+      "learning_rate": 2.158667916568924e-07,
+      "loss": 0.8943,
+      "step": 39718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6082470711133547,
+      "learning_rate": 2.1570913188755925e-07,
+      "loss": 0.8121,
+      "step": 39719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5503078346778711,
+      "learning_rate": 2.1555152908546484e-07,
+      "loss": 0.8357,
+      "step": 39720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7306105483158254,
+      "learning_rate": 2.1539398325152839e-07,
+      "loss": 0.7633,
+      "step": 39721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.705808845083486,
+      "learning_rate": 2.1523649438666472e-07,
+      "loss": 0.8389,
+      "step": 39722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5608760351722675,
+      "learning_rate": 2.15079062491792e-07,
+      "loss": 0.8031,
+      "step": 39723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6012545548166617,
+      "learning_rate": 2.1492168756782728e-07,
+      "loss": 0.9056,
+      "step": 39724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7803047846053792,
+      "learning_rate": 2.1476436961568648e-07,
+      "loss": 0.7416,
+      "step": 39725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6308362403816297,
+      "learning_rate": 2.1460710863628443e-07,
+      "loss": 0.8979,
+      "step": 39726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.408590748940956,
+      "learning_rate": 2.1444990463053817e-07,
+      "loss": 0.8493,
+      "step": 39727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.493101902196317,
+      "learning_rate": 2.1429275759936363e-07,
+      "loss": 0.7415,
+      "step": 39728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0721588930986505,
+      "learning_rate": 2.1413566754367454e-07,
+      "loss": 0.9745,
+      "step": 39729
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8784103724038759,
+      "learning_rate": 2.139786344643846e-07,
+      "loss": 0.7933,
+      "step": 39730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6749198079512355,
+      "learning_rate": 2.1382165836240977e-07,
+      "loss": 0.859,
+      "step": 39731
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.529057853856288,
+      "learning_rate": 2.1366473923866372e-07,
+      "loss": 0.7431,
+      "step": 39732
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6554663956929143,
+      "learning_rate": 2.135078770940602e-07,
+      "loss": 0.7953,
+      "step": 39733
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.543238085050914,
+      "learning_rate": 2.1335107192951176e-07,
+      "loss": 0.9024,
+      "step": 39734
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8061040144682228,
+      "learning_rate": 2.1319432374593218e-07,
+      "loss": 0.8849,
+      "step": 39735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8069458766435358,
+      "learning_rate": 2.1303763254423293e-07,
+      "loss": 0.8759,
+      "step": 39736
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7401216582025438,
+      "learning_rate": 2.128809983253277e-07,
+      "loss": 0.6571,
+      "step": 39737
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4869449002470327,
+      "learning_rate": 2.127244210901269e-07,
+      "loss": 0.9526,
+      "step": 39738
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.525955115513265,
+      "learning_rate": 2.1256790083954537e-07,
+      "loss": 0.8863,
+      "step": 39739
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5187589315036336,
+      "learning_rate": 2.1241143757449012e-07,
+      "loss": 0.8299,
+      "step": 39740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5885816616961363,
+      "learning_rate": 2.1225503129587488e-07,
+      "loss": 0.8017,
+      "step": 39741
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.898643982225513,
+      "learning_rate": 2.1209868200461004e-07,
+      "loss": 0.8807,
+      "step": 39742
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0377560036991578,
+      "learning_rate": 2.1194238970160376e-07,
+      "loss": 0.7618,
+      "step": 39743
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.602433451339659,
+      "learning_rate": 2.117861543877686e-07,
+      "loss": 0.7333,
+      "step": 39744
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.578158938689464,
+      "learning_rate": 2.1162997606401392e-07,
+      "loss": 0.8995,
+      "step": 39745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4391993185621468,
+      "learning_rate": 2.1147385473124892e-07,
+      "loss": 0.843,
+      "step": 39746
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7993327950102607,
+      "learning_rate": 2.113177903903807e-07,
+      "loss": 0.895,
+      "step": 39747
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7161231405557762,
+      "learning_rate": 2.111617830423196e-07,
+      "loss": 0.7737,
+      "step": 39748
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6281343468309235,
+      "learning_rate": 2.1100583268797493e-07,
+      "loss": 0.6844,
+      "step": 39749
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7205405190088436,
+      "learning_rate": 2.1084993932825259e-07,
+      "loss": 0.7717,
+      "step": 39750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7470871104972935,
+      "learning_rate": 2.1069410296406079e-07,
+      "loss": 0.6872,
+      "step": 39751
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.713463712829776,
+      "learning_rate": 2.1053832359630655e-07,
+      "loss": 0.6987,
+      "step": 39752
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.330361660463121,
+      "learning_rate": 2.1038260122589915e-07,
+      "loss": 0.9283,
+      "step": 39753
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6154492335802577,
+      "learning_rate": 2.1022693585374343e-07,
+      "loss": 0.8517,
+      "step": 39754
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7961088511718875,
+      "learning_rate": 2.1007132748074533e-07,
+      "loss": 0.8235,
+      "step": 39755
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6790519032128357,
+      "learning_rate": 2.0991577610781188e-07,
+      "loss": 0.8787,
+      "step": 39756
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4764051200946475,
+      "learning_rate": 2.0976028173584683e-07,
+      "loss": 0.717,
+      "step": 39757
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4202921467728342,
+      "learning_rate": 2.0960484436575835e-07,
+      "loss": 0.861,
+      "step": 39758
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8309582277639744,
+      "learning_rate": 2.09449463998449e-07,
+      "loss": 0.9062,
+      "step": 39759
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6092089659524866,
+      "learning_rate": 2.092941406348259e-07,
+      "loss": 0.8305,
+      "step": 39760
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4644895425948308,
+      "learning_rate": 2.0913887427579048e-07,
+      "loss": 0.808,
+      "step": 39761
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6329914030941703,
+      "learning_rate": 2.0898366492224986e-07,
+      "loss": 0.7665,
+      "step": 39762
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.469649513892793,
+      "learning_rate": 2.0882851257510551e-07,
+      "loss": 0.7758,
+      "step": 39763
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4216961431720267,
+      "learning_rate": 2.0867341723526002e-07,
+      "loss": 0.7761,
+      "step": 39764
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.099422852692251,
+      "learning_rate": 2.0851837890361936e-07,
+      "loss": 0.8032,
+      "step": 39765
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.555524656012374,
+      "learning_rate": 2.0836339758108393e-07,
+      "loss": 0.7833,
+      "step": 39766
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.530094131716238,
+      "learning_rate": 2.082084732685574e-07,
+      "loss": 0.7904,
+      "step": 39767
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8037108643521337,
+      "learning_rate": 2.0805360596694024e-07,
+      "loss": 0.7379,
+      "step": 39768
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.880477197703746,
+      "learning_rate": 2.078987956771361e-07,
+      "loss": 0.6386,
+      "step": 39769
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5588662981358947,
+      "learning_rate": 2.0774404240004432e-07,
+      "loss": 0.8712,
+      "step": 39770
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7923497308481906,
+      "learning_rate": 2.0758934613656856e-07,
+      "loss": 0.7527,
+      "step": 39771
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.651068002077142,
+      "learning_rate": 2.0743470688760593e-07,
+      "loss": 0.7456,
+      "step": 39772
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4693292397378186,
+      "learning_rate": 2.07280124654059e-07,
+      "loss": 0.8462,
+      "step": 39773
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.590124604209701,
+      "learning_rate": 2.071255994368282e-07,
+      "loss": 0.9895,
+      "step": 39774
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7976882683055369,
+      "learning_rate": 2.069711312368128e-07,
+      "loss": 0.9499,
+      "step": 39775
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9666530084233842,
+      "learning_rate": 2.0681672005491093e-07,
+      "loss": 0.8238,
+      "step": 39776
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7279377138251726,
+      "learning_rate": 2.0666236589202305e-07,
+      "loss": 0.8525,
+      "step": 39777
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.650822640516239,
+      "learning_rate": 2.0650806874904727e-07,
+      "loss": 0.7324,
+      "step": 39778
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.747559839967932,
+      "learning_rate": 2.063538286268818e-07,
+      "loss": 0.8019,
+      "step": 39779
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7891885542586095,
+      "learning_rate": 2.061996455264259e-07,
+      "loss": 0.8423,
+      "step": 39780
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5301138793201325,
+      "learning_rate": 2.0604551944857665e-07,
+      "loss": 0.8906,
+      "step": 39781
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7401067661241436,
+      "learning_rate": 2.0589145039422887e-07,
+      "loss": 0.7129,
+      "step": 39782
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7760381936465892,
+      "learning_rate": 2.0573743836428294e-07,
+      "loss": 0.7529,
+      "step": 39783
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6182732509282471,
+      "learning_rate": 2.0558348335963595e-07,
+      "loss": 0.748,
+      "step": 39784
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5313382524159713,
+      "learning_rate": 2.0542958538118162e-07,
+      "loss": 0.8371,
+      "step": 39785
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7952703914042553,
+      "learning_rate": 2.05275744429817e-07,
+      "loss": 0.751,
+      "step": 39786
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8118357658566313,
+      "learning_rate": 2.0512196050643697e-07,
+      "loss": 0.961,
+      "step": 39787
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8256052911913696,
+      "learning_rate": 2.0496823361194074e-07,
+      "loss": 0.8901,
+      "step": 39788
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.902643326160134,
+      "learning_rate": 2.0481456374721763e-07,
+      "loss": 0.7885,
+      "step": 39789
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7080411477847708,
+      "learning_rate": 2.0466095091316585e-07,
+      "loss": 0.8846,
+      "step": 39790
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6625984056929022,
+      "learning_rate": 2.0450739511067907e-07,
+      "loss": 0.7402,
+      "step": 39791
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7513549525014998,
+      "learning_rate": 2.0435389634065328e-07,
+      "loss": 0.7593,
+      "step": 39792
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6702992403282018,
+      "learning_rate": 2.0420045460397885e-07,
+      "loss": 0.8619,
+      "step": 39793
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4631047466193583,
+      "learning_rate": 2.0404706990155065e-07,
+      "loss": 0.7236,
+      "step": 39794
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5347320666219084,
+      "learning_rate": 2.038937422342624e-07,
+      "loss": 0.773,
+      "step": 39795
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9924988854782275,
+      "learning_rate": 2.0374047160300557e-07,
+      "loss": 0.8122,
+      "step": 39796
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.577378197260362,
+      "learning_rate": 2.0358725800867395e-07,
+      "loss": 0.7105,
+      "step": 39797
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.741477268893975,
+      "learning_rate": 2.0343410145215793e-07,
+      "loss": 0.8029,
+      "step": 39798
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.849965389051184,
+      "learning_rate": 2.0328100193435006e-07,
+      "loss": 0.7406,
+      "step": 39799
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4481662490710685,
+      "learning_rate": 2.0312795945614193e-07,
+      "loss": 0.792,
+      "step": 39800
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4884377380699174,
+      "learning_rate": 2.0297497401842502e-07,
+      "loss": 0.8072,
+      "step": 39801
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6374240274263938,
+      "learning_rate": 2.0282204562208974e-07,
+      "loss": 0.7613,
+      "step": 39802
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7417022225407783,
+      "learning_rate": 2.0266917426802536e-07,
+      "loss": 0.7302,
+      "step": 39803
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4795115794195115,
+      "learning_rate": 2.0251635995712226e-07,
+      "loss": 0.8158,
+      "step": 39804
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7787995832653773,
+      "learning_rate": 2.02363602690272e-07,
+      "loss": 0.827,
+      "step": 39805
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7862128617646424,
+      "learning_rate": 2.0221090246836274e-07,
+      "loss": 0.7288,
+      "step": 39806
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.738173736170002,
+      "learning_rate": 2.0205825929228262e-07,
+      "loss": 0.7581,
+      "step": 39807
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.590064940904878,
+      "learning_rate": 2.019056731629221e-07,
+      "loss": 0.9,
+      "step": 39808
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.525555218772956,
+      "learning_rate": 2.017531440811682e-07,
+      "loss": 0.8126,
+      "step": 39809
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7538646819549517,
+      "learning_rate": 2.0160067204791023e-07,
+      "loss": 0.7944,
+      "step": 39810
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5175727803065289,
+      "learning_rate": 2.0144825706403526e-07,
+      "loss": 0.7195,
+      "step": 39811
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.48193841792803,
+      "learning_rate": 2.0129589913042924e-07,
+      "loss": 0.8267,
+      "step": 39812
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6460389377395486,
+      "learning_rate": 2.0114359824798257e-07,
+      "loss": 0.7846,
+      "step": 39813
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6928296624962575,
+      "learning_rate": 2.00991354417579e-07,
+      "loss": 0.8395,
+      "step": 39814
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5022988705648734,
+      "learning_rate": 2.008391676401067e-07,
+      "loss": 0.8141,
+      "step": 39815
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8348089379011665,
+      "learning_rate": 2.0068703791645162e-07,
+      "loss": 0.7839,
+      "step": 39816
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6402383249130064,
+      "learning_rate": 2.005349652474975e-07,
+      "loss": 0.8493,
+      "step": 39817
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6018612753765127,
+      "learning_rate": 2.0038294963413251e-07,
+      "loss": 0.7583,
+      "step": 39818
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7571018282776616,
+      "learning_rate": 2.002309910772393e-07,
+      "loss": 0.8211,
+      "step": 39819
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5944286985214717,
+      "learning_rate": 2.0007908957770493e-07,
+      "loss": 0.7878,
+      "step": 39820
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7641852971327275,
+      "learning_rate": 1.9992724513641204e-07,
+      "loss": 0.6702,
+      "step": 39821
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7045085735173788,
+      "learning_rate": 1.9977545775424545e-07,
+      "loss": 0.8497,
+      "step": 39822
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7708187252199395,
+      "learning_rate": 1.9962372743209e-07,
+      "loss": 0.7365,
+      "step": 39823
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.926545081875244,
+      "learning_rate": 1.9947205417082615e-07,
+      "loss": 0.6751,
+      "step": 39824
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4171194737047834,
+      "learning_rate": 1.993204379713398e-07,
+      "loss": 0.8097,
+      "step": 39825
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9796249653167393,
+      "learning_rate": 1.991688788345125e-07,
+      "loss": 0.8523,
+      "step": 39826
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6060535262593187,
+      "learning_rate": 1.99017376761228e-07,
+      "loss": 0.8244,
+      "step": 39827
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7545059412721733,
+      "learning_rate": 1.9886593175236557e-07,
+      "loss": 0.6827,
+      "step": 39828
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4616645540820947,
+      "learning_rate": 1.9871454380880894e-07,
+      "loss": 0.9442,
+      "step": 39829
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8159541662627554,
+      "learning_rate": 1.9856321293143965e-07,
+      "loss": 0.792,
+      "step": 39830
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6141668936732316,
+      "learning_rate": 1.9841193912113922e-07,
+      "loss": 0.7883,
+      "step": 39831
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5029083353689126,
+      "learning_rate": 1.982607223787858e-07,
+      "loss": 0.873,
+      "step": 39832
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5956191164254065,
+      "learning_rate": 1.9810956270526204e-07,
+      "loss": 0.911,
+      "step": 39833
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7659310523000843,
+      "learning_rate": 1.9795846010144837e-07,
+      "loss": 0.7351,
+      "step": 39834
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6025397698736752,
+      "learning_rate": 1.9780741456822404e-07,
+      "loss": 0.8239,
+      "step": 39835
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1339834877580017,
+      "learning_rate": 1.976564261064684e-07,
+      "loss": 0.893,
+      "step": 39836
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6232204019034757,
+      "learning_rate": 1.975054947170596e-07,
+      "loss": 0.8386,
+      "step": 39837
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5719706378007459,
+      "learning_rate": 1.9735462040087693e-07,
+      "loss": 0.944,
+      "step": 39838
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7737339496158101,
+      "learning_rate": 1.9720380315880083e-07,
+      "loss": 0.6994,
+      "step": 39839
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4732366208471093,
+      "learning_rate": 1.9705304299170613e-07,
+      "loss": 0.719,
+      "step": 39840
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.641362159156318,
+      "learning_rate": 1.9690233990047325e-07,
+      "loss": 0.9047,
+      "step": 39841
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8574130294710272,
+      "learning_rate": 1.9675169388597705e-07,
+      "loss": 0.7573,
+      "step": 39842
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5083639776561453,
+      "learning_rate": 1.966011049490979e-07,
+      "loss": 0.8295,
+      "step": 39843
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9634442880176768,
+      "learning_rate": 1.9645057309070958e-07,
+      "loss": 0.8975,
+      "step": 39844
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7025266131379084,
+      "learning_rate": 1.9630009831169027e-07,
+      "loss": 0.9176,
+      "step": 39845
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7206699241145151,
+      "learning_rate": 1.9614968061291484e-07,
+      "loss": 0.8539,
+      "step": 39846
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6202663218224482,
+      "learning_rate": 1.9599931999526035e-07,
+      "loss": 0.7742,
+      "step": 39847
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7542544164685825,
+      "learning_rate": 1.9584901645960275e-07,
+      "loss": 0.8065,
+      "step": 39848
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.526682658014568,
+      "learning_rate": 1.956987700068147e-07,
+      "loss": 0.8094,
+      "step": 39849
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7285423899891172,
+      "learning_rate": 1.9554858063777215e-07,
+      "loss": 0.8683,
+      "step": 39850
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7881906641235537,
+      "learning_rate": 1.953984483533511e-07,
+      "loss": 0.8077,
+      "step": 39851
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.680954558191113,
+      "learning_rate": 1.9524837315442414e-07,
+      "loss": 0.9566,
+      "step": 39852
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5629471998892124,
+      "learning_rate": 1.9509835504186504e-07,
+      "loss": 0.7245,
+      "step": 39853
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4719872101755038,
+      "learning_rate": 1.9494839401654642e-07,
+      "loss": 0.8631,
+      "step": 39854
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3120983311483156,
+      "learning_rate": 1.9479849007934425e-07,
+      "loss": 0.8563,
+      "step": 39855
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4294368214991164,
+      "learning_rate": 1.9464864323112897e-07,
+      "loss": 0.7726,
+      "step": 39856
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4701176311605775,
+      "learning_rate": 1.944988534727732e-07,
+      "loss": 0.8312,
+      "step": 39857
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5707365877739707,
+      "learning_rate": 1.9434912080514955e-07,
+      "loss": 0.8563,
+      "step": 39858
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.723816263199966,
+      "learning_rate": 1.941994452291296e-07,
+      "loss": 0.8428,
+      "step": 39859
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5524036234201235,
+      "learning_rate": 1.9404982674558593e-07,
+      "loss": 0.8609,
+      "step": 39860
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.102322011793471,
+      "learning_rate": 1.9390026535538675e-07,
+      "loss": 0.7728,
+      "step": 39861
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4338411180205601,
+      "learning_rate": 1.9375076105940693e-07,
+      "loss": 0.8282,
+      "step": 39862
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5426119585014535,
+      "learning_rate": 1.9360131385851356e-07,
+      "loss": 0.8097,
+      "step": 39863
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4198923729564488,
+      "learning_rate": 1.9345192375357812e-07,
+      "loss": 0.8605,
+      "step": 39864
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6012381805873575,
+      "learning_rate": 1.933025907454711e-07,
+      "loss": 0.8663,
+      "step": 39865
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7409715979497554,
+      "learning_rate": 1.9315331483506062e-07,
+      "loss": 0.673,
+      "step": 39866
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8790753493801675,
+      "learning_rate": 1.930040960232149e-07,
+      "loss": 0.9018,
+      "step": 39867
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9352582849658129,
+      "learning_rate": 1.928549343108055e-07,
+      "loss": 0.7571,
+      "step": 39868
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6132372713473613,
+      "learning_rate": 1.9270582969870056e-07,
+      "loss": 0.6173,
+      "step": 39869
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.618756161984482,
+      "learning_rate": 1.9255678218776608e-07,
+      "loss": 0.9352,
+      "step": 39870
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4439136289209067,
+      "learning_rate": 1.9240779177887027e-07,
+      "loss": 0.807,
+      "step": 39871
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6363983287686603,
+      "learning_rate": 1.922588584728824e-07,
+      "loss": 0.8859,
+      "step": 39872
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.603180656921642,
+      "learning_rate": 1.9210998227066846e-07,
+      "loss": 0.9035,
+      "step": 39873
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7548411036790421,
+      "learning_rate": 1.9196116317309554e-07,
+      "loss": 0.7263,
+      "step": 39874
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.608907166269319,
+      "learning_rate": 1.918124011810285e-07,
+      "loss": 0.6706,
+      "step": 39875
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8945770767687626,
+      "learning_rate": 1.9166369629533664e-07,
+      "loss": 0.9523,
+      "step": 39876
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9928152054532002,
+      "learning_rate": 1.9151504851688263e-07,
+      "loss": 0.8521,
+      "step": 39877
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7711636000702139,
+      "learning_rate": 1.9136645784653464e-07,
+      "loss": 0.6845,
+      "step": 39878
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.646371998379733,
+      "learning_rate": 1.912179242851564e-07,
+      "loss": 0.9085,
+      "step": 39879
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6437199263246,
+      "learning_rate": 1.910694478336117e-07,
+      "loss": 0.8063,
+      "step": 39880
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6070272779656138,
+      "learning_rate": 1.9092102849276651e-07,
+      "loss": 0.7695,
+      "step": 39881
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.739374137472242,
+      "learning_rate": 1.9077266626348457e-07,
+      "loss": 0.7048,
+      "step": 39882
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6001990177497396,
+      "learning_rate": 1.9062436114663073e-07,
+      "loss": 0.8206,
+      "step": 39883
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4837650296914142,
+      "learning_rate": 1.9047611314306546e-07,
+      "loss": 0.7147,
+      "step": 39884
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8344438550659015,
+      "learning_rate": 1.903279222536547e-07,
+      "loss": 0.7685,
+      "step": 39885
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9815350134330167,
+      "learning_rate": 1.9017978847926222e-07,
+      "loss": 0.8869,
+      "step": 39886
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6038695926548814,
+      "learning_rate": 1.9003171182074732e-07,
+      "loss": 0.7446,
+      "step": 39887
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0746349329960814,
+      "learning_rate": 1.8988369227897263e-07,
+      "loss": 0.8958,
+      "step": 39888
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6796276791588038,
+      "learning_rate": 1.8973572985480193e-07,
+      "loss": 0.863,
+      "step": 39889
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7809224190066282,
+      "learning_rate": 1.8958782454909563e-07,
+      "loss": 0.985,
+      "step": 39890
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7561782236715932,
+      "learning_rate": 1.8943997636271528e-07,
+      "loss": 0.7104,
+      "step": 39891
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5918566309172435,
+      "learning_rate": 1.8929218529652128e-07,
+      "loss": 0.8194,
+      "step": 39892
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8293496361841277,
+      "learning_rate": 1.8914445135137405e-07,
+      "loss": 0.7062,
+      "step": 39893
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5392412841105751,
+      "learning_rate": 1.889967745281329e-07,
+      "loss": 0.7151,
+      "step": 39894
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6591278989558271,
+      "learning_rate": 1.8884915482766054e-07,
+      "loss": 0.8841,
+      "step": 39895
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6562599716583608,
+      "learning_rate": 1.8870159225081285e-07,
+      "loss": 0.7584,
+      "step": 39896
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0510860620767084,
+      "learning_rate": 1.8855408679845143e-07,
+      "loss": 0.763,
+      "step": 39897
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8038967253068585,
+      "learning_rate": 1.884066384714345e-07,
+      "loss": 0.8629,
+      "step": 39898
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5261569056372994,
+      "learning_rate": 1.8825924727062016e-07,
+      "loss": 0.8056,
+      "step": 39899
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0412218038372063,
+      "learning_rate": 1.8811191319686783e-07,
+      "loss": 0.7663,
+      "step": 39900
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6071125297177922,
+      "learning_rate": 1.8796463625103233e-07,
+      "loss": 0.8746,
+      "step": 39901
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1121944428426858,
+      "learning_rate": 1.8781741643397412e-07,
+      "loss": 0.816,
+      "step": 39902
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7871055354827516,
+      "learning_rate": 1.8767025374655023e-07,
+      "loss": 0.6677,
+      "step": 39903
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.613550369963563,
+      "learning_rate": 1.875231481896167e-07,
+      "loss": 0.7709,
+      "step": 39904
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.582508172027847,
+      "learning_rate": 1.8737609976402836e-07,
+      "loss": 0.9239,
+      "step": 39905
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9044300620105035,
+      "learning_rate": 1.8722910847064346e-07,
+      "loss": 0.6918,
+      "step": 39906
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7560627094116756,
+      "learning_rate": 1.8708217431031793e-07,
+      "loss": 0.9,
+      "step": 39907
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4931329425672786,
+      "learning_rate": 1.869352972839067e-07,
+      "loss": 0.8824,
+      "step": 39908
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6249709753772241,
+      "learning_rate": 1.8678847739226459e-07,
+      "loss": 0.851,
+      "step": 39909
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5116717861240068,
+      "learning_rate": 1.8664171463624648e-07,
+      "loss": 0.737,
+      "step": 39910
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5810335317224773,
+      "learning_rate": 1.8649500901670837e-07,
+      "loss": 0.7906,
+      "step": 39911
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5920590753003354,
+      "learning_rate": 1.863483605345029e-07,
+      "loss": 0.9422,
+      "step": 39912
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6945642605361955,
+      "learning_rate": 1.8620176919048273e-07,
+      "loss": 0.7844,
+      "step": 39913
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5937497555734585,
+      "learning_rate": 1.860552349855038e-07,
+      "loss": 0.9289,
+      "step": 39914
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6027204956330114,
+      "learning_rate": 1.8590875792041884e-07,
+      "loss": 0.8334,
+      "step": 39915
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.712431087903932,
+      "learning_rate": 1.8576233799608046e-07,
+      "loss": 0.7824,
+      "step": 39916
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4736042045997853,
+      "learning_rate": 1.8561597521334018e-07,
+      "loss": 0.8089,
+      "step": 39917
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.702029839275282,
+      "learning_rate": 1.854696695730507e-07,
+      "loss": 0.7305,
+      "step": 39918
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8839299034123307,
+      "learning_rate": 1.853234210760635e-07,
+      "loss": 0.8293,
+      "step": 39919
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5836077006813356,
+      "learning_rate": 1.8517722972323239e-07,
+      "loss": 0.921,
+      "step": 39920
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7638922725386292,
+      "learning_rate": 1.8503109551540556e-07,
+      "loss": 0.7299,
+      "step": 39921
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7219609942495897,
+      "learning_rate": 1.8488501845343565e-07,
+      "loss": 0.8101,
+      "step": 39922
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6932273316077626,
+      "learning_rate": 1.847389985381709e-07,
+      "loss": 0.7345,
+      "step": 39923
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.655148406233236,
+      "learning_rate": 1.8459303577046505e-07,
+      "loss": 0.7708,
+      "step": 39924
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5245594409569279,
+      "learning_rate": 1.8444713015116632e-07,
+      "loss": 0.5899,
+      "step": 39925
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4688618685455972,
+      "learning_rate": 1.8430128168112293e-07,
+      "loss": 0.7832,
+      "step": 39926
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6764671482027957,
+      "learning_rate": 1.841554903611842e-07,
+      "loss": 0.7494,
+      "step": 39927
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9654424520239353,
+      "learning_rate": 1.8400975619220163e-07,
+      "loss": 0.8764,
+      "step": 39928
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.735057029915579,
+      "learning_rate": 1.8386407917502126e-07,
+      "loss": 0.817,
+      "step": 39929
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5725341490751923,
+      "learning_rate": 1.8371845931049126e-07,
+      "loss": 0.9304,
+      "step": 39930
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7293022194199326,
+      "learning_rate": 1.8357289659946098e-07,
+      "loss": 0.9689,
+      "step": 39931
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7469415400319661,
+      "learning_rate": 1.8342739104277753e-07,
+      "loss": 0.8783,
+      "step": 39932
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7807328822805935,
+      "learning_rate": 1.8328194264128686e-07,
+      "loss": 0.7099,
+      "step": 39933
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4836013648002566,
+      "learning_rate": 1.831365513958361e-07,
+      "loss": 0.8127,
+      "step": 39934
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4388925244080035,
+      "learning_rate": 1.8299121730727342e-07,
+      "loss": 0.6782,
+      "step": 39935
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6247471846414132,
+      "learning_rate": 1.8284594037644377e-07,
+      "loss": 0.8173,
+      "step": 39936
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.707956302081064,
+      "learning_rate": 1.8270072060419308e-07,
+      "loss": 0.8511,
+      "step": 39937
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.215855908586041,
+      "learning_rate": 1.8255555799136627e-07,
+      "loss": 0.7286,
+      "step": 39938
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5861447786917142,
+      "learning_rate": 1.824104525388104e-07,
+      "loss": 0.8214,
+      "step": 39939
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8931125874177508,
+      "learning_rate": 1.8226540424736706e-07,
+      "loss": 0.7722,
+      "step": 39940
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.74124006617663,
+      "learning_rate": 1.8212041311788442e-07,
+      "loss": 0.8177,
+      "step": 39941
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7594071622052209,
+      "learning_rate": 1.8197547915120407e-07,
+      "loss": 0.704,
+      "step": 39942
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7481762349765858,
+      "learning_rate": 1.8183060234817196e-07,
+      "loss": 0.7174,
+      "step": 39943
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4676146966632952,
+      "learning_rate": 1.8168578270962968e-07,
+      "loss": 0.8353,
+      "step": 39944
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7128734606113396,
+      "learning_rate": 1.8154102023642205e-07,
+      "loss": 0.8377,
+      "step": 39945
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4732355093300487,
+      "learning_rate": 1.813963149293907e-07,
+      "loss": 0.7392,
+      "step": 39946
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.703917690371485,
+      "learning_rate": 1.812516667893771e-07,
+      "loss": 0.8655,
+      "step": 39947
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4360626120806437,
+      "learning_rate": 1.811070758172262e-07,
+      "loss": 0.9468,
+      "step": 39948
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6471221482521679,
+      "learning_rate": 1.8096254201377838e-07,
+      "loss": 0.7673,
+      "step": 39949
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6939574208753783,
+      "learning_rate": 1.8081806537987524e-07,
+      "loss": 0.9074,
+      "step": 39950
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8188818203514519,
+      "learning_rate": 1.8067364591635828e-07,
+      "loss": 0.8402,
+      "step": 39951
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6630909644249436,
+      "learning_rate": 1.8052928362406686e-07,
+      "loss": 0.7387,
+      "step": 39952
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8518840497477598,
+      "learning_rate": 1.803849785038436e-07,
+      "loss": 0.7363,
+      "step": 39953
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5724186522563264,
+      "learning_rate": 1.8024073055652903e-07,
+      "loss": 0.7556,
+      "step": 39954
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.845236786693327,
+      "learning_rate": 1.8009653978296017e-07,
+      "loss": 0.8616,
+      "step": 39955
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8720958638138954,
+      "learning_rate": 1.7995240618397859e-07,
+      "loss": 0.7999,
+      "step": 39956
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6960442055552025,
+      "learning_rate": 1.798083297604236e-07,
+      "loss": 0.7986,
+      "step": 39957
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.042714490907615,
+      "learning_rate": 1.7966431051313349e-07,
+      "loss": 0.7643,
+      "step": 39958
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8307836684245367,
+      "learning_rate": 1.795203484429453e-07,
+      "loss": 0.8859,
+      "step": 39959
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.620279968768812,
+      "learning_rate": 1.793764435507006e-07,
+      "loss": 0.8254,
+      "step": 39960
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6284678141451583,
+      "learning_rate": 1.792325958372343e-07,
+      "loss": 0.9178,
+      "step": 39961
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6491771792429302,
+      "learning_rate": 1.7908880530338457e-07,
+      "loss": 0.8019,
+      "step": 39962
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.428399758117875,
+      "learning_rate": 1.7894507194999077e-07,
+      "loss": 0.7934,
+      "step": 39963
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8904930306798216,
+      "learning_rate": 1.7880139577788779e-07,
+      "loss": 0.8468,
+      "step": 39964
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7628091656064802,
+      "learning_rate": 1.7865777678791163e-07,
+      "loss": 0.7264,
+      "step": 39965
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8686113000733706,
+      "learning_rate": 1.7851421498089826e-07,
+      "loss": 0.776,
+      "step": 39966
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.654327950771917,
+      "learning_rate": 1.7837071035768704e-07,
+      "loss": 1.0034,
+      "step": 39967
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5429836656884652,
+      "learning_rate": 1.7822726291910952e-07,
+      "loss": 0.7223,
+      "step": 39968
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.774650720377051,
+      "learning_rate": 1.780838726660028e-07,
+      "loss": 0.8681,
+      "step": 39969
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9515681320545626,
+      "learning_rate": 1.7794053959920065e-07,
+      "loss": 0.8694,
+      "step": 39970
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5772212655713906,
+      "learning_rate": 1.777972637195402e-07,
+      "loss": 0.7677,
+      "step": 39971
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5712086643535446,
+      "learning_rate": 1.776540450278519e-07,
+      "loss": 0.8788,
+      "step": 39972
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5423571374528047,
+      "learning_rate": 1.775108835249717e-07,
+      "loss": 0.791,
+      "step": 39973
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5864805302076948,
+      "learning_rate": 1.7736777921173343e-07,
+      "loss": 0.8367,
+      "step": 39974
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9211038751341896,
+      "learning_rate": 1.7722473208896862e-07,
+      "loss": 0.8703,
+      "step": 39975
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7894816765006012,
+      "learning_rate": 1.7708174215751217e-07,
+      "loss": 0.7335,
+      "step": 39976
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.772787788726163,
+      "learning_rate": 1.7693880941819564e-07,
+      "loss": 0.8254,
+      "step": 39977
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7514195519177643,
+      "learning_rate": 1.7679593387185169e-07,
+      "loss": 0.7217,
+      "step": 39978
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7064599338612734,
+      "learning_rate": 1.7665311551931075e-07,
+      "loss": 0.7616,
+      "step": 39979
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6415001227914348,
+      "learning_rate": 1.7651035436140662e-07,
+      "loss": 0.9821,
+      "step": 39980
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4444253637028956,
+      "learning_rate": 1.7636765039896865e-07,
+      "loss": 0.7698,
+      "step": 39981
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4685013173652226,
+      "learning_rate": 1.7622500363282836e-07,
+      "loss": 0.8458,
+      "step": 39982
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.822594139359328,
+      "learning_rate": 1.7608241406381509e-07,
+      "loss": 0.7018,
+      "step": 39983
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6306868987156973,
+      "learning_rate": 1.7593988169276156e-07,
+      "loss": 0.8482,
+      "step": 39984
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4444754926760777,
+      "learning_rate": 1.7579740652049705e-07,
+      "loss": 0.8219,
+      "step": 39985
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.865564295852521,
+      "learning_rate": 1.7565498854784868e-07,
+      "loss": 0.8415,
+      "step": 39986
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7432733063382845,
+      "learning_rate": 1.7551262777564693e-07,
+      "loss": 0.6642,
+      "step": 39987
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.707698598553083,
+      "learning_rate": 1.7537032420472334e-07,
+      "loss": 0.783,
+      "step": 39988
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5546336482363836,
+      "learning_rate": 1.7522807783590278e-07,
+      "loss": 0.9301,
+      "step": 39989
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4517843985294732,
+      "learning_rate": 1.7508588867001463e-07,
+      "loss": 0.8268,
+      "step": 39990
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5835623029255657,
+      "learning_rate": 1.749437567078871e-07,
+      "loss": 0.7518,
+      "step": 39991
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6464752935568339,
+      "learning_rate": 1.748016819503473e-07,
+      "loss": 0.9529,
+      "step": 39992
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7541125319509507,
+      "learning_rate": 1.7465966439822346e-07,
+      "loss": 0.7702,
+      "step": 39993
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7169005361465772,
+      "learning_rate": 1.745177040523416e-07,
+      "loss": 0.8882,
+      "step": 39994
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6522079130049432,
+      "learning_rate": 1.7437580091352878e-07,
+      "loss": 0.8239,
+      "step": 39995
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5296500374826967,
+      "learning_rate": 1.7423395498260887e-07,
+      "loss": 0.8831,
+      "step": 39996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5571424955637523,
+      "learning_rate": 1.7409216626041225e-07,
+      "loss": 0.9393,
+      "step": 39997
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5830586546293859,
+      "learning_rate": 1.7395043474775942e-07,
+      "loss": 0.8555,
+      "step": 39998
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7179809142281546,
+      "learning_rate": 1.7380876044547966e-07,
+      "loss": 0.866,
+      "step": 39999
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7579745390640262,
+      "learning_rate": 1.736671433543957e-07,
+      "loss": 0.7286,
+      "step": 40000
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4332087883268145,
+      "learning_rate": 1.7352558347533243e-07,
+      "loss": 0.7032,
+      "step": 40001
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7498169810295443,
+      "learning_rate": 1.7338408080911473e-07,
+      "loss": 0.6165,
+      "step": 40002
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.641278270069481,
+      "learning_rate": 1.7324263535656528e-07,
+      "loss": 0.8738,
+      "step": 40003
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7751640824888448,
+      "learning_rate": 1.7310124711850894e-07,
+      "loss": 0.7438,
+      "step": 40004
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6725494202863795,
+      "learning_rate": 1.7295991609576845e-07,
+      "loss": 0.8257,
+      "step": 40005
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.534262150946677,
+      "learning_rate": 1.7281864228916644e-07,
+      "loss": 0.825,
+      "step": 40006
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5976681651280853,
+      "learning_rate": 1.726774256995245e-07,
+      "loss": 0.9571,
+      "step": 40007
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8287589873073025,
+      "learning_rate": 1.725362663276653e-07,
+      "loss": 0.77,
+      "step": 40008
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5397761839961448,
+      "learning_rate": 1.723951641744137e-07,
+      "loss": 0.8361,
+      "step": 40009
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7100797258054625,
+      "learning_rate": 1.72254119240588e-07,
+      "loss": 0.7716,
+      "step": 40010
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.477205906989825,
+      "learning_rate": 1.721131315270097e-07,
+      "loss": 0.8404,
+      "step": 40011
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.591659462810437,
+      "learning_rate": 1.719722010344993e-07,
+      "loss": 0.8174,
+      "step": 40012
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4950993198595368,
+      "learning_rate": 1.7183132776387945e-07,
+      "loss": 0.8939,
+      "step": 40013
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6564247774628635,
+      "learning_rate": 1.716905117159695e-07,
+      "loss": 0.8239,
+      "step": 40014
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8984662354914448,
+      "learning_rate": 1.715497528915888e-07,
+      "loss": 0.8082,
+      "step": 40015
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5875036736555777,
+      "learning_rate": 1.714090512915567e-07,
+      "loss": 0.9182,
+      "step": 40016
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5185378122771982,
+      "learning_rate": 1.7126840691669255e-07,
+      "loss": 0.7271,
+      "step": 40017
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.072931766643773,
+      "learning_rate": 1.7112781976781567e-07,
+      "loss": 0.9383,
+      "step": 40018
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5113647528767145,
+      "learning_rate": 1.709872898457443e-07,
+      "loss": 0.8616,
+      "step": 40019
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8290167462633257,
+      "learning_rate": 1.708468171512978e-07,
+      "loss": 0.7137,
+      "step": 40020
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4965055049416671,
+      "learning_rate": 1.7070640168529107e-07,
+      "loss": 0.8435,
+      "step": 40021
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4690339331652371,
+      "learning_rate": 1.7056604344854565e-07,
+      "loss": 0.7797,
+      "step": 40022
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1868471681156416,
+      "learning_rate": 1.7042574244187537e-07,
+      "loss": 0.7706,
+      "step": 40023
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0750689669462417,
+      "learning_rate": 1.7028549866609845e-07,
+      "loss": 0.8181,
+      "step": 40024
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4762941753241734,
+      "learning_rate": 1.7014531212203088e-07,
+      "loss": 0.8852,
+      "step": 40025
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5686320390223107,
+      "learning_rate": 1.7000518281049093e-07,
+      "loss": 0.7745,
+      "step": 40026
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6472061077230196,
+      "learning_rate": 1.698651107322924e-07,
+      "loss": 0.8735,
+      "step": 40027
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.603432763255213,
+      "learning_rate": 1.6972509588825015e-07,
+      "loss": 0.8175,
+      "step": 40028
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0124136935600005,
+      "learning_rate": 1.6958513827918134e-07,
+      "loss": 0.8337,
+      "step": 40029
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.594556348629907,
+      "learning_rate": 1.6944523790590084e-07,
+      "loss": 0.8542,
+      "step": 40030
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.98283408118147,
+      "learning_rate": 1.6930539476922248e-07,
+      "loss": 0.8579,
+      "step": 40031
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7689540309552692,
+      "learning_rate": 1.6916560886995893e-07,
+      "loss": 0.7961,
+      "step": 40032
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7814480650252832,
+      "learning_rate": 1.690258802089262e-07,
+      "loss": 0.6662,
+      "step": 40033
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7399143794175018,
+      "learning_rate": 1.6888620878693806e-07,
+      "loss": 0.6509,
+      "step": 40034
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4597929048483762,
+      "learning_rate": 1.6874659460480726e-07,
+      "loss": 0.773,
+      "step": 40035
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7794314990360047,
+      "learning_rate": 1.686070376633453e-07,
+      "loss": 0.9206,
+      "step": 40036
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6781549536154086,
+      "learning_rate": 1.6846753796336491e-07,
+      "loss": 0.7916,
+      "step": 40037
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9439706467307414,
+      "learning_rate": 1.6832809550568097e-07,
+      "loss": 0.8972,
+      "step": 40038
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5840246693770874,
+      "learning_rate": 1.6818871029110284e-07,
+      "loss": 0.8366,
+      "step": 40039
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5730882814080536,
+      "learning_rate": 1.680493823204421e-07,
+      "loss": 0.6658,
+      "step": 40040
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6470598027207708,
+      "learning_rate": 1.6791011159451141e-07,
+      "loss": 0.7985,
+      "step": 40041
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5625668890985598,
+      "learning_rate": 1.6777089811412017e-07,
+      "loss": 0.8312,
+      "step": 40042
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.995430922484778,
+      "learning_rate": 1.676317418800799e-07,
+      "loss": 0.8927,
+      "step": 40043
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6253042758139213,
+      "learning_rate": 1.6749264289320111e-07,
+      "loss": 0.7574,
+      "step": 40044
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8763539094490855,
+      "learning_rate": 1.67353601154292e-07,
+      "loss": 0.7422,
+      "step": 40045
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7879196738733929,
+      "learning_rate": 1.6721461666416305e-07,
+      "loss": 0.7528,
+      "step": 40046
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6845171784922248,
+      "learning_rate": 1.6707568942362473e-07,
+      "loss": 0.8275,
+      "step": 40047
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6794189647037436,
+      "learning_rate": 1.6693681943348527e-07,
+      "loss": 0.8133,
+      "step": 40048
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5286547674016533,
+      "learning_rate": 1.6679800669455072e-07,
+      "loss": 0.724,
+      "step": 40049
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5337154744584571,
+      "learning_rate": 1.666592512076326e-07,
+      "loss": 0.8178,
+      "step": 40050
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6542028014529173,
+      "learning_rate": 1.6652055297353698e-07,
+      "loss": 0.9018,
+      "step": 40051
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9893573919733538,
+      "learning_rate": 1.663819119930732e-07,
+      "loss": 0.777,
+      "step": 40052
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.563653617826187,
+      "learning_rate": 1.6624332826704616e-07,
+      "loss": 0.6663,
+      "step": 40053
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.613733939688744,
+      "learning_rate": 1.66104801796263e-07,
+      "loss": 0.7701,
+      "step": 40054
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.231253506410946,
+      "learning_rate": 1.6596633258153193e-07,
+      "loss": 0.7956,
+      "step": 40055
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5247269191176647,
+      "learning_rate": 1.6582792062365905e-07,
+      "loss": 0.7607,
+      "step": 40056
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7370930239736964,
+      "learning_rate": 1.6568956592344808e-07,
+      "loss": 0.7126,
+      "step": 40057
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8317176531004222,
+      "learning_rate": 1.6555126848170622e-07,
+      "loss": 0.7674,
+      "step": 40058
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5474482149735678,
+      "learning_rate": 1.6541302829923945e-07,
+      "loss": 0.8328,
+      "step": 40059
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7805531565036652,
+      "learning_rate": 1.6527484537684935e-07,
+      "loss": 0.8724,
+      "step": 40060
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.660233765012612,
+      "learning_rate": 1.6513671971534528e-07,
+      "loss": 0.9838,
+      "step": 40061
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.592624446986101,
+      "learning_rate": 1.6499865131552773e-07,
+      "loss": 0.8071,
+      "step": 40062
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.345814469739683,
+      "learning_rate": 1.6486064017820157e-07,
+      "loss": 0.8288,
+      "step": 40063
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8123156695252085,
+      "learning_rate": 1.6472268630417065e-07,
+      "loss": 0.9362,
+      "step": 40064
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.708904893252999,
+      "learning_rate": 1.6458478969423876e-07,
+      "loss": 0.8431,
+      "step": 40065
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5983459606024182,
+      "learning_rate": 1.6444695034920743e-07,
+      "loss": 0.8768,
+      "step": 40066
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4435731252424229,
+      "learning_rate": 1.643091682698794e-07,
+      "loss": 0.7322,
+      "step": 40067
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4878300487736722,
+      "learning_rate": 1.6417144345705738e-07,
+      "loss": 0.8641,
+      "step": 40068
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6569312088465484,
+      "learning_rate": 1.6403377591154512e-07,
+      "loss": 0.8979,
+      "step": 40069
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6266867437378068,
+      "learning_rate": 1.6389616563413978e-07,
+      "loss": 0.7738,
+      "step": 40070
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9671050539697759,
+      "learning_rate": 1.637586126256463e-07,
+      "loss": 0.7862,
+      "step": 40071
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.474592618657586,
+      "learning_rate": 1.6362111688686288e-07,
+      "loss": 0.603,
+      "step": 40072
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5235314951152077,
+      "learning_rate": 1.6348367841859337e-07,
+      "loss": 0.8036,
+      "step": 40073
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6035085840185457,
+      "learning_rate": 1.6334629722163597e-07,
+      "loss": 0.8193,
+      "step": 40074
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6612812981862426,
+      "learning_rate": 1.6320897329678899e-07,
+      "loss": 0.8315,
+      "step": 40075
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8269192581566188,
+      "learning_rate": 1.6307170664485506e-07,
+      "loss": 0.6385,
+      "step": 40076
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4861059904780107,
+      "learning_rate": 1.6293449726663136e-07,
+      "loss": 0.8432,
+      "step": 40077
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.008206888885445,
+      "learning_rate": 1.6279734516291724e-07,
+      "loss": 0.8004,
+      "step": 40078
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6916917616781064,
+      "learning_rate": 1.6266025033451206e-07,
+      "loss": 0.7551,
+      "step": 40079
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7085665934470153,
+      "learning_rate": 1.6252321278221294e-07,
+      "loss": 0.9425,
+      "step": 40080
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6092909972132914,
+      "learning_rate": 1.6238623250681817e-07,
+      "loss": 0.7843,
+      "step": 40081
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.611293960736515,
+      "learning_rate": 1.6224930950912487e-07,
+      "loss": 0.9853,
+      "step": 40082
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6446028082081434,
+      "learning_rate": 1.621124437899324e-07,
+      "loss": 0.7473,
+      "step": 40083
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6834596580649825,
+      "learning_rate": 1.6197563535003347e-07,
+      "loss": 0.8219,
+      "step": 40084
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.075212566261932,
+      "learning_rate": 1.6183888419022854e-07,
+      "loss": 0.7573,
+      "step": 40085
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4428041575730841,
+      "learning_rate": 1.617021903113114e-07,
+      "loss": 0.7408,
+      "step": 40086
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6038001406262536,
+      "learning_rate": 1.6156555371408033e-07,
+      "loss": 0.7532,
+      "step": 40087
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8215678225413907,
+      "learning_rate": 1.6142897439932804e-07,
+      "loss": 0.8395,
+      "step": 40088
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5819149674585604,
+      "learning_rate": 1.6129245236785162e-07,
+      "loss": 0.7621,
+      "step": 40089
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6842020652312435,
+      "learning_rate": 1.6115598762044493e-07,
+      "loss": 0.8234,
+      "step": 40090
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0597740300126364,
+      "learning_rate": 1.6101958015790397e-07,
+      "loss": 0.7573,
+      "step": 40091
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6316347185756994,
+      "learning_rate": 1.6088322998102147e-07,
+      "loss": 0.8612,
+      "step": 40092
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.107965155597637,
+      "learning_rate": 1.607469370905912e-07,
+      "loss": 0.8452,
+      "step": 40093
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.644054630630894,
+      "learning_rate": 1.606107014874081e-07,
+      "loss": 0.8116,
+      "step": 40094
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5115199698026884,
+      "learning_rate": 1.6047452317226485e-07,
+      "loss": 0.9116,
+      "step": 40095
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9937627509178937,
+      "learning_rate": 1.6033840214595309e-07,
+      "loss": 0.8704,
+      "step": 40096
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7257384184692761,
+      "learning_rate": 1.602023384092677e-07,
+      "loss": 0.8463,
+      "step": 40097
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4959349413943805,
+      "learning_rate": 1.6006633196299804e-07,
+      "loss": 0.7687,
+      "step": 40098
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.693817589829888,
+      "learning_rate": 1.5993038280793905e-07,
+      "loss": 0.8874,
+      "step": 40099
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8754283340338715,
+      "learning_rate": 1.5979449094487898e-07,
+      "loss": 0.8421,
+      "step": 40100
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5659536495015454,
+      "learning_rate": 1.5965865637461163e-07,
+      "loss": 0.935,
+      "step": 40101
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8156388251573479,
+      "learning_rate": 1.595228790979264e-07,
+      "loss": 0.7096,
+      "step": 40102
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7927696503419375,
+      "learning_rate": 1.5938715911561596e-07,
+      "loss": 0.9196,
+      "step": 40103
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5862148806412473,
+      "learning_rate": 1.592514964284686e-07,
+      "loss": 0.7155,
+      "step": 40104
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.73779241907752,
+      "learning_rate": 1.5911589103727253e-07,
+      "loss": 0.899,
+      "step": 40105
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1907894831921793,
+      "learning_rate": 1.589803429428205e-07,
+      "loss": 0.8091,
+      "step": 40106
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7849597981865688,
+      "learning_rate": 1.5884485214590072e-07,
+      "loss": 0.6861,
+      "step": 40107
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4649436713550772,
+      "learning_rate": 1.587094186473026e-07,
+      "loss": 0.8346,
+      "step": 40108
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.687977337356906,
+      "learning_rate": 1.5857404244781326e-07,
+      "loss": 0.9171,
+      "step": 40109
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6901742557907242,
+      "learning_rate": 1.5843872354822099e-07,
+      "loss": 0.8703,
+      "step": 40110
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5084025380740294,
+      "learning_rate": 1.5830346194931513e-07,
+      "loss": 0.8202,
+      "step": 40111
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.371188223149212,
+      "learning_rate": 1.5816825765188171e-07,
+      "loss": 0.8971,
+      "step": 40112
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7636655747744922,
+      "learning_rate": 1.58033110656709e-07,
+      "loss": 0.7163,
+      "step": 40113
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.718722053745637,
+      "learning_rate": 1.5789802096458306e-07,
+      "loss": 0.7665,
+      "step": 40114
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6521989581261793,
+      "learning_rate": 1.5776298857629103e-07,
+      "loss": 0.8282,
+      "step": 40115
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.151660625788899,
+      "learning_rate": 1.576280134926189e-07,
+      "loss": 0.8121,
+      "step": 40116
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.503150169473856,
+      "learning_rate": 1.5749309571435277e-07,
+      "loss": 0.8734,
+      "step": 40117
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3642161035869091,
+      "learning_rate": 1.5735823524227755e-07,
+      "loss": 0.7502,
+      "step": 40118
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9999761068784914,
+      "learning_rate": 1.5722343207717816e-07,
+      "loss": 0.812,
+      "step": 40119
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7591018302561141,
+      "learning_rate": 1.5708868621984175e-07,
+      "loss": 0.7345,
+      "step": 40120
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4520768199515968,
+      "learning_rate": 1.569539976710499e-07,
+      "loss": 0.9266,
+      "step": 40121
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.203760716799924,
+      "learning_rate": 1.5681936643158867e-07,
+      "loss": 0.8222,
+      "step": 40122
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5217664712134862,
+      "learning_rate": 1.566847925022408e-07,
+      "loss": 0.9126,
+      "step": 40123
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4936262087844079,
+      "learning_rate": 1.5655027588379113e-07,
+      "loss": 0.7367,
+      "step": 40124
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6385366769772192,
+      "learning_rate": 1.5641581657702244e-07,
+      "loss": 0.8443,
+      "step": 40125
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6232648257119173,
+      "learning_rate": 1.5628141458271628e-07,
+      "loss": 0.8439,
+      "step": 40126
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7911825826493322,
+      "learning_rate": 1.561470699016565e-07,
+      "loss": 0.9296,
+      "step": 40127
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0947600066076095,
+      "learning_rate": 1.5601278253462582e-07,
+      "loss": 0.8243,
+      "step": 40128
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6918837013868206,
+      "learning_rate": 1.558785524824047e-07,
+      "loss": 0.8765,
+      "step": 40129
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5130036790030776,
+      "learning_rate": 1.5574437974577473e-07,
+      "loss": 0.727,
+      "step": 40130
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7060355844967598,
+      "learning_rate": 1.5561026432551752e-07,
+      "loss": 0.9529,
+      "step": 40131
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7892029055392737,
+      "learning_rate": 1.5547620622241467e-07,
+      "loss": 0.6712,
+      "step": 40132
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.518323989383268,
+      "learning_rate": 1.5534220543724666e-07,
+      "loss": 0.7476,
+      "step": 40133
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7711356213359765,
+      "learning_rate": 1.5520826197079176e-07,
+      "loss": 0.6706,
+      "step": 40134
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5444351216859522,
+      "learning_rate": 1.5507437582383156e-07,
+      "loss": 0.8365,
+      "step": 40135
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.833215335837641,
+      "learning_rate": 1.5494054699714545e-07,
+      "loss": 0.9321,
+      "step": 40136
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7666725834974456,
+      "learning_rate": 1.548067754915128e-07,
+      "loss": 0.689,
+      "step": 40137
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.695928201136358,
+      "learning_rate": 1.5467306130771077e-07,
+      "loss": 0.7773,
+      "step": 40138
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6746037478059883,
+      "learning_rate": 1.5453940444652092e-07,
+      "loss": 0.8279,
+      "step": 40139
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5656757925025113,
+      "learning_rate": 1.5440580490871827e-07,
+      "loss": 0.8346,
+      "step": 40140
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7318147338926968,
+      "learning_rate": 1.5427226269508322e-07,
+      "loss": 0.8609,
+      "step": 40141
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8034692110107715,
+      "learning_rate": 1.5413877780639076e-07,
+      "loss": 0.7565,
+      "step": 40142
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5980242953990396,
+      "learning_rate": 1.5400535024342023e-07,
+      "loss": 0.8066,
+      "step": 40143
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6330656702448114,
+      "learning_rate": 1.538719800069477e-07,
+      "loss": 0.7531,
+      "step": 40144
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.120967958680326,
+      "learning_rate": 1.537386670977503e-07,
+      "loss": 0.9147,
+      "step": 40145
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5278609550206412,
+      "learning_rate": 1.5360541151660303e-07,
+      "loss": 0.9275,
+      "step": 40146
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.259062732069086,
+      "learning_rate": 1.5347221326428187e-07,
+      "loss": 0.8757,
+      "step": 40147
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8730519785086812,
+      "learning_rate": 1.5333907234156288e-07,
+      "loss": 0.8305,
+      "step": 40148
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6268058135190588,
+      "learning_rate": 1.532059887492221e-07,
+      "loss": 0.6734,
+      "step": 40149
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5143323535994493,
+      "learning_rate": 1.530729624880345e-07,
+      "loss": 0.8741,
+      "step": 40150
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5400880413560893,
+      "learning_rate": 1.5293999355877164e-07,
+      "loss": 0.967,
+      "step": 40151
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5986565012019993,
+      "learning_rate": 1.528070819622096e-07,
+      "loss": 0.7255,
+      "step": 40152
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.719393799273309,
+      "learning_rate": 1.526742276991222e-07,
+      "loss": 0.8773,
+      "step": 40153
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.736333534319035,
+      "learning_rate": 1.525414307702855e-07,
+      "loss": 0.7423,
+      "step": 40154
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5099236396514544,
+      "learning_rate": 1.5240869117646774e-07,
+      "loss": 0.7161,
+      "step": 40155
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.648286456233155,
+      "learning_rate": 1.5227600891844495e-07,
+      "loss": 0.8571,
+      "step": 40156
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6619598992659235,
+      "learning_rate": 1.521433839969888e-07,
+      "loss": 0.8715,
+      "step": 40157
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5549809908780658,
+      "learning_rate": 1.5201081641287196e-07,
+      "loss": 0.9063,
+      "step": 40158
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8342778012969798,
+      "learning_rate": 1.5187830616686605e-07,
+      "loss": 0.6515,
+      "step": 40159
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7825897811874989,
+      "learning_rate": 1.5174585325974266e-07,
+      "loss": 0.7052,
+      "step": 40160
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5420694693270622,
+      "learning_rate": 1.516134576922723e-07,
+      "loss": 0.8086,
+      "step": 40161
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7307511674545102,
+      "learning_rate": 1.5148111946522659e-07,
+      "loss": 0.6457,
+      "step": 40162
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.53465306793368,
+      "learning_rate": 1.51348838579376e-07,
+      "loss": 0.8351,
+      "step": 40163
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.641264868510209,
+      "learning_rate": 1.51216615035491e-07,
+      "loss": 0.8612,
+      "step": 40164
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6165497315986972,
+      "learning_rate": 1.5108444883433994e-07,
+      "loss": 0.9275,
+      "step": 40165
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7145151586571037,
+      "learning_rate": 1.5095233997669323e-07,
+      "loss": 0.8053,
+      "step": 40166
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6826653143624835,
+      "learning_rate": 1.508202884633203e-07,
+      "loss": 0.7934,
+      "step": 40167
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7185246080708183,
+      "learning_rate": 1.5068829429499054e-07,
+      "loss": 0.8145,
+      "step": 40168
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5493630128684175,
+      "learning_rate": 1.505563574724711e-07,
+      "loss": 0.7399,
+      "step": 40169
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.951650120111319,
+      "learning_rate": 1.5042447799653136e-07,
+      "loss": 0.7465,
+      "step": 40170
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7372261167333525,
+      "learning_rate": 1.5029265586793851e-07,
+      "loss": 0.7646,
+      "step": 40171
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6266649970374043,
+      "learning_rate": 1.501608910874608e-07,
+      "loss": 0.8171,
+      "step": 40172
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.106766876293942,
+      "learning_rate": 1.5002918365586317e-07,
+      "loss": 0.8295,
+      "step": 40173
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4069077863578272,
+      "learning_rate": 1.4989753357391502e-07,
+      "loss": 0.798,
+      "step": 40174
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7877559406920008,
+      "learning_rate": 1.4976594084238127e-07,
+      "loss": 0.6745,
+      "step": 40175
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7586866297128239,
+      "learning_rate": 1.4963440546203023e-07,
+      "loss": 0.832,
+      "step": 40176
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7850116463986154,
+      "learning_rate": 1.495029274336246e-07,
+      "loss": 0.6798,
+      "step": 40177
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5442290564121417,
+      "learning_rate": 1.4937150675793266e-07,
+      "loss": 0.8993,
+      "step": 40178
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4152459911589113,
+      "learning_rate": 1.4924014343571824e-07,
+      "loss": 0.7651,
+      "step": 40179
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4748660352507368,
+      "learning_rate": 1.491088374677463e-07,
+      "loss": 0.8677,
+      "step": 40180
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6926902270055542,
+      "learning_rate": 1.4897758885478064e-07,
+      "loss": 0.77,
+      "step": 40181
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9164778549859565,
+      "learning_rate": 1.4884639759758735e-07,
+      "loss": 0.75,
+      "step": 40182
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4700804982840525,
+      "learning_rate": 1.48715263696928e-07,
+      "loss": 0.822,
+      "step": 40183
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6864905675822612,
+      "learning_rate": 1.485841871535687e-07,
+      "loss": 0.8444,
+      "step": 40184
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3635301920081413,
+      "learning_rate": 1.4845316796827103e-07,
+      "loss": 0.8184,
+      "step": 40185
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5562643071128517,
+      "learning_rate": 1.483222061417966e-07,
+      "loss": 0.8345,
+      "step": 40186
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6953928679371197,
+      "learning_rate": 1.4819130167491036e-07,
+      "loss": 0.8229,
+      "step": 40187
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7535308446157912,
+      "learning_rate": 1.4806045456837393e-07,
+      "loss": 0.6935,
+      "step": 40188
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.476158601254064,
+      "learning_rate": 1.479296648229478e-07,
+      "loss": 0.8119,
+      "step": 40189
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6244663562230024,
+      "learning_rate": 1.4779893243939358e-07,
+      "loss": 0.8492,
+      "step": 40190
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.511342754452503,
+      "learning_rate": 1.47668257418474e-07,
+      "loss": 0.7751,
+      "step": 40191
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6395840702669862,
+      "learning_rate": 1.4753763976094848e-07,
+      "loss": 0.7778,
+      "step": 40192
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6582146040480623,
+      "learning_rate": 1.474070794675797e-07,
+      "loss": 0.9054,
+      "step": 40193
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5738264606100554,
+      "learning_rate": 1.4727657653912374e-07,
+      "loss": 0.8252,
+      "step": 40194
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.783124892846059,
+      "learning_rate": 1.4714613097634445e-07,
+      "loss": 0.8906,
+      "step": 40195
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.247380040950665,
+      "learning_rate": 1.4701574277999896e-07,
+      "loss": 0.8737,
+      "step": 40196
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5255776512307753,
+      "learning_rate": 1.468854119508478e-07,
+      "loss": 0.9093,
+      "step": 40197
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7409488504767605,
+      "learning_rate": 1.4675513848964928e-07,
+      "loss": 0.7792,
+      "step": 40198
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7964087306153158,
+      "learning_rate": 1.4662492239716164e-07,
+      "loss": 0.8568,
+      "step": 40199
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4299299129135665,
+      "learning_rate": 1.4649476367414318e-07,
+      "loss": 0.9421,
+      "step": 40200
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.705396564485718,
+      "learning_rate": 1.4636466232135217e-07,
+      "loss": 0.919,
+      "step": 40201
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4297970763753025,
+      "learning_rate": 1.4623461833954467e-07,
+      "loss": 0.7034,
+      "step": 40202
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5987517952188335,
+      "learning_rate": 1.4610463172948008e-07,
+      "loss": 0.8368,
+      "step": 40203
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7857945923264458,
+      "learning_rate": 1.4597470249191225e-07,
+      "loss": 1.0255,
+      "step": 40204
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.788029699812321,
+      "learning_rate": 1.458448306276017e-07,
+      "loss": 0.8582,
+      "step": 40205
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7416276643019948,
+      "learning_rate": 1.457150161373011e-07,
+      "loss": 0.8355,
+      "step": 40206
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5198826259674603,
+      "learning_rate": 1.455852590217677e-07,
+      "loss": 0.859,
+      "step": 40207
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8490355901929743,
+      "learning_rate": 1.4545555928175637e-07,
+      "loss": 0.9616,
+      "step": 40208
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.686748206133242,
+      "learning_rate": 1.4532591691802323e-07,
+      "loss": 0.8637,
+      "step": 40209
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6151411828459366,
+      "learning_rate": 1.451963319313221e-07,
+      "loss": 0.7957,
+      "step": 40210
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5232456629810434,
+      "learning_rate": 1.4506680432240794e-07,
+      "loss": 0.7963,
+      "step": 40211
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.534409866387086,
+      "learning_rate": 1.449373340920346e-07,
+      "loss": 0.9029,
+      "step": 40212
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6485993432949815,
+      "learning_rate": 1.4480792124095698e-07,
+      "loss": 0.7564,
+      "step": 40213
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.593912655697192,
+      "learning_rate": 1.446785657699279e-07,
+      "loss": 0.8792,
+      "step": 40214
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6381182331194535,
+      "learning_rate": 1.445492676796989e-07,
+      "loss": 0.7364,
+      "step": 40215
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.568102974515601,
+      "learning_rate": 1.4442002697102497e-07,
+      "loss": 0.7727,
+      "step": 40216
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5498560090474747,
+      "learning_rate": 1.4429084364465885e-07,
+      "loss": 0.7392,
+      "step": 40217
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8109401712785326,
+      "learning_rate": 1.44161717701351e-07,
+      "loss": 0.7039,
+      "step": 40218
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7987525638312085,
+      "learning_rate": 1.440326491418531e-07,
+      "loss": 0.8464,
+      "step": 40219
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.444155813404087,
+      "learning_rate": 1.4390363796691897e-07,
+      "loss": 0.8813,
+      "step": 40220
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6509764787968277,
+      "learning_rate": 1.437746841772969e-07,
+      "loss": 0.8527,
+      "step": 40221
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.463785085087896,
+      "learning_rate": 1.4364578777373962e-07,
+      "loss": 1.0219,
+      "step": 40222
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7771296791930447,
+      "learning_rate": 1.4351694875699763e-07,
+      "loss": 0.8976,
+      "step": 40223
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5711972425293321,
+      "learning_rate": 1.4338816712781921e-07,
+      "loss": 0.8124,
+      "step": 40224
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8658849845364083,
+      "learning_rate": 1.4325944288695603e-07,
+      "loss": 0.7485,
+      "step": 40225
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6689032933122414,
+      "learning_rate": 1.4313077603515745e-07,
+      "loss": 0.7497,
+      "step": 40226
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6857608104826993,
+      "learning_rate": 1.4300216657317178e-07,
+      "loss": 0.7979,
+      "step": 40227
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7859666728105625,
+      "learning_rate": 1.4287361450174843e-07,
+      "loss": 0.821,
+      "step": 40228
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5408129763747405,
+      "learning_rate": 1.4274511982163454e-07,
+      "loss": 0.9256,
+      "step": 40229
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6814112372863794,
+      "learning_rate": 1.4261668253358062e-07,
+      "loss": 0.8756,
+      "step": 40230
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8652575399672955,
+      "learning_rate": 1.424883026383328e-07,
+      "loss": 0.7176,
+      "step": 40231
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8864093071485115,
+      "learning_rate": 1.423599801366382e-07,
+      "loss": 0.7442,
+      "step": 40232
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.71835497388764,
+      "learning_rate": 1.4223171502924516e-07,
+      "loss": 0.6969,
+      "step": 40233
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4598262931699442,
+      "learning_rate": 1.421035073168997e-07,
+      "loss": 0.8709,
+      "step": 40234
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7774133019146652,
+      "learning_rate": 1.4197535700035014e-07,
+      "loss": 0.6293,
+      "step": 40235
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6918480170724397,
+      "learning_rate": 1.418472640803392e-07,
+      "loss": 0.9258,
+      "step": 40236
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9606987665501068,
+      "learning_rate": 1.417192285576141e-07,
+      "loss": 0.8803,
+      "step": 40237
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4868130704219085,
+      "learning_rate": 1.415912504329231e-07,
+      "loss": 0.8799,
+      "step": 40238
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5504060553996337,
+      "learning_rate": 1.4146332970700782e-07,
+      "loss": 0.8042,
+      "step": 40239
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8073011723568504,
+      "learning_rate": 1.4133546638061323e-07,
+      "loss": 0.7323,
+      "step": 40240
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.844298033946689,
+      "learning_rate": 1.4120766045448653e-07,
+      "loss": 0.6801,
+      "step": 40241
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.437893241893318,
+      "learning_rate": 1.4107991192936821e-07,
+      "loss": 0.8034,
+      "step": 40242
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7639402160007398,
+      "learning_rate": 1.4095222080600546e-07,
+      "loss": 0.7928,
+      "step": 40243
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7292209131265532,
+      "learning_rate": 1.408245870851388e-07,
+      "loss": 0.8667,
+      "step": 40244
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4685460004092656,
+      "learning_rate": 1.406970107675143e-07,
+      "loss": 0.8911,
+      "step": 40245
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6459972178921558,
+      "learning_rate": 1.4056949185387136e-07,
+      "loss": 0.8478,
+      "step": 40246
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.541840717806876,
+      "learning_rate": 1.4044203034495497e-07,
+      "loss": 0.8497,
+      "step": 40247
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5225700768039638,
+      "learning_rate": 1.4031462624150782e-07,
+      "loss": 0.7387,
+      "step": 40248
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6286560915910266,
+      "learning_rate": 1.4018727954426937e-07,
+      "loss": 0.9035,
+      "step": 40249
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.588416451730293,
+      "learning_rate": 1.4005999025398231e-07,
+      "loss": 0.9567,
+      "step": 40250
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5526468824944621,
+      "learning_rate": 1.399327583713872e-07,
+      "loss": 0.7818,
+      "step": 40251
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7899660831513109,
+      "learning_rate": 1.3980558389722677e-07,
+      "loss": 0.6743,
+      "step": 40252
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9786663836742802,
+      "learning_rate": 1.3967846683223818e-07,
+      "loss": 0.7446,
+      "step": 40253
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6869226942504596,
+      "learning_rate": 1.3955140717716308e-07,
+      "loss": 0.7577,
+      "step": 40254
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7907438386539086,
+      "learning_rate": 1.3942440493274313e-07,
+      "loss": 0.735,
+      "step": 40255
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.710135564337519,
+      "learning_rate": 1.3929746009971434e-07,
+      "loss": 0.8281,
+      "step": 40256
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4525506206449867,
+      "learning_rate": 1.3917057267881951e-07,
+      "loss": 0.7792,
+      "step": 40257
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0169922997249174,
+      "learning_rate": 1.3904374267079356e-07,
+      "loss": 0.8747,
+      "step": 40258
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1953822314747615,
+      "learning_rate": 1.3891697007637817e-07,
+      "loss": 0.8031,
+      "step": 40259
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4949859854574328,
+      "learning_rate": 1.3879025489630936e-07,
+      "loss": 0.8241,
+      "step": 40260
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3894056911684385,
+      "learning_rate": 1.386635971313266e-07,
+      "loss": 0.7291,
+      "step": 40261
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7245554539424557,
+      "learning_rate": 1.3853699678216592e-07,
+      "loss": 0.8617,
+      "step": 40262
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6362094426718028,
+      "learning_rate": 1.3841045384956453e-07,
+      "loss": 0.8773,
+      "step": 40263
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4998731908803302,
+      "learning_rate": 1.3828396833425962e-07,
+      "loss": 0.6397,
+      "step": 40264
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.984773896845459,
+      "learning_rate": 1.3815754023698834e-07,
+      "loss": 0.8307,
+      "step": 40265
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.664705795309815,
+      "learning_rate": 1.3803116955848573e-07,
+      "loss": 0.8269,
+      "step": 40266
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2974364451221527,
+      "learning_rate": 1.379048562994878e-07,
+      "loss": 0.8832,
+      "step": 40267
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6138641970067593,
+      "learning_rate": 1.3777860046072955e-07,
+      "loss": 0.8352,
+      "step": 40268
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.796257556504015,
+      "learning_rate": 1.3765240204294706e-07,
+      "loss": 0.7123,
+      "step": 40269
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0370758377962384,
+      "learning_rate": 1.3752626104687528e-07,
+      "loss": 0.825,
+      "step": 40270
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.57839313259831,
+      "learning_rate": 1.3740017747324698e-07,
+      "loss": 0.8027,
+      "step": 40271
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3873974878915978,
+      "learning_rate": 1.372741513227982e-07,
+      "loss": 0.7008,
+      "step": 40272
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.976262394364212,
+      "learning_rate": 1.3714818259626173e-07,
+      "loss": 0.7609,
+      "step": 40273
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7916037171270329,
+      "learning_rate": 1.3702227129437028e-07,
+      "loss": 0.6897,
+      "step": 40274
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7609496128511406,
+      "learning_rate": 1.3689641741785775e-07,
+      "loss": 0.8155,
+      "step": 40275
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8416346464034827,
+      "learning_rate": 1.3677062096745686e-07,
+      "loss": 0.7994,
+      "step": 40276
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.139307466866867,
+      "learning_rate": 1.3664488194390147e-07,
+      "loss": 0.7875,
+      "step": 40277
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7321138490310457,
+      "learning_rate": 1.365192003479221e-07,
+      "loss": 0.7559,
+      "step": 40278
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5694118122302294,
+      "learning_rate": 1.363935761802493e-07,
+      "loss": 0.7844,
+      "step": 40279
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2825848775344877,
+      "learning_rate": 1.362680094416169e-07,
+      "loss": 0.8645,
+      "step": 40280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6424469423413404,
+      "learning_rate": 1.3614250013275433e-07,
+      "loss": 0.7917,
+      "step": 40281
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6650698921664906,
+      "learning_rate": 1.3601704825439323e-07,
+      "loss": 0.8177,
+      "step": 40282
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7727338809442573,
+      "learning_rate": 1.35891653807263e-07,
+      "loss": 0.6758,
+      "step": 40283
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5397702448250923,
+      "learning_rate": 1.357663167920953e-07,
+      "loss": 0.8233,
+      "step": 40284
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5756025087520549,
+      "learning_rate": 1.356410372096184e-07,
+      "loss": 0.7384,
+      "step": 40285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.928000146683634,
+      "learning_rate": 1.35515815060564e-07,
+      "loss": 0.8217,
+      "step": 40286
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8507358292792673,
+      "learning_rate": 1.3539065034565813e-07,
+      "loss": 0.8029,
+      "step": 40287
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5929861049356548,
+      "learning_rate": 1.3526554306563134e-07,
+      "loss": 0.6804,
+      "step": 40288
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5434761006398559,
+      "learning_rate": 1.351404932212119e-07,
+      "loss": 0.852,
+      "step": 40289
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7208923501753024,
+      "learning_rate": 1.3501550081312707e-07,
+      "loss": 0.8075,
+      "step": 40290
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5428055121592144,
+      "learning_rate": 1.3489056584210626e-07,
+      "loss": 0.8144,
+      "step": 40291
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6740171090797977,
+      "learning_rate": 1.3476568830887439e-07,
+      "loss": 0.7956,
+      "step": 40292
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.054966743714859,
+      "learning_rate": 1.346408682141609e-07,
+      "loss": 0.7938,
+      "step": 40293
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5387493370668661,
+      "learning_rate": 1.345161055586919e-07,
+      "loss": 0.9951,
+      "step": 40294
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7861381005875785,
+      "learning_rate": 1.3439140034319342e-07,
+      "loss": 0.874,
+      "step": 40295
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7467700755817701,
+      "learning_rate": 1.342667525683905e-07,
+      "loss": 0.5754,
+      "step": 40296
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7819829948832888,
+      "learning_rate": 1.3414216223501031e-07,
+      "loss": 0.7021,
+      "step": 40297
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.58324745880892,
+      "learning_rate": 1.3401762934377894e-07,
+      "loss": 0.7874,
+      "step": 40298
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5506033867092521,
+      "learning_rate": 1.3389315389541911e-07,
+      "loss": 0.8632,
+      "step": 40299
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7933318440472095,
+      "learning_rate": 1.3376873589065809e-07,
+      "loss": 0.9585,
+      "step": 40300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4654803644295165,
+      "learning_rate": 1.3364437533021856e-07,
+      "loss": 0.9397,
+      "step": 40301
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.738672342005081,
+      "learning_rate": 1.3352007221482443e-07,
+      "loss": 0.7149,
+      "step": 40302
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.465276217398694,
+      "learning_rate": 1.3339582654520177e-07,
+      "loss": 0.83,
+      "step": 40303
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.034976937157265,
+      "learning_rate": 1.3327163832207114e-07,
+      "loss": 0.7382,
+      "step": 40304
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1260031661151815,
+      "learning_rate": 1.3314750754615634e-07,
+      "loss": 0.8993,
+      "step": 40305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.66123816192832,
+      "learning_rate": 1.330234342181813e-07,
+      "loss": 0.8193,
+      "step": 40306
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7128436489634047,
+      "learning_rate": 1.3289941833886767e-07,
+      "loss": 0.7728,
+      "step": 40307
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5411816718181814,
+      "learning_rate": 1.3277545990893813e-07,
+      "loss": 0.7936,
+      "step": 40308
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6560292954250857,
+      "learning_rate": 1.326515589291122e-07,
+      "loss": 0.9424,
+      "step": 40309
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.813044020961752,
+      "learning_rate": 1.3252771540011367e-07,
+      "loss": 0.9536,
+      "step": 40310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6511404479023764,
+      "learning_rate": 1.324039293226631e-07,
+      "loss": 0.8441,
+      "step": 40311
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5411712931001713,
+      "learning_rate": 1.3228020069748105e-07,
+      "loss": 0.736,
+      "step": 40312
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9541804407446977,
+      "learning_rate": 1.321565295252869e-07,
+      "loss": 0.8384,
+      "step": 40313
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5529457067680035,
+      "learning_rate": 1.3203291580680122e-07,
+      "loss": 0.6978,
+      "step": 40314
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6898302926505613,
+      "learning_rate": 1.3190935954274565e-07,
+      "loss": 0.7223,
+      "step": 40315
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8027945164395007,
+      "learning_rate": 1.3178586073383738e-07,
+      "loss": 0.7968,
+      "step": 40316
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8015750198824072,
+      "learning_rate": 1.3166241938079472e-07,
+      "loss": 0.9085,
+      "step": 40317
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6402288269352048,
+      "learning_rate": 1.3153903548433822e-07,
+      "loss": 0.93,
+      "step": 40318
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5712056697586516,
+      "learning_rate": 1.314157090451873e-07,
+      "loss": 0.7884,
+      "step": 40319
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6089950962242263,
+      "learning_rate": 1.3129244006405805e-07,
+      "loss": 0.8346,
+      "step": 40320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8203535353208633,
+      "learning_rate": 1.311692285416666e-07,
+      "loss": 0.8205,
+      "step": 40321
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8829434306216015,
+      "learning_rate": 1.3104607447873452e-07,
+      "loss": 0.9585,
+      "step": 40322
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6117056992703824,
+      "learning_rate": 1.3092297787597576e-07,
+      "loss": 0.7345,
+      "step": 40323
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5405330181584083,
+      "learning_rate": 1.3079993873410857e-07,
+      "loss": 0.7702,
+      "step": 40324
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5685943499160038,
+      "learning_rate": 1.3067695705384687e-07,
+      "loss": 0.7818,
+      "step": 40325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0360462949345113,
+      "learning_rate": 1.3055403283591006e-07,
+      "loss": 0.8844,
+      "step": 40326
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9306425841467663,
+      "learning_rate": 1.3043116608101204e-07,
+      "loss": 0.8036,
+      "step": 40327
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7352239698911875,
+      "learning_rate": 1.3030835678986886e-07,
+      "loss": 0.8141,
+      "step": 40328
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6125870796853123,
+      "learning_rate": 1.3018560496319443e-07,
+      "loss": 0.8823,
+      "step": 40329
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.576150691315191,
+      "learning_rate": 1.3006291060170373e-07,
+      "loss": 0.8616,
+      "step": 40330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2156191471295403,
+      "learning_rate": 1.2994027370611173e-07,
+      "loss": 0.8963,
+      "step": 40331
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.084116259677854,
+      "learning_rate": 1.298176942771323e-07,
+      "loss": 0.8645,
+      "step": 40332
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.794970614804213,
+      "learning_rate": 1.2969517231548045e-07,
+      "loss": 0.7938,
+      "step": 40333
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6216629429404925,
+      "learning_rate": 1.295727078218667e-07,
+      "loss": 0.8181,
+      "step": 40334
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4266177287106516,
+      "learning_rate": 1.2945030079700606e-07,
+      "loss": 0.8932,
+      "step": 40335
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.659667256755794,
+      "learning_rate": 1.2932795124161124e-07,
+      "loss": 0.8787,
+      "step": 40336
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.779079153064469,
+      "learning_rate": 1.2920565915639395e-07,
+      "loss": 0.9169,
+      "step": 40337
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6125236013288824,
+      "learning_rate": 1.290834245420658e-07,
+      "loss": 0.8533,
+      "step": 40338
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6370455754069075,
+      "learning_rate": 1.2896124739933846e-07,
+      "loss": 0.7884,
+      "step": 40339
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3274645231768183,
+      "learning_rate": 1.288391277289247e-07,
+      "loss": 0.6816,
+      "step": 40340
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.891030451094919,
+      "learning_rate": 1.2871706553153395e-07,
+      "loss": 0.8492,
+      "step": 40341
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7408951093750835,
+      "learning_rate": 1.2859506080787898e-07,
+      "loss": 0.8997,
+      "step": 40342
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7084319168590165,
+      "learning_rate": 1.284731135586681e-07,
+      "loss": 0.8341,
+      "step": 40343
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7244417042318783,
+      "learning_rate": 1.2835122378461186e-07,
+      "loss": 0.8817,
+      "step": 40344
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3720274295757495,
+      "learning_rate": 1.2822939148641967e-07,
+      "loss": 0.8361,
+      "step": 40345
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7142217124408081,
+      "learning_rate": 1.2810761666480321e-07,
+      "loss": 0.7748,
+      "step": 40346
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.351721506182192,
+      "learning_rate": 1.2798589932046856e-07,
+      "loss": 0.7696,
+      "step": 40347
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.816884427457722,
+      "learning_rate": 1.2786423945412407e-07,
+      "loss": 0.7562,
+      "step": 40348
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6695434597478855,
+      "learning_rate": 1.2774263706648027e-07,
+      "loss": 0.8251,
+      "step": 40349
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.739588142451531,
+      "learning_rate": 1.2762109215824547e-07,
+      "loss": 0.9089,
+      "step": 40350
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4700560306223167,
+      "learning_rate": 1.274996047301258e-07,
+      "loss": 0.7708,
+      "step": 40351
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5053776280907516,
+      "learning_rate": 1.2737817478282845e-07,
+      "loss": 0.7661,
+      "step": 40352
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.623514422987008,
+      "learning_rate": 1.2725680231706061e-07,
+      "loss": 0.8193,
+      "step": 40353
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4910319788632822,
+      "learning_rate": 1.2713548733352955e-07,
+      "loss": 0.8167,
+      "step": 40354
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.487681810444338,
+      "learning_rate": 1.2701422983294243e-07,
+      "loss": 0.8262,
+      "step": 40355
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.393587023676831,
+      "learning_rate": 1.2689302981600316e-07,
+      "loss": 0.8097,
+      "step": 40356
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1360987506040154,
+      "learning_rate": 1.2677188728341895e-07,
+      "loss": 0.6956,
+      "step": 40357
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4884286077385327,
+      "learning_rate": 1.2665080223589366e-07,
+      "loss": 0.9591,
+      "step": 40358
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3525312549360555,
+      "learning_rate": 1.2652977467413453e-07,
+      "loss": 0.7187,
+      "step": 40359
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5953404513335983,
+      "learning_rate": 1.264088045988443e-07,
+      "loss": 0.8087,
+      "step": 40360
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.745128259312513,
+      "learning_rate": 1.26287892010728e-07,
+      "loss": 0.677,
+      "step": 40361
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.73053436827289,
+      "learning_rate": 1.2616703691048947e-07,
+      "loss": 0.8402,
+      "step": 40362
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5409043699342964,
+      "learning_rate": 1.2604623929883265e-07,
+      "loss": 0.7413,
+      "step": 40363
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7666419670842994,
+      "learning_rate": 1.2592549917646136e-07,
+      "loss": 0.7992,
+      "step": 40364
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8120092073223726,
+      "learning_rate": 1.2580481654407616e-07,
+      "loss": 0.9348,
+      "step": 40365
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6815182080934863,
+      "learning_rate": 1.2568419140238208e-07,
+      "loss": 0.8334,
+      "step": 40366
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5716045998116883,
+      "learning_rate": 1.2556362375208186e-07,
+      "loss": 0.7742,
+      "step": 40367
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.750268878664168,
+      "learning_rate": 1.2544311359387606e-07,
+      "loss": 0.7644,
+      "step": 40368
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6621014663693603,
+      "learning_rate": 1.2532266092846524e-07,
+      "loss": 0.7387,
+      "step": 40369
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.638964655258957,
+      "learning_rate": 1.2520226575655325e-07,
+      "loss": 0.7558,
+      "step": 40370
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7507378107051026,
+      "learning_rate": 1.2508192807883956e-07,
+      "loss": 0.9018,
+      "step": 40371
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8162374493067062,
+      "learning_rate": 1.2496164789602583e-07,
+      "loss": 0.7162,
+      "step": 40372
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5619234363543062,
+      "learning_rate": 1.248414252088115e-07,
+      "loss": 0.7813,
+      "step": 40373
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6621895817780137,
+      "learning_rate": 1.2472126001789708e-07,
+      "loss": 0.8781,
+      "step": 40374
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4996559306005681,
+      "learning_rate": 1.2460115232398207e-07,
+      "loss": 0.7713,
+      "step": 40375
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6726562608129116,
+      "learning_rate": 1.2448110212776477e-07,
+      "loss": 0.8172,
+      "step": 40376
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.48579156849211,
+      "learning_rate": 1.2436110942994573e-07,
+      "loss": 0.7676,
+      "step": 40377
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7291286958955172,
+      "learning_rate": 1.242411742312233e-07,
+      "loss": 0.7219,
+      "step": 40378
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.61871919594594,
+      "learning_rate": 1.2412129653229355e-07,
+      "loss": 0.6848,
+      "step": 40379
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5093106600839996,
+      "learning_rate": 1.2400147633385817e-07,
+      "loss": 0.8165,
+      "step": 40380
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7012503179508696,
+      "learning_rate": 1.2388171363661216e-07,
+      "loss": 0.8453,
+      "step": 40381
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.650250081526725,
+      "learning_rate": 1.2376200844125387e-07,
+      "loss": 0.8705,
+      "step": 40382
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6074845656242207,
+      "learning_rate": 1.2364236074847934e-07,
+      "loss": 0.8775,
+      "step": 40383
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7018078419243565,
+      "learning_rate": 1.2352277055898587e-07,
+      "loss": 0.8055,
+      "step": 40384
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5911264786659942,
+      "learning_rate": 1.2340323787347064e-07,
+      "loss": 0.8295,
+      "step": 40385
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5146982178587138,
+      "learning_rate": 1.2328376269262754e-07,
+      "loss": 0.7687,
+      "step": 40386
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6368624230730997,
+      "learning_rate": 1.231643450171527e-07,
+      "loss": 0.7126,
+      "step": 40387
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4748308067810472,
+      "learning_rate": 1.2304498484774329e-07,
+      "loss": 0.8515,
+      "step": 40388
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6316457176474348,
+      "learning_rate": 1.2292568218509325e-07,
+      "loss": 0.9353,
+      "step": 40389
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8687512758080806,
+      "learning_rate": 1.2280643702989647e-07,
+      "loss": 0.7782,
+      "step": 40390
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.67806120900042,
+      "learning_rate": 1.2268724938284682e-07,
+      "loss": 0.8741,
+      "step": 40391
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6772351178354283,
+      "learning_rate": 1.225681192446404e-07,
+      "loss": 0.8026,
+      "step": 40392
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4813314604049892,
+      "learning_rate": 1.224490466159689e-07,
+      "loss": 0.7645,
+      "step": 40393
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6989017476224733,
+      "learning_rate": 1.223300314975262e-07,
+      "loss": 0.914,
+      "step": 40394
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.539077339669637,
+      "learning_rate": 1.2221107389000508e-07,
+      "loss": 0.8259,
+      "step": 40395
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6921868352340268,
+      "learning_rate": 1.2209217379409834e-07,
+      "loss": 0.8412,
+      "step": 40396
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5656947878702092,
+      "learning_rate": 1.219733312104987e-07,
+      "loss": 0.7417,
+      "step": 40397
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8509678646253909,
+      "learning_rate": 1.2185454613989678e-07,
+      "loss": 0.8317,
+      "step": 40398
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7567693201287293,
+      "learning_rate": 1.2173581858298644e-07,
+      "loss": 0.6616,
+      "step": 40399
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6077597620821482,
+      "learning_rate": 1.2161714854045713e-07,
+      "loss": 0.9155,
+      "step": 40400
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.717458360240078,
+      "learning_rate": 1.2149853601299945e-07,
+      "loss": 0.9155,
+      "step": 40401
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9000347727968967,
+      "learning_rate": 1.21379981001305e-07,
+      "loss": 0.9269,
+      "step": 40402
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9075195069638458,
+      "learning_rate": 1.2126148350606438e-07,
+      "loss": 0.7407,
+      "step": 40403
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6443855031946593,
+      "learning_rate": 1.2114304352796703e-07,
+      "loss": 0.7227,
+      "step": 40404
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.207043773928909,
+      "learning_rate": 1.210246610677024e-07,
+      "loss": 0.8917,
+      "step": 40405
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4806875533729558,
+      "learning_rate": 1.2090633612595881e-07,
+      "loss": 0.8444,
+      "step": 40406
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.771405171487588,
+      "learning_rate": 1.2078806870342684e-07,
+      "loss": 0.7743,
+      "step": 40407
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7596528535873264,
+      "learning_rate": 1.2066985880079484e-07,
+      "loss": 0.7401,
+      "step": 40408
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6701525302832414,
+      "learning_rate": 1.2055170641875003e-07,
+      "loss": 0.7305,
+      "step": 40409
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.658359385009155,
+      "learning_rate": 1.2043361155798184e-07,
+      "loss": 0.7495,
+      "step": 40410
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6617789473057984,
+      "learning_rate": 1.2031557421917639e-07,
+      "loss": 0.885,
+      "step": 40411
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8380444272261252,
+      "learning_rate": 1.2019759440302093e-07,
+      "loss": 0.8576,
+      "step": 40412
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7333285567475911,
+      "learning_rate": 1.2007967211020378e-07,
+      "loss": 0.8306,
+      "step": 40413
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9209875875256768,
+      "learning_rate": 1.1996180734141104e-07,
+      "loss": 0.8221,
+      "step": 40414
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7992531450543108,
+      "learning_rate": 1.1984400009732776e-07,
+      "loss": 0.9671,
+      "step": 40415
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2845782677948216,
+      "learning_rate": 1.1972625037864117e-07,
+      "loss": 0.7356,
+      "step": 40416
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.864376023121889,
+      "learning_rate": 1.1960855818603624e-07,
+      "loss": 0.7685,
+      "step": 40417
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8054350941404943,
+      "learning_rate": 1.194909235201991e-07,
+      "loss": 0.707,
+      "step": 40418
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5840113106133968,
+      "learning_rate": 1.1937334638181252e-07,
+      "loss": 0.7827,
+      "step": 40419
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5801296931909312,
+      "learning_rate": 1.1925582677156268e-07,
+      "loss": 0.9252,
+      "step": 40420
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5901559140209924,
+      "learning_rate": 1.1913836469013451e-07,
+      "loss": 0.8811,
+      "step": 40421
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5736447319293918,
+      "learning_rate": 1.1902096013821085e-07,
+      "loss": 0.8281,
+      "step": 40422
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.686619094236789,
+      "learning_rate": 1.1890361311647558e-07,
+      "loss": 0.9103,
+      "step": 40423
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6191161680912516,
+      "learning_rate": 1.1878632362561149e-07,
+      "loss": 0.8145,
+      "step": 40424
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.632527459683214,
+      "learning_rate": 1.1866909166630136e-07,
+      "loss": 0.8584,
+      "step": 40425
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5284135292078744,
+      "learning_rate": 1.185519172392291e-07,
+      "loss": 0.9483,
+      "step": 40426
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6990436195142242,
+      "learning_rate": 1.1843480034507415e-07,
+      "loss": 0.9348,
+      "step": 40427
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.640313130564672,
+      "learning_rate": 1.1831774098452153e-07,
+      "loss": 0.7726,
+      "step": 40428
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5998213721989778,
+      "learning_rate": 1.1820073915825181e-07,
+      "loss": 0.9585,
+      "step": 40429
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.476848171599736,
+      "learning_rate": 1.1808379486694444e-07,
+      "loss": 0.8555,
+      "step": 40430
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7207443653383854,
+      "learning_rate": 1.1796690811128442e-07,
+      "loss": 0.7899,
+      "step": 40431
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7538747676127127,
+      "learning_rate": 1.1785007889194788e-07,
+      "loss": 0.7665,
+      "step": 40432
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9276162834095012,
+      "learning_rate": 1.1773330720961652e-07,
+      "loss": 0.8798,
+      "step": 40433
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5718963075266295,
+      "learning_rate": 1.1761659306497086e-07,
+      "loss": 0.8475,
+      "step": 40434
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.691398372429449,
+      "learning_rate": 1.1749993645869151e-07,
+      "loss": 0.8548,
+      "step": 40435
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7233687254565657,
+      "learning_rate": 1.1738333739145457e-07,
+      "loss": 0.7751,
+      "step": 40436
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8069903508247944,
+      "learning_rate": 1.1726679586394064e-07,
+      "loss": 0.6506,
+      "step": 40437
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0743467230194548,
+      "learning_rate": 1.1715031187682912e-07,
+      "loss": 0.7208,
+      "step": 40438
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7836167260934152,
+      "learning_rate": 1.1703388543079619e-07,
+      "loss": 0.7334,
+      "step": 40439
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7595214350835213,
+      "learning_rate": 1.169175165265224e-07,
+      "loss": 0.8035,
+      "step": 40440
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7804003502716259,
+      "learning_rate": 1.1680120516468274e-07,
+      "loss": 0.7076,
+      "step": 40441
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.62718917758102,
+      "learning_rate": 1.1668495134595558e-07,
+      "loss": 0.7366,
+      "step": 40442
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6159140699207233,
+      "learning_rate": 1.1656875507101706e-07,
+      "loss": 0.7213,
+      "step": 40443
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5216511466571236,
+      "learning_rate": 1.164526163405455e-07,
+      "loss": 0.8951,
+      "step": 40444
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5021123274670332,
+      "learning_rate": 1.1633653515521593e-07,
+      "loss": 0.8441,
+      "step": 40445
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.784951640330543,
+      "learning_rate": 1.1622051151570335e-07,
+      "loss": 0.7892,
+      "step": 40446
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7679811630884197,
+      "learning_rate": 1.161045454226839e-07,
+      "loss": 0.7175,
+      "step": 40447
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.662615011141436,
+      "learning_rate": 1.1598863687683371e-07,
+      "loss": 0.8015,
+      "step": 40448
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.736332986207895,
+      "learning_rate": 1.1587278587882667e-07,
+      "loss": 0.8612,
+      "step": 40449
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5761109773939896,
+      "learning_rate": 1.157569924293367e-07,
+      "loss": 0.719,
+      "step": 40450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.756985614188928,
+      "learning_rate": 1.156412565290399e-07,
+      "loss": 0.774,
+      "step": 40451
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4724847696665677,
+      "learning_rate": 1.155255781786091e-07,
+      "loss": 0.8038,
+      "step": 40452
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0526494375029753,
+      "learning_rate": 1.1540995737871708e-07,
+      "loss": 0.835,
+      "step": 40453
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5684502929595663,
+      "learning_rate": 1.1529439413003773e-07,
+      "loss": 0.7096,
+      "step": 40454
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.770221218198924,
+      "learning_rate": 1.1517888843324387e-07,
+      "loss": 0.6863,
+      "step": 40455
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9440827768292126,
+      "learning_rate": 1.1506344028900829e-07,
+      "loss": 0.877,
+      "step": 40456
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6864025893433015,
+      "learning_rate": 1.1494804969800377e-07,
+      "loss": 0.877,
+      "step": 40457
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6933631013861685,
+      "learning_rate": 1.148327166608998e-07,
+      "loss": 0.7982,
+      "step": 40458
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5303221081406237,
+      "learning_rate": 1.1471744117837024e-07,
+      "loss": 0.7971,
+      "step": 40459
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5287688476149934,
+      "learning_rate": 1.1460222325108461e-07,
+      "loss": 0.869,
+      "step": 40460
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6462235955522293,
+      "learning_rate": 1.1448706287971567e-07,
+      "loss": 0.6569,
+      "step": 40461
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6236777009985885,
+      "learning_rate": 1.1437196006493179e-07,
+      "loss": 0.8666,
+      "step": 40462
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6598153171836405,
+      "learning_rate": 1.1425691480740353e-07,
+      "loss": 0.8983,
+      "step": 40463
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4323508745882811,
+      "learning_rate": 1.1414192710780259e-07,
+      "loss": 0.7564,
+      "step": 40464
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7459833537742748,
+      "learning_rate": 1.140269969667962e-07,
+      "loss": 0.7187,
+      "step": 40465
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3405101913288324,
+      "learning_rate": 1.1391212438505605e-07,
+      "loss": 0.7483,
+      "step": 40466
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6209571274168018,
+      "learning_rate": 1.1379730936324717e-07,
+      "loss": 0.8551,
+      "step": 40467
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.664560478767157,
+      "learning_rate": 1.1368255190204125e-07,
+      "loss": 0.7942,
+      "step": 40468
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5228890836233813,
+      "learning_rate": 1.1356785200210552e-07,
+      "loss": 0.6747,
+      "step": 40469
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5754990373936615,
+      "learning_rate": 1.1345320966410833e-07,
+      "loss": 0.8177,
+      "step": 40470
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6550430511677383,
+      "learning_rate": 1.1333862488871583e-07,
+      "loss": 0.8204,
+      "step": 40471
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5899484799294317,
+      "learning_rate": 1.1322409767659526e-07,
+      "loss": 0.8435,
+      "step": 40472
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7656317020910246,
+      "learning_rate": 1.1310962802841607e-07,
+      "loss": 0.7033,
+      "step": 40473
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1212134998839214,
+      "learning_rate": 1.129952159448422e-07,
+      "loss": 0.7797,
+      "step": 40474
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5251203631957417,
+      "learning_rate": 1.1288086142653865e-07,
+      "loss": 0.7962,
+      "step": 40475
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5705397613661425,
+      "learning_rate": 1.1276656447417378e-07,
+      "loss": 0.8203,
+      "step": 40476
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8907881113573441,
+      "learning_rate": 1.1265232508841262e-07,
+      "loss": 0.7326,
+      "step": 40477
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.702286472515828,
+      "learning_rate": 1.1253814326992019e-07,
+      "loss": 0.8873,
+      "step": 40478
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5691733959695933,
+      "learning_rate": 1.124240190193604e-07,
+      "loss": 0.7941,
+      "step": 40479
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7511417329263395,
+      "learning_rate": 1.1230995233739939e-07,
+      "loss": 0.9311,
+      "step": 40480
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6014863687031502,
+      "learning_rate": 1.1219594322469885e-07,
+      "loss": 0.8299,
+      "step": 40481
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.556165622213003,
+      "learning_rate": 1.1208199168192491e-07,
+      "loss": 0.7745,
+      "step": 40482
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6734154341593714,
+      "learning_rate": 1.1196809770973926e-07,
+      "loss": 0.7817,
+      "step": 40483
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5058859621090028,
+      "learning_rate": 1.1185426130880694e-07,
+      "loss": 0.8877,
+      "step": 40484
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7139075048267576,
+      "learning_rate": 1.1174048247978852e-07,
+      "loss": 0.6397,
+      "step": 40485
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5721994076395398,
+      "learning_rate": 1.1162676122334793e-07,
+      "loss": 0.8321,
+      "step": 40486
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7980796969446542,
+      "learning_rate": 1.1151309754014683e-07,
+      "loss": 0.8497,
+      "step": 40487
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.632862039976945,
+      "learning_rate": 1.1139949143084804e-07,
+      "loss": 0.7505,
+      "step": 40488
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7214177501260586,
+      "learning_rate": 1.1128594289611105e-07,
+      "loss": 0.8946,
+      "step": 40489
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9742106397442005,
+      "learning_rate": 1.1117245193659864e-07,
+      "loss": 0.8155,
+      "step": 40490
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6472489126611727,
+      "learning_rate": 1.110590185529703e-07,
+      "loss": 0.9181,
+      "step": 40491
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7815643305542098,
+      "learning_rate": 1.109456427458877e-07,
+      "loss": 0.6833,
+      "step": 40492
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7615990589483432,
+      "learning_rate": 1.1083232451600922e-07,
+      "loss": 0.8831,
+      "step": 40493
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8129378725495666,
+      "learning_rate": 1.1071906386399767e-07,
+      "loss": 0.9294,
+      "step": 40494
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7507567417476797,
+      "learning_rate": 1.1060586079050917e-07,
+      "loss": 0.9272,
+      "step": 40495
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4822969288182755,
+      "learning_rate": 1.1049271529620542e-07,
+      "loss": 0.8807,
+      "step": 40496
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4760723115775884,
+      "learning_rate": 1.1037962738174257e-07,
+      "loss": 0.8184,
+      "step": 40497
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7404901977680959,
+      "learning_rate": 1.102665970477823e-07,
+      "loss": 0.694,
+      "step": 40498
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4752601029903951,
+      "learning_rate": 1.1015362429497967e-07,
+      "loss": 0.7707,
+      "step": 40499
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6178908806880183,
+      "learning_rate": 1.1004070912399411e-07,
+      "loss": 0.8432,
+      "step": 40500
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6527402342727064,
+      "learning_rate": 1.099278515354829e-07,
+      "loss": 0.8154,
+      "step": 40501
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5027771081694554,
+      "learning_rate": 1.0981505153010218e-07,
+      "loss": 0.8747,
+      "step": 40502
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5248872031216913,
+      "learning_rate": 1.0970230910851031e-07,
+      "loss": 0.9512,
+      "step": 40503
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7235604364204464,
+      "learning_rate": 1.0958962427136232e-07,
+      "loss": 0.867,
+      "step": 40504
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.642865246061573,
+      "learning_rate": 1.0947699701931436e-07,
+      "loss": 0.77,
+      "step": 40505
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5850261847887246,
+      "learning_rate": 1.0936442735302255e-07,
+      "loss": 0.7606,
+      "step": 40506
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8520679911822848,
+      "learning_rate": 1.0925191527314416e-07,
+      "loss": 0.7803,
+      "step": 40507
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7560844330193024,
+      "learning_rate": 1.0913946078033089e-07,
+      "loss": 0.8935,
+      "step": 40508
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5608138205889361,
+      "learning_rate": 1.0902706387523887e-07,
+      "loss": 0.9634,
+      "step": 40509
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.519073048852936,
+      "learning_rate": 1.0891472455852315e-07,
+      "loss": 0.8882,
+      "step": 40510
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8521155238369793,
+      "learning_rate": 1.0880244283083762e-07,
+      "loss": 0.8343,
+      "step": 40511
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.630164009599729,
+      "learning_rate": 1.0869021869283625e-07,
+      "loss": 0.7829,
+      "step": 40512
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6303691376489853,
+      "learning_rate": 1.085780521451707e-07,
+      "loss": 0.7646,
+      "step": 40513
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.811364112499563,
+      "learning_rate": 1.0846594318849602e-07,
+      "loss": 0.7546,
+      "step": 40514
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.696169916008361,
+      "learning_rate": 1.0835389182346389e-07,
+      "loss": 0.8601,
+      "step": 40515
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.526262576810265,
+      "learning_rate": 1.0824189805072827e-07,
+      "loss": 0.8414,
+      "step": 40516
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8311752453990683,
+      "learning_rate": 1.081299618709386e-07,
+      "loss": 0.7027,
+      "step": 40517
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6959137107003104,
+      "learning_rate": 1.0801808328474772e-07,
+      "loss": 0.8299,
+      "step": 40518
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5619092893651259,
+      "learning_rate": 1.0790626229280842e-07,
+      "loss": 0.7899,
+      "step": 40519
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7662731256814928,
+      "learning_rate": 1.077944988957702e-07,
+      "loss": 0.8403,
+      "step": 40520
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6921703865978401,
+      "learning_rate": 1.0768279309428365e-07,
+      "loss": 0.745,
+      "step": 40521
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5351856697290163,
+      "learning_rate": 1.0757114488900156e-07,
+      "loss": 0.7556,
+      "step": 40522
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.625718155759782,
+      "learning_rate": 1.0745955428057009e-07,
+      "loss": 0.8138,
+      "step": 40523
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7771835919198478,
+      "learning_rate": 1.0734802126964206e-07,
+      "loss": 0.7527,
+      "step": 40524
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6169402761478913,
+      "learning_rate": 1.0723654585686582e-07,
+      "loss": 0.8167,
+      "step": 40525
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5103089183888456,
+      "learning_rate": 1.0712512804288977e-07,
+      "loss": 0.8202,
+      "step": 40526
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7212255995816979,
+      "learning_rate": 1.0701376782836337e-07,
+      "loss": 0.8191,
+      "step": 40527
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8185244139721182,
+      "learning_rate": 1.0690246521393499e-07,
+      "loss": 0.7849,
+      "step": 40528
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5539278666667633,
+      "learning_rate": 1.0679122020025301e-07,
+      "loss": 0.9197,
+      "step": 40529
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7035802110455007,
+      "learning_rate": 1.0668003278796468e-07,
+      "loss": 0.8564,
+      "step": 40530
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6748694479282868,
+      "learning_rate": 1.0656890297771727e-07,
+      "loss": 0.8065,
+      "step": 40531
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5339953450393107,
+      "learning_rate": 1.0645783077015692e-07,
+      "loss": 0.786,
+      "step": 40532
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7370682737405699,
+      "learning_rate": 1.0634681616593312e-07,
+      "loss": 0.6777,
+      "step": 40533
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7923242496671301,
+      "learning_rate": 1.0623585916568979e-07,
+      "loss": 0.7198,
+      "step": 40534
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7804162894544255,
+      "learning_rate": 1.0612495977007197e-07,
+      "loss": 0.8387,
+      "step": 40535
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.488457372438401,
+      "learning_rate": 1.0601411797972804e-07,
+      "loss": 0.8682,
+      "step": 40536
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6166598172053972,
+      "learning_rate": 1.0590333379530304e-07,
+      "loss": 0.8168,
+      "step": 40537
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7829214434025629,
+      "learning_rate": 1.0579260721744089e-07,
+      "loss": 0.7894,
+      "step": 40538
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8649279852382359,
+      "learning_rate": 1.0568193824678663e-07,
+      "loss": 0.8523,
+      "step": 40539
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7104926845488753,
+      "learning_rate": 1.055713268839853e-07,
+      "loss": 0.753,
+      "step": 40540
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.647085964079751,
+      "learning_rate": 1.0546077312967862e-07,
+      "loss": 0.8165,
+      "step": 40541
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7836301034918618,
+      "learning_rate": 1.0535027698451272e-07,
+      "loss": 0.7987,
+      "step": 40542
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8024278647653047,
+      "learning_rate": 1.0523983844913043e-07,
+      "loss": 0.7614,
+      "step": 40543
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8953725061711786,
+      "learning_rate": 1.0512945752417347e-07,
+      "loss": 0.9043,
+      "step": 40544
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7973558376815408,
+      "learning_rate": 1.0501913421028575e-07,
+      "loss": 0.9481,
+      "step": 40545
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.678212903361801,
+      "learning_rate": 1.049088685081101e-07,
+      "loss": 0.8034,
+      "step": 40546
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5430033471281839,
+      "learning_rate": 1.0479866041828824e-07,
+      "loss": 0.81,
+      "step": 40547
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7639846745377534,
+      "learning_rate": 1.0468850994145962e-07,
+      "loss": 0.762,
+      "step": 40548
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.576186067581241,
+      "learning_rate": 1.0457841707826822e-07,
+      "loss": 0.8094,
+      "step": 40549
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.715028331815283,
+      "learning_rate": 1.0446838182935349e-07,
+      "loss": 0.7752,
+      "step": 40550
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4640491096691415,
+      "learning_rate": 1.0435840419535826e-07,
+      "loss": 0.7841,
+      "step": 40551
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0003835149563516,
+      "learning_rate": 1.042484841769198e-07,
+      "loss": 0.8862,
+      "step": 40552
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9203607728107122,
+      "learning_rate": 1.0413862177467982e-07,
+      "loss": 0.9134,
+      "step": 40553
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7828685763236833,
+      "learning_rate": 1.040288169892778e-07,
+      "loss": 0.8254,
+      "step": 40554
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8539077422656063,
+      "learning_rate": 1.0391906982135324e-07,
+      "loss": 0.6898,
+      "step": 40555
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8601782082841152,
+      "learning_rate": 1.0380938027154452e-07,
+      "loss": 0.8369,
+      "step": 40556
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3822008195843902,
+      "learning_rate": 1.0369974834049112e-07,
+      "loss": 0.8009,
+      "step": 40557
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6322088920443112,
+      "learning_rate": 1.0359017402883143e-07,
+      "loss": 0.8057,
+      "step": 40558
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5246600249167341,
+      "learning_rate": 1.0348065733720269e-07,
+      "loss": 0.7211,
+      "step": 40559
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.579411044127634,
+      "learning_rate": 1.0337119826624109e-07,
+      "loss": 0.8164,
+      "step": 40560
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4562378216081582,
+      "learning_rate": 1.0326179681658721e-07,
+      "loss": 0.8428,
+      "step": 40561
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5173075099762223,
+      "learning_rate": 1.031524529888761e-07,
+      "loss": 0.8075,
+      "step": 40562
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5105223813084439,
+      "learning_rate": 1.0304316678374504e-07,
+      "loss": 0.8245,
+      "step": 40563
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5797119202896448,
+      "learning_rate": 1.0293393820182907e-07,
+      "loss": 0.8238,
+      "step": 40564
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4798045310268693,
+      "learning_rate": 1.0282476724376544e-07,
+      "loss": 0.8215,
+      "step": 40565
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.561360131971744,
+      "learning_rate": 1.0271565391018922e-07,
+      "loss": 0.8735,
+      "step": 40566
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4459278864629963,
+      "learning_rate": 1.0260659820173658e-07,
+      "loss": 0.8644,
+      "step": 40567
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4754511137799589,
+      "learning_rate": 1.0249760011904253e-07,
+      "loss": 0.7707,
+      "step": 40568
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2493422414694972,
+      "learning_rate": 1.0238865966273881e-07,
+      "loss": 0.9004,
+      "step": 40569
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7771508356803576,
+      "learning_rate": 1.0227977683346268e-07,
+      "loss": 0.8851,
+      "step": 40570
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5320815358263042,
+      "learning_rate": 1.021709516318481e-07,
+      "loss": 0.8362,
+      "step": 40571
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4284419707491471,
+      "learning_rate": 1.0206218405852786e-07,
+      "loss": 0.7442,
+      "step": 40572
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5143186378598785,
+      "learning_rate": 1.019534741141337e-07,
+      "loss": 0.8852,
+      "step": 40573
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5418069267960088,
+      "learning_rate": 1.0184482179930066e-07,
+      "loss": 0.7038,
+      "step": 40574
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0415693075643206,
+      "learning_rate": 1.0173622711466157e-07,
+      "loss": 0.7556,
+      "step": 40575
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5700386060040747,
+      "learning_rate": 1.0162769006084705e-07,
+      "loss": 0.8483,
+      "step": 40576
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6756952751803385,
+      "learning_rate": 1.0151921063848991e-07,
+      "loss": 0.8862,
+      "step": 40577
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8314753561785004,
+      "learning_rate": 1.0141078884822187e-07,
+      "loss": 0.7967,
+      "step": 40578
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4727393715289339,
+      "learning_rate": 1.0130242469067464e-07,
+      "loss": 0.8797,
+      "step": 40579
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6423765295431674,
+      "learning_rate": 1.0119411816647884e-07,
+      "loss": 0.8369,
+      "step": 40580
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3595396165414766,
+      "learning_rate": 1.0108586927626285e-07,
+      "loss": 0.694,
+      "step": 40581
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.264471432733326,
+      "learning_rate": 1.009776780206606e-07,
+      "loss": 0.677,
+      "step": 40582
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.902211970244713,
+      "learning_rate": 1.008695444002994e-07,
+      "loss": 0.9014,
+      "step": 40583
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0150386359884824,
+      "learning_rate": 1.0076146841581091e-07,
+      "loss": 0.9123,
+      "step": 40584
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5684537354069434,
+      "learning_rate": 1.0065345006782246e-07,
+      "loss": 0.8233,
+      "step": 40585
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4094260536386405,
+      "learning_rate": 1.0054548935696351e-07,
+      "loss": 0.8957,
+      "step": 40586
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4831290833759387,
+      "learning_rate": 1.0043758628386247e-07,
+      "loss": 0.8216,
+      "step": 40587
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9138707584322876,
+      "learning_rate": 1.0032974084914882e-07,
+      "loss": 0.714,
+      "step": 40588
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7275536039673325,
+      "learning_rate": 1.0022195305344984e-07,
+      "loss": 0.6845,
+      "step": 40589
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7661802416918537,
+      "learning_rate": 1.0011422289739169e-07,
+      "loss": 0.6872,
+      "step": 40590
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5805901539836014,
+      "learning_rate": 1.0000655038160278e-07,
+      "loss": 0.8721,
+      "step": 40591
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7175315908092696,
+      "learning_rate": 9.989893550671148e-08,
+      "loss": 0.848,
+      "step": 40592
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5817153418594712,
+      "learning_rate": 9.979137827334173e-08,
+      "loss": 0.8471,
+      "step": 40593
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5594375132398814,
+      "learning_rate": 9.968387868212192e-08,
+      "loss": 0.8438,
+      "step": 40594
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5572437391765577,
+      "learning_rate": 9.95764367336749e-08,
+      "loss": 0.8126,
+      "step": 40595
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6261852758894262,
+      "learning_rate": 9.946905242863014e-08,
+      "loss": 0.7615,
+      "step": 40596
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.897492176120007,
+      "learning_rate": 9.93617257676116e-08,
+      "loss": 0.6878,
+      "step": 40597
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7170690123586068,
+      "learning_rate": 9.925445675124212e-08,
+      "loss": 0.7835,
+      "step": 40598
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5559890379985382,
+      "learning_rate": 9.914724538014787e-08,
+      "loss": 0.8989,
+      "step": 40599
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3439818659720175,
+      "learning_rate": 9.904009165495276e-08,
+      "loss": 0.822,
+      "step": 40600
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.845789196893733,
+      "learning_rate": 9.893299557628188e-08,
+      "loss": 0.8081,
+      "step": 40601
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6463374677966633,
+      "learning_rate": 9.882595714475696e-08,
+      "loss": 0.8164,
+      "step": 40602
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6299508053213814,
+      "learning_rate": 9.871897636100191e-08,
+      "loss": 0.8024,
+      "step": 40603
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5771531135420378,
+      "learning_rate": 9.86120532256396e-08,
+      "loss": 0.7631,
+      "step": 40604
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8923946215132867,
+      "learning_rate": 9.850518773929286e-08,
+      "loss": 0.7937,
+      "step": 40605
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7297316296405285,
+      "learning_rate": 9.839837990258338e-08,
+      "loss": 0.7445,
+      "step": 40606
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7560011224005536,
+      "learning_rate": 9.829162971613405e-08,
+      "loss": 0.7811,
+      "step": 40607
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5604361692277509,
+      "learning_rate": 9.818493718056543e-08,
+      "loss": 0.8752,
+      "step": 40608
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6784446489606815,
+      "learning_rate": 9.807830229649929e-08,
+      "loss": 0.7752,
+      "step": 40609
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5305556033623107,
+      "learning_rate": 9.79717250645551e-08,
+      "loss": 0.8433,
+      "step": 40610
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3710935493225238,
+      "learning_rate": 9.78652054853546e-08,
+      "loss": 0.8392,
+      "step": 40611
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6141881458054435,
+      "learning_rate": 9.775874355951842e-08,
+      "loss": 0.7752,
+      "step": 40612
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5448548568078315,
+      "learning_rate": 9.765233928766493e-08,
+      "loss": 0.9718,
+      "step": 40613
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6399998033488417,
+      "learning_rate": 9.754599267041697e-08,
+      "loss": 0.7339,
+      "step": 40614
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5183131153451774,
+      "learning_rate": 9.743970370838963e-08,
+      "loss": 0.9053,
+      "step": 40615
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.497544551047514,
+      "learning_rate": 9.733347240220237e-08,
+      "loss": 0.9322,
+      "step": 40616
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.852439246334656,
+      "learning_rate": 9.722729875247694e-08,
+      "loss": 0.8165,
+      "step": 40617
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6144394148250338,
+      "learning_rate": 9.71211827598284e-08,
+      "loss": 0.9979,
+      "step": 40618
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6011097798240317,
+      "learning_rate": 9.701512442487515e-08,
+      "loss": 0.7442,
+      "step": 40619
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6500795570812992,
+      "learning_rate": 9.690912374823446e-08,
+      "loss": 0.8824,
+      "step": 40620
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5519330172893309,
+      "learning_rate": 9.680318073052474e-08,
+      "loss": 0.9003,
+      "step": 40621
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4813989090268616,
+      "learning_rate": 9.669729537236327e-08,
+      "loss": 0.6919,
+      "step": 40622
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7805249301005127,
+      "learning_rate": 9.6591467674364e-08,
+      "loss": 0.7148,
+      "step": 40623
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6377347314170279,
+      "learning_rate": 9.648569763714532e-08,
+      "loss": 0.7261,
+      "step": 40624
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0647635157485262,
+      "learning_rate": 9.637998526132231e-08,
+      "loss": 0.8414,
+      "step": 40625
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6039792346181898,
+      "learning_rate": 9.627433054751001e-08,
+      "loss": 0.993,
+      "step": 40626
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4439714425343073,
+      "learning_rate": 9.616873349632461e-08,
+      "loss": 0.9017,
+      "step": 40627
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5255617297494724,
+      "learning_rate": 9.606319410838006e-08,
+      "loss": 0.8725,
+      "step": 40628
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.483672268924914,
+      "learning_rate": 9.595771238429142e-08,
+      "loss": 0.7734,
+      "step": 40629
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.671798313767153,
+      "learning_rate": 9.585228832467153e-08,
+      "loss": 0.7592,
+      "step": 40630
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4554925972126895,
+      "learning_rate": 9.574692193013769e-08,
+      "loss": 0.7754,
+      "step": 40631
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.545596992598292,
+      "learning_rate": 9.564161320129828e-08,
+      "loss": 0.8339,
+      "step": 40632
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7580805691606656,
+      "learning_rate": 9.553636213876949e-08,
+      "loss": 0.717,
+      "step": 40633
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5474839253760189,
+      "learning_rate": 9.543116874316416e-08,
+      "loss": 0.8465,
+      "step": 40634
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8148598734267571,
+      "learning_rate": 9.532603301509514e-08,
+      "loss": 0.8431,
+      "step": 40635
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7731484138703991,
+      "learning_rate": 9.522095495517303e-08,
+      "loss": 0.6569,
+      "step": 40636
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4530020837950406,
+      "learning_rate": 9.511593456400958e-08,
+      "loss": 0.7916,
+      "step": 40637
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7328681607000043,
+      "learning_rate": 9.501097184221763e-08,
+      "loss": 0.8926,
+      "step": 40638
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7701208950330063,
+      "learning_rate": 9.490606679040782e-08,
+      "loss": 0.6718,
+      "step": 40639
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7824091361532287,
+      "learning_rate": 9.480121940919074e-08,
+      "loss": 0.7137,
+      "step": 40640
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7200542465268941,
+      "learning_rate": 9.469642969917592e-08,
+      "loss": 0.6684,
+      "step": 40641
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9333616324803697,
+      "learning_rate": 9.45916976609762e-08,
+      "loss": 0.8332,
+      "step": 40642
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7483348506499939,
+      "learning_rate": 9.448702329519777e-08,
+      "loss": 0.6625,
+      "step": 40643
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.655904388615971,
+      "learning_rate": 9.438240660245346e-08,
+      "loss": 0.9743,
+      "step": 40644
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.483705896738853,
+      "learning_rate": 9.427784758335056e-08,
+      "loss": 0.8954,
+      "step": 40645
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7351505799821616,
+      "learning_rate": 9.41733462384975e-08,
+      "loss": 0.8195,
+      "step": 40646
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8644077929224288,
+      "learning_rate": 9.406890256850376e-08,
+      "loss": 0.9264,
+      "step": 40647
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.633897102660517,
+      "learning_rate": 9.396451657397666e-08,
+      "loss": 0.8577,
+      "step": 40648
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7646269676246107,
+      "learning_rate": 9.386018825552568e-08,
+      "loss": 0.8632,
+      "step": 40649
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7481254826654479,
+      "learning_rate": 9.37559176137548e-08,
+      "loss": 0.7014,
+      "step": 40650
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5554449874329428,
+      "learning_rate": 9.365170464927464e-08,
+      "loss": 0.8356,
+      "step": 40651
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5632241478757165,
+      "learning_rate": 9.354754936269029e-08,
+      "loss": 0.9675,
+      "step": 40652
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4841827885352759,
+      "learning_rate": 9.344345175460901e-08,
+      "loss": 0.9357,
+      "step": 40653
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8172670615740991,
+      "learning_rate": 9.333941182563477e-08,
+      "loss": 0.8228,
+      "step": 40654
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6613052086401048,
+      "learning_rate": 9.323542957637599e-08,
+      "loss": 0.7803,
+      "step": 40655
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6967237265780897,
+      "learning_rate": 9.313150500743661e-08,
+      "loss": 0.8067,
+      "step": 40656
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5025836283660172,
+      "learning_rate": 9.302763811942283e-08,
+      "loss": 0.868,
+      "step": 40657
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6741388458967592,
+      "learning_rate": 9.29238289129375e-08,
+      "loss": 0.7864,
+      "step": 40658
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.875165074718047,
+      "learning_rate": 9.282007738858679e-08,
+      "loss": 0.9086,
+      "step": 40659
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6511599763118776,
+      "learning_rate": 9.271638354697354e-08,
+      "loss": 0.8146,
+      "step": 40660
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0370382856199476,
+      "learning_rate": 9.261274738870286e-08,
+      "loss": 0.8142,
+      "step": 40661
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5109276406122167,
+      "learning_rate": 9.250916891437756e-08,
+      "loss": 0.7841,
+      "step": 40662
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7736404357220867,
+      "learning_rate": 9.24056481245994e-08,
+      "loss": 0.7836,
+      "step": 40663
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4000215608113882,
+      "learning_rate": 9.230218501997346e-08,
+      "loss": 0.7431,
+      "step": 40664
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6900024522900337,
+      "learning_rate": 9.219877960110036e-08,
+      "loss": 0.7717,
+      "step": 40665
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7682802758172986,
+      "learning_rate": 9.209543186858294e-08,
+      "loss": 0.6862,
+      "step": 40666
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5752180566257252,
+      "learning_rate": 9.199214182302186e-08,
+      "loss": 0.6967,
+      "step": 40667
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5169842532960196,
+      "learning_rate": 9.188890946501993e-08,
+      "loss": 0.8607,
+      "step": 40668
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5229392887252227,
+      "learning_rate": 9.178573479517783e-08,
+      "loss": 0.8006,
+      "step": 40669
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.947507295120373,
+      "learning_rate": 9.168261781409726e-08,
+      "loss": 0.8647,
+      "step": 40670
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6628163077347553,
+      "learning_rate": 9.157955852237554e-08,
+      "loss": 0.8123,
+      "step": 40671
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.708594563185539,
+      "learning_rate": 9.147655692061442e-08,
+      "loss": 0.8949,
+      "step": 40672
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6746747211501218,
+      "learning_rate": 9.13736130094145e-08,
+      "loss": 0.8777,
+      "step": 40673
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.194571510635171,
+      "learning_rate": 9.127072678937532e-08,
+      "loss": 0.7055,
+      "step": 40674
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7081533942529488,
+      "learning_rate": 9.116789826109417e-08,
+      "loss": 0.879,
+      "step": 40675
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5787115810530545,
+      "learning_rate": 9.106512742516948e-08,
+      "loss": 0.8964,
+      "step": 40676
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7658439269132062,
+      "learning_rate": 9.096241428220299e-08,
+      "loss": 0.7244,
+      "step": 40677
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6266622980695928,
+      "learning_rate": 9.085975883278864e-08,
+      "loss": 0.8001,
+      "step": 40678
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5542353262176567,
+      "learning_rate": 9.075716107752708e-08,
+      "loss": 0.7194,
+      "step": 40679
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7703748885500692,
+      "learning_rate": 9.065462101701339e-08,
+      "loss": 0.7148,
+      "step": 40680
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7649492243687712,
+      "learning_rate": 9.055213865184708e-08,
+      "loss": 0.7946,
+      "step": 40681
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6341235778671555,
+      "learning_rate": 9.044971398262325e-08,
+      "loss": 0.7564,
+      "step": 40682
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.50234599280441,
+      "learning_rate": 9.034734700993696e-08,
+      "loss": 0.7879,
+      "step": 40683
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.540087234715628,
+      "learning_rate": 9.024503773438776e-08,
+      "loss": 0.8424,
+      "step": 40684
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6983443300802277,
+      "learning_rate": 9.014278615656735e-08,
+      "loss": 0.8231,
+      "step": 40685
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7455432372571125,
+      "learning_rate": 9.004059227707528e-08,
+      "loss": 0.7351,
+      "step": 40686
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7256198899213584,
+      "learning_rate": 8.99384560965022e-08,
+      "loss": 0.8922,
+      "step": 40687
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.562747691999976,
+      "learning_rate": 8.98363776154465e-08,
+      "loss": 0.7688,
+      "step": 40688
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7373577393637012,
+      "learning_rate": 8.973435683449994e-08,
+      "loss": 0.8927,
+      "step": 40689
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.465419377893086,
+      "learning_rate": 8.96323937542587e-08,
+      "loss": 0.8375,
+      "step": 40690
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9284093600060115,
+      "learning_rate": 8.953048837531453e-08,
+      "loss": 0.6869,
+      "step": 40691
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4249928807817136,
+      "learning_rate": 8.94286406982603e-08,
+      "loss": 0.7791,
+      "step": 40692
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5816175743579,
+      "learning_rate": 8.932685072369107e-08,
+      "loss": 0.9119,
+      "step": 40693
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6072879561621074,
+      "learning_rate": 8.922511845219972e-08,
+      "loss": 0.7943,
+      "step": 40694
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5268810288261843,
+      "learning_rate": 8.912344388437577e-08,
+      "loss": 0.8166,
+      "step": 40695
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.011482454225133,
+      "learning_rate": 8.902182702081318e-08,
+      "loss": 0.821,
+      "step": 40696
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.665076781832182,
+      "learning_rate": 8.892026786210372e-08,
+      "loss": 0.7242,
+      "step": 40697
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.912990662712854,
+      "learning_rate": 8.881876640883802e-08,
+      "loss": 0.8255,
+      "step": 40698
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.755619467053956,
+      "learning_rate": 8.871732266160782e-08,
+      "loss": 0.8345,
+      "step": 40699
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5758647256409852,
+      "learning_rate": 8.861593662100265e-08,
+      "loss": 0.8623,
+      "step": 40700
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5694207797850084,
+      "learning_rate": 8.851460828761316e-08,
+      "loss": 0.7452,
+      "step": 40701
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7980905122934532,
+      "learning_rate": 8.841333766202997e-08,
+      "loss": 0.8341,
+      "step": 40702
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5949817891672085,
+      "learning_rate": 8.831212474484263e-08,
+      "loss": 0.6525,
+      "step": 40703
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8984335846851979,
+      "learning_rate": 8.821096953663955e-08,
+      "loss": 0.8941,
+      "step": 40704
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6769351392651612,
+      "learning_rate": 8.810987203801136e-08,
+      "loss": 0.7899,
+      "step": 40705
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5591553394314686,
+      "learning_rate": 8.800883224954315e-08,
+      "loss": 0.8248,
+      "step": 40706
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7065676000254175,
+      "learning_rate": 8.790785017182779e-08,
+      "loss": 0.8512,
+      "step": 40707
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6617152334756626,
+      "learning_rate": 8.780692580545036e-08,
+      "loss": 0.7931,
+      "step": 40708
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5448209382823026,
+      "learning_rate": 8.770605915099927e-08,
+      "loss": 0.8216,
+      "step": 40709
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7803144112418687,
+      "learning_rate": 8.760525020906074e-08,
+      "loss": 0.7179,
+      "step": 40710
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4441479274582152,
+      "learning_rate": 8.750449898022317e-08,
+      "loss": 0.7791,
+      "step": 40711
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.403910782080111,
+      "learning_rate": 8.740380546507388e-08,
+      "loss": 0.8069,
+      "step": 40712
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7854801062637309,
+      "learning_rate": 8.730316966419682e-08,
+      "loss": 0.9066,
+      "step": 40713
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7133431408306803,
+      "learning_rate": 8.720259157817823e-08,
+      "loss": 0.8975,
+      "step": 40714
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.635762593636693,
+      "learning_rate": 8.71020712076054e-08,
+      "loss": 0.8365,
+      "step": 40715
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.578892449349061,
+      "learning_rate": 8.70016085530634e-08,
+      "loss": 0.9019,
+      "step": 40716
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7024724392955572,
+      "learning_rate": 8.69012036151351e-08,
+      "loss": 0.9309,
+      "step": 40717
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4095528060140587,
+      "learning_rate": 8.680085639440672e-08,
+      "loss": 0.8948,
+      "step": 40718
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.643281594542638,
+      "learning_rate": 8.670056689146334e-08,
+      "loss": 0.8489,
+      "step": 40719
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7096265139581612,
+      "learning_rate": 8.66003351068867e-08,
+      "loss": 0.8188,
+      "step": 40720
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5938270778072852,
+      "learning_rate": 8.650016104126192e-08,
+      "loss": 0.712,
+      "step": 40721
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6545063493692982,
+      "learning_rate": 8.640004469517072e-08,
+      "loss": 0.8557,
+      "step": 40722
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6661456509698631,
+      "learning_rate": 8.62999860691982e-08,
+      "loss": 0.8845,
+      "step": 40723
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5582818972889159,
+      "learning_rate": 8.619998516392503e-08,
+      "loss": 0.9304,
+      "step": 40724
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7853268888990985,
+      "learning_rate": 8.610004197993516e-08,
+      "loss": 0.7143,
+      "step": 40725
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6635876498248394,
+      "learning_rate": 8.600015651780924e-08,
+      "loss": 0.9499,
+      "step": 40726
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5370102194184583,
+      "learning_rate": 8.590032877812792e-08,
+      "loss": 0.8145,
+      "step": 40727
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7812396165073001,
+      "learning_rate": 8.580055876147408e-08,
+      "loss": 0.7905,
+      "step": 40728
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5238031930364113,
+      "learning_rate": 8.570084646842946e-08,
+      "loss": 0.8673,
+      "step": 40729
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6712066350409207,
+      "learning_rate": 8.56011918995725e-08,
+      "loss": 0.9116,
+      "step": 40730
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.540278600324446,
+      "learning_rate": 8.550159505548383e-08,
+      "loss": 0.7902,
+      "step": 40731
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5918955347065482,
+      "learning_rate": 8.540205593674411e-08,
+      "loss": 0.7615,
+      "step": 40732
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5802662094872877,
+      "learning_rate": 8.530257454393287e-08,
+      "loss": 0.8374,
+      "step": 40733
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7725271777118001,
+      "learning_rate": 8.520315087762964e-08,
+      "loss": 0.7801,
+      "step": 40734
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4107363494955978,
+      "learning_rate": 8.510378493841065e-08,
+      "loss": 0.7642,
+      "step": 40735
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.63221970264429,
+      "learning_rate": 8.500447672685764e-08,
+      "loss": 0.8997,
+      "step": 40736
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.671435539987321,
+      "learning_rate": 8.490522624354792e-08,
+      "loss": 0.9008,
+      "step": 40737
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6583370899788954,
+      "learning_rate": 8.48060334890588e-08,
+      "loss": 0.8052,
+      "step": 40738
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.957498239551888,
+      "learning_rate": 8.470689846396762e-08,
+      "loss": 0.8234,
+      "step": 40739
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5891044015172284,
+      "learning_rate": 8.460782116885169e-08,
+      "loss": 0.887,
+      "step": 40740
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.680832845235594,
+      "learning_rate": 8.450880160428831e-08,
+      "loss": 0.7799,
+      "step": 40741
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5550330674459956,
+      "learning_rate": 8.440983977085481e-08,
+      "loss": 0.8456,
+      "step": 40742
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7755216474037407,
+      "learning_rate": 8.43109356691263e-08,
+      "loss": 0.6674,
+      "step": 40743
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.725991042721648,
+      "learning_rate": 8.421208929967784e-08,
+      "loss": 0.751,
+      "step": 40744
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5024672098806235,
+      "learning_rate": 8.411330066308676e-08,
+      "loss": 0.8673,
+      "step": 40745
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0347785436758774,
+      "learning_rate": 8.401456975992706e-08,
+      "loss": 0.9398,
+      "step": 40746
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5899866139375656,
+      "learning_rate": 8.391589659077381e-08,
+      "loss": 0.7953,
+      "step": 40747
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.787242524585133,
+      "learning_rate": 8.381728115620213e-08,
+      "loss": 0.8151,
+      "step": 40748
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5709864991595126,
+      "learning_rate": 8.371872345678378e-08,
+      "loss": 0.8654,
+      "step": 40749
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7265892150677186,
+      "learning_rate": 8.362022349309718e-08,
+      "loss": 0.7527,
+      "step": 40750
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9505312331953841,
+      "learning_rate": 8.352178126571186e-08,
+      "loss": 0.7989,
+      "step": 40751
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2806795926646481,
+      "learning_rate": 8.342339677520184e-08,
+      "loss": 0.7754,
+      "step": 40752
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7505304131442752,
+      "learning_rate": 8.332507002213996e-08,
+      "loss": 0.7378,
+      "step": 40753
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5686968616252488,
+      "learning_rate": 8.322680100710023e-08,
+      "loss": 0.8764,
+      "step": 40754
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4316270768733217,
+      "learning_rate": 8.31285897306533e-08,
+      "loss": 0.7969,
+      "step": 40755
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7644532996313861,
+      "learning_rate": 8.303043619337093e-08,
+      "loss": 0.7112,
+      "step": 40756
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7592441911911175,
+      "learning_rate": 8.293234039582487e-08,
+      "loss": 0.6149,
+      "step": 40757
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.612108486061542,
+      "learning_rate": 8.283430233858691e-08,
+      "loss": 0.8729,
+      "step": 40758
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5708324717212934,
+      "learning_rate": 8.27363220222277e-08,
+      "loss": 0.7922,
+      "step": 40759
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6995789603920803,
+      "learning_rate": 8.263839944731677e-08,
+      "loss": 0.8423,
+      "step": 40760
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7090529622892343,
+      "learning_rate": 8.254053461442591e-08,
+      "loss": 0.839,
+      "step": 40761
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7406614069776511,
+      "learning_rate": 8.244272752412241e-08,
+      "loss": 0.7953,
+      "step": 40762
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.689440854254024,
+      "learning_rate": 8.234497817697806e-08,
+      "loss": 0.8272,
+      "step": 40763
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3040575375505052,
+      "learning_rate": 8.224728657356129e-08,
+      "loss": 0.8055,
+      "step": 40764
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7070025304469594,
+      "learning_rate": 8.214965271444053e-08,
+      "loss": 0.7311,
+      "step": 40765
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5259337571775862,
+      "learning_rate": 8.205207660018311e-08,
+      "loss": 0.971,
+      "step": 40766
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3700052375227612,
+      "learning_rate": 8.195455823136078e-08,
+      "loss": 0.7493,
+      "step": 40767
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.766133097812843,
+      "learning_rate": 8.185709760853754e-08,
+      "loss": 0.7221,
+      "step": 40768
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7969806747616266,
+      "learning_rate": 8.175969473228296e-08,
+      "loss": 0.7843,
+      "step": 40769
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7738740570175551,
+      "learning_rate": 8.16623496031621e-08,
+      "loss": 0.7236,
+      "step": 40770
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6603685785649935,
+      "learning_rate": 8.156506222174455e-08,
+      "loss": 0.7945,
+      "step": 40771
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5289652730066583,
+      "learning_rate": 8.146783258859536e-08,
+      "loss": 0.7705,
+      "step": 40772
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6495281031346876,
+      "learning_rate": 8.137066070427968e-08,
+      "loss": 0.77,
+      "step": 40773
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5261644770044402,
+      "learning_rate": 8.127354656936481e-08,
+      "loss": 0.873,
+      "step": 40774
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6007912469142038,
+      "learning_rate": 8.117649018441587e-08,
+      "loss": 0.8881,
+      "step": 40775
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5870983050937582,
+      "learning_rate": 8.107949154999684e-08,
+      "loss": 0.8508,
+      "step": 40776
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5081864316851523,
+      "learning_rate": 8.098255066667393e-08,
+      "loss": 0.7172,
+      "step": 40777
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5154761479485201,
+      "learning_rate": 8.088566753501004e-08,
+      "loss": 0.7346,
+      "step": 40778
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7325670396591595,
+      "learning_rate": 8.078884215557137e-08,
+      "loss": 0.7958,
+      "step": 40779
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6929334331809582,
+      "learning_rate": 8.06920745289197e-08,
+      "loss": 0.899,
+      "step": 40780
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.136597314425962,
+      "learning_rate": 8.059536465561901e-08,
+      "loss": 0.839,
+      "step": 40781
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6498658701198907,
+      "learning_rate": 8.049871253623332e-08,
+      "loss": 0.7798,
+      "step": 40782
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5724259770852735,
+      "learning_rate": 8.040211817132438e-08,
+      "loss": 0.8802,
+      "step": 40783
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5594301611641568,
+      "learning_rate": 8.030558156145507e-08,
+      "loss": 0.7964,
+      "step": 40784
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5581501310227135,
+      "learning_rate": 8.020910270718607e-08,
+      "loss": 0.6658,
+      "step": 40785
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7445543149725657,
+      "learning_rate": 8.011268160908137e-08,
+      "loss": 0.9273,
+      "step": 40786
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.612075963705588,
+      "learning_rate": 8.00163182676994e-08,
+      "loss": 0.7918,
+      "step": 40787
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5998732791042583,
+      "learning_rate": 7.992001268360528e-08,
+      "loss": 0.8684,
+      "step": 40788
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5655045479912055,
+      "learning_rate": 7.982376485735632e-08,
+      "loss": 0.8509,
+      "step": 40789
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7572882586961767,
+      "learning_rate": 7.972757478951432e-08,
+      "loss": 0.8152,
+      "step": 40790
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6487280176292272,
+      "learning_rate": 7.963144248063881e-08,
+      "loss": 0.7379,
+      "step": 40791
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4286836572162456,
+      "learning_rate": 7.953536793129046e-08,
+      "loss": 0.77,
+      "step": 40792
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4791494122915527,
+      "learning_rate": 7.943935114202772e-08,
+      "loss": 0.7453,
+      "step": 40793
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6628035087592214,
+      "learning_rate": 7.934339211340902e-08,
+      "loss": 0.9931,
+      "step": 40794
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4570362806912485,
+      "learning_rate": 7.924749084599393e-08,
+      "loss": 0.7307,
+      "step": 40795
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.71149637755889,
+      "learning_rate": 7.915164734034198e-08,
+      "loss": 0.8851,
+      "step": 40796
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7701892264213127,
+      "learning_rate": 7.90558615970094e-08,
+      "loss": 0.6761,
+      "step": 40797
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5303045096906562,
+      "learning_rate": 7.896013361655464e-08,
+      "loss": 0.744,
+      "step": 40798
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7075326265294362,
+      "learning_rate": 7.886446339953502e-08,
+      "loss": 0.9346,
+      "step": 40799
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7827145492367487,
+      "learning_rate": 7.876885094650677e-08,
+      "loss": 0.8381,
+      "step": 40800
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6947395121994742,
+      "learning_rate": 7.867329625802833e-08,
+      "loss": 1.0144,
+      "step": 40801
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5912254489326763,
+      "learning_rate": 7.857779933465482e-08,
+      "loss": 0.855,
+      "step": 40802
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4729905583407565,
+      "learning_rate": 7.848236017694133e-08,
+      "loss": 0.7567,
+      "step": 40803
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1308413044996324,
+      "learning_rate": 7.838697878544632e-08,
+      "loss": 0.8387,
+      "step": 40804
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6258270772208903,
+      "learning_rate": 7.829165516072268e-08,
+      "loss": 0.8076,
+      "step": 40805
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6639459636969252,
+      "learning_rate": 7.819638930332552e-08,
+      "loss": 0.7693,
+      "step": 40806
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8235675189550625,
+      "learning_rate": 7.810118121381105e-08,
+      "loss": 0.7139,
+      "step": 40807
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.642773374030371,
+      "learning_rate": 7.800603089273218e-08,
+      "loss": 0.8406,
+      "step": 40808
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.383284164501496,
+      "learning_rate": 7.791093834064289e-08,
+      "loss": 0.7547,
+      "step": 40809
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7421069189296664,
+      "learning_rate": 7.781590355809831e-08,
+      "loss": 0.6617,
+      "step": 40810
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.90571223997595,
+      "learning_rate": 7.772092654565133e-08,
+      "loss": 0.8258,
+      "step": 40811
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7138670549773032,
+      "learning_rate": 7.762600730385373e-08,
+      "loss": 0.9048,
+      "step": 40812
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5681930405503117,
+      "learning_rate": 7.753114583325838e-08,
+      "loss": 0.9768,
+      "step": 40813
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7727599734237607,
+      "learning_rate": 7.74363421344193e-08,
+      "loss": 0.7795,
+      "step": 40814
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5636917790917533,
+      "learning_rate": 7.734159620788717e-08,
+      "loss": 0.715,
+      "step": 40815
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.037810574898187,
+      "learning_rate": 7.724690805421265e-08,
+      "loss": 0.8382,
+      "step": 40816
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7202199078364064,
+      "learning_rate": 7.715227767394862e-08,
+      "loss": 0.9491,
+      "step": 40817
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4804191459555716,
+      "learning_rate": 7.705770506764576e-08,
+      "loss": 0.75,
+      "step": 40818
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6570220438501613,
+      "learning_rate": 7.696319023585475e-08,
+      "loss": 0.9765,
+      "step": 40819
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.57871876858112,
+      "learning_rate": 7.686873317912514e-08,
+      "loss": 0.7983,
+      "step": 40820
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.00512684419212,
+      "learning_rate": 7.677433389800759e-08,
+      "loss": 0.9763,
+      "step": 40821
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6571451104351334,
+      "learning_rate": 7.667999239305057e-08,
+      "loss": 0.7523,
+      "step": 40822
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5529035148664323,
+      "learning_rate": 7.658570866480586e-08,
+      "loss": 0.8553,
+      "step": 40823
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4913954901058888,
+      "learning_rate": 7.649148271381967e-08,
+      "loss": 0.87,
+      "step": 40824
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.157716369451576,
+      "learning_rate": 7.63973145406427e-08,
+      "loss": 0.73,
+      "step": 40825
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6775851798659511,
+      "learning_rate": 7.630320414582115e-08,
+      "loss": 0.7592,
+      "step": 40826
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4582454572520478,
+      "learning_rate": 7.620915152990571e-08,
+      "loss": 0.8159,
+      "step": 40827
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8728163112863847,
+      "learning_rate": 7.61151566934415e-08,
+      "loss": 0.9438,
+      "step": 40828
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.73161153726954,
+      "learning_rate": 7.602121963697695e-08,
+      "loss": 0.9091,
+      "step": 40829
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4356516965642907,
+      "learning_rate": 7.592734036105831e-08,
+      "loss": 0.8249,
+      "step": 40830
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5060551903912067,
+      "learning_rate": 7.583351886623291e-08,
+      "loss": 0.7139,
+      "step": 40831
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.633104571697533,
+      "learning_rate": 7.5739755153047e-08,
+      "loss": 0.8991,
+      "step": 40832
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3828850358979021,
+      "learning_rate": 7.564604922204566e-08,
+      "loss": 0.81,
+      "step": 40833
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6998231488265723,
+      "learning_rate": 7.555240107377404e-08,
+      "loss": 0.6921,
+      "step": 40834
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4430636294025785,
+      "learning_rate": 7.545881070878059e-08,
+      "loss": 0.8927,
+      "step": 40835
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5574612113009103,
+      "learning_rate": 7.536527812760708e-08,
+      "loss": 0.6976,
+      "step": 40836
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0011892161801454,
+      "learning_rate": 7.527180333079865e-08,
+      "loss": 0.9071,
+      "step": 40837
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8983387225997614,
+      "learning_rate": 7.517838631889929e-08,
+      "loss": 0.9629,
+      "step": 40838
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5974397607169661,
+      "learning_rate": 7.508502709245524e-08,
+      "loss": 0.8025,
+      "step": 40839
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5981463117033041,
+      "learning_rate": 7.499172565200718e-08,
+      "loss": 0.8674,
+      "step": 40840
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.203827846653412,
+      "learning_rate": 7.489848199809913e-08,
+      "loss": 0.873,
+      "step": 40841
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9241372246970718,
+      "learning_rate": 7.480529613127507e-08,
+      "loss": 0.8778,
+      "step": 40842
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7348730678252786,
+      "learning_rate": 7.471216805207682e-08,
+      "loss": 0.7212,
+      "step": 40843
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5497393588878423,
+      "learning_rate": 7.461909776104725e-08,
+      "loss": 0.8255,
+      "step": 40844
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.065081620535506,
+      "learning_rate": 7.452608525872595e-08,
+      "loss": 0.7931,
+      "step": 40845
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6273252208292261,
+      "learning_rate": 7.443313054565804e-08,
+      "loss": 0.9081,
+      "step": 40846
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8234568696396334,
+      "learning_rate": 7.434023362238086e-08,
+      "loss": 0.8049,
+      "step": 40847
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7443332063103814,
+      "learning_rate": 7.42473944894384e-08,
+      "loss": 0.7373,
+      "step": 40848
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6072047036976636,
+      "learning_rate": 7.415461314737026e-08,
+      "loss": 0.6334,
+      "step": 40849
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5910125663122012,
+      "learning_rate": 7.406188959671601e-08,
+      "loss": 0.8312,
+      "step": 40850
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5680524334855457,
+      "learning_rate": 7.39692238380163e-08,
+      "loss": 0.8635,
+      "step": 40851
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7632799280704212,
+      "learning_rate": 7.387661587180961e-08,
+      "loss": 0.9115,
+      "step": 40852
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7712550416002264,
+      "learning_rate": 7.37840656986366e-08,
+      "loss": 0.672,
+      "step": 40853
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2875823214056656,
+      "learning_rate": 7.369157331903465e-08,
+      "loss": 0.9066,
+      "step": 40854
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1183210923020526,
+      "learning_rate": 7.35991387335433e-08,
+      "loss": 0.8391,
+      "step": 40855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6821572759703893,
+      "learning_rate": 7.350676194270101e-08,
+      "loss": 0.8521,
+      "step": 40856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6726627733245034,
+      "learning_rate": 7.341444294704403e-08,
+      "loss": 0.8756,
+      "step": 40857
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6446944323600852,
+      "learning_rate": 7.332218174711081e-08,
+      "loss": 0.7646,
+      "step": 40858
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.371596963055272,
+      "learning_rate": 7.322997834343871e-08,
+      "loss": 0.7266,
+      "step": 40859
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6682461954803756,
+      "learning_rate": 7.313783273656506e-08,
+      "loss": 0.8489,
+      "step": 40860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.940997311062022,
+      "learning_rate": 7.304574492702609e-08,
+      "loss": 0.7347,
+      "step": 40861
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7973075357454545,
+      "learning_rate": 7.295371491535697e-08,
+      "loss": 0.7322,
+      "step": 40862
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.712821097144461,
+      "learning_rate": 7.28617427020939e-08,
+      "loss": 0.8711,
+      "step": 40863
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4690592729813348,
+      "learning_rate": 7.276982828777312e-08,
+      "loss": 0.7633,
+      "step": 40864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.583870945173741,
+      "learning_rate": 7.267797167292867e-08,
+      "loss": 0.7078,
+      "step": 40865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7156097754859485,
+      "learning_rate": 7.258617285809677e-08,
+      "loss": 0.8251,
+      "step": 40866
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7270830138334143,
+      "learning_rate": 7.249443184381033e-08,
+      "loss": 0.6989,
+      "step": 40867
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8175421525754294,
+      "learning_rate": 7.240274863060448e-08,
+      "loss": 0.9189,
+      "step": 40868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6429427290603258,
+      "learning_rate": 7.231112321901435e-08,
+      "loss": 0.8109,
+      "step": 40869
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6806306425603155,
+      "learning_rate": 7.221955560957062e-08,
+      "loss": 0.8905,
+      "step": 40870
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8757378087058887,
+      "learning_rate": 7.21280458028073e-08,
+      "loss": 0.9518,
+      "step": 40871
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6878315977640415,
+      "learning_rate": 7.203659379925843e-08,
+      "loss": 0.8962,
+      "step": 40872
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7933564046361459,
+      "learning_rate": 7.194519959945467e-08,
+      "loss": 0.6645,
+      "step": 40873
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5268926431825924,
+      "learning_rate": 7.185386320393006e-08,
+      "loss": 0.7976,
+      "step": 40874
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7799697646181724,
+      "learning_rate": 7.176258461321528e-08,
+      "loss": 0.7167,
+      "step": 40875
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8611394368861305,
+      "learning_rate": 7.167136382784212e-08,
+      "loss": 0.8368,
+      "step": 40876
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4574975651654478,
+      "learning_rate": 7.158020084834128e-08,
+      "loss": 0.779,
+      "step": 40877
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4656662912675706,
+      "learning_rate": 7.148909567524343e-08,
+      "loss": 0.7717,
+      "step": 40878
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.213815795302881,
+      "learning_rate": 7.139804830908037e-08,
+      "loss": 0.7632,
+      "step": 40879
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9187962856070084,
+      "learning_rate": 7.130705875037946e-08,
+      "loss": 0.7904,
+      "step": 40880
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5641259768391056,
+      "learning_rate": 7.121612699967362e-08,
+      "loss": 0.7107,
+      "step": 40881
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6522478057886956,
+      "learning_rate": 7.112525305749129e-08,
+      "loss": 0.8345,
+      "step": 40882
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3548932903184898,
+      "learning_rate": 7.103443692435985e-08,
+      "loss": 0.7125,
+      "step": 40883
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5467233244436738,
+      "learning_rate": 7.094367860081108e-08,
+      "loss": 0.7583,
+      "step": 40884
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7733518414781078,
+      "learning_rate": 7.085297808737013e-08,
+      "loss": 0.8701,
+      "step": 40885
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5730758143511119,
+      "learning_rate": 7.076233538456767e-08,
+      "loss": 0.8962,
+      "step": 40886
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.633300549780825,
+      "learning_rate": 7.067175049292884e-08,
+      "loss": 0.7976,
+      "step": 40887
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5637076839646549,
+      "learning_rate": 7.058122341298435e-08,
+      "loss": 0.7352,
+      "step": 40888
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7017578708897023,
+      "learning_rate": 7.049075414525818e-08,
+      "loss": 0.8808,
+      "step": 40889
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6599751266907798,
+      "learning_rate": 7.040034269027995e-08,
+      "loss": 0.8525,
+      "step": 40890
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5391365510519006,
+      "learning_rate": 7.030998904857478e-08,
+      "loss": 0.8547,
+      "step": 40891
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.870994294409382,
+      "learning_rate": 7.021969322066668e-08,
+      "loss": 0.8311,
+      "step": 40892
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.800462979277788,
+      "learning_rate": 7.012945520708414e-08,
+      "loss": 0.8336,
+      "step": 40893
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9123005774908155,
+      "learning_rate": 7.003927500835118e-08,
+      "loss": 0.7923,
+      "step": 40894
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.707239968682334,
+      "learning_rate": 6.994915262499513e-08,
+      "loss": 0.8191,
+      "step": 40895
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5750708504794435,
+      "learning_rate": 6.985908805753672e-08,
+      "loss": 0.9083,
+      "step": 40896
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7910130999298177,
+      "learning_rate": 6.976908130650329e-08,
+      "loss": 0.6645,
+      "step": 40897
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6753573072975492,
+      "learning_rate": 6.967913237241885e-08,
+      "loss": 0.8933,
+      "step": 40898
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9128719525860187,
+      "learning_rate": 6.958924125580635e-08,
+      "loss": 0.8596,
+      "step": 40899
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7796324150257425,
+      "learning_rate": 6.949940795718868e-08,
+      "loss": 0.8564,
+      "step": 40900
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7263556553795008,
+      "learning_rate": 6.940963247708877e-08,
+      "loss": 0.8339,
+      "step": 40901
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6116312602311407,
+      "learning_rate": 6.931991481603062e-08,
+      "loss": 0.7232,
+      "step": 40902
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5717262934591687,
+      "learning_rate": 6.923025497453606e-08,
+      "loss": 0.8419,
+      "step": 40903
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6679044089998412,
+      "learning_rate": 6.914065295312689e-08,
+      "loss": 0.8332,
+      "step": 40904
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6945127357838468,
+      "learning_rate": 6.90511087523249e-08,
+      "loss": 0.891,
+      "step": 40905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6255252952234598,
+      "learning_rate": 6.896162237265192e-08,
+      "loss": 0.9869,
+      "step": 40906
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4795381905545215,
+      "learning_rate": 6.887219381462751e-08,
+      "loss": 0.8141,
+      "step": 40907
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7019304473035124,
+      "learning_rate": 6.878282307877349e-08,
+      "loss": 0.8222,
+      "step": 40908
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7808889862869743,
+      "learning_rate": 6.869351016561055e-08,
+      "loss": 0.7053,
+      "step": 40909
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5317076498564626,
+      "learning_rate": 6.860425507565826e-08,
+      "loss": 0.7505,
+      "step": 40910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4517276737508658,
+      "learning_rate": 6.851505780943512e-08,
+      "loss": 0.859,
+      "step": 40911
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8606662151082192,
+      "learning_rate": 6.842591836746293e-08,
+      "loss": 0.7665,
+      "step": 40912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7160225916847418,
+      "learning_rate": 6.833683675025904e-08,
+      "loss": 0.7846,
+      "step": 40913
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7016601375733078,
+      "learning_rate": 6.824781295834193e-08,
+      "loss": 0.8972,
+      "step": 40914
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3504582142006862,
+      "learning_rate": 6.815884699223007e-08,
+      "loss": 0.8599,
+      "step": 40915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.490752853457243,
+      "learning_rate": 6.806993885244307e-08,
+      "loss": 0.9114,
+      "step": 40916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.703958975461202,
+      "learning_rate": 6.798108853949603e-08,
+      "loss": 0.8916,
+      "step": 40917
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6459435932972064,
+      "learning_rate": 6.789229605390856e-08,
+      "loss": 0.7857,
+      "step": 40918
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6049085020261342,
+      "learning_rate": 6.780356139619582e-08,
+      "loss": 0.881,
+      "step": 40919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7330811098266865,
+      "learning_rate": 6.771488456687514e-08,
+      "loss": 0.8576,
+      "step": 40920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6348824988010364,
+      "learning_rate": 6.76262655664639e-08,
+      "loss": 0.9282,
+      "step": 40921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5981436905201636,
+      "learning_rate": 6.753770439547614e-08,
+      "loss": 0.8264,
+      "step": 40922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8580871885235706,
+      "learning_rate": 6.74492010544281e-08,
+      "loss": 0.9243,
+      "step": 40923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7430519121186843,
+      "learning_rate": 6.736075554383604e-08,
+      "loss": 0.9366,
+      "step": 40924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.676543743177162,
+      "learning_rate": 6.727236786421509e-08,
+      "loss": 0.8116,
+      "step": 40925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7391391489910346,
+      "learning_rate": 6.718403801607709e-08,
+      "loss": 0.8372,
+      "step": 40926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4806066871225017,
+      "learning_rate": 6.70957659999405e-08,
+      "loss": 0.7646,
+      "step": 40927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.704814900982906,
+      "learning_rate": 6.700755181631491e-08,
+      "loss": 0.7676,
+      "step": 40928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5672154892273384,
+      "learning_rate": 6.691939546571769e-08,
+      "loss": 0.8752,
+      "step": 40929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7695946195710567,
+      "learning_rate": 6.683129694866063e-08,
+      "loss": 0.7137,
+      "step": 40930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9458493332866127,
+      "learning_rate": 6.674325626565558e-08,
+      "loss": 0.8293,
+      "step": 40931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9466327295871584,
+      "learning_rate": 6.665527341721544e-08,
+      "loss": 0.8789,
+      "step": 40932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6007916186590854,
+      "learning_rate": 6.656734840385425e-08,
+      "loss": 0.8211,
+      "step": 40933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7137185331823723,
+      "learning_rate": 6.647948122608383e-08,
+      "loss": 0.8512,
+      "step": 40934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.567474892658164,
+      "learning_rate": 6.639167188441264e-08,
+      "loss": 0.8207,
+      "step": 40935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.607366835342359,
+      "learning_rate": 6.630392037935474e-08,
+      "loss": 0.7687,
+      "step": 40936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7215693954949896,
+      "learning_rate": 6.621622671142191e-08,
+      "loss": 0.9254,
+      "step": 40937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4625552236430355,
+      "learning_rate": 6.612859088112155e-08,
+      "loss": 0.8524,
+      "step": 40938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4214539049673858,
+      "learning_rate": 6.604101288896548e-08,
+      "loss": 0.7132,
+      "step": 40939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3874367237933007,
+      "learning_rate": 6.595349273546437e-08,
+      "loss": 0.8223,
+      "step": 40940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.976022611219974,
+      "learning_rate": 6.586603042112671e-08,
+      "loss": 0.8996,
+      "step": 40941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.657231397880239,
+      "learning_rate": 6.577862594646323e-08,
+      "loss": 0.8847,
+      "step": 40942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.631643449230298,
+      "learning_rate": 6.569127931198016e-08,
+      "loss": 0.8051,
+      "step": 40943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.921661662953926,
+      "learning_rate": 6.560399051818822e-08,
+      "loss": 0.8811,
+      "step": 40944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6672814713857538,
+      "learning_rate": 6.551675956559478e-08,
+      "loss": 0.8478,
+      "step": 40945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9350951895729418,
+      "learning_rate": 6.542958645470832e-08,
+      "loss": 0.794,
+      "step": 40946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.464500188394018,
+      "learning_rate": 6.534247118603621e-08,
+      "loss": 0.8816,
+      "step": 40947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7215054834193528,
+      "learning_rate": 6.525541376008471e-08,
+      "loss": 0.948,
+      "step": 40948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6342448115423216,
+      "learning_rate": 6.51684141773623e-08,
+      "loss": 0.7807,
+      "step": 40949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.487824555687859,
+      "learning_rate": 6.508147243837415e-08,
+      "loss": 0.8682,
+      "step": 40950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3314284996775092,
+      "learning_rate": 6.499458854362761e-08,
+      "loss": 0.6458,
+      "step": 40951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7505324989251325,
+      "learning_rate": 6.490776249362785e-08,
+      "loss": 0.7899,
+      "step": 40952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.94190996089337,
+      "learning_rate": 6.482099428888e-08,
+      "loss": 0.7992,
+      "step": 40953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7788011121978673,
+      "learning_rate": 6.473428392989034e-08,
+      "loss": 0.8443,
+      "step": 40954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7585315783361136,
+      "learning_rate": 6.46476314171629e-08,
+      "loss": 0.6837,
+      "step": 40955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9157172224760206,
+      "learning_rate": 6.456103675120173e-08,
+      "loss": 1.0117,
+      "step": 40956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.507841634654189,
+      "learning_rate": 6.447449993251197e-08,
+      "loss": 0.823,
+      "step": 40957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.512773830385134,
+      "learning_rate": 6.438802096159769e-08,
+      "loss": 0.8352,
+      "step": 40958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7748212642713834,
+      "learning_rate": 6.430159983896179e-08,
+      "loss": 0.7218,
+      "step": 40959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.528904577078383,
+      "learning_rate": 6.42152365651072e-08,
+      "loss": 0.8174,
+      "step": 40960
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.669456926600054,
+      "learning_rate": 6.412893114053798e-08,
+      "loss": 0.7574,
+      "step": 40961
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7170522348554575,
+      "learning_rate": 6.404268356575483e-08,
+      "loss": 0.8088,
+      "step": 40962
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6649924366818762,
+      "learning_rate": 6.395649384126179e-08,
+      "loss": 0.7873,
+      "step": 40963
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5814569846791144,
+      "learning_rate": 6.387036196755847e-08,
+      "loss": 0.9323,
+      "step": 40964
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7789831681043209,
+      "learning_rate": 6.378428794514891e-08,
+      "loss": 0.786,
+      "step": 40965
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5951664795125078,
+      "learning_rate": 6.369827177453158e-08,
+      "loss": 0.8904,
+      "step": 40966
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4813945307708216,
+      "learning_rate": 6.361231345621055e-08,
+      "loss": 0.7585,
+      "step": 40967
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8489913599650418,
+      "learning_rate": 6.352641299068318e-08,
+      "loss": 0.8001,
+      "step": 40968
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8889557690318135,
+      "learning_rate": 6.344057037845131e-08,
+      "loss": 0.8713,
+      "step": 40969
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.514383972640567,
+      "learning_rate": 6.33547856200134e-08,
+      "loss": 0.8376,
+      "step": 40970
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6699944759747891,
+      "learning_rate": 6.32690587158713e-08,
+      "loss": 0.7406,
+      "step": 40971
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5648074759492094,
+      "learning_rate": 6.318338966652127e-08,
+      "loss": 0.8421,
+      "step": 40972
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6175858618788563,
+      "learning_rate": 6.309777847246401e-08,
+      "loss": 0.9265,
+      "step": 40973
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8921480057606175,
+      "learning_rate": 6.301222513419692e-08,
+      "loss": 0.876,
+      "step": 40974
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5185822018663628,
+      "learning_rate": 6.292672965221958e-08,
+      "loss": 0.9423,
+      "step": 40975
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.039080262820432,
+      "learning_rate": 6.284129202702938e-08,
+      "loss": 0.8862,
+      "step": 40976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8091618274127432,
+      "learning_rate": 6.275591225912148e-08,
+      "loss": 0.7228,
+      "step": 40977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5502568545227609,
+      "learning_rate": 6.267059034899547e-08,
+      "loss": 0.8157,
+      "step": 40978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8000712105202699,
+      "learning_rate": 6.258532629714765e-08,
+      "loss": 0.812,
+      "step": 40979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.791816162603197,
+      "learning_rate": 6.250012010407425e-08,
+      "loss": 0.8292,
+      "step": 40980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.572728542187309,
+      "learning_rate": 6.241497177027156e-08,
+      "loss": 0.7849,
+      "step": 40981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6116721233278108,
+      "learning_rate": 6.232988129623475e-08,
+      "loss": 0.7916,
+      "step": 40982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9645582328422206,
+      "learning_rate": 6.224484868245895e-08,
+      "loss": 0.801,
+      "step": 40983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8372741569479721,
+      "learning_rate": 6.215987392944045e-08,
+      "loss": 0.7938,
+      "step": 40984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6590566933525264,
+      "learning_rate": 6.20749570376733e-08,
+      "loss": 0.7786,
+      "step": 40985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7273028866884168,
+      "learning_rate": 6.199009800765265e-08,
+      "loss": 0.7949,
+      "step": 40986
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0650688793569225,
+      "learning_rate": 6.190529683987145e-08,
+      "loss": 0.8637,
+      "step": 40987
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.439876559665893,
+      "learning_rate": 6.182055353482375e-08,
+      "loss": 0.9039,
+      "step": 40988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8304232166518186,
+      "learning_rate": 6.173586809300358e-08,
+      "loss": 0.7641,
+      "step": 40989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7112164251985815,
+      "learning_rate": 6.16512405149039e-08,
+      "loss": 0.8496,
+      "step": 40990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.704437297762237,
+      "learning_rate": 6.156667080101652e-08,
+      "loss": 0.8032,
+      "step": 40991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7478532689916695,
+      "learning_rate": 6.14821589518344e-08,
+      "loss": 0.8032,
+      "step": 40992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.706519138561365,
+      "learning_rate": 6.139770496785047e-08,
+      "loss": 0.9788,
+      "step": 40993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4848186650820625,
+      "learning_rate": 6.131330884955433e-08,
+      "loss": 0.8287,
+      "step": 40994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5821814584580802,
+      "learning_rate": 6.122897059743894e-08,
+      "loss": 0.7945,
+      "step": 40995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5611334610883332,
+      "learning_rate": 6.114469021199498e-08,
+      "loss": 0.7808,
+      "step": 40996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5966400794762043,
+      "learning_rate": 6.106046769371432e-08,
+      "loss": 0.8552,
+      "step": 40997
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.949756678555296,
+      "learning_rate": 6.097630304308543e-08,
+      "loss": 0.7713,
+      "step": 40998
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7514418107818848,
+      "learning_rate": 6.089219626059905e-08,
+      "loss": 0.8573,
+      "step": 40999
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7252019239427672,
+      "learning_rate": 6.080814734674478e-08,
+      "loss": 0.8094,
+      "step": 41000
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4911770013440138,
+      "learning_rate": 6.072415630201333e-08,
+      "loss": 0.7285,
+      "step": 41001
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4817089191728825,
+      "learning_rate": 6.0640223126891e-08,
+      "loss": 0.8653,
+      "step": 41002
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8341113160746256,
+      "learning_rate": 6.055634782186848e-08,
+      "loss": 0.7244,
+      "step": 41003
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5836416890733147,
+      "learning_rate": 6.047253038743428e-08,
+      "loss": 0.7078,
+      "step": 41004
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7791908284326151,
+      "learning_rate": 6.03887708240758e-08,
+      "loss": 0.7514,
+      "step": 41005
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8758259728451299,
+      "learning_rate": 6.03050691322804e-08,
+      "loss": 0.8165,
+      "step": 41006
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4231174878679567,
+      "learning_rate": 6.02214253125366e-08,
+      "loss": 0.8281,
+      "step": 41007
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6530902546205442,
+      "learning_rate": 6.013783936532846e-08,
+      "loss": 0.8646,
+      "step": 41008
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5183414724477202,
+      "learning_rate": 6.00543112911467e-08,
+      "loss": 0.8983,
+      "step": 41009
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6685287352132703,
+      "learning_rate": 5.997084109047424e-08,
+      "loss": 0.8396,
+      "step": 41010
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7014725376963822,
+      "learning_rate": 5.98874287637996e-08,
+      "loss": 0.7219,
+      "step": 41011
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8706057575158892,
+      "learning_rate": 5.980407431160684e-08,
+      "loss": 0.7978,
+      "step": 41012
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5461623015554484,
+      "learning_rate": 5.972077773438223e-08,
+      "loss": 0.8935,
+      "step": 41013
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.620809547771287,
+      "learning_rate": 5.963753903260983e-08,
+      "loss": 0.8018,
+      "step": 41014
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.802355112919252,
+      "learning_rate": 5.95543582067748e-08,
+      "loss": 0.8146,
+      "step": 41015
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6784317725081936,
+      "learning_rate": 5.94712352573612e-08,
+      "loss": 0.7496,
+      "step": 41016
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4983838095980615,
+      "learning_rate": 5.938817018485421e-08,
+      "loss": 0.8936,
+      "step": 41017
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1910241326772675,
+      "learning_rate": 5.930516298973565e-08,
+      "loss": 0.8087,
+      "step": 41018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7562787379759879,
+      "learning_rate": 5.922221367248959e-08,
+      "loss": 0.8743,
+      "step": 41019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.606373891467492,
+      "learning_rate": 5.913932223359897e-08,
+      "loss": 0.7866,
+      "step": 41020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.808717587379972,
+      "learning_rate": 5.905648867354563e-08,
+      "loss": 0.8316,
+      "step": 41021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7838324149414354,
+      "learning_rate": 5.897371299281473e-08,
+      "loss": 0.7429,
+      "step": 41022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6454757958807917,
+      "learning_rate": 5.8890995191884795e-08,
+      "loss": 0.8864,
+      "step": 41023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.766966173915765,
+      "learning_rate": 5.880833527123875e-08,
+      "loss": 0.7177,
+      "step": 41024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7330389984923396,
+      "learning_rate": 5.872573323135844e-08,
+      "loss": 0.6781,
+      "step": 41025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7131559897587154,
+      "learning_rate": 5.8643189072723485e-08,
+      "loss": 0.8899,
+      "step": 41026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4665424729534848,
+      "learning_rate": 5.856070279581572e-08,
+      "loss": 0.9085,
+      "step": 41027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9138365795286707,
+      "learning_rate": 5.847827440111475e-08,
+      "loss": 0.8678,
+      "step": 41028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6495275492096537,
+      "learning_rate": 5.839590388910132e-08,
+      "loss": 0.9227,
+      "step": 41029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5195070759871698,
+      "learning_rate": 5.831359126025393e-08,
+      "loss": 0.837,
+      "step": 41030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4806155869385056,
+      "learning_rate": 5.8231336515052196e-08,
+      "loss": 0.831,
+      "step": 41031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.715538906480099,
+      "learning_rate": 5.814913965397573e-08,
+      "loss": 0.9029,
+      "step": 41032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.802878339347037,
+      "learning_rate": 5.806700067750193e-08,
+      "loss": 0.816,
+      "step": 41033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5483611637020536,
+      "learning_rate": 5.79849195861093e-08,
+      "loss": 0.8926,
+      "step": 41034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7474683065321559,
+      "learning_rate": 5.7902896380276354e-08,
+      "loss": 0.7062,
+      "step": 41035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3971716714829476,
+      "learning_rate": 5.782093106048159e-08,
+      "loss": 0.7441,
+      "step": 41036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.595188038164802,
+      "learning_rate": 5.7739023627199074e-08,
+      "loss": 0.815,
+      "step": 41037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6572142854782603,
+      "learning_rate": 5.765717408090843e-08,
+      "loss": 0.7608,
+      "step": 41038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6830686306532574,
+      "learning_rate": 5.7575382422085935e-08,
+      "loss": 0.7886,
+      "step": 41039
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.724238730741248,
+      "learning_rate": 5.749364865120677e-08,
+      "loss": 0.7198,
+      "step": 41040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.660492474118983,
+      "learning_rate": 5.7411972768747213e-08,
+      "loss": 0.7799,
+      "step": 41041
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6677900685376095,
+      "learning_rate": 5.733035477518245e-08,
+      "loss": 0.8804,
+      "step": 41042
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4317601101079773,
+      "learning_rate": 5.7248794670989874e-08,
+      "loss": 0.8457,
+      "step": 41043
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.526615800649673,
+      "learning_rate": 5.7167292456641324e-08,
+      "loss": 0.7857,
+      "step": 41044
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5333811131773734,
+      "learning_rate": 5.708584813261308e-08,
+      "loss": 0.7656,
+      "step": 41045
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7551385525037446,
+      "learning_rate": 5.700446169937812e-08,
+      "loss": 0.8209,
+      "step": 41046
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7408696056585273,
+      "learning_rate": 5.69231331574116e-08,
+      "loss": 0.818,
+      "step": 41047
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8872104135059826,
+      "learning_rate": 5.6841862507186484e-08,
+      "loss": 0.8292,
+      "step": 41048
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5313083812564883,
+      "learning_rate": 5.6760649749175726e-08,
+      "loss": 0.7787,
+      "step": 41049
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.066755783905316,
+      "learning_rate": 5.667949488385338e-08,
+      "loss": 0.8674,
+      "step": 41050
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6890005139186053,
+      "learning_rate": 5.6598397911690194e-08,
+      "loss": 0.8225,
+      "step": 41051
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7298119702083679,
+      "learning_rate": 5.651735883316023e-08,
+      "loss": 0.7318,
+      "step": 41052
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5572389200525363,
+      "learning_rate": 5.643637764873311e-08,
+      "loss": 0.7909,
+      "step": 41053
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5763038488010634,
+      "learning_rate": 5.635545435888179e-08,
+      "loss": 0.8456,
+      "step": 41054
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5419348800797432,
+      "learning_rate": 5.6274588964077006e-08,
+      "loss": 0.7752,
+      "step": 41055
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6356067626498187,
+      "learning_rate": 5.619378146478949e-08,
+      "loss": 0.7978,
+      "step": 41056
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6066475853205775,
+      "learning_rate": 5.6113031861491086e-08,
+      "loss": 0.8688,
+      "step": 41057
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9742332488143408,
+      "learning_rate": 5.60323401546492e-08,
+      "loss": 0.775,
+      "step": 41058
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.689554815663477,
+      "learning_rate": 5.595170634473568e-08,
+      "loss": 0.7553,
+      "step": 41059
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6815905897215424,
+      "learning_rate": 5.587113043222015e-08,
+      "loss": 0.8497,
+      "step": 41060
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5685199384651514,
+      "learning_rate": 5.579061241757111e-08,
+      "loss": 0.8498,
+      "step": 41061
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9730967801979982,
+      "learning_rate": 5.571015230125709e-08,
+      "loss": 0.7917,
+      "step": 41062
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7671738022758776,
+      "learning_rate": 5.5629750083746606e-08,
+      "loss": 0.6632,
+      "step": 41063
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.697382673957747,
+      "learning_rate": 5.5549405765509265e-08,
+      "loss": 0.849,
+      "step": 41064
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5877067050969145,
+      "learning_rate": 5.546911934701027e-08,
+      "loss": 0.8339,
+      "step": 41065
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4822257567579253,
+      "learning_rate": 5.538889082871923e-08,
+      "loss": 0.7285,
+      "step": 41066
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5520553830577877,
+      "learning_rate": 5.5308720211102454e-08,
+      "loss": 0.8688,
+      "step": 41067
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5770848831821873,
+      "learning_rate": 5.522860749462622e-08,
+      "loss": 0.782,
+      "step": 41068
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6270723877698767,
+      "learning_rate": 5.514855267975794e-08,
+      "loss": 0.9217,
+      "step": 41069
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.13087787322486,
+      "learning_rate": 5.506855576696279e-08,
+      "loss": 0.779,
+      "step": 41070
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6958028595414913,
+      "learning_rate": 5.498861675670819e-08,
+      "loss": 0.7551,
+      "step": 41071
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6648719949191402,
+      "learning_rate": 5.490873564945709e-08,
+      "loss": 0.7817,
+      "step": 41072
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7650405038049943,
+      "learning_rate": 5.482891244567689e-08,
+      "loss": 0.6962,
+      "step": 41073
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.259248622733783,
+      "learning_rate": 5.474914714583057e-08,
+      "loss": 0.8376,
+      "step": 41074
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.770217125380984,
+      "learning_rate": 5.46694397503833e-08,
+      "loss": 0.6961,
+      "step": 41075
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7194835734468454,
+      "learning_rate": 5.458979025979916e-08,
+      "loss": 0.905,
+      "step": 41076
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.509292562391661,
+      "learning_rate": 5.451019867454221e-08,
+      "loss": 0.9374,
+      "step": 41077
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.592747828017793,
+      "learning_rate": 5.4430664995076543e-08,
+      "loss": 0.6857,
+      "step": 41078
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7685090837801023,
+      "learning_rate": 5.435118922186289e-08,
+      "loss": 0.6859,
+      "step": 41079
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.549497856226855,
+      "learning_rate": 5.427177135536532e-08,
+      "loss": 0.7398,
+      "step": 41080
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5198759823182932,
+      "learning_rate": 5.4192411396046806e-08,
+      "loss": 0.9042,
+      "step": 41081
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5051644802829414,
+      "learning_rate": 5.41131093443692e-08,
+      "loss": 0.7203,
+      "step": 41082
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.526058694394717,
+      "learning_rate": 5.4033865200793236e-08,
+      "loss": 0.7913,
+      "step": 41083
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5285192188801473,
+      "learning_rate": 5.395467896578077e-08,
+      "loss": 0.7546,
+      "step": 41084
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7527998194098326,
+      "learning_rate": 5.387555063979366e-08,
+      "loss": 0.8269,
+      "step": 41085
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8078262051129013,
+      "learning_rate": 5.3796480223292643e-08,
+      "loss": 0.7169,
+      "step": 41086
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.67390818034783,
+      "learning_rate": 5.371746771673625e-08,
+      "loss": 0.8424,
+      "step": 41087
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6047291553115894,
+      "learning_rate": 5.363851312058632e-08,
+      "loss": 0.7996,
+      "step": 41088
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4877355149446077,
+      "learning_rate": 5.355961643530139e-08,
+      "loss": 0.9161,
+      "step": 41089
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.238519850981052,
+      "learning_rate": 5.3480777661341077e-08,
+      "loss": 0.906,
+      "step": 41090
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7090810190701264,
+      "learning_rate": 5.340199679916502e-08,
+      "loss": 0.7706,
+      "step": 41091
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5492075139982806,
+      "learning_rate": 5.332327384923175e-08,
+      "loss": 0.7804,
+      "step": 41092
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4285133548313476,
+      "learning_rate": 5.3244608811999776e-08,
+      "loss": 0.7378,
+      "step": 41093
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.59458905712432,
+      "learning_rate": 5.3166001687926515e-08,
+      "loss": 0.8606,
+      "step": 41094
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.671482060720957,
+      "learning_rate": 5.308745247747049e-08,
+      "loss": 0.8706,
+      "step": 41095
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.098494169818741,
+      "learning_rate": 5.3008961181086895e-08,
+      "loss": 0.8155,
+      "step": 41096
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5809514949993424,
+      "learning_rate": 5.293052779923535e-08,
+      "loss": 0.8272,
+      "step": 41097
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5397540023410017,
+      "learning_rate": 5.285215233237107e-08,
+      "loss": 0.7629,
+      "step": 41098
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5197366981674851,
+      "learning_rate": 5.277383478095144e-08,
+      "loss": 0.9767,
+      "step": 41099
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5265335722145534,
+      "learning_rate": 5.2695575145431664e-08,
+      "loss": 0.7207,
+      "step": 41100
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4443545866614058,
+      "learning_rate": 5.261737342626694e-08,
+      "loss": 0.8246,
+      "step": 41101
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5533359865343672,
+      "learning_rate": 5.253922962391356e-08,
+      "loss": 0.7552,
+      "step": 41102
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.494641455918641,
+      "learning_rate": 5.2461143738825605e-08,
+      "loss": 0.7875,
+      "step": 41103
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.244158439092424,
+      "learning_rate": 5.238311577145827e-08,
+      "loss": 0.8086,
+      "step": 41104
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.559419628896069,
+      "learning_rate": 5.230514572226675e-08,
+      "loss": 0.9106,
+      "step": 41105
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.862426908767693,
+      "learning_rate": 5.22272335917029e-08,
+      "loss": 0.668,
+      "step": 41106
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.003446705086436,
+      "learning_rate": 5.214937938022191e-08,
+      "loss": 0.941,
+      "step": 41107
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5217597038828314,
+      "learning_rate": 5.2071583088276755e-08,
+      "loss": 0.7895,
+      "step": 41108
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7266795830167478,
+      "learning_rate": 5.1993844716321515e-08,
+      "loss": 0.8408,
+      "step": 41109
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3988244766691709,
+      "learning_rate": 5.191616426480584e-08,
+      "loss": 0.6924,
+      "step": 41110
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6419579518601346,
+      "learning_rate": 5.183854173418379e-08,
+      "loss": 0.8138,
+      "step": 41111
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7357721557812962,
+      "learning_rate": 5.1760977124908354e-08,
+      "loss": 0.8237,
+      "step": 41112
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.366633457869744,
+      "learning_rate": 5.168347043742916e-08,
+      "loss": 0.7551,
+      "step": 41113
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8006127148329205,
+      "learning_rate": 5.160602167219919e-08,
+      "loss": 0.7438,
+      "step": 41114
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6128216073996933,
+      "learning_rate": 5.152863082966808e-08,
+      "loss": 0.8407,
+      "step": 41115
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6159424696401,
+      "learning_rate": 5.145129791028658e-08,
+      "loss": 0.9058,
+      "step": 41116
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.519823522743918,
+      "learning_rate": 5.137402291450655e-08,
+      "loss": 0.8412,
+      "step": 41117
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8403865391490086,
+      "learning_rate": 5.129680584277541e-08,
+      "loss": 0.8485,
+      "step": 41118
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8660378144851004,
+      "learning_rate": 5.1219646695543915e-08,
+      "loss": 0.8484,
+      "step": 41119
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9250563843463306,
+      "learning_rate": 5.1142545473261694e-08,
+      "loss": 0.7631,
+      "step": 41120
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4530038115936106,
+      "learning_rate": 5.10655021763784e-08,
+      "loss": 0.784,
+      "step": 41121
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7875978184410528,
+      "learning_rate": 5.0988516805340337e-08,
+      "loss": 0.9568,
+      "step": 41122
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8217396829063903,
+      "learning_rate": 5.091158936059715e-08,
+      "loss": 0.8638,
+      "step": 41123
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6921126713780484,
+      "learning_rate": 5.0834719842596246e-08,
+      "loss": 0.7015,
+      "step": 41124
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5369580419225606,
+      "learning_rate": 5.075790825178617e-08,
+      "loss": 0.8434,
+      "step": 41125
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7747327242908006,
+      "learning_rate": 5.0681154588612114e-08,
+      "loss": 0.9444,
+      "step": 41126
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.705854205426184,
+      "learning_rate": 5.0604458853523716e-08,
+      "loss": 0.9903,
+      "step": 41127
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.450664874233554,
+      "learning_rate": 5.052782104696508e-08,
+      "loss": 0.8716,
+      "step": 41128
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6171227511967832,
+      "learning_rate": 5.045124116938471e-08,
+      "loss": 0.8639,
+      "step": 41129
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6267101061370695,
+      "learning_rate": 5.037471922122561e-08,
+      "loss": 0.8425,
+      "step": 41130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.634482012643192,
+      "learning_rate": 5.02982552029363e-08,
+      "loss": 0.8117,
+      "step": 41131
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5788350077150433,
+      "learning_rate": 5.022184911495864e-08,
+      "loss": 0.784,
+      "step": 41132
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7540928060182497,
+      "learning_rate": 5.0145500957741176e-08,
+      "loss": 0.8488,
+      "step": 41133
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7934764806674518,
+      "learning_rate": 5.006921073172688e-08,
+      "loss": 0.6521,
+      "step": 41134
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.566712413765833,
+      "learning_rate": 4.999297843735873e-08,
+      "loss": 0.9211,
+      "step": 41135
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6992223920454257,
+      "learning_rate": 4.991680407508193e-08,
+      "loss": 0.7839,
+      "step": 41136
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5570103895011496,
+      "learning_rate": 4.9840687645339446e-08,
+      "loss": 0.7453,
+      "step": 41137
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6833194535288118,
+      "learning_rate": 4.976462914857538e-08,
+      "loss": 0.765,
+      "step": 41138
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8027613879941029,
+      "learning_rate": 4.968862858523049e-08,
+      "loss": 0.7096,
+      "step": 41139
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6486936023591174,
+      "learning_rate": 4.961268595574886e-08,
+      "loss": 0.7597,
+      "step": 41140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.642654574133425,
+      "learning_rate": 4.9536801260573475e-08,
+      "loss": 0.88,
+      "step": 41141
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.025785928656936,
+      "learning_rate": 4.946097450014509e-08,
+      "loss": 0.6792,
+      "step": 41142
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.673986834350542,
+      "learning_rate": 4.9385205674904455e-08,
+      "loss": 0.9405,
+      "step": 41143
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.672458774127032,
+      "learning_rate": 4.930949478529346e-08,
+      "loss": 0.8036,
+      "step": 41144
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5184909502257835,
+      "learning_rate": 4.923384183175284e-08,
+      "loss": 0.6799,
+      "step": 41145
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8035379597858322,
+      "learning_rate": 4.9158246814723367e-08,
+      "loss": 0.6901,
+      "step": 41146
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0009646276306103,
+      "learning_rate": 4.90827097346458e-08,
+      "loss": 0.8285,
+      "step": 41147
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0620443515196247,
+      "learning_rate": 4.9007230591957555e-08,
+      "loss": 0.8161,
+      "step": 41148
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7526339770975965,
+      "learning_rate": 4.8931809387100515e-08,
+      "loss": 0.851,
+      "step": 41149
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1701271286404036,
+      "learning_rate": 4.8856446120513215e-08,
+      "loss": 0.838,
+      "step": 41150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7665928715620209,
+      "learning_rate": 4.8781140792633074e-08,
+      "loss": 0.6621,
+      "step": 41151
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8103021648229884,
+      "learning_rate": 4.870589340390086e-08,
+      "loss": 0.7681,
+      "step": 41152
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6782791229403953,
+      "learning_rate": 4.863070395475289e-08,
+      "loss": 0.7619,
+      "step": 41153
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.53651434753673,
+      "learning_rate": 4.85555724456277e-08,
+      "loss": 0.772,
+      "step": 41154
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7211115094504852,
+      "learning_rate": 4.848049887696271e-08,
+      "loss": 0.7256,
+      "step": 41155
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.776952554809742,
+      "learning_rate": 4.8405483249194255e-08,
+      "loss": 0.6853,
+      "step": 41156
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7996522303970091,
+      "learning_rate": 4.8330525562759745e-08,
+      "loss": 0.7155,
+      "step": 41157
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7362910947334758,
+      "learning_rate": 4.8255625818096616e-08,
+      "loss": 0.7768,
+      "step": 41158
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6536614911979661,
+      "learning_rate": 4.818078401563897e-08,
+      "loss": 0.7169,
+      "step": 41159
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.53735437914906,
+      "learning_rate": 4.810600015582312e-08,
+      "loss": 0.8401,
+      "step": 41160
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5995315967438553,
+      "learning_rate": 4.803127423908538e-08,
+      "loss": 0.8777,
+      "step": 41161
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.708180514680619,
+      "learning_rate": 4.795660626586096e-08,
+      "loss": 0.6658,
+      "step": 41162
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5797959889644608,
+      "learning_rate": 4.788199623658285e-08,
+      "loss": 0.7241,
+      "step": 41163
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8206089649362309,
+      "learning_rate": 4.7807444151687363e-08,
+      "loss": 0.8039,
+      "step": 41164
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8312498196895834,
+      "learning_rate": 4.773295001160638e-08,
+      "loss": 0.9716,
+      "step": 41165
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7228981330729078,
+      "learning_rate": 4.765851381677622e-08,
+      "loss": 0.6344,
+      "step": 41166
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.934657694960392,
+      "learning_rate": 4.758413556762986e-08,
+      "loss": 0.7511,
+      "step": 41167
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.58663652109454,
+      "learning_rate": 4.750981526459808e-08,
+      "loss": 0.939,
+      "step": 41168
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.506603627061087,
+      "learning_rate": 4.7435552908114966e-08,
+      "loss": 0.7337,
+      "step": 41169
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5745860593371854,
+      "learning_rate": 4.7361348498613516e-08,
+      "loss": 0.7899,
+      "step": 41170
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5308277196131737,
+      "learning_rate": 4.728720203652559e-08,
+      "loss": 0.8774,
+      "step": 41171
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6778933850243178,
+      "learning_rate": 4.721311352228197e-08,
+      "loss": 0.9042,
+      "step": 41172
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6673253436665796,
+      "learning_rate": 4.713908295631564e-08,
+      "loss": 0.8515,
+      "step": 41173
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6517033750086005,
+      "learning_rate": 4.706511033905514e-08,
+      "loss": 0.8573,
+      "step": 41174
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8092176144130478,
+      "learning_rate": 4.6991195670932355e-08,
+      "loss": 0.8583,
+      "step": 41175
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6328062025895866,
+      "learning_rate": 4.6917338952380266e-08,
+      "loss": 0.9293,
+      "step": 41176
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6202455942054028,
+      "learning_rate": 4.684354018382409e-08,
+      "loss": 0.7953,
+      "step": 41177
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9405371171887693,
+      "learning_rate": 4.6769799365696814e-08,
+      "loss": 0.753,
+      "step": 41178
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7888376475195373,
+      "learning_rate": 4.669611649842698e-08,
+      "loss": 0.9059,
+      "step": 41179
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4329713096973138,
+      "learning_rate": 4.662249158244425e-08,
+      "loss": 0.7512,
+      "step": 41180
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5417481694255428,
+      "learning_rate": 4.6548924618176064e-08,
+      "loss": 0.8044,
+      "step": 41181
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.685171823709823,
+      "learning_rate": 4.6475415606049845e-08,
+      "loss": 0.9886,
+      "step": 41182
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.641180358602358,
+      "learning_rate": 4.640196454649637e-08,
+      "loss": 0.927,
+      "step": 41183
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4716168845387856,
+      "learning_rate": 4.632857143994196e-08,
+      "loss": 0.8204,
+      "step": 41184
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8077292072074191,
+      "learning_rate": 4.625523628681405e-08,
+      "loss": 0.7439,
+      "step": 41185
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9151169354033255,
+      "learning_rate": 4.6181959087538975e-08,
+      "loss": 0.8615,
+      "step": 41186
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4789341531600262,
+      "learning_rate": 4.610873984254416e-08,
+      "loss": 0.8482,
+      "step": 41187
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6921932694679862,
+      "learning_rate": 4.603557855225482e-08,
+      "loss": 0.7638,
+      "step": 41188
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.002160381340914,
+      "learning_rate": 4.59624752170984e-08,
+      "loss": 0.8456,
+      "step": 41189
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7307096983939929,
+      "learning_rate": 4.588942983750011e-08,
+      "loss": 0.6397,
+      "step": 41190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5379763791634835,
+      "learning_rate": 4.5816442413884055e-08,
+      "loss": 0.853,
+      "step": 41191
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.637283713789487,
+      "learning_rate": 4.574351294667656e-08,
+      "loss": 0.8822,
+      "step": 41192
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8005088990959541,
+      "learning_rate": 4.5670641436301735e-08,
+      "loss": 0.7393,
+      "step": 41193
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6512831719991459,
+      "learning_rate": 4.559782788318479e-08,
+      "loss": 0.8579,
+      "step": 41194
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9895835608523371,
+      "learning_rate": 4.552507228774761e-08,
+      "loss": 0.7026,
+      "step": 41195
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.197679430302863,
+      "learning_rate": 4.545237465041541e-08,
+      "loss": 0.9151,
+      "step": 41196
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7454292985243642,
+      "learning_rate": 4.5379734971611186e-08,
+      "loss": 0.7991,
+      "step": 41197
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5309250122794051,
+      "learning_rate": 4.530715325175794e-08,
+      "loss": 0.7738,
+      "step": 41198
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.672680563518362,
+      "learning_rate": 4.5234629491277546e-08,
+      "loss": 0.7227,
+      "step": 41199
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7833938743385456,
+      "learning_rate": 4.516216369059301e-08,
+      "loss": 0.8764,
+      "step": 41200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8516599154762658,
+      "learning_rate": 4.5089755850126206e-08,
+      "loss": 0.9983,
+      "step": 41201
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8855214639999085,
+      "learning_rate": 4.501740597029791e-08,
+      "loss": 0.9206,
+      "step": 41202
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.795647367730506,
+      "learning_rate": 4.494511405153002e-08,
+      "loss": 0.7886,
+      "step": 41203
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7546641901239162,
+      "learning_rate": 4.4872880094244395e-08,
+      "loss": 0.8699,
+      "step": 41204
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9286519228230115,
+      "learning_rate": 4.4800704098859614e-08,
+      "loss": 0.7536,
+      "step": 41205
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8421900015446167,
+      "learning_rate": 4.472858606579755e-08,
+      "loss": 0.7347,
+      "step": 41206
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5652799040500815,
+      "learning_rate": 4.465652599547676e-08,
+      "loss": 0.8479,
+      "step": 41207
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4929989118789875,
+      "learning_rate": 4.4584523888318023e-08,
+      "loss": 0.6896,
+      "step": 41208
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9146201213717473,
+      "learning_rate": 4.451257974473988e-08,
+      "loss": 0.9871,
+      "step": 41209
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8179451056345928,
+      "learning_rate": 4.4440693565160895e-08,
+      "loss": 0.8647,
+      "step": 41210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6617332259554716,
+      "learning_rate": 4.4368865350000736e-08,
+      "loss": 0.8819,
+      "step": 41211
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6274706546104887,
+      "learning_rate": 4.4297095099676836e-08,
+      "loss": 0.8338,
+      "step": 41212
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5573140471383649,
+      "learning_rate": 4.4225382814606643e-08,
+      "loss": 0.8402,
+      "step": 41213
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5282511627683866,
+      "learning_rate": 4.415372849520871e-08,
+      "loss": 0.644,
+      "step": 41214
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4966508048343532,
+      "learning_rate": 4.408213214189938e-08,
+      "loss": 0.8207,
+      "step": 41215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5528934850646678,
+      "learning_rate": 4.4010593755096086e-08,
+      "loss": 0.8675,
+      "step": 41216
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.500109529217766,
+      "learning_rate": 4.393911333521517e-08,
+      "loss": 0.7167,
+      "step": 41217
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5095985986498515,
+      "learning_rate": 4.386769088267295e-08,
+      "loss": 0.8962,
+      "step": 41218
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6612806062418788,
+      "learning_rate": 4.379632639788467e-08,
+      "loss": 0.8591,
+      "step": 41219
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6533274521211632,
+      "learning_rate": 4.372501988126665e-08,
+      "loss": 0.8796,
+      "step": 41220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4134885174997305,
+      "learning_rate": 4.365377133323301e-08,
+      "loss": 0.7751,
+      "step": 41221
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8403722945390489,
+      "learning_rate": 4.3582580754200074e-08,
+      "loss": 0.7291,
+      "step": 41222
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4432375042065102,
+      "learning_rate": 4.3511448144580857e-08,
+      "loss": 0.6959,
+      "step": 41223
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0562609521510677,
+      "learning_rate": 4.3440373504790576e-08,
+      "loss": 0.8219,
+      "step": 41224
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6455798177217984,
+      "learning_rate": 4.336935683524224e-08,
+      "loss": 0.885,
+      "step": 41225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7152845536524788,
+      "learning_rate": 4.3298398136349954e-08,
+      "loss": 0.8742,
+      "step": 41226
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6213364015942022,
+      "learning_rate": 4.322749740852672e-08,
+      "loss": 0.7387,
+      "step": 41227
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8733331555494355,
+      "learning_rate": 4.315665465218555e-08,
+      "loss": 0.816,
+      "step": 41228
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7813060633969195,
+      "learning_rate": 4.308586986773833e-08,
+      "loss": 0.9461,
+      "step": 41229
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7561642376174481,
+      "learning_rate": 4.3015143055596954e-08,
+      "loss": 0.7258,
+      "step": 41230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6911201629658634,
+      "learning_rate": 4.2944474216174425e-08,
+      "loss": 0.9301,
+      "step": 41231
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.495802571844526,
+      "learning_rate": 4.287386334988153e-08,
+      "loss": 0.7881,
+      "step": 41232
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6025197435411154,
+      "learning_rate": 4.2803310457129046e-08,
+      "loss": 0.8415,
+      "step": 41233
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7687613145300285,
+      "learning_rate": 4.273281553832775e-08,
+      "loss": 0.7482,
+      "step": 41234
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6810451089252714,
+      "learning_rate": 4.266237859388955e-08,
+      "loss": 0.792,
+      "step": 41235
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.561846914690639,
+      "learning_rate": 4.2591999624223e-08,
+      "loss": 0.7774,
+      "step": 41236
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4865688380554123,
+      "learning_rate": 4.2521678629737774e-08,
+      "loss": 0.7423,
+      "step": 41237
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7679342927118351,
+      "learning_rate": 4.245141561084354e-08,
+      "loss": 0.6875,
+      "step": 41238
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.640871573904555,
+      "learning_rate": 4.238121056794997e-08,
+      "loss": 0.7382,
+      "step": 41239
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8949806731691936,
+      "learning_rate": 4.231106350146563e-08,
+      "loss": 0.8483,
+      "step": 41240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.127800964288928,
+      "learning_rate": 4.224097441179908e-08,
+      "loss": 0.7876,
+      "step": 41241
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6134710436935435,
+      "learning_rate": 4.217094329935778e-08,
+      "loss": 0.7943,
+      "step": 41242
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6243426708413975,
+      "learning_rate": 4.210097016455028e-08,
+      "loss": 0.8383,
+      "step": 41243
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6092374268529959,
+      "learning_rate": 4.2031055007784036e-08,
+      "loss": 0.7055,
+      "step": 41244
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5597614756095965,
+      "learning_rate": 4.196119782946429e-08,
+      "loss": 0.8065,
+      "step": 41245
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.826562711607949,
+      "learning_rate": 4.1891398629999583e-08,
+      "loss": 0.7837,
+      "step": 41246
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0344534048176834,
+      "learning_rate": 4.182165740979627e-08,
+      "loss": 0.9193,
+      "step": 41247
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7626643189349829,
+      "learning_rate": 4.175197416925958e-08,
+      "loss": 0.7836,
+      "step": 41248
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6651698185424533,
+      "learning_rate": 4.168234890879586e-08,
+      "loss": 0.9542,
+      "step": 41249
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4573879953157367,
+      "learning_rate": 4.161278162881033e-08,
+      "loss": 0.7456,
+      "step": 41250
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4430571609184002,
+      "learning_rate": 4.1543272329707116e-08,
+      "loss": 0.838,
+      "step": 41251
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.637430695049814,
+      "learning_rate": 4.1473821011892565e-08,
+      "loss": 0.9169,
+      "step": 41252
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7064293365693887,
+      "learning_rate": 4.14044276757708e-08,
+      "loss": 0.6622,
+      "step": 41253
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7383165654770036,
+      "learning_rate": 4.133509232174482e-08,
+      "loss": 0.8524,
+      "step": 41254
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5574762321501934,
+      "learning_rate": 4.126581495021764e-08,
+      "loss": 0.8985,
+      "step": 41255
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3930272171386073,
+      "learning_rate": 4.119659556159561e-08,
+      "loss": 0.6872,
+      "step": 41256
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.089825094219969,
+      "learning_rate": 4.11274341562784e-08,
+      "loss": 0.7478,
+      "step": 41257
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3880396249019884,
+      "learning_rate": 4.105833073467125e-08,
+      "loss": 0.7909,
+      "step": 41258
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6834674117803394,
+      "learning_rate": 4.0989285297174944e-08,
+      "loss": 0.9532,
+      "step": 41259
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6971889017995874,
+      "learning_rate": 4.092029784419138e-08,
+      "loss": 0.7709,
+      "step": 41260
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6060426228156912,
+      "learning_rate": 4.085136837612358e-08,
+      "loss": 0.777,
+      "step": 41261
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6740667269300642,
+      "learning_rate": 4.0782496893372327e-08,
+      "loss": 0.8539,
+      "step": 41262
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5978504271451355,
+      "learning_rate": 4.071368339633841e-08,
+      "loss": 0.8544,
+      "step": 41263
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6973026115596952,
+      "learning_rate": 4.064492788542152e-08,
+      "loss": 0.7858,
+      "step": 41264
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0622428682593683,
+      "learning_rate": 4.057623036102465e-08,
+      "loss": 0.7555,
+      "step": 41265
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7274393529457623,
+      "learning_rate": 4.0507590823544164e-08,
+      "loss": 0.8402,
+      "step": 41266
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6302596090452255,
+      "learning_rate": 4.043900927338307e-08,
+      "loss": 0.7891,
+      "step": 41267
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5936002200941695,
+      "learning_rate": 4.037048571093771e-08,
+      "loss": 0.9111,
+      "step": 41268
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7032750193090036,
+      "learning_rate": 4.0302020136609995e-08,
+      "loss": 0.8425,
+      "step": 41269
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7826088189699447,
+      "learning_rate": 4.0233612550796273e-08,
+      "loss": 0.7104,
+      "step": 41270
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5824008989579998,
+      "learning_rate": 4.0165262953896225e-08,
+      "loss": 0.8886,
+      "step": 41271
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7889935895586232,
+      "learning_rate": 4.00969713463073e-08,
+      "loss": 0.6473,
+      "step": 41272
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.295303724383056,
+      "learning_rate": 4.002873772842586e-08,
+      "loss": 0.9745,
+      "step": 41273
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8210522995869791,
+      "learning_rate": 3.996056210065158e-08,
+      "loss": 0.7827,
+      "step": 41274
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.713629023973177,
+      "learning_rate": 3.989244446338081e-08,
+      "loss": 0.9421,
+      "step": 41275
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.848975289846571,
+      "learning_rate": 3.982438481700768e-08,
+      "loss": 0.9507,
+      "step": 41276
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.509067390095068,
+      "learning_rate": 3.975638316193076e-08,
+      "loss": 0.7816,
+      "step": 41277
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4544048139021601,
+      "learning_rate": 3.96884394985475e-08,
+      "loss": 0.8029,
+      "step": 41278
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9273575641689238,
+      "learning_rate": 3.962055382724983e-08,
+      "loss": 0.9193,
+      "step": 41279
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5408538557166915,
+      "learning_rate": 3.955272614843408e-08,
+      "loss": 0.7917,
+      "step": 41280
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6323503930917111,
+      "learning_rate": 3.948495646249662e-08,
+      "loss": 0.7215,
+      "step": 41281
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6537506298656846,
+      "learning_rate": 3.941724476983155e-08,
+      "loss": 0.7439,
+      "step": 41282
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3680302887146454,
+      "learning_rate": 3.934959107083303e-08,
+      "loss": 0.8022,
+      "step": 41283
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7451975828760475,
+      "learning_rate": 3.9281995365892944e-08,
+      "loss": 0.7231,
+      "step": 41284
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.603422490981674,
+      "learning_rate": 3.921445765540877e-08,
+      "loss": 0.7331,
+      "step": 41285
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6630243757699794,
+      "learning_rate": 3.914697793977018e-08,
+      "loss": 0.7618,
+      "step": 41286
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6511132341216261,
+      "learning_rate": 3.9079556219371317e-08,
+      "loss": 0.8438,
+      "step": 41287
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8029284726302437,
+      "learning_rate": 3.9012192494605197e-08,
+      "loss": 0.8681,
+      "step": 41288
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7565218775203195,
+      "learning_rate": 3.894488676586372e-08,
+      "loss": 0.7738,
+      "step": 41289
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7178278049197955,
+      "learning_rate": 3.8877639033538805e-08,
+      "loss": 0.7987,
+      "step": 41290
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.841734317029878,
+      "learning_rate": 3.881044929802236e-08,
+      "loss": 0.7916,
+      "step": 41291
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6533989338043575,
+      "learning_rate": 3.8743317559705174e-08,
+      "loss": 0.8755,
+      "step": 41292
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5635042809829582,
+      "learning_rate": 3.867624381897694e-08,
+      "loss": 0.7189,
+      "step": 41293
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7233015276810888,
+      "learning_rate": 3.8609228076229576e-08,
+      "loss": 0.8018,
+      "step": 41294
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0707828434489035,
+      "learning_rate": 3.854227033185387e-08,
+      "loss": 0.8372,
+      "step": 41295
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6825615536873626,
+      "learning_rate": 3.84753705862384e-08,
+      "loss": 0.6306,
+      "step": 41296
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.170799031876969,
+      "learning_rate": 3.840852883977286e-08,
+      "loss": 0.8187,
+      "step": 41297
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4841073753121394,
+      "learning_rate": 3.8341745092846936e-08,
+      "loss": 0.9493,
+      "step": 41298
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5357127665906618,
+      "learning_rate": 3.827501934584921e-08,
+      "loss": 0.8004,
+      "step": 41299
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5395148254482385,
+      "learning_rate": 3.8208351599167135e-08,
+      "loss": 0.7992,
+      "step": 41300
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.605338286936358,
+      "learning_rate": 3.814174185319153e-08,
+      "loss": 0.8342,
+      "step": 41301
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6197102945002535,
+      "learning_rate": 3.8075190108307624e-08,
+      "loss": 0.7868,
+      "step": 41302
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7280959350568408,
+      "learning_rate": 3.8008696364904006e-08,
+      "loss": 0.7694,
+      "step": 41303
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.536947388683971,
+      "learning_rate": 3.7942260623367035e-08,
+      "loss": 0.8144,
+      "step": 41304
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8495961075875925,
+      "learning_rate": 3.787588288408528e-08,
+      "loss": 0.6665,
+      "step": 41305
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7598627721010376,
+      "learning_rate": 3.7809563147443993e-08,
+      "loss": 0.8181,
+      "step": 41306
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5712637600504813,
+      "learning_rate": 3.7743301413828424e-08,
+      "loss": 0.7788,
+      "step": 41307
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7902799181601797,
+      "learning_rate": 3.767709768362604e-08,
+      "loss": 0.7292,
+      "step": 41308
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.460547448962806,
+      "learning_rate": 3.7610951957220975e-08,
+      "loss": 0.8509,
+      "step": 41309
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1491057053373126,
+      "learning_rate": 3.754486423499959e-08,
+      "loss": 0.8817,
+      "step": 41310
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7750666349314251,
+      "learning_rate": 3.747883451734602e-08,
+      "loss": 0.7235,
+      "step": 41311
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5352595584262776,
+      "learning_rate": 3.7412862804645514e-08,
+      "loss": 0.9201,
+      "step": 41312
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6303070092680596,
+      "learning_rate": 3.7346949097279984e-08,
+      "loss": 0.8174,
+      "step": 41313
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5448558341664071,
+      "learning_rate": 3.728109339563579e-08,
+      "loss": 0.8802,
+      "step": 41314
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2224659214049587,
+      "learning_rate": 3.721529570009486e-08,
+      "loss": 0.8684,
+      "step": 41315
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9309588855199766,
+      "learning_rate": 3.71495560110402e-08,
+      "loss": 0.8987,
+      "step": 41316
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6592848377597103,
+      "learning_rate": 3.7083874328855965e-08,
+      "loss": 0.8715,
+      "step": 41317
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.008107947778199,
+      "learning_rate": 3.701825065392184e-08,
+      "loss": 0.9806,
+      "step": 41318
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.607126900162198,
+      "learning_rate": 3.6952684986623076e-08,
+      "loss": 0.8397,
+      "step": 41319
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7678981981879538,
+      "learning_rate": 3.6887177327340486e-08,
+      "loss": 0.7663,
+      "step": 41320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6029184753920716,
+      "learning_rate": 3.682172767645375e-08,
+      "loss": 0.7707,
+      "step": 41321
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.418685077801361,
+      "learning_rate": 3.675633603434592e-08,
+      "loss": 0.7822,
+      "step": 41322
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.475912358756703,
+      "learning_rate": 3.6691002401396666e-08,
+      "loss": 0.7801,
+      "step": 41323
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5545737374716149,
+      "learning_rate": 3.662572677798682e-08,
+      "loss": 0.7648,
+      "step": 41324
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5101549651763158,
+      "learning_rate": 3.656050916449716e-08,
+      "loss": 1.0428,
+      "step": 41325
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6311523950728706,
+      "learning_rate": 3.649534956130518e-08,
+      "loss": 0.8237,
+      "step": 41326
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4186708918384239,
+      "learning_rate": 3.64302479687928e-08,
+      "loss": 0.7937,
+      "step": 41327
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7886851363374364,
+      "learning_rate": 3.636520438733748e-08,
+      "loss": 0.6362,
+      "step": 41328
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4954809738873134,
+      "learning_rate": 3.6300218817317825e-08,
+      "loss": 0.8174,
+      "step": 41329
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4788578634684624,
+      "learning_rate": 3.6235291259113516e-08,
+      "loss": 0.6627,
+      "step": 41330
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6361132135144631,
+      "learning_rate": 3.6170421713102034e-08,
+      "loss": 0.8055,
+      "step": 41331
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5688407603503207,
+      "learning_rate": 3.6105610179659745e-08,
+      "loss": 0.8732,
+      "step": 41332
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8487346416333803,
+      "learning_rate": 3.6040856659166345e-08,
+      "loss": 0.7976,
+      "step": 41333
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.661288110749146,
+      "learning_rate": 3.597616115199709e-08,
+      "loss": 0.8669,
+      "step": 41334
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6367988126210467,
+      "learning_rate": 3.5911523658528347e-08,
+      "loss": 0.786,
+      "step": 41335
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4325541226952003,
+      "learning_rate": 3.58469441791387e-08,
+      "loss": 0.7385,
+      "step": 41336
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2161707596256313,
+      "learning_rate": 3.578242271420229e-08,
+      "loss": 0.7374,
+      "step": 41337
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7513717101380002,
+      "learning_rate": 3.571795926409549e-08,
+      "loss": 0.8726,
+      "step": 41338
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7626668274540104,
+      "learning_rate": 3.5653553829192444e-08,
+      "loss": 0.7086,
+      "step": 41339
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6188895296099288,
+      "learning_rate": 3.558920640986952e-08,
+      "loss": 0.7376,
+      "step": 41340
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7051724314998093,
+      "learning_rate": 3.5524917006500844e-08,
+      "loss": 0.8625,
+      "step": 41341
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6375981315650718,
+      "learning_rate": 3.546068561946059e-08,
+      "loss": 0.7549,
+      "step": 41342
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5627990104815892,
+      "learning_rate": 3.5396512249123996e-08,
+      "loss": 0.8065,
+      "step": 41343
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6192816958586982,
+      "learning_rate": 3.533239689586188e-08,
+      "loss": 0.8811,
+      "step": 41344
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7356197644368956,
+      "learning_rate": 3.52683395600506e-08,
+      "loss": 0.8831,
+      "step": 41345
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4857431816944862,
+      "learning_rate": 3.5204340242062096e-08,
+      "loss": 0.8566,
+      "step": 41346
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.428756907260328,
+      "learning_rate": 3.5140398942268286e-08,
+      "loss": 0.808,
+      "step": 41347
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5991042267659676,
+      "learning_rate": 3.50765156610422e-08,
+      "loss": 0.8422,
+      "step": 41348
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.516587573900852,
+      "learning_rate": 3.5012690398754654e-08,
+      "loss": 0.9087,
+      "step": 41349
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4118779760239604,
+      "learning_rate": 3.494892315577869e-08,
+      "loss": 0.7019,
+      "step": 41350
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7105862870277209,
+      "learning_rate": 3.488521393248401e-08,
+      "loss": 0.8306,
+      "step": 41351
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6656697043821653,
+      "learning_rate": 3.4821562729243644e-08,
+      "loss": 0.7222,
+      "step": 41352
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.612171570085174,
+      "learning_rate": 3.475796954642729e-08,
+      "loss": 0.8824,
+      "step": 41353
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5130127596181517,
+      "learning_rate": 3.469443438440467e-08,
+      "loss": 0.7415,
+      "step": 41354
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.422130946859379,
+      "learning_rate": 3.463095724354548e-08,
+      "loss": 0.8852,
+      "step": 41355
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.523896118321193,
+      "learning_rate": 3.456753812421943e-08,
+      "loss": 0.8484,
+      "step": 41356
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6350139404012598,
+      "learning_rate": 3.450417702679731e-08,
+      "loss": 0.7486,
+      "step": 41357
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5302036514220678,
+      "learning_rate": 3.444087395164664e-08,
+      "loss": 0.6737,
+      "step": 41358
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6594729849443461,
+      "learning_rate": 3.437762889913598e-08,
+      "loss": 0.8254,
+      "step": 41359
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.540930706287381,
+      "learning_rate": 3.4314441869633954e-08,
+      "loss": 0.8255,
+      "step": 41360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5262346934756086,
+      "learning_rate": 3.425131286350802e-08,
+      "loss": 0.9654,
+      "step": 41361
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7237651694435043,
+      "learning_rate": 3.418824188112679e-08,
+      "loss": 0.781,
+      "step": 41362
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7835839464327824,
+      "learning_rate": 3.412522892285663e-08,
+      "loss": 0.6889,
+      "step": 41363
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7681547999103941,
+      "learning_rate": 3.4062273989063923e-08,
+      "loss": 0.7418,
+      "step": 41364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4503051194047438,
+      "learning_rate": 3.3999377080115026e-08,
+      "loss": 0.7917,
+      "step": 41365
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6587911134192548,
+      "learning_rate": 3.393653819637854e-08,
+      "loss": 0.9536,
+      "step": 41366
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.453984724396006,
+      "learning_rate": 3.387375733821863e-08,
+      "loss": 0.7876,
+      "step": 41367
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5303952685571456,
+      "learning_rate": 3.381103450600054e-08,
+      "loss": 0.6555,
+      "step": 41368
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.396672679532365,
+      "learning_rate": 3.3748369700089547e-08,
+      "loss": 0.7988,
+      "step": 41369
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.514426476270653,
+      "learning_rate": 3.368576292085091e-08,
+      "loss": 0.8781,
+      "step": 41370
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5572635011034006,
+      "learning_rate": 3.362321416864878e-08,
+      "loss": 0.876,
+      "step": 41371
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5769310114338901,
+      "learning_rate": 3.356072344384842e-08,
+      "loss": 1.0127,
+      "step": 41372
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7667852073529184,
+      "learning_rate": 3.3498290746812876e-08,
+      "loss": 0.8404,
+      "step": 41373
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4586249926467334,
+      "learning_rate": 3.3435916077905196e-08,
+      "loss": 0.848,
+      "step": 41374
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2291495859870873,
+      "learning_rate": 3.3373599437488414e-08,
+      "loss": 0.7276,
+      "step": 41375
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5915117290878573,
+      "learning_rate": 3.331134082592669e-08,
+      "loss": 0.8321,
+      "step": 41376
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6800823610431321,
+      "learning_rate": 3.3249140243581947e-08,
+      "loss": 0.8003,
+      "step": 41377
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7353493287163881,
+      "learning_rate": 3.318699769081501e-08,
+      "loss": 0.8591,
+      "step": 41378
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.209124005824355,
+      "learning_rate": 3.312491316798894e-08,
+      "loss": 0.8967,
+      "step": 41379
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5513742972935274,
+      "learning_rate": 3.3062886675466754e-08,
+      "loss": 0.8494,
+      "step": 41380
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8912848925800607,
+      "learning_rate": 3.300091821360596e-08,
+      "loss": 0.8319,
+      "step": 41381
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6242037252351749,
+      "learning_rate": 3.29390077827707e-08,
+      "loss": 0.893,
+      "step": 41382
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4683213360796203,
+      "learning_rate": 3.287715538331959e-08,
+      "loss": 0.8296,
+      "step": 41383
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.624855411343558,
+      "learning_rate": 3.281536101561233e-08,
+      "loss": 0.8695,
+      "step": 41384
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7252553053185105,
+      "learning_rate": 3.2753624680010864e-08,
+      "loss": 0.7067,
+      "step": 41385
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.495314401629862,
+      "learning_rate": 3.2691946376872674e-08,
+      "loss": 0.7834,
+      "step": 41386
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.533156358710257,
+      "learning_rate": 3.2630326106558584e-08,
+      "loss": 0.8448,
+      "step": 41387
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8116563734630192,
+      "learning_rate": 3.256876386942498e-08,
+      "loss": 0.7325,
+      "step": 41388
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.452911193686095,
+      "learning_rate": 3.250725966583268e-08,
+      "loss": 0.7234,
+      "step": 41389
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4357948220064212,
+      "learning_rate": 3.244581349613807e-08,
+      "loss": 0.8535,
+      "step": 41390
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.88430885542271,
+      "learning_rate": 3.238442536069974e-08,
+      "loss": 0.813,
+      "step": 41391
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6453799004885825,
+      "learning_rate": 3.2323095259875205e-08,
+      "loss": 0.8706,
+      "step": 41392
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7596035412837042,
+      "learning_rate": 3.226182319402194e-08,
+      "loss": 0.7015,
+      "step": 41393
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4082739803882955,
+      "learning_rate": 3.220060916349521e-08,
+      "loss": 0.8833,
+      "step": 41394
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5375564090932918,
+      "learning_rate": 3.213945316865252e-08,
+      "loss": 0.7196,
+      "step": 41395
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6861657546870332,
+      "learning_rate": 3.2078355209849145e-08,
+      "loss": 0.7251,
+      "step": 41396
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.567174299466258,
+      "learning_rate": 3.201731528744145e-08,
+      "loss": 0.8559,
+      "step": 41397
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7234841199867164,
+      "learning_rate": 3.195633340178583e-08,
+      "loss": 0.7838,
+      "step": 41398
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6969015815760262,
+      "learning_rate": 3.189540955323534e-08,
+      "loss": 0.9189,
+      "step": 41399
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3893790385199756,
+      "learning_rate": 3.183454374214634e-08,
+      "loss": 0.7681,
+      "step": 41400
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5874870432840686,
+      "learning_rate": 3.17737359688719e-08,
+      "loss": 0.6911,
+      "step": 41401
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4760379630056593,
+      "learning_rate": 3.1712986233767286e-08,
+      "loss": 0.8347,
+      "step": 41402
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6462417808244285,
+      "learning_rate": 3.1652294537184436e-08,
+      "loss": 0.8991,
+      "step": 41403
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7756935979177608,
+      "learning_rate": 3.1591660879478624e-08,
+      "loss": 0.9008,
+      "step": 41404
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5889093364735678,
+      "learning_rate": 3.153108526100179e-08,
+      "loss": 0.8669,
+      "step": 41405
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.793864003147156,
+      "learning_rate": 3.147056768210699e-08,
+      "loss": 0.7283,
+      "step": 41406
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3757832391267306,
+      "learning_rate": 3.141010814314727e-08,
+      "loss": 0.7593,
+      "step": 41407
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.845228128235252,
+      "learning_rate": 3.1349706644473454e-08,
+      "loss": 0.7948,
+      "step": 41408
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5584738263869633,
+      "learning_rate": 3.1289363186437494e-08,
+      "loss": 0.7659,
+      "step": 41409
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3571865143177138,
+      "learning_rate": 3.122907776939022e-08,
+      "loss": 0.7327,
+      "step": 41410
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6657899822414008,
+      "learning_rate": 3.1168850393684666e-08,
+      "loss": 0.8469,
+      "step": 41411
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7770339770769504,
+      "learning_rate": 3.1108681059669466e-08,
+      "loss": 0.6524,
+      "step": 41412
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8416833168289963,
+      "learning_rate": 3.1048569767695435e-08,
+      "loss": 0.9867,
+      "step": 41413
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7378209079807376,
+      "learning_rate": 3.098851651811341e-08,
+      "loss": 0.9176,
+      "step": 41414
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.432107013457182,
+      "learning_rate": 3.0928521311271997e-08,
+      "loss": 0.8818,
+      "step": 41415
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.62012899009284,
+      "learning_rate": 3.086858414752092e-08,
+      "loss": 0.8323,
+      "step": 41416
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6194445333065046,
+      "learning_rate": 3.0808705027208786e-08,
+      "loss": 0.8884,
+      "step": 41417
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6175869202841076,
+      "learning_rate": 3.07488839506842e-08,
+      "loss": 0.7822,
+      "step": 41418
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7033565503381347,
+      "learning_rate": 3.06891209182969e-08,
+      "loss": 0.8518,
+      "step": 41419
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6288646885231717,
+      "learning_rate": 3.062941593039326e-08,
+      "loss": 0.8828,
+      "step": 41420
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7008624430199729,
+      "learning_rate": 3.056976898732189e-08,
+      "loss": 0.6814,
+      "step": 41421
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.996390927805562,
+      "learning_rate": 3.051018008942919e-08,
+      "loss": 0.8782,
+      "step": 41422
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5912092865331557,
+      "learning_rate": 3.045064923706265e-08,
+      "loss": 0.7386,
+      "step": 41423
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7760823088154425,
+      "learning_rate": 3.039117643056755e-08,
+      "loss": 0.7205,
+      "step": 41424
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8040497773233024,
+      "learning_rate": 3.0331761670292505e-08,
+      "loss": 0.7355,
+      "step": 41425
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6167654943361456,
+      "learning_rate": 3.027240495658279e-08,
+      "loss": 0.8755,
+      "step": 41426
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5593810793939047,
+      "learning_rate": 3.021310628978258e-08,
+      "loss": 0.8141,
+      "step": 41427
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7685707416140157,
+      "learning_rate": 3.015386567023826e-08,
+      "loss": 0.7444,
+      "step": 41428
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6437605227045877,
+      "learning_rate": 3.009468309829511e-08,
+      "loss": 0.8367,
+      "step": 41429
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8221714844425194,
+      "learning_rate": 3.003555857429619e-08,
+      "loss": 0.8558,
+      "step": 41430
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6742818691826398,
+      "learning_rate": 2.9976492098586775e-08,
+      "loss": 0.8944,
+      "step": 41431
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6960358086632605,
+      "learning_rate": 2.991748367151104e-08,
+      "loss": 0.8726,
+      "step": 41432
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6231136858232769,
+      "learning_rate": 2.985853329341315e-08,
+      "loss": 0.7585,
+      "step": 41433
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7011952856580717,
+      "learning_rate": 2.9799640964633946e-08,
+      "loss": 0.8752,
+      "step": 41434
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8980255807771187,
+      "learning_rate": 2.9740806685517597e-08,
+      "loss": 0.8571,
+      "step": 41435
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.598248866876645,
+      "learning_rate": 2.9682030456408272e-08,
+      "loss": 0.9744,
+      "step": 41436
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7514838317335197,
+      "learning_rate": 2.96233122776457e-08,
+      "loss": 0.7129,
+      "step": 41437
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5623315109925064,
+      "learning_rate": 2.9564652149572938e-08,
+      "loss": 0.8162,
+      "step": 41438
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5270331032844549,
+      "learning_rate": 2.950605007253082e-08,
+      "loss": 0.8178,
+      "step": 41439
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4734225507458345,
+      "learning_rate": 2.9447506046862417e-08,
+      "loss": 0.8392,
+      "step": 41440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.448553886470413,
+      "learning_rate": 2.938902007290634e-08,
+      "loss": 0.7805,
+      "step": 41441
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4608276765652197,
+      "learning_rate": 2.9330592151003424e-08,
+      "loss": 0.9046,
+      "step": 41442
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.529557155982922,
+      "learning_rate": 2.927222228149562e-08,
+      "loss": 0.8079,
+      "step": 41443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.971656402559501,
+      "learning_rate": 2.921391046472155e-08,
+      "loss": 0.8731,
+      "step": 41444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6418647695873678,
+      "learning_rate": 2.915565670101983e-08,
+      "loss": 0.7596,
+      "step": 41445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.948020222455546,
+      "learning_rate": 2.9097460990731296e-08,
+      "loss": 0.6343,
+      "step": 41446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7154007725897986,
+      "learning_rate": 2.9039323334193458e-08,
+      "loss": 0.7831,
+      "step": 41447
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7485639616509148,
+      "learning_rate": 2.8981243731746044e-08,
+      "loss": 0.8231,
+      "step": 41448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5278021881568054,
+      "learning_rate": 2.8923222183726563e-08,
+      "loss": 0.9455,
+      "step": 41449
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7890051770637836,
+      "learning_rate": 2.886525869047363e-08,
+      "loss": 0.7299,
+      "step": 41450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6567023147353424,
+      "learning_rate": 2.8807353252322533e-08,
+      "loss": 0.8698,
+      "step": 41451
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9963731366795425,
+      "learning_rate": 2.8749505869613003e-08,
+      "loss": 0.898,
+      "step": 41452
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.698549055042238,
+      "learning_rate": 2.8691716542679215e-08,
+      "loss": 0.6855,
+      "step": 41453
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6532580053726156,
+      "learning_rate": 2.8633985271860898e-08,
+      "loss": 0.7752,
+      "step": 41454
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8884425549366348,
+      "learning_rate": 2.8576312057491118e-08,
+      "loss": 0.9029,
+      "step": 41455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5144003656538632,
+      "learning_rate": 2.8518696899907382e-08,
+      "loss": 0.8131,
+      "step": 41456
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6126668609994221,
+      "learning_rate": 2.8461139799444982e-08,
+      "loss": 0.9713,
+      "step": 41457
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.027684304920911,
+      "learning_rate": 2.840364075643809e-08,
+      "loss": 0.8418,
+      "step": 41458
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.870135452250413,
+      "learning_rate": 2.8346199771221995e-08,
+      "loss": 0.909,
+      "step": 41459
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7619058211942371,
+      "learning_rate": 2.8288816844131982e-08,
+      "loss": 0.687,
+      "step": 41460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6998192355076795,
+      "learning_rate": 2.8231491975502233e-08,
+      "loss": 0.7457,
+      "step": 41461
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6532373800449593,
+      "learning_rate": 2.8174225165664703e-08,
+      "loss": 0.8162,
+      "step": 41462
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7276444999495497,
+      "learning_rate": 2.8117016414953567e-08,
+      "loss": 0.8239,
+      "step": 41463
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4972857732877982,
+      "learning_rate": 2.8059865723703006e-08,
+      "loss": 0.7961,
+      "step": 41464
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6535672538281059,
+      "learning_rate": 2.800277309224497e-08,
+      "loss": 0.7905,
+      "step": 41465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7026336719533814,
+      "learning_rate": 2.7945738520911426e-08,
+      "loss": 0.7798,
+      "step": 41466
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.433390026459964,
+      "learning_rate": 2.788876201003432e-08,
+      "loss": 0.8841,
+      "step": 41467
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1379811607637667,
+      "learning_rate": 2.7831843559946724e-08,
+      "loss": 0.9307,
+      "step": 41468
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5172821984004754,
+      "learning_rate": 2.7774983170979487e-08,
+      "loss": 0.7755,
+      "step": 41469
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5914263841716756,
+      "learning_rate": 2.771818084346345e-08,
+      "loss": 0.7407,
+      "step": 41470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8056945365761727,
+      "learning_rate": 2.766143657772835e-08,
+      "loss": 0.8098,
+      "step": 41471
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0878998521717116,
+      "learning_rate": 2.7604750374106148e-08,
+      "loss": 0.9017,
+      "step": 41472
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4746924751515833,
+      "learning_rate": 2.7548122232926578e-08,
+      "loss": 0.7895,
+      "step": 41473
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6365338997612306,
+      "learning_rate": 2.7491552154519375e-08,
+      "loss": 0.8647,
+      "step": 41474
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6239783292754908,
+      "learning_rate": 2.7435040139213164e-08,
+      "loss": 0.7302,
+      "step": 41475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.715775644089216,
+      "learning_rate": 2.7378586187336576e-08,
+      "loss": 0.9509,
+      "step": 41476
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5129157294247317,
+      "learning_rate": 2.732219029921934e-08,
+      "loss": 0.8247,
+      "step": 41477
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4558913198732433,
+      "learning_rate": 2.72658524751912e-08,
+      "loss": 0.9382,
+      "step": 41478
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6418349370284364,
+      "learning_rate": 2.7209572715577447e-08,
+      "loss": 0.7552,
+      "step": 41479
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7557069979318747,
+      "learning_rate": 2.7153351020706707e-08,
+      "loss": 0.6801,
+      "step": 41480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5177496972821118,
+      "learning_rate": 2.7097187390906497e-08,
+      "loss": 0.7446,
+      "step": 41481
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0595734723492174,
+      "learning_rate": 2.7041081826504333e-08,
+      "loss": 0.8041,
+      "step": 41482
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7491115072522347,
+      "learning_rate": 2.698503432782662e-08,
+      "loss": 0.7934,
+      "step": 41483
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.593038874005151,
+      "learning_rate": 2.6929044895198652e-08,
+      "loss": 0.8893,
+      "step": 41484
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7389211161596798,
+      "learning_rate": 2.6873113528947946e-08,
+      "loss": 0.7772,
+      "step": 41485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6685292758530925,
+      "learning_rate": 2.681724022939869e-08,
+      "loss": 0.879,
+      "step": 41486
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8091735495636192,
+      "learning_rate": 2.6761424996877284e-08,
+      "loss": 0.7241,
+      "step": 41487
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5845966762268895,
+      "learning_rate": 2.6705667831707915e-08,
+      "loss": 0.8393,
+      "step": 41488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5738205298638812,
+      "learning_rate": 2.664996873421588e-08,
+      "loss": 1.0172,
+      "step": 41489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6646538565198759,
+      "learning_rate": 2.6594327704725365e-08,
+      "loss": 0.8809,
+      "step": 41490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7957573346180566,
+      "learning_rate": 2.6538744743560552e-08,
+      "loss": 0.7392,
+      "step": 41491
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.57382112852053,
+      "learning_rate": 2.648321985104452e-08,
+      "loss": 0.8495,
+      "step": 41492
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6352673592780076,
+      "learning_rate": 2.6427753027500335e-08,
+      "loss": 0.8414,
+      "step": 41493
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0961644858677184,
+      "learning_rate": 2.6372344273251082e-08,
+      "loss": 0.8341,
+      "step": 41494
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7324951721907269,
+      "learning_rate": 2.6316993588620944e-08,
+      "loss": 0.6937,
+      "step": 41495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.149465380024187,
+      "learning_rate": 2.626170097392966e-08,
+      "loss": 0.8192,
+      "step": 41496
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5907418255349421,
+      "learning_rate": 2.620646642950142e-08,
+      "loss": 0.7027,
+      "step": 41497
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6619278539208524,
+      "learning_rate": 2.6151289955655966e-08,
+      "loss": 0.845,
+      "step": 41498
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.508895554418533,
+      "learning_rate": 2.6096171552715265e-08,
+      "loss": 0.9338,
+      "step": 41499
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5831043801101858,
+      "learning_rate": 2.604111122100128e-08,
+      "loss": 0.8176,
+      "step": 41500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7305715905553284,
+      "learning_rate": 2.5986108960832644e-08,
+      "loss": 0.6557,
+      "step": 41501
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1490257250533262,
+      "learning_rate": 2.5931164772530214e-08,
+      "loss": 0.8799,
+      "step": 41502
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5897046398900807,
+      "learning_rate": 2.5876278656415954e-08,
+      "loss": 0.9291,
+      "step": 41503
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.676585601567934,
+      "learning_rate": 2.5821450612806276e-08,
+      "loss": 0.7648,
+      "step": 41504
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8040433153639694,
+      "learning_rate": 2.5766680642023147e-08,
+      "loss": 0.8926,
+      "step": 41505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5048678041032466,
+      "learning_rate": 2.5711968744382975e-08,
+      "loss": 0.8402,
+      "step": 41506
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.530341995246481,
+      "learning_rate": 2.565731492020662e-08,
+      "loss": 0.7872,
+      "step": 41507
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4148401997186038,
+      "learning_rate": 2.5602719169810496e-08,
+      "loss": 0.9555,
+      "step": 41508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5967484199643445,
+      "learning_rate": 2.5548181493513235e-08,
+      "loss": 0.8525,
+      "step": 41509
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3996198618939077,
+      "learning_rate": 2.5493701891632362e-08,
+      "loss": 0.8877,
+      "step": 41510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8937328315013575,
+      "learning_rate": 2.5439280364485398e-08,
+      "loss": 0.8757,
+      "step": 41511
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5420514208279004,
+      "learning_rate": 2.538491691238765e-08,
+      "loss": 0.9306,
+      "step": 41512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5239903047289425,
+      "learning_rate": 2.5330611535658857e-08,
+      "loss": 0.886,
+      "step": 41513
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.13926457445258,
+      "learning_rate": 2.5276364234610995e-08,
+      "loss": 0.833,
+      "step": 41514
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5414072133631689,
+      "learning_rate": 2.5222175009563808e-08,
+      "loss": 0.8173,
+      "step": 41515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.52341969071579,
+      "learning_rate": 2.516804386083038e-08,
+      "loss": 0.7074,
+      "step": 41516
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.515318094720344,
+      "learning_rate": 2.5113970788726018e-08,
+      "loss": 0.9679,
+      "step": 41517
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.25750877419671,
+      "learning_rate": 2.5059955793567125e-08,
+      "loss": 0.7611,
+      "step": 41518
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4626785462882732,
+      "learning_rate": 2.5005998875665683e-08,
+      "loss": 0.8174,
+      "step": 41519
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.555190307685583,
+      "learning_rate": 2.495210003533921e-08,
+      "loss": 0.89,
+      "step": 41520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.834029910542488,
+      "learning_rate": 2.4898259272898573e-08,
+      "loss": 0.8273,
+      "step": 41521
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5528666924983492,
+      "learning_rate": 2.4844476588659073e-08,
+      "loss": 0.78,
+      "step": 41522
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7665452438291267,
+      "learning_rate": 2.4790751982931572e-08,
+      "loss": 1.0164,
+      "step": 41523
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7592089813534317,
+      "learning_rate": 2.4737085456032483e-08,
+      "loss": 0.7488,
+      "step": 41524
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3923495668211194,
+      "learning_rate": 2.4683477008270453e-08,
+      "loss": 0.9797,
+      "step": 41525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9778986585358482,
+      "learning_rate": 2.4629926639960777e-08,
+      "loss": 0.8671,
+      "step": 41526
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.940806749534156,
+      "learning_rate": 2.4576434351412103e-08,
+      "loss": 0.7924,
+      "step": 41527
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9590978211041872,
+      "learning_rate": 2.4523000142939735e-08,
+      "loss": 0.7983,
+      "step": 41528
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5785513737081007,
+      "learning_rate": 2.44696240148512e-08,
+      "loss": 0.7909,
+      "step": 41529
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5956614084989096,
+      "learning_rate": 2.4416305967458476e-08,
+      "loss": 0.8358,
+      "step": 41530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6261665673008083,
+      "learning_rate": 2.436304600107353e-08,
+      "loss": 0.7312,
+      "step": 41531
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8641345462847694,
+      "learning_rate": 2.43098441160039e-08,
+      "loss": 0.8888,
+      "step": 41532
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9980548404923333,
+      "learning_rate": 2.4256700312561554e-08,
+      "loss": 0.8419,
+      "step": 41533
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.766783671679107,
+      "learning_rate": 2.4203614591054024e-08,
+      "loss": 0.8503,
+      "step": 41534
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5680555966733962,
+      "learning_rate": 2.4150586951792175e-08,
+      "loss": 0.7476,
+      "step": 41535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5971077805548675,
+      "learning_rate": 2.409761739508354e-08,
+      "loss": 0.8094,
+      "step": 41536
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5737644252958856,
+      "learning_rate": 2.4044705921236756e-08,
+      "loss": 0.9191,
+      "step": 41537
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4724369821604515,
+      "learning_rate": 2.3991852530560467e-08,
+      "loss": 0.8375,
+      "step": 41538
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.647459411221138,
+      "learning_rate": 2.3939057223362206e-08,
+      "loss": 0.7645,
+      "step": 41539
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6021024123556764,
+      "learning_rate": 2.3886319999948393e-08,
+      "loss": 0.7661,
+      "step": 41540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.184281368793094,
+      "learning_rate": 2.383364086062656e-08,
+      "loss": 0.7898,
+      "step": 41541
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5776234780880238,
+      "learning_rate": 2.3781019805704243e-08,
+      "loss": 0.7975,
+      "step": 41542
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6037125737675286,
+      "learning_rate": 2.3728456835487857e-08,
+      "loss": 0.8678,
+      "step": 41543
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4059130201557615,
+      "learning_rate": 2.3675951950281605e-08,
+      "loss": 0.7436,
+      "step": 41544
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7963090266683253,
+      "learning_rate": 2.3623505150393024e-08,
+      "loss": 0.8077,
+      "step": 41545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4517920552965622,
+      "learning_rate": 2.3571116436127418e-08,
+      "loss": 0.6448,
+      "step": 41546
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.790200463514226,
+      "learning_rate": 2.351878580778788e-08,
+      "loss": 0.7309,
+      "step": 41547
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5900794937698974,
+      "learning_rate": 2.3466513265680835e-08,
+      "loss": 0.8082,
+      "step": 41548
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7589587931311634,
+      "learning_rate": 2.341429881011159e-08,
+      "loss": 0.7497,
+      "step": 41549
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9091295051869919,
+      "learning_rate": 2.336214244138102e-08,
+      "loss": 0.7576,
+      "step": 41550
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4561785233160978,
+      "learning_rate": 2.331004415979554e-08,
+      "loss": 0.8563,
+      "step": 41551
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4640987981950593,
+      "learning_rate": 2.325800396565825e-08,
+      "loss": 0.8739,
+      "step": 41552
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5431510507127144,
+      "learning_rate": 2.3206021859270012e-08,
+      "loss": 0.7471,
+      "step": 41553
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9716753172406807,
+      "learning_rate": 2.315409784093503e-08,
+      "loss": 0.8042,
+      "step": 41554
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5459888925267784,
+      "learning_rate": 2.3102231910956398e-08,
+      "loss": 0.8551,
+      "step": 41555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.05071265037788,
+      "learning_rate": 2.305042406963498e-08,
+      "loss": 0.8025,
+      "step": 41556
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9201532472293865,
+      "learning_rate": 2.2998674317272762e-08,
+      "loss": 0.6508,
+      "step": 41557
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5470114381586957,
+      "learning_rate": 2.2946982654170614e-08,
+      "loss": 0.92,
+      "step": 41558
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6255459505766143,
+      "learning_rate": 2.2895349080630512e-08,
+      "loss": 0.808,
+      "step": 41559
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4919530718085732,
+      "learning_rate": 2.2843773596951115e-08,
+      "loss": 0.9498,
+      "step": 41560
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.421930238084233,
+      "learning_rate": 2.2792256203434393e-08,
+      "loss": 0.7159,
+      "step": 41561
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.697118031429135,
+      "learning_rate": 2.2740796900380114e-08,
+      "loss": 1.0203,
+      "step": 41562
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5916790688617986,
+      "learning_rate": 2.2689395688088034e-08,
+      "loss": 0.8929,
+      "step": 41563
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6644133279283293,
+      "learning_rate": 2.2638052566856806e-08,
+      "loss": 0.7364,
+      "step": 41564
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9979332044271856,
+      "learning_rate": 2.2586767536985078e-08,
+      "loss": 0.747,
+      "step": 41565
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8073292737018383,
+      "learning_rate": 2.253554059877261e-08,
+      "loss": 0.7706,
+      "step": 41566
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6139176604433045,
+      "learning_rate": 2.2484371752516942e-08,
+      "loss": 0.8726,
+      "step": 41567
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6126506987656493,
+      "learning_rate": 2.2433260998515615e-08,
+      "loss": 0.8199,
+      "step": 41568
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9446026981843711,
+      "learning_rate": 2.238220833706728e-08,
+      "loss": 0.798,
+      "step": 41569
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9395086460069353,
+      "learning_rate": 2.2331213768468363e-08,
+      "loss": 0.9446,
+      "step": 41570
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.751952329340938,
+      "learning_rate": 2.2280277293016404e-08,
+      "loss": 0.8718,
+      "step": 41571
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7044454908618554,
+      "learning_rate": 2.2229398911006726e-08,
+      "loss": 0.7049,
+      "step": 41572
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5980767344728144,
+      "learning_rate": 2.2178578622736867e-08,
+      "loss": 0.7211,
+      "step": 41573
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7154680921105805,
+      "learning_rate": 2.2127816428502148e-08,
+      "loss": 0.7724,
+      "step": 41574
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6437000586058292,
+      "learning_rate": 2.207711232859788e-08,
+      "loss": 0.9515,
+      "step": 41575
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5752460496101672,
+      "learning_rate": 2.2026466323320504e-08,
+      "loss": 0.8462,
+      "step": 41576
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7862404512860085,
+      "learning_rate": 2.1975878412964225e-08,
+      "loss": 0.7625,
+      "step": 41577
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7914629259878305,
+      "learning_rate": 2.1925348597822137e-08,
+      "loss": 0.6774,
+      "step": 41578
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.856286129794031,
+      "learning_rate": 2.1874876878189564e-08,
+      "loss": 0.7696,
+      "step": 41579
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0163073979823762,
+      "learning_rate": 2.1824463254361826e-08,
+      "loss": 0.796,
+      "step": 41580
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.571757892387998,
+      "learning_rate": 2.177410772663091e-08,
+      "loss": 0.8199,
+      "step": 41581
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5466079630056433,
+      "learning_rate": 2.1723810295288807e-08,
+      "loss": 0.7618,
+      "step": 41582
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7822253410126145,
+      "learning_rate": 2.167357096063083e-08,
+      "loss": 0.7994,
+      "step": 41583
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4627362892277884,
+      "learning_rate": 2.1623389722947864e-08,
+      "loss": 0.9384,
+      "step": 41584
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8247188860282564,
+      "learning_rate": 2.1573266582533002e-08,
+      "loss": 0.783,
+      "step": 41585
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5520630952034797,
+      "learning_rate": 2.152320153967824e-08,
+      "loss": 0.8597,
+      "step": 41586
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.721370607678776,
+      "learning_rate": 2.147319459467334e-08,
+      "loss": 0.8733,
+      "step": 41587
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6550366733795738,
+      "learning_rate": 2.1423245747811407e-08,
+      "loss": 0.7337,
+      "step": 41588
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6731379489513711,
+      "learning_rate": 2.13733549993822e-08,
+      "loss": 0.6767,
+      "step": 41589
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.803440076028565,
+      "learning_rate": 2.1323522349676607e-08,
+      "loss": 0.7478,
+      "step": 41590
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4824089787447476,
+      "learning_rate": 2.12737477989855e-08,
+      "loss": 0.8298,
+      "step": 41591
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.422735468268066,
+      "learning_rate": 2.122403134759643e-08,
+      "loss": 0.8461,
+      "step": 41592
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6791224110790095,
+      "learning_rate": 2.1174372995801385e-08,
+      "loss": 0.9112,
+      "step": 41593
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4869255280735352,
+      "learning_rate": 2.112477274388791e-08,
+      "loss": 0.9095,
+      "step": 41594
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6885887567425333,
+      "learning_rate": 2.107523059214467e-08,
+      "loss": 0.8482,
+      "step": 41595
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0384610855095775,
+      "learning_rate": 2.1025746540861425e-08,
+      "loss": 0.8497,
+      "step": 41596
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.566651163023725,
+      "learning_rate": 2.0976320590325727e-08,
+      "loss": 0.8075,
+      "step": 41597
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5121214932786307,
+      "learning_rate": 2.0926952740825124e-08,
+      "loss": 0.8364,
+      "step": 41598
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7873293056388411,
+      "learning_rate": 2.0877642992647164e-08,
+      "loss": 0.7271,
+      "step": 41599
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.506127937466413,
+      "learning_rate": 2.082839134607828e-08,
+      "loss": 0.6735,
+      "step": 41600
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.75282029695164,
+      "learning_rate": 2.0779197801406026e-08,
+      "loss": 0.7107,
+      "step": 41601
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6765215872871735,
+      "learning_rate": 2.0730062358916835e-08,
+      "loss": 0.8252,
+      "step": 41602
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6546614549812433,
+      "learning_rate": 2.068098501889715e-08,
+      "loss": 0.7761,
+      "step": 41603
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7963209179053563,
+      "learning_rate": 2.063196578163118e-08,
+      "loss": 0.9267,
+      "step": 41604
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7171583905409729,
+      "learning_rate": 2.058300464740648e-08,
+      "loss": 0.9415,
+      "step": 41605
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.675194445681317,
+      "learning_rate": 2.0534101616506154e-08,
+      "loss": 0.8558,
+      "step": 41606
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.407146723474149,
+      "learning_rate": 2.0485256689215525e-08,
+      "loss": 0.7682,
+      "step": 41607
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2221761124243113,
+      "learning_rate": 2.043646986581993e-08,
+      "loss": 0.8893,
+      "step": 41608
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4252604130930744,
+      "learning_rate": 2.038774114660136e-08,
+      "loss": 0.7824,
+      "step": 41609
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7510509975784518,
+      "learning_rate": 2.033907053184625e-08,
+      "loss": 0.8285,
+      "step": 41610
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8805379206944044,
+      "learning_rate": 2.029045802183549e-08,
+      "loss": 0.8247,
+      "step": 41611
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.495209509129837,
+      "learning_rate": 2.024190361685441e-08,
+      "loss": 0.7718,
+      "step": 41612
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6046086115355054,
+      "learning_rate": 2.019340731718389e-08,
+      "loss": 0.9085,
+      "step": 41613
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1474507891008376,
+      "learning_rate": 2.014496912310704e-08,
+      "loss": 0.7246,
+      "step": 41614
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.538370016672645,
+      "learning_rate": 2.0096589034905856e-08,
+      "loss": 0.9425,
+      "step": 41615
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5889662619296294,
+      "learning_rate": 2.0048267052861226e-08,
+      "loss": 0.8483,
+      "step": 41616
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5726582080720621,
+      "learning_rate": 2.0000003177255145e-08,
+      "loss": 0.7244,
+      "step": 41617
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7003204379361352,
+      "learning_rate": 1.9951797408369612e-08,
+      "loss": 0.8648,
+      "step": 41618
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.568358943126062,
+      "learning_rate": 1.99036497464844e-08,
+      "loss": 0.7965,
+      "step": 41619
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4640899570451722,
+      "learning_rate": 1.9855560191879285e-08,
+      "loss": 0.7173,
+      "step": 41620
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.456175398893764,
+      "learning_rate": 1.9807528744834048e-08,
+      "loss": 0.8145,
+      "step": 41621
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6306037068548533,
+      "learning_rate": 1.975955540563068e-08,
+      "loss": 0.8649,
+      "step": 41622
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7739147967317159,
+      "learning_rate": 1.9711640174545633e-08,
+      "loss": 0.6583,
+      "step": 41623
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8338079208777769,
+      "learning_rate": 1.9663783051858676e-08,
+      "loss": 0.7618,
+      "step": 41624
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.396585540487495,
+      "learning_rate": 1.9615984037849588e-08,
+      "loss": 0.7487,
+      "step": 41625
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7102774135239016,
+      "learning_rate": 1.9568243132797036e-08,
+      "loss": 0.739,
+      "step": 41626
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4356112816855637,
+      "learning_rate": 1.9520560336976357e-08,
+      "loss": 0.6955,
+      "step": 41627
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8149693063331562,
+      "learning_rate": 1.9472935650667325e-08,
+      "loss": 0.8407,
+      "step": 41628
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.626268023352364,
+      "learning_rate": 1.9425369074146384e-08,
+      "loss": 0.6508,
+      "step": 41629
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.373255206998429,
+      "learning_rate": 1.937786060769109e-08,
+      "loss": 0.9087,
+      "step": 41630
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5025825005556646,
+      "learning_rate": 1.933041025157789e-08,
+      "loss": 0.8203,
+      "step": 41631
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5389036406955598,
+      "learning_rate": 1.928301800608212e-08,
+      "loss": 0.7732,
+      "step": 41632
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7329834669319154,
+      "learning_rate": 1.9235683871481338e-08,
+      "loss": 0.7183,
+      "step": 41633
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3602793836177824,
+      "learning_rate": 1.918840784804976e-08,
+      "loss": 0.7139,
+      "step": 41634
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7051001739844676,
+      "learning_rate": 1.9141189936063842e-08,
+      "loss": 0.768,
+      "step": 41635
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6865022581835347,
+      "learning_rate": 1.9094030135796694e-08,
+      "loss": 0.782,
+      "step": 41636
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5087808786636296,
+      "learning_rate": 1.904692844752476e-08,
+      "loss": 0.8791,
+      "step": 41637
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.86286015706552,
+      "learning_rate": 1.8999884871521157e-08,
+      "loss": 0.8846,
+      "step": 41638
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6121427097885117,
+      "learning_rate": 1.895289940806011e-08,
+      "loss": 0.8946,
+      "step": 41639
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6095204474957878,
+      "learning_rate": 1.8905972057415845e-08,
+      "loss": 1.0084,
+      "step": 41640
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.60587746079308,
+      "learning_rate": 1.8859102819860365e-08,
+      "loss": 0.8149,
+      "step": 41641
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.709466898612084,
+      "learning_rate": 1.8812291695666783e-08,
+      "loss": 0.7466,
+      "step": 41642
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6882110458159894,
+      "learning_rate": 1.876553868510822e-08,
+      "loss": 0.8238,
+      "step": 41643
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6773894648514305,
+      "learning_rate": 1.8718843788457787e-08,
+      "loss": 1.0024,
+      "step": 41644
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4713949693088466,
+      "learning_rate": 1.8672207005985267e-08,
+      "loss": 0.7533,
+      "step": 41645
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.609471428275949,
+      "learning_rate": 1.8625628337963775e-08,
+      "loss": 0.7619,
+      "step": 41646
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6347042173719835,
+      "learning_rate": 1.8579107784663097e-08,
+      "loss": 0.8872,
+      "step": 41647
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.435643905254322,
+      "learning_rate": 1.853264534635635e-08,
+      "loss": 0.8003,
+      "step": 41648
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.670454216568774,
+      "learning_rate": 1.8486241023311093e-08,
+      "loss": 0.8345,
+      "step": 41649
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7467465257098801,
+      "learning_rate": 1.843989481579933e-08,
+      "loss": 0.7067,
+      "step": 41650
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6200532846353926,
+      "learning_rate": 1.8393606724090852e-08,
+      "loss": 0.9029,
+      "step": 41651
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5535819382526892,
+      "learning_rate": 1.8347376748455438e-08,
+      "loss": 0.811,
+      "step": 41652
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5984923443786052,
+      "learning_rate": 1.8301204889160652e-08,
+      "loss": 0.7481,
+      "step": 41653
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6203831711537326,
+      "learning_rate": 1.8255091146477387e-08,
+      "loss": 0.8323,
+      "step": 41654
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.615772112883891,
+      "learning_rate": 1.8209035520673213e-08,
+      "loss": 0.7635,
+      "step": 41655
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5452202796863022,
+      "learning_rate": 1.816303801201569e-08,
+      "loss": 0.8992,
+      "step": 41656
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6817871837005696,
+      "learning_rate": 1.8117098620773487e-08,
+      "loss": 0.8529,
+      "step": 41657
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4698666646390075,
+      "learning_rate": 1.8071217347213066e-08,
+      "loss": 0.8359,
+      "step": 41658
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6154171901001715,
+      "learning_rate": 1.8025394191603095e-08,
+      "loss": 0.7962,
+      "step": 41659
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6488723017258045,
+      "learning_rate": 1.7979629154208923e-08,
+      "loss": 0.776,
+      "step": 41660
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6546661093742978,
+      "learning_rate": 1.7933922235298108e-08,
+      "loss": 0.7989,
+      "step": 41661
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4757051259603642,
+      "learning_rate": 1.7888273435135994e-08,
+      "loss": 0.8019,
+      "step": 41662
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8424483730300292,
+      "learning_rate": 1.7842682753987927e-08,
+      "loss": 0.8243,
+      "step": 41663
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5089253364828326,
+      "learning_rate": 1.779715019212036e-08,
+      "loss": 0.8267,
+      "step": 41664
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4946208058769879,
+      "learning_rate": 1.7751675749798635e-08,
+      "loss": 0.8153,
+      "step": 41665
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5408742922800043,
+      "learning_rate": 1.7706259427286988e-08,
+      "loss": 0.7063,
+      "step": 41666
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.86103323652468,
+      "learning_rate": 1.7660901224849647e-08,
+      "loss": 0.8512,
+      "step": 41667
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6269648688257135,
+      "learning_rate": 1.761560114275085e-08,
+      "loss": 0.7083,
+      "step": 41668
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5459975290981558,
+      "learning_rate": 1.7570359181254827e-08,
+      "loss": 0.826,
+      "step": 41669
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5765080259324635,
+      "learning_rate": 1.752517534062359e-08,
+      "loss": 0.7401,
+      "step": 41670
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4109315183724351,
+      "learning_rate": 1.7480049621122485e-08,
+      "loss": 0.8286,
+      "step": 41671
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7418361504103279,
+      "learning_rate": 1.7434982023012414e-08,
+      "loss": 0.6848,
+      "step": 41672
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5295232200576785,
+      "learning_rate": 1.738997254655539e-08,
+      "loss": 0.7052,
+      "step": 41673
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6721008266777773,
+      "learning_rate": 1.7345021192015643e-08,
+      "loss": 0.7544,
+      "step": 41674
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5375664358378014,
+      "learning_rate": 1.730012795965297e-08,
+      "loss": 0.8305,
+      "step": 41675
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6309020754940386,
+      "learning_rate": 1.7255292849729376e-08,
+      "loss": 0.8411,
+      "step": 41676
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8321552186356445,
+      "learning_rate": 1.7210515862505773e-08,
+      "loss": 0.7615,
+      "step": 41677
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5837907064935983,
+      "learning_rate": 1.7165796998244168e-08,
+      "loss": 0.8598,
+      "step": 41678
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6404784341466887,
+      "learning_rate": 1.712113625720324e-08,
+      "loss": 0.8109,
+      "step": 41679
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.732111173094501,
+      "learning_rate": 1.7076533639642788e-08,
+      "loss": 0.692,
+      "step": 41680
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7532455177414994,
+      "learning_rate": 1.7031989145823713e-08,
+      "loss": 0.9706,
+      "step": 41681
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7972877542628162,
+      "learning_rate": 1.6987502776005803e-08,
+      "loss": 0.707,
+      "step": 41682
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4218884088940487,
+      "learning_rate": 1.6943074530446636e-08,
+      "loss": 0.9039,
+      "step": 41683
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7423930206631586,
+      "learning_rate": 1.6898704409404888e-08,
+      "loss": 0.8629,
+      "step": 41684
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6300718158875613,
+      "learning_rate": 1.6854392413139244e-08,
+      "loss": 0.8816,
+      "step": 41685
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.695160402621397,
+      "learning_rate": 1.6810138541909492e-08,
+      "loss": 0.8184,
+      "step": 41686
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.476744184909176,
+      "learning_rate": 1.676594279597099e-08,
+      "loss": 0.8851,
+      "step": 41687
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7414076993502445,
+      "learning_rate": 1.6721805175581307e-08,
+      "loss": 0.9943,
+      "step": 41688
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.569343680502579,
+      "learning_rate": 1.6677725680999124e-08,
+      "loss": 0.9206,
+      "step": 41689
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9489446576727234,
+      "learning_rate": 1.6633704312478683e-08,
+      "loss": 0.8127,
+      "step": 41690
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7547885329989803,
+      "learning_rate": 1.6589741070277555e-08,
+      "loss": 0.7637,
+      "step": 41691
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6234152607817525,
+      "learning_rate": 1.65458359546522e-08,
+      "loss": 0.7469,
+      "step": 41692
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6015363025448914,
+      "learning_rate": 1.6501988965857972e-08,
+      "loss": 0.7896,
+      "step": 41693
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6630315771230413,
+      "learning_rate": 1.6458200104149115e-08,
+      "loss": 0.8567,
+      "step": 41694
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7377454058367534,
+      "learning_rate": 1.641446936978208e-08,
+      "loss": 0.8614,
+      "step": 41695
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.961305493147844,
+      "learning_rate": 1.637079676301001e-08,
+      "loss": 0.8616,
+      "step": 41696
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5357524882396487,
+      "learning_rate": 1.632718228408825e-08,
+      "loss": 0.9109,
+      "step": 41697
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6088935588704292,
+      "learning_rate": 1.6283625933269932e-08,
+      "loss": 0.8532,
+      "step": 41698
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8007025938765835,
+      "learning_rate": 1.6240127710810405e-08,
+      "loss": 0.8418,
+      "step": 41699
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9085227830099414,
+      "learning_rate": 1.6196687616960583e-08,
+      "loss": 0.796,
+      "step": 41700
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5631366047000694,
+      "learning_rate": 1.6153305651974703e-08,
+      "loss": 0.9129,
+      "step": 41701
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5885721231369538,
+      "learning_rate": 1.610998181610479e-08,
+      "loss": 0.8052,
+      "step": 41702
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6722603757028665,
+      "learning_rate": 1.6066716109603975e-08,
+      "loss": 0.6413,
+      "step": 41703
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5961126388014684,
+      "learning_rate": 1.6023508532724273e-08,
+      "loss": 0.7841,
+      "step": 41704
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6968358423248309,
+      "learning_rate": 1.5980359085715493e-08,
+      "loss": 0.8922,
+      "step": 41705
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.496873002116147,
+      "learning_rate": 1.5937267768829646e-08,
+      "loss": 0.8998,
+      "step": 41706
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5807709463969817,
+      "learning_rate": 1.5894234582318758e-08,
+      "loss": 0.821,
+      "step": 41707
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6896610870699178,
+      "learning_rate": 1.585125952643263e-08,
+      "loss": 0.7799,
+      "step": 41708
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7671147262937487,
+      "learning_rate": 1.5808342601421055e-08,
+      "loss": 0.6753,
+      "step": 41709
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6153114538226607,
+      "learning_rate": 1.5765483807533843e-08,
+      "loss": 0.8771,
+      "step": 41710
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7448091510937809,
+      "learning_rate": 1.5722683145020788e-08,
+      "loss": 0.7053,
+      "step": 41711
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5624295717996655,
+      "learning_rate": 1.5679940614131696e-08,
+      "loss": 0.865,
+      "step": 41712
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.861718514202844,
+      "learning_rate": 1.563725621511525e-08,
+      "loss": 0.7949,
+      "step": 41713
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6734683535604529,
+      "learning_rate": 1.559462994822014e-08,
+      "loss": 0.9426,
+      "step": 41714
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.546136486919246,
+      "learning_rate": 1.5552061813692843e-08,
+      "loss": 0.813,
+      "step": 41715
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6764279803395001,
+      "learning_rate": 1.5509551811783152e-08,
+      "loss": 0.8001,
+      "step": 41716
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6510802488281215,
+      "learning_rate": 1.546709994273865e-08,
+      "loss": 0.8714,
+      "step": 41717
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7429246000120062,
+      "learning_rate": 1.5424706206804695e-08,
+      "loss": 0.7571,
+      "step": 41718
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.736388614866179,
+      "learning_rate": 1.538237060422887e-08,
+      "loss": 0.6933,
+      "step": 41719
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4496112269380983,
+      "learning_rate": 1.5340093135258748e-08,
+      "loss": 0.7722,
+      "step": 41720
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6427880593193964,
+      "learning_rate": 1.5297873800139696e-08,
+      "loss": 0.8124,
+      "step": 41721
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4688851522676127,
+      "learning_rate": 1.5255712599117066e-08,
+      "loss": 0.7112,
+      "step": 41722
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9590160805071655,
+      "learning_rate": 1.5213609532436223e-08,
+      "loss": 0.928,
+      "step": 41723
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.940132938908784,
+      "learning_rate": 1.5171564600343636e-08,
+      "loss": 0.7902,
+      "step": 41724
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.76446413108869,
+      "learning_rate": 1.512957780308355e-08,
+      "loss": 0.8524,
+      "step": 41725
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4289134219670994,
+      "learning_rate": 1.5087649140900218e-08,
+      "loss": 0.7132,
+      "step": 41726
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7810961061847428,
+      "learning_rate": 1.5045778614036775e-08,
+      "loss": 0.749,
+      "step": 41727
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6971293502448606,
+      "learning_rate": 1.5003966222739697e-08,
+      "loss": 0.758,
+      "step": 41728
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6101309755061604,
+      "learning_rate": 1.4962211967249895e-08,
+      "loss": 0.7553,
+      "step": 41729
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.535418068963271,
+      "learning_rate": 1.4920515847810513e-08,
+      "loss": 0.8117,
+      "step": 41730
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4554827379791337,
+      "learning_rate": 1.4878877864665797e-08,
+      "loss": 0.8171,
+      "step": 41731
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7775865232950692,
+      "learning_rate": 1.4837298018057778e-08,
+      "loss": 0.748,
+      "step": 41732
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5767995753917157,
+      "learning_rate": 1.4795776308228482e-08,
+      "loss": 0.8562,
+      "step": 41733
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5325086679447923,
+      "learning_rate": 1.475431273541883e-08,
+      "loss": 0.6613,
+      "step": 41734
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5635135341613307,
+      "learning_rate": 1.4712907299870848e-08,
+      "loss": 0.8432,
+      "step": 41735
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.507345826833167,
+      "learning_rate": 1.4671560001825458e-08,
+      "loss": 0.8121,
+      "step": 41736
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5219365312952176,
+      "learning_rate": 1.4630270841523575e-08,
+      "loss": 0.8428,
+      "step": 41737
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8606864854600672,
+      "learning_rate": 1.4589039819206118e-08,
+      "loss": 0.6891,
+      "step": 41738
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.962448495636934,
+      "learning_rate": 1.4547866935111787e-08,
+      "loss": 0.8599,
+      "step": 41739
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6250986307120439,
+      "learning_rate": 1.45067521894815e-08,
+      "loss": 0.7746,
+      "step": 41740
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5840843298093559,
+      "learning_rate": 1.4465695582553951e-08,
+      "loss": 0.7635,
+      "step": 41741
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6688136154973106,
+      "learning_rate": 1.4424697114570063e-08,
+      "loss": 0.8642,
+      "step": 41742
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4838431130065373,
+      "learning_rate": 1.4383756785765201e-08,
+      "loss": 0.7669,
+      "step": 41743
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6408593421455324,
+      "learning_rate": 1.4342874596380285e-08,
+      "loss": 0.7918,
+      "step": 41744
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.850411645358254,
+      "learning_rate": 1.4302050546651791e-08,
+      "loss": 0.7949,
+      "step": 41745
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5255354315606555,
+      "learning_rate": 1.4261284636819527e-08,
+      "loss": 0.8401,
+      "step": 41746
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6123891887508799,
+      "learning_rate": 1.4220576867118862e-08,
+      "loss": 0.7043,
+      "step": 41747
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5311229491394223,
+      "learning_rate": 1.4179927237786272e-08,
+      "loss": 0.8438,
+      "step": 41748
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.686533012336037,
+      "learning_rate": 1.4139335749061567e-08,
+      "loss": 0.8156,
+      "step": 41749
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5898228210348984,
+      "learning_rate": 1.4098802401177891e-08,
+      "loss": 0.8437,
+      "step": 41750
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4635324153606513,
+      "learning_rate": 1.4058327194372834e-08,
+      "loss": 0.8326,
+      "step": 41751
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7513581621630065,
+      "learning_rate": 1.4017910128881762e-08,
+      "loss": 0.714,
+      "step": 41752
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6162696547156872,
+      "learning_rate": 1.3977551204938932e-08,
+      "loss": 0.933,
+      "step": 41753
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1224553749176214,
+      "learning_rate": 1.3937250422781934e-08,
+      "loss": 0.6899,
+      "step": 41754
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.74357296461334,
+      "learning_rate": 1.3897007782642802e-08,
+      "loss": 0.8843,
+      "step": 41755
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5691440967931116,
+      "learning_rate": 1.3856823284756904e-08,
+      "loss": 0.8266,
+      "step": 41756
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4520987762614062,
+      "learning_rate": 1.3816696929358498e-08,
+      "loss": 0.7234,
+      "step": 41757
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6661901830809405,
+      "learning_rate": 1.377662871668073e-08,
+      "loss": 0.9046,
+      "step": 41758
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.878696902580624,
+      "learning_rate": 1.3736618646956746e-08,
+      "loss": 0.9297,
+      "step": 41759
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7396033246848694,
+      "learning_rate": 1.3696666720419693e-08,
+      "loss": 0.6929,
+      "step": 41760
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7008918204649297,
+      "learning_rate": 1.365677293730272e-08,
+      "loss": 0.8562,
+      "step": 41761
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7664529464257273,
+      "learning_rate": 1.3616937297837863e-08,
+      "loss": 0.7253,
+      "step": 41762
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.081750715790304,
+      "learning_rate": 1.3577159802256045e-08,
+      "loss": 0.7602,
+      "step": 41763
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.599984469137709,
+      "learning_rate": 1.3537440450790417e-08,
+      "loss": 0.9565,
+      "step": 41764
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5151512425631979,
+      "learning_rate": 1.3497779243670794e-08,
+      "loss": 0.8642,
+      "step": 41765
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.514279797401002,
+      "learning_rate": 1.3458176181129211e-08,
+      "loss": 0.8163,
+      "step": 41766
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.505583430114054,
+      "learning_rate": 1.3418631263395487e-08,
+      "loss": 0.8058,
+      "step": 41767
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6208652784706579,
+      "learning_rate": 1.3379144490700546e-08,
+      "loss": 0.8205,
+      "step": 41768
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6741584410315344,
+      "learning_rate": 1.3339715863273094e-08,
+      "loss": 0.8278,
+      "step": 41769
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3766498747685099,
+      "learning_rate": 1.330034538134517e-08,
+      "loss": 0.8851,
+      "step": 41770
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5673511949251333,
+      "learning_rate": 1.3261033045143257e-08,
+      "loss": 0.826,
+      "step": 41771
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.770956985127694,
+      "learning_rate": 1.3221778854897171e-08,
+      "loss": 0.6826,
+      "step": 41772
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0939528849092284,
+      "learning_rate": 1.318258281083562e-08,
+      "loss": 0.7752,
+      "step": 41773
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5953912445702998,
+      "learning_rate": 1.3143444913187309e-08,
+      "loss": 0.7945,
+      "step": 41774
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5860671621085303,
+      "learning_rate": 1.3104365162179833e-08,
+      "loss": 0.7202,
+      "step": 41775
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.47842307227049,
+      "learning_rate": 1.3065343558040789e-08,
+      "loss": 0.8445,
+      "step": 41776
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6531876977255968,
+      "learning_rate": 1.302638010099666e-08,
+      "loss": 0.904,
+      "step": 41777
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.787510177854233,
+      "learning_rate": 1.2987474791273936e-08,
+      "loss": 0.84,
+      "step": 41778
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7773640668313034,
+      "learning_rate": 1.2948627629101318e-08,
+      "loss": 0.73,
+      "step": 41779
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7723555908997133,
+      "learning_rate": 1.2909838614703073e-08,
+      "loss": 0.6962,
+      "step": 41780
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7641394592611792,
+      "learning_rate": 1.287110774830569e-08,
+      "loss": 0.8708,
+      "step": 41781
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4182773841135774,
+      "learning_rate": 1.2832435030133427e-08,
+      "loss": 0.7146,
+      "step": 41782
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5594697974296043,
+      "learning_rate": 1.2793820460413886e-08,
+      "loss": 0.952,
+      "step": 41783
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9136225943876188,
+      "learning_rate": 1.275526403937022e-08,
+      "loss": 0.8105,
+      "step": 41784
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7903445733173156,
+      "learning_rate": 1.2716765767227801e-08,
+      "loss": 0.8372,
+      "step": 41785
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7099515194466044,
+      "learning_rate": 1.2678325644209789e-08,
+      "loss": 0.7248,
+      "step": 41786
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6244074965330875,
+      "learning_rate": 1.2639943670540445e-08,
+      "loss": 0.8816,
+      "step": 41787
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9565409824431703,
+      "learning_rate": 1.2601619846444035e-08,
+      "loss": 0.8607,
+      "step": 41788
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5092825529478509,
+      "learning_rate": 1.2563354172142605e-08,
+      "loss": 0.7643,
+      "step": 41789
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0063076610331754,
+      "learning_rate": 1.252514664785931e-08,
+      "loss": 0.7116,
+      "step": 41790
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4911911709542738,
+      "learning_rate": 1.2486997273816192e-08,
+      "loss": 0.7233,
+      "step": 41791
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9158075823609098,
+      "learning_rate": 1.2448906050236408e-08,
+      "loss": 0.8668,
+      "step": 41792
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7390170201415678,
+      "learning_rate": 1.2410872977340893e-08,
+      "loss": 0.7507,
+      "step": 41793
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7717696628142487,
+      "learning_rate": 1.2372898055350579e-08,
+      "loss": 0.6875,
+      "step": 41794
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8091833864668897,
+      "learning_rate": 1.2334981284487513e-08,
+      "loss": 0.6861,
+      "step": 41795
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.483154160476019,
+      "learning_rate": 1.2297122664972627e-08,
+      "loss": 0.8359,
+      "step": 41796
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4792422523556608,
+      "learning_rate": 1.2259322197025747e-08,
+      "loss": 0.8359,
+      "step": 41797
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5527333366230125,
+      "learning_rate": 1.2221579880867807e-08,
+      "loss": 0.7863,
+      "step": 41798
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5511495158495103,
+      "learning_rate": 1.218389571671752e-08,
+      "loss": 0.8506,
+      "step": 41799
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7512172762991352,
+      "learning_rate": 1.2146269704793601e-08,
+      "loss": 0.7505,
+      "step": 41800
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6020943743786564,
+      "learning_rate": 1.2108701845318094e-08,
+      "loss": 0.9411,
+      "step": 41801
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.302668697518668,
+      "learning_rate": 1.2071192138506382e-08,
+      "loss": 0.8074,
+      "step": 41802
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.497045016374426,
+      "learning_rate": 1.203374058457829e-08,
+      "loss": 0.8995,
+      "step": 41803
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7194853497576028,
+      "learning_rate": 1.1996347183752532e-08,
+      "loss": 0.7279,
+      "step": 41804
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6635018321188675,
+      "learning_rate": 1.1959011936246712e-08,
+      "loss": 0.7703,
+      "step": 41805
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.826780710576579,
+      "learning_rate": 1.1921734842277321e-08,
+      "loss": 0.7173,
+      "step": 41806
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7857924369923905,
+      "learning_rate": 1.1884515902060856e-08,
+      "loss": 0.8969,
+      "step": 41807
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.440825054739088,
+      "learning_rate": 1.1847355115814918e-08,
+      "loss": 0.9136,
+      "step": 41808
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.568029226012399,
+      "learning_rate": 1.1810252483757112e-08,
+      "loss": 0.8575,
+      "step": 41809
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5735941628549122,
+      "learning_rate": 1.177320800610171e-08,
+      "loss": 0.853,
+      "step": 41810
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7558280079631566,
+      "learning_rate": 1.1736221683064097e-08,
+      "loss": 0.6791,
+      "step": 41811
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.963477297104431,
+      "learning_rate": 1.1699293514861877e-08,
+      "loss": 0.7929,
+      "step": 41812
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4411581610844129,
+      "learning_rate": 1.1662423501707099e-08,
+      "loss": 0.7739,
+      "step": 41813
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7679486971994107,
+      "learning_rate": 1.162561164381737e-08,
+      "loss": 0.9203,
+      "step": 41814
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5044642287650345,
+      "learning_rate": 1.158885794140474e-08,
+      "loss": 0.9054,
+      "step": 41815
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5440968089715525,
+      "learning_rate": 1.1552162394684596e-08,
+      "loss": 0.7499,
+      "step": 41816
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5838764775106966,
+      "learning_rate": 1.1515525003868988e-08,
+      "loss": 0.889,
+      "step": 41817
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.489219188990583,
+      "learning_rate": 1.14789457691733e-08,
+      "loss": 0.6913,
+      "step": 41818
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.657735605418852,
+      "learning_rate": 1.1442424690809584e-08,
+      "loss": 0.8796,
+      "step": 41819
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3825507282444414,
+      "learning_rate": 1.1405961768989893e-08,
+      "loss": 0.8159,
+      "step": 41820
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4930097565014244,
+      "learning_rate": 1.1369557003927389e-08,
+      "loss": 0.8827,
+      "step": 41821
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7069829282942917,
+      "learning_rate": 1.1333210395833017e-08,
+      "loss": 0.8071,
+      "step": 41822
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.205907812755376,
+      "learning_rate": 1.1296921944919936e-08,
+      "loss": 0.8258,
+      "step": 41823
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9859198414205894,
+      "learning_rate": 1.126069165139798e-08,
+      "loss": 0.9403,
+      "step": 41824
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4430082497825907,
+      "learning_rate": 1.1224519515478095e-08,
+      "loss": 0.8616,
+      "step": 41825
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.274874287457045,
+      "learning_rate": 1.1188405537372327e-08,
+      "loss": 0.8372,
+      "step": 41826
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7410609857142372,
+      "learning_rate": 1.1152349717290512e-08,
+      "loss": 0.8634,
+      "step": 41827
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4896803805729928,
+      "learning_rate": 1.111635205544137e-08,
+      "loss": 0.81,
+      "step": 41828
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6649523642731918,
+      "learning_rate": 1.1080412552034736e-08,
+      "loss": 0.8228,
+      "step": 41829
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5943933429419541,
+      "learning_rate": 1.104453120728155e-08,
+      "loss": 0.8003,
+      "step": 41830
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5365457851387168,
+      "learning_rate": 1.1008708021389426e-08,
+      "loss": 0.8894,
+      "step": 41831
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5207232624949443,
+      "learning_rate": 1.0972942994565971e-08,
+      "loss": 0.7796,
+      "step": 41832
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6213608153732515,
+      "learning_rate": 1.0937236127022133e-08,
+      "loss": 0.8403,
+      "step": 41833
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.591735351962779,
+      "learning_rate": 1.0901587418962189e-08,
+      "loss": 0.8718,
+      "step": 41834
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7799956973582244,
+      "learning_rate": 1.0865996870597084e-08,
+      "loss": 0.7088,
+      "step": 41835
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5531959745946426,
+      "learning_rate": 1.0830464482132208e-08,
+      "loss": 0.8031,
+      "step": 41836
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6893701468364837,
+      "learning_rate": 1.0794990253775172e-08,
+      "loss": 0.7609,
+      "step": 41837
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7193990039996612,
+      "learning_rate": 1.075957418573248e-08,
+      "loss": 0.6748,
+      "step": 41838
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4909796165716593,
+      "learning_rate": 1.072421627820952e-08,
+      "loss": 0.7789,
+      "step": 41839
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5940231082739216,
+      "learning_rate": 1.0688916531412796e-08,
+      "loss": 0.8354,
+      "step": 41840
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7278837487797885,
+      "learning_rate": 1.0653674945547699e-08,
+      "loss": 0.8936,
+      "step": 41841
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8083565261078165,
+      "learning_rate": 1.0618491520819619e-08,
+      "loss": 0.855,
+      "step": 41842
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5414987243188247,
+      "learning_rate": 1.0583366257432836e-08,
+      "loss": 0.8962,
+      "step": 41843
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4804398615957626,
+      "learning_rate": 1.0548299155593855e-08,
+      "loss": 0.8693,
+      "step": 41844
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5093458071273713,
+      "learning_rate": 1.0513290215503624e-08,
+      "loss": 0.8485,
+      "step": 41845
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.700737714214359,
+      "learning_rate": 1.0478339437368645e-08,
+      "loss": 0.8377,
+      "step": 41846
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6853186686598494,
+      "learning_rate": 1.044344682139098e-08,
+      "loss": 0.8598,
+      "step": 41847
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4866875396514914,
+      "learning_rate": 1.0408612367773796e-08,
+      "loss": 0.7418,
+      "step": 41848
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5033244779420236,
+      "learning_rate": 1.0373836076720267e-08,
+      "loss": 0.7074,
+      "step": 41849
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6497303407926267,
+      "learning_rate": 1.0339117948433564e-08,
+      "loss": 0.8233,
+      "step": 41850
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8799858348657497,
+      "learning_rate": 1.0304457983114636e-08,
+      "loss": 0.8587,
+      "step": 41851
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7630237809581055,
+      "learning_rate": 1.0269856180966653e-08,
+      "loss": 0.7579,
+      "step": 41852
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6670863201826625,
+      "learning_rate": 1.0235312542189458e-08,
+      "loss": 0.8046,
+      "step": 41853
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.538942930418896,
+      "learning_rate": 1.0200827066986219e-08,
+      "loss": 0.8679,
+      "step": 41854
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5430123242090847,
+      "learning_rate": 1.0166399755554557e-08,
+      "loss": 0.7431,
+      "step": 41855
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7459468229953221,
+      "learning_rate": 1.0132030608098752e-08,
+      "loss": 0.72,
+      "step": 41856
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6209300827513866,
+      "learning_rate": 1.0097719624815316e-08,
+      "loss": 0.8657,
+      "step": 41857
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6039018163932437,
+      "learning_rate": 1.0063466805907418e-08,
+      "loss": 0.8369,
+      "step": 41858
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7999463376941315,
+      "learning_rate": 1.0029272151571567e-08,
+      "loss": 0.6811,
+      "step": 41859
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7802888269734918,
+      "learning_rate": 9.995135662007604e-09,
+      "loss": 0.7922,
+      "step": 41860
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5808047499789728,
+      "learning_rate": 9.96105733741648e-09,
+      "loss": 0.8611,
+      "step": 41861
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7530616195581469,
+      "learning_rate": 9.927037177993593e-09,
+      "loss": 0.7375,
+      "step": 41862
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6377257072102507,
+      "learning_rate": 9.893075183938783e-09,
+      "loss": 0.878,
+      "step": 41863
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6523500227780914,
+      "learning_rate": 9.859171355448561e-09,
+      "loss": 0.7517,
+      "step": 41864
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6049837807483927,
+      "learning_rate": 9.825325692721654e-09,
+      "loss": 0.781,
+      "step": 41865
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7063274343792814,
+      "learning_rate": 9.791538195954575e-09,
+      "loss": 0.9295,
+      "step": 41866
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1929138570046223,
+      "learning_rate": 9.757808865343831e-09,
+      "loss": 0.9369,
+      "step": 41867
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6912130503666614,
+      "learning_rate": 9.724137701085933e-09,
+      "loss": 0.7955,
+      "step": 41868
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7204453327680194,
+      "learning_rate": 9.690524703377391e-09,
+      "loss": 1.0012,
+      "step": 41869
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6433619315980652,
+      "learning_rate": 9.656969872413601e-09,
+      "loss": 0.8682,
+      "step": 41870
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4880636097846032,
+      "learning_rate": 9.623473208388855e-09,
+      "loss": 0.8735,
+      "step": 41871
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5631332750661575,
+      "learning_rate": 9.590034711499663e-09,
+      "loss": 0.9165,
+      "step": 41872
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6581816368597866,
+      "learning_rate": 9.556654381940311e-09,
+      "loss": 0.7218,
+      "step": 41873
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.594711909125336,
+      "learning_rate": 9.523332219905091e-09,
+      "loss": 0.6793,
+      "step": 41874
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6107479452022697,
+      "learning_rate": 9.490068225588289e-09,
+      "loss": 0.7793,
+      "step": 41875
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6651167315884885,
+      "learning_rate": 9.456862399183087e-09,
+      "loss": 0.8699,
+      "step": 41876
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5926780768067188,
+      "learning_rate": 9.423714740882661e-09,
+      "loss": 0.8986,
+      "step": 41877
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7478886855230067,
+      "learning_rate": 9.390625250880191e-09,
+      "loss": 0.9168,
+      "step": 41878
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.800321635559154,
+      "learning_rate": 9.357593929368857e-09,
+      "loss": 0.8619,
+      "step": 41879
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6120234456393763,
+      "learning_rate": 9.324620776540727e-09,
+      "loss": 0.932,
+      "step": 41880
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8144654555326503,
+      "learning_rate": 9.291705792586758e-09,
+      "loss": 0.8277,
+      "step": 41881
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4241758457996683,
+      "learning_rate": 9.25884897770013e-09,
+      "loss": 0.7159,
+      "step": 41882
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5843784439088398,
+      "learning_rate": 9.226050332071801e-09,
+      "loss": 0.8455,
+      "step": 41883
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8638277087814255,
+      "learning_rate": 9.193309855891619e-09,
+      "loss": 0.7336,
+      "step": 41884
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6471199706081168,
+      "learning_rate": 9.160627549351652e-09,
+      "loss": 0.7834,
+      "step": 41885
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7261456402114468,
+      "learning_rate": 9.12800341264175e-09,
+      "loss": 0.9011,
+      "step": 41886
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7269119908162536,
+      "learning_rate": 9.095437445950651e-09,
+      "loss": 0.798,
+      "step": 41887
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.152220822013732,
+      "learning_rate": 9.06292964947042e-09,
+      "loss": 0.8323,
+      "step": 41888
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5208600270792056,
+      "learning_rate": 9.030480023387579e-09,
+      "loss": 0.9764,
+      "step": 41889
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.825497451724092,
+      "learning_rate": 8.998088567893081e-09,
+      "loss": 0.8893,
+      "step": 41890
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.193770268210044,
+      "learning_rate": 8.96575528317456e-09,
+      "loss": 0.7398,
+      "step": 41891
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5672470653618171,
+      "learning_rate": 8.93348016942075e-09,
+      "loss": 0.7696,
+      "step": 41892
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7893129917086192,
+      "learning_rate": 8.90126322681928e-09,
+      "loss": 0.7807,
+      "step": 41893
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5652899618974498,
+      "learning_rate": 8.869104455557776e-09,
+      "loss": 0.8135,
+      "step": 41894
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7927124173548392,
+      "learning_rate": 8.837003855822756e-09,
+      "loss": 0.878,
+      "step": 41895
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.875730332103803,
+      "learning_rate": 8.804961427802961e-09,
+      "loss": 0.7955,
+      "step": 41896
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7434252834276223,
+      "learning_rate": 8.772977171682684e-09,
+      "loss": 0.7038,
+      "step": 41897
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8379959346015657,
+      "learning_rate": 8.741051087650665e-09,
+      "loss": 0.8462,
+      "step": 41898
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4267828996337253,
+      "learning_rate": 8.70918317589009e-09,
+      "loss": 0.6702,
+      "step": 41899
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.989018280368812,
+      "learning_rate": 8.677373436588588e-09,
+      "loss": 0.8926,
+      "step": 41900
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7690685624450557,
+      "learning_rate": 8.645621869930453e-09,
+      "loss": 0.7064,
+      "step": 41901
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.905596774365902,
+      "learning_rate": 8.613928476101097e-09,
+      "loss": 0.8478,
+      "step": 41902
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9461582365098065,
+      "learning_rate": 8.582293255284813e-09,
+      "loss": 0.8164,
+      "step": 41903
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6601419079385469,
+      "learning_rate": 8.550716207664788e-09,
+      "loss": 0.9181,
+      "step": 41904
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7615748612042184,
+      "learning_rate": 8.519197333426431e-09,
+      "loss": 0.9252,
+      "step": 41905
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6707142349727042,
+      "learning_rate": 8.48773663275182e-09,
+      "loss": 0.9549,
+      "step": 41906
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1421303850952187,
+      "learning_rate": 8.456334105825248e-09,
+      "loss": 0.7236,
+      "step": 41907
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7777113867511471,
+      "learning_rate": 8.424989752829904e-09,
+      "loss": 0.6998,
+      "step": 41908
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4748783991160097,
+      "learning_rate": 8.393703573946754e-09,
+      "loss": 0.837,
+      "step": 41909
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.742148068876152,
+      "learning_rate": 8.362475569358985e-09,
+      "loss": 0.7714,
+      "step": 41910
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.04838410661873,
+      "learning_rate": 8.331305739247565e-09,
+      "loss": 0.7956,
+      "step": 41911
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8553210162025475,
+      "learning_rate": 8.300194083795676e-09,
+      "loss": 0.7465,
+      "step": 41912
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7560498216507465,
+      "learning_rate": 8.269140603182069e-09,
+      "loss": 0.8179,
+      "step": 41913
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.697039096624039,
+      "learning_rate": 8.238145297589927e-09,
+      "loss": 0.6841,
+      "step": 41914
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.535412404141047,
+      "learning_rate": 8.207208167198e-09,
+      "loss": 0.7341,
+      "step": 41915
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7767927537456556,
+      "learning_rate": 8.176329212187251e-09,
+      "loss": 0.8861,
+      "step": 41916
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5945348144552562,
+      "learning_rate": 8.145508432737537e-09,
+      "loss": 0.9093,
+      "step": 41917
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5663619310900436,
+      "learning_rate": 8.114745829027604e-09,
+      "loss": 0.797,
+      "step": 41918
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5328654095740561,
+      "learning_rate": 8.084041401237308e-09,
+      "loss": 0.9,
+      "step": 41919
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9297411007533904,
+      "learning_rate": 8.053395149545395e-09,
+      "loss": 0.7581,
+      "step": 41920
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6483682791924432,
+      "learning_rate": 8.022807074129501e-09,
+      "loss": 0.9214,
+      "step": 41921
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.939298238516844,
+      "learning_rate": 7.992277175168373e-09,
+      "loss": 0.7522,
+      "step": 41922
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6047272257907268,
+      "learning_rate": 7.961805452839644e-09,
+      "loss": 0.8724,
+      "step": 41923
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.655936855846607,
+      "learning_rate": 7.931391907320951e-09,
+      "loss": 0.808,
+      "step": 41924
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6597993441469099,
+      "learning_rate": 7.901036538788819e-09,
+      "loss": 0.8568,
+      "step": 41925
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5578262255724826,
+      "learning_rate": 7.870739347420886e-09,
+      "loss": 0.7799,
+      "step": 41926
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6863981912384889,
+      "learning_rate": 7.840500333392565e-09,
+      "loss": 0.7154,
+      "step": 41927
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4828156405451518,
+      "learning_rate": 7.81031949688038e-09,
+      "loss": 0.9089,
+      "step": 41928
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7611093082519496,
+      "learning_rate": 7.78019683805975e-09,
+      "loss": 0.8068,
+      "step": 41929
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4417432808807757,
+      "learning_rate": 7.750132357106089e-09,
+      "loss": 0.8828,
+      "step": 41930
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5978511178883397,
+      "learning_rate": 7.72012605419481e-09,
+      "loss": 0.7744,
+      "step": 41931
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7461684686580133,
+      "learning_rate": 7.69017792950022e-09,
+      "loss": 0.6518,
+      "step": 41932
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.548792307624905,
+      "learning_rate": 7.660287983197735e-09,
+      "loss": 0.8971,
+      "step": 41933
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.532557098215765,
+      "learning_rate": 7.630456215460546e-09,
+      "loss": 0.7483,
+      "step": 41934
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7211683818785448,
+      "learning_rate": 7.600682626461853e-09,
+      "loss": 0.817,
+      "step": 41935
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6492565005064208,
+      "learning_rate": 7.570967216374847e-09,
+      "loss": 0.6878,
+      "step": 41936
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8267000380934253,
+      "learning_rate": 7.541309985373834e-09,
+      "loss": 0.7692,
+      "step": 41937
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.760880614854285,
+      "learning_rate": 7.51171093363201e-09,
+      "loss": 0.8254,
+      "step": 41938
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0403560320965703,
+      "learning_rate": 7.482170061319238e-09,
+      "loss": 0.9892,
+      "step": 41939
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.825487752671865,
+      "learning_rate": 7.452687368608713e-09,
+      "loss": 0.8799,
+      "step": 41940
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.608576667588643,
+      "learning_rate": 7.42326285567363e-09,
+      "loss": 0.8642,
+      "step": 41941
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5503941592660735,
+      "learning_rate": 7.393896522682742e-09,
+      "loss": 0.8697,
+      "step": 41942
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.545473521731658,
+      "learning_rate": 7.364588369808135e-09,
+      "loss": 0.9147,
+      "step": 41943
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6694502539583016,
+      "learning_rate": 7.335338397220782e-09,
+      "loss": 0.851,
+      "step": 41944
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7850567203928347,
+      "learning_rate": 7.306146605090547e-09,
+      "loss": 0.739,
+      "step": 41945
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5764003821824117,
+      "learning_rate": 7.277012993588406e-09,
+      "loss": 0.7573,
+      "step": 41946
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7625140071124823,
+      "learning_rate": 7.247937562882001e-09,
+      "loss": 0.7982,
+      "step": 41947
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5782034719346079,
+      "learning_rate": 7.218920313142308e-09,
+      "loss": 0.8349,
+      "step": 41948
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5928420214911378,
+      "learning_rate": 7.1899612445380795e-09,
+      "loss": 0.9107,
+      "step": 41949
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5315856906040244,
+      "learning_rate": 7.16106035723696e-09,
+      "loss": 0.8793,
+      "step": 41950
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9511578369100437,
+      "learning_rate": 7.132217651407702e-09,
+      "loss": 0.7668,
+      "step": 41951
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8453358178445545,
+      "learning_rate": 7.103433127219061e-09,
+      "loss": 0.7536,
+      "step": 41952
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.338734503312036,
+      "learning_rate": 7.074706784837571e-09,
+      "loss": 0.8466,
+      "step": 41953
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.731644689046437,
+      "learning_rate": 7.0460386244308735e-09,
+      "loss": 0.9123,
+      "step": 41954
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5413688471965805,
+      "learning_rate": 7.017428646165503e-09,
+      "loss": 0.8488,
+      "step": 41955
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7010266652499269,
+      "learning_rate": 6.988876850209103e-09,
+      "loss": 0.8788,
+      "step": 41956
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5969658230317705,
+      "learning_rate": 6.960383236727098e-09,
+      "loss": 0.8973,
+      "step": 41957
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5688416152250688,
+      "learning_rate": 6.931947805884909e-09,
+      "loss": 0.7635,
+      "step": 41958
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7734033156233038,
+      "learning_rate": 6.903570557849071e-09,
+      "loss": 0.6726,
+      "step": 41959
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.575006427709941,
+      "learning_rate": 6.875251492783897e-09,
+      "loss": 0.83,
+      "step": 41960
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.576319255790347,
+      "learning_rate": 6.8469906108559195e-09,
+      "loss": 0.8118,
+      "step": 41961
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.576407000500549,
+      "learning_rate": 6.8187879122272315e-09,
+      "loss": 0.7654,
+      "step": 41962
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5964130524030458,
+      "learning_rate": 6.790643397064367e-09,
+      "loss": 0.8384,
+      "step": 41963
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7394951425594846,
+      "learning_rate": 6.762557065529418e-09,
+      "loss": 0.7773,
+      "step": 41964
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4795064079589548,
+      "learning_rate": 6.734528917786698e-09,
+      "loss": 0.7277,
+      "step": 41965
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4399657933980936,
+      "learning_rate": 6.706558953999409e-09,
+      "loss": 0.7273,
+      "step": 41966
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5635406703556178,
+      "learning_rate": 6.678647174330755e-09,
+      "loss": 0.893,
+      "step": 41967
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7593951573342264,
+      "learning_rate": 6.650793578942827e-09,
+      "loss": 0.6539,
+      "step": 41968
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7796193960428932,
+      "learning_rate": 6.622998167997718e-09,
+      "loss": 0.7507,
+      "step": 41969
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.926229691534594,
+      "learning_rate": 6.595260941657522e-09,
+      "loss": 0.7918,
+      "step": 41970
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8012444955740998,
+      "learning_rate": 6.56758190008322e-09,
+      "loss": 0.9175,
+      "step": 41971
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.106170614193945,
+      "learning_rate": 6.539961043436904e-09,
+      "loss": 0.8742,
+      "step": 41972
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6535227739836635,
+      "learning_rate": 6.512398371877338e-09,
+      "loss": 0.8792,
+      "step": 41973
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.737138964879444,
+      "learning_rate": 6.484893885567722e-09,
+      "loss": 0.7315,
+      "step": 41974
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.613828741706436,
+      "learning_rate": 6.457447584666821e-09,
+      "loss": 0.8071,
+      "step": 41975
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6353938154141876,
+      "learning_rate": 6.430059469334504e-09,
+      "loss": 0.8622,
+      "step": 41976
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6123664234805481,
+      "learning_rate": 6.4027295397306455e-09,
+      "loss": 0.7494,
+      "step": 41977
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5210229633056638,
+      "learning_rate": 6.375457796014006e-09,
+      "loss": 0.7925,
+      "step": 41978
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6174267254700743,
+      "learning_rate": 6.3482442383422385e-09,
+      "loss": 0.8786,
+      "step": 41979
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7956070073129142,
+      "learning_rate": 6.3210888668763235e-09,
+      "loss": 0.7157,
+      "step": 41980
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6448918439100395,
+      "learning_rate": 6.293991681771694e-09,
+      "loss": 0.7514,
+      "step": 41981
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9009945354531366,
+      "learning_rate": 6.266952683188221e-09,
+      "loss": 0.8137,
+      "step": 41982
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4032595012960454,
+      "learning_rate": 6.2399718712824465e-09,
+      "loss": 0.7441,
+      "step": 41983
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8685044967332287,
+      "learning_rate": 6.2130492462120214e-09,
+      "loss": 0.953,
+      "step": 41984
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.507436854653997,
+      "learning_rate": 6.186184808132378e-09,
+      "loss": 0.7108,
+      "step": 41985
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.918893878181222,
+      "learning_rate": 6.159378557201168e-09,
+      "loss": 0.8543,
+      "step": 41986
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7480280513478889,
+      "learning_rate": 6.1326304935738215e-09,
+      "loss": 0.6686,
+      "step": 41987
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6895498411727639,
+      "learning_rate": 6.10594061740577e-09,
+      "loss": 0.8125,
+      "step": 41988
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7422256713230387,
+      "learning_rate": 6.0793089288535555e-09,
+      "loss": 0.6618,
+      "step": 41989
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7628640235791582,
+      "learning_rate": 6.052735428071499e-09,
+      "loss": 0.773,
+      "step": 41990
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3045173040683573,
+      "learning_rate": 6.0262201152150315e-09,
+      "loss": 0.7633,
+      "step": 41991
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.812994534129076,
+      "learning_rate": 5.999762990437363e-09,
+      "loss": 0.7606,
+      "step": 41992
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4548546143975325,
+      "learning_rate": 5.973364053892816e-09,
+      "loss": 0.8083,
+      "step": 41993
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9682016095126047,
+      "learning_rate": 5.947023305735711e-09,
+      "loss": 0.8459,
+      "step": 41994
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.841050926938972,
+      "learning_rate": 5.920740746119258e-09,
+      "loss": 0.7342,
+      "step": 41995
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6292282833068548,
+      "learning_rate": 5.894516375196668e-09,
+      "loss": 0.8368,
+      "step": 41996
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6376517153693662,
+      "learning_rate": 5.868350193120043e-09,
+      "loss": 0.957,
+      "step": 41997
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5677425754853822,
+      "learning_rate": 5.842242200041481e-09,
+      "loss": 0.8523,
+      "step": 41998
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7163458193890406,
+      "learning_rate": 5.816192396114195e-09,
+      "loss": 0.7124,
+      "step": 41999
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5959636266286108,
+      "learning_rate": 5.790200781488065e-09,
+      "loss": 0.8304,
+      "step": 42000
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.60047236498428,
+      "learning_rate": 5.7642673563163e-09,
+      "loss": 0.8714,
+      "step": 42001
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8086512786733929,
+      "learning_rate": 5.738392120749892e-09,
+      "loss": 0.708,
+      "step": 42002
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6360507396157238,
+      "learning_rate": 5.712575074937609e-09,
+      "loss": 0.7868,
+      "step": 42003
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6875467415530974,
+      "learning_rate": 5.686816219030444e-09,
+      "loss": 0.8125,
+      "step": 42004
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5564414259088013,
+      "learning_rate": 5.661115553179386e-09,
+      "loss": 0.7974,
+      "step": 42005
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9714894960538927,
+      "learning_rate": 5.6354730775332046e-09,
+      "loss": 0.8733,
+      "step": 42006
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7253191268369965,
+      "learning_rate": 5.60988879224178e-09,
+      "loss": 0.8116,
+      "step": 42007
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5733209569509938,
+      "learning_rate": 5.584362697453882e-09,
+      "loss": 0.9592,
+      "step": 42008
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4898842023871535,
+      "learning_rate": 5.558894793318281e-09,
+      "loss": 0.7801,
+      "step": 42009
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6238206246737126,
+      "learning_rate": 5.533485079982637e-09,
+      "loss": 0.8502,
+      "step": 42010
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5461110578736694,
+      "learning_rate": 5.508133557594608e-09,
+      "loss": 0.9267,
+      "step": 42011
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5745190894173169,
+      "learning_rate": 5.482840226302966e-09,
+      "loss": 0.9146,
+      "step": 42012
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.613508267199235,
+      "learning_rate": 5.457605086254258e-09,
+      "loss": 0.7565,
+      "step": 42013
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7957235311448905,
+      "learning_rate": 5.4324281375961466e-09,
+      "loss": 0.911,
+      "step": 42014
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.916848708740567,
+      "learning_rate": 5.4073093804740685e-09,
+      "loss": 0.907,
+      "step": 42015
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3623906113375608,
+      "learning_rate": 5.382248815034574e-09,
+      "loss": 0.741,
+      "step": 42016
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6360614399766014,
+      "learning_rate": 5.357246441424213e-09,
+      "loss": 0.9546,
+      "step": 42017
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6900423943106193,
+      "learning_rate": 5.332302259788424e-09,
+      "loss": 0.7028,
+      "step": 42018
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7555677249594162,
+      "learning_rate": 5.307416270272647e-09,
+      "loss": 0.8327,
+      "step": 42019
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6542947213497803,
+      "learning_rate": 5.282588473021211e-09,
+      "loss": 0.7745,
+      "step": 42020
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9200205924427867,
+      "learning_rate": 5.257818868178444e-09,
+      "loss": 0.9044,
+      "step": 42021
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7971206608869363,
+      "learning_rate": 5.2331074558886756e-09,
+      "loss": 0.7941,
+      "step": 42022
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.432204400360843,
+      "learning_rate": 5.208454236296234e-09,
+      "loss": 0.844,
+      "step": 42023
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8276887138666706,
+      "learning_rate": 5.18385920954434e-09,
+      "loss": 0.8795,
+      "step": 42024
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.681793831725793,
+      "learning_rate": 5.1593223757762104e-09,
+      "loss": 0.8211,
+      "step": 42025
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4557018697470403,
+      "learning_rate": 5.134843735135065e-09,
+      "loss": 0.9139,
+      "step": 42026
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4942313427359197,
+      "learning_rate": 5.110423287764121e-09,
+      "loss": 0.7925,
+      "step": 42027
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7702556873427607,
+      "learning_rate": 5.086061033803269e-09,
+      "loss": 0.7072,
+      "step": 42028
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6213438235066502,
+      "learning_rate": 5.061756973396836e-09,
+      "loss": 0.9244,
+      "step": 42029
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6146090975759584,
+      "learning_rate": 5.037511106684712e-09,
+      "loss": 0.8488,
+      "step": 42030
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6574207149959963,
+      "learning_rate": 5.013323433807893e-09,
+      "loss": 0.8622,
+      "step": 42031
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7012173891753748,
+      "learning_rate": 4.989193954908489e-09,
+      "loss": 0.7764,
+      "step": 42032
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.492394742104991,
+      "learning_rate": 4.9651226701263874e-09,
+      "loss": 0.9072,
+      "step": 42033
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6604185550871489,
+      "learning_rate": 4.9411095796014775e-09,
+      "loss": 0.7709,
+      "step": 42034
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0112511067993952,
+      "learning_rate": 4.917154683473646e-09,
+      "loss": 0.8986,
+      "step": 42035
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.734765086688506,
+      "learning_rate": 4.8932579818827816e-09,
+      "loss": 0.8164,
+      "step": 42036
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4146781479876835,
+      "learning_rate": 4.869419474967663e-09,
+      "loss": 0.8488,
+      "step": 42037
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4107350438895805,
+      "learning_rate": 4.845639162868177e-09,
+      "loss": 0.7561,
+      "step": 42038
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5407813969043773,
+      "learning_rate": 4.821917045720881e-09,
+      "loss": 0.8033,
+      "step": 42039
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.836594986334505,
+      "learning_rate": 4.7982531236645535e-09,
+      "loss": 0.883,
+      "step": 42040
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.490446151738191,
+      "learning_rate": 4.774647396837973e-09,
+      "loss": 0.6762,
+      "step": 42041
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5223946831635171,
+      "learning_rate": 4.751099865376585e-09,
+      "loss": 0.8638,
+      "step": 42042
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6582059643971607,
+      "learning_rate": 4.727610529419169e-09,
+      "loss": 0.7387,
+      "step": 42043
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6188291301567288,
+      "learning_rate": 4.704179389103391e-09,
+      "loss": 0.9632,
+      "step": 42044
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6519178207796725,
+      "learning_rate": 4.6808064445635905e-09,
+      "loss": 0.8252,
+      "step": 42045
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5685927902775283,
+      "learning_rate": 4.6574916959363226e-09,
+      "loss": 0.8694,
+      "step": 42046
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.860574218549373,
+      "learning_rate": 4.634235143357036e-09,
+      "loss": 0.7808,
+      "step": 42047
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4648592592054135,
+      "learning_rate": 4.611036786962286e-09,
+      "loss": 0.7645,
+      "step": 42048
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1396406404446973,
+      "learning_rate": 4.587896626886412e-09,
+      "loss": 0.9102,
+      "step": 42049
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6898657395335692,
+      "learning_rate": 4.5648146632648605e-09,
+      "loss": 0.6843,
+      "step": 42050
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9186603201503372,
+      "learning_rate": 4.541790896230858e-09,
+      "loss": 0.8907,
+      "step": 42051
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8115638369974165,
+      "learning_rate": 4.518825325919851e-09,
+      "loss": 0.8985,
+      "step": 42052
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6726601303601845,
+      "learning_rate": 4.495917952463957e-09,
+      "loss": 0.7959,
+      "step": 42053
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7267300860021146,
+      "learning_rate": 4.473068775997513e-09,
+      "loss": 0.7816,
+      "step": 42054
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5750722919682387,
+      "learning_rate": 4.450277796653746e-09,
+      "loss": 0.8147,
+      "step": 42055
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5946372219424638,
+      "learning_rate": 4.427545014564771e-09,
+      "loss": 0.8895,
+      "step": 42056
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5130485629711143,
+      "learning_rate": 4.404870429862706e-09,
+      "loss": 0.8083,
+      "step": 42057
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7053948873194787,
+      "learning_rate": 4.382254042680778e-09,
+      "loss": 0.718,
+      "step": 42058
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5045918871785915,
+      "learning_rate": 4.359695853149992e-09,
+      "loss": 0.7785,
+      "step": 42059
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8110061428741544,
+      "learning_rate": 4.337195861400245e-09,
+      "loss": 0.7316,
+      "step": 42060
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6243104617076798,
+      "learning_rate": 4.314754067565874e-09,
+      "loss": 0.8336,
+      "step": 42061
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5893870664237537,
+      "learning_rate": 4.292370471774554e-09,
+      "loss": 0.9552,
+      "step": 42062
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.939462732398592,
+      "learning_rate": 4.270045074157292e-09,
+      "loss": 0.8126,
+      "step": 42063
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6179622474367747,
+      "learning_rate": 4.247777874845094e-09,
+      "loss": 0.7666,
+      "step": 42064
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7164274535221344,
+      "learning_rate": 4.225568873966746e-09,
+      "loss": 0.8096,
+      "step": 42065
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5840615804994322,
+      "learning_rate": 4.203418071652143e-09,
+      "loss": 0.9026,
+      "step": 42066
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5917213574513505,
+      "learning_rate": 4.181325468028963e-09,
+      "loss": 0.8132,
+      "step": 42067
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7102750912942513,
+      "learning_rate": 4.1592910632271e-09,
+      "loss": 0.7834,
+      "step": 42068
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5090199973356018,
+      "learning_rate": 4.13731485737423e-09,
+      "loss": 0.9013,
+      "step": 42069
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5655927347096446,
+      "learning_rate": 4.11539685059914e-09,
+      "loss": 0.7007,
+      "step": 42070
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7168617456999442,
+      "learning_rate": 4.093537043028395e-09,
+      "loss": 0.6611,
+      "step": 42071
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7696031302794394,
+      "learning_rate": 4.07173543478967e-09,
+      "loss": 0.739,
+      "step": 42072
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.649854025037273,
+      "learning_rate": 4.04999202601064e-09,
+      "loss": 0.7276,
+      "step": 42073
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8493766840180599,
+      "learning_rate": 4.028306816816763e-09,
+      "loss": 0.8045,
+      "step": 42074
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5838138631898973,
+      "learning_rate": 4.006679807334601e-09,
+      "loss": 0.7464,
+      "step": 42075
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6824954649295156,
+      "learning_rate": 3.985110997690722e-09,
+      "loss": 0.8517,
+      "step": 42076
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.629527128220467,
+      "learning_rate": 3.963600388009469e-09,
+      "loss": 0.7851,
+      "step": 42077
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6636963557476276,
+      "learning_rate": 3.94214797841741e-09,
+      "loss": 0.8401,
+      "step": 42078
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5988226621922081,
+      "learning_rate": 3.920753769038888e-09,
+      "loss": 0.855,
+      "step": 42079
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7094595351590027,
+      "learning_rate": 3.899417759998248e-09,
+      "loss": 0.8503,
+      "step": 42080
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6455601177335613,
+      "learning_rate": 3.878139951419835e-09,
+      "loss": 0.708,
+      "step": 42081
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6456125853749777,
+      "learning_rate": 3.856920343427994e-09,
+      "loss": 0.8274,
+      "step": 42082
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5610921799248412,
+      "learning_rate": 3.835758936145961e-09,
+      "loss": 0.8567,
+      "step": 42083
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.279416668083877,
+      "learning_rate": 3.81465572969586e-09,
+      "loss": 0.8126,
+      "step": 42084
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.496761557016281,
+      "learning_rate": 3.793610724202035e-09,
+      "loss": 0.8963,
+      "step": 42085
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7139423250813832,
+      "learning_rate": 3.772623919786611e-09,
+      "loss": 0.7131,
+      "step": 42086
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5348082579794529,
+      "learning_rate": 3.751695316572823e-09,
+      "loss": 0.8163,
+      "step": 42087
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7206914423653952,
+      "learning_rate": 3.730824914679465e-09,
+      "loss": 0.8193,
+      "step": 42088
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7890083445702105,
+      "learning_rate": 3.7100127142308816e-09,
+      "loss": 0.6845,
+      "step": 42089
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.632871977828484,
+      "learning_rate": 3.6892587153480874e-09,
+      "loss": 0.8003,
+      "step": 42090
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.573910722161535,
+      "learning_rate": 3.6685629181509864e-09,
+      "loss": 0.8504,
+      "step": 42091
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6597648241905727,
+      "learning_rate": 3.6479253227594825e-09,
+      "loss": 0.6972,
+      "step": 42092
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8632808429514642,
+      "learning_rate": 3.62734592929459e-09,
+      "loss": 0.7189,
+      "step": 42093
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5340071473539283,
+      "learning_rate": 3.6068247378773236e-09,
+      "loss": 0.8076,
+      "step": 42094
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6725556762039764,
+      "learning_rate": 3.5863617486242565e-09,
+      "loss": 0.7629,
+      "step": 42095
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5513561390902737,
+      "learning_rate": 3.565956961657513e-09,
+      "loss": 0.7395,
+      "step": 42096
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7879562561870199,
+      "learning_rate": 3.5456103770936668e-09,
+      "loss": 0.6874,
+      "step": 42097
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.886645090055294,
+      "learning_rate": 3.525321995052622e-09,
+      "loss": 0.7246,
+      "step": 42098
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.716124197545648,
+      "learning_rate": 3.5050918156520618e-09,
+      "loss": 0.906,
+      "step": 42099
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.619730369188966,
+      "learning_rate": 3.4849198390096705e-09,
+      "loss": 0.9721,
+      "step": 42100
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0151043846966914,
+      "learning_rate": 3.4648060652431314e-09,
+      "loss": 0.8242,
+      "step": 42101
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7681106459395923,
+      "learning_rate": 3.444750494469018e-09,
+      "loss": 0.7814,
+      "step": 42102
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5606227225849767,
+      "learning_rate": 3.4247531268039034e-09,
+      "loss": 0.862,
+      "step": 42103
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.598888194745378,
+      "learning_rate": 3.404813962365472e-09,
+      "loss": 0.835,
+      "step": 42104
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5410013153616873,
+      "learning_rate": 3.384933001269186e-09,
+      "loss": 0.8494,
+      "step": 42105
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.648972875977617,
+      "learning_rate": 3.3651102436293993e-09,
+      "loss": 0.7645,
+      "step": 42106
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.496496869405415,
+      "learning_rate": 3.345345689563795e-09,
+      "loss": 0.8668,
+      "step": 42107
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5377025391877217,
+      "learning_rate": 3.325639339186726e-09,
+      "loss": 0.7904,
+      "step": 42108
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7590166129522792,
+      "learning_rate": 3.3059911926125455e-09,
+      "loss": 0.6746,
+      "step": 42109
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.638636988204103,
+      "learning_rate": 3.2864012499556066e-09,
+      "loss": 0.9098,
+      "step": 42110
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.85440765570375,
+      "learning_rate": 3.266869511330262e-09,
+      "loss": 0.9248,
+      "step": 42111
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7280356778032524,
+      "learning_rate": 3.2473959768497542e-09,
+      "loss": 0.846,
+      "step": 42112
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7867079587832477,
+      "learning_rate": 3.227980646628437e-09,
+      "loss": 0.7227,
+      "step": 42113
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5488762805388343,
+      "learning_rate": 3.2086235207784423e-09,
+      "loss": 0.8064,
+      "step": 42114
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8208351407897256,
+      "learning_rate": 3.189324599413013e-09,
+      "loss": 0.8355,
+      "step": 42115
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4767771999923338,
+      "learning_rate": 3.170083882643171e-09,
+      "loss": 0.7314,
+      "step": 42116
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6294776443868724,
+      "learning_rate": 3.1509013705832703e-09,
+      "loss": 0.8717,
+      "step": 42117
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4475305679288326,
+      "learning_rate": 3.131777063343222e-09,
+      "loss": 0.8556,
+      "step": 42118
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7309691619212593,
+      "learning_rate": 3.1127109610351593e-09,
+      "loss": 0.6849,
+      "step": 42119
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7891592796887521,
+      "learning_rate": 3.0937030637701037e-09,
+      "loss": 0.6889,
+      "step": 42120
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6321058779676643,
+      "learning_rate": 3.074753371657968e-09,
+      "loss": 0.8884,
+      "step": 42121
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.416749884254004,
+      "learning_rate": 3.0558618848097744e-09,
+      "loss": 0.7119,
+      "step": 42122
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.777745466139594,
+      "learning_rate": 3.0370286033354348e-09,
+      "loss": 0.7092,
+      "step": 42123
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5561903327844975,
+      "learning_rate": 3.018253527343751e-09,
+      "loss": 0.7556,
+      "step": 42124
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.765372504121034,
+      "learning_rate": 2.999536656945745e-09,
+      "loss": 0.8775,
+      "step": 42125
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9017997404063829,
+      "learning_rate": 2.980877992249109e-09,
+      "loss": 0.8871,
+      "step": 42126
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.406898496094147,
+      "learning_rate": 2.9622775333626453e-09,
+      "loss": 0.8889,
+      "step": 42127
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7169002223256826,
+      "learning_rate": 2.9437352803951546e-09,
+      "loss": 0.7223,
+      "step": 42128
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6027375522716145,
+      "learning_rate": 2.9252512334543293e-09,
+      "loss": 0.7242,
+      "step": 42129
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5007992894467623,
+      "learning_rate": 2.9068253926478605e-09,
+      "loss": 0.9813,
+      "step": 42130
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.612779544114847,
+      "learning_rate": 2.8884577580823304e-09,
+      "loss": 0.8319,
+      "step": 42131
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6613997772640432,
+      "learning_rate": 2.87014832986543e-09,
+      "loss": 0.8391,
+      "step": 42132
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8595565502968958,
+      "learning_rate": 2.8518971081037403e-09,
+      "loss": 0.8165,
+      "step": 42133
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9474869490096633,
+      "learning_rate": 2.8337040929027336e-09,
+      "loss": 0.8813,
+      "step": 42134
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4186606926970071,
+      "learning_rate": 2.8155692843689907e-09,
+      "loss": 0.7658,
+      "step": 42135
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7767970545008607,
+      "learning_rate": 2.797492682609093e-09,
+      "loss": 1.0313,
+      "step": 42136
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9050809083136673,
+      "learning_rate": 2.7794742877262914e-09,
+      "loss": 0.881,
+      "step": 42137
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.74724581830692,
+      "learning_rate": 2.761514099826057e-09,
+      "loss": 0.7817,
+      "step": 42138
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2794938643076073,
+      "learning_rate": 2.743612119014971e-09,
+      "loss": 0.8526,
+      "step": 42139
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.351408526310065,
+      "learning_rate": 2.725768345394064e-09,
+      "loss": 0.8021,
+      "step": 42140
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7442771404585786,
+      "learning_rate": 2.7079827790699177e-09,
+      "loss": 0.932,
+      "step": 42141
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5981714327140704,
+      "learning_rate": 2.690255420144672e-09,
+      "loss": 0.8744,
+      "step": 42142
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8554735796953388,
+      "learning_rate": 2.6725862687215774e-09,
+      "loss": 0.8384,
+      "step": 42143
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7940699748360345,
+      "learning_rate": 2.6549753249038858e-09,
+      "loss": 0.919,
+      "step": 42144
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6694068955195627,
+      "learning_rate": 2.6374225887937365e-09,
+      "loss": 0.8758,
+      "step": 42145
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5523892031362272,
+      "learning_rate": 2.619928060494381e-09,
+      "loss": 0.7663,
+      "step": 42146
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4984805264805499,
+      "learning_rate": 2.6024917401068495e-09,
+      "loss": 0.7905,
+      "step": 42147
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5914017674025287,
+      "learning_rate": 2.585113627732172e-09,
+      "loss": 0.8077,
+      "step": 42148
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7947136026987764,
+      "learning_rate": 2.567793723471379e-09,
+      "loss": 0.7987,
+      "step": 42149
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7882944635729524,
+      "learning_rate": 2.550532027427721e-09,
+      "loss": 0.7306,
+      "step": 42150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4807020954324805,
+      "learning_rate": 2.533328539698898e-09,
+      "loss": 0.8604,
+      "step": 42151
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4439546813880915,
+      "learning_rate": 2.5161832603870506e-09,
+      "loss": 0.7264,
+      "step": 42152
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8123681933380591,
+      "learning_rate": 2.499096189589878e-09,
+      "loss": 0.7087,
+      "step": 42153
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.646078458518604,
+      "learning_rate": 2.4820673274095207e-09,
+      "loss": 0.7621,
+      "step": 42154
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5728331077996829,
+      "learning_rate": 2.4650966739425686e-09,
+      "loss": 0.7399,
+      "step": 42155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6663202608396253,
+      "learning_rate": 2.4481842292900517e-09,
+      "loss": 0.7746,
+      "step": 42156
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7555283748559,
+      "learning_rate": 2.431329993548559e-09,
+      "loss": 0.7323,
+      "step": 42157
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4958176515271293,
+      "learning_rate": 2.414533966816901e-09,
+      "loss": 0.8003,
+      "step": 42158
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.274521636984133,
+      "learning_rate": 2.3977961491938874e-09,
+      "loss": 0.742,
+      "step": 42159
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7125181244841998,
+      "learning_rate": 2.381116540774997e-09,
+      "loss": 0.8051,
+      "step": 42160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5786489768569147,
+      "learning_rate": 2.36449514165904e-09,
+      "loss": 0.8589,
+      "step": 42161
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7644619998364341,
+      "learning_rate": 2.347931951942606e-09,
+      "loss": 0.7062,
+      "step": 42162
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.536095314714831,
+      "learning_rate": 2.3314269717211735e-09,
+      "loss": 0.7282,
+      "step": 42163
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6880516832975563,
+      "learning_rate": 2.3149802010913323e-09,
+      "loss": 0.8925,
+      "step": 42164
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.741761141728455,
+      "learning_rate": 2.298591640149672e-09,
+      "loss": 0.932,
+      "step": 42165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9544051355502077,
+      "learning_rate": 2.2822612889894513e-09,
+      "loss": 0.8079,
+      "step": 42166
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5960722788497315,
+      "learning_rate": 2.2659891477083694e-09,
+      "loss": 0.7724,
+      "step": 42167
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.451434604034068,
+      "learning_rate": 2.2497752163996855e-09,
+      "loss": 0.8096,
+      "step": 42168
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9069247583664048,
+      "learning_rate": 2.2336194951577685e-09,
+      "loss": 0.898,
+      "step": 42169
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5762068586975495,
+      "learning_rate": 2.217521984076987e-09,
+      "loss": 0.7634,
+      "step": 42170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9540440330432223,
+      "learning_rate": 2.2014826832506e-09,
+      "loss": 0.8717,
+      "step": 42171
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5239298699207087,
+      "learning_rate": 2.185501592772976e-09,
+      "loss": 0.8049,
+      "step": 42172
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5970437557304056,
+      "learning_rate": 2.1695787127362644e-09,
+      "loss": 0.9367,
+      "step": 42173
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4294868943721604,
+      "learning_rate": 2.1537140432348335e-09,
+      "loss": 0.8172,
+      "step": 42174
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8360997977325373,
+      "learning_rate": 2.1379075843586116e-09,
+      "loss": 0.8032,
+      "step": 42175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7422123141163925,
+      "learning_rate": 2.1221593362008575e-09,
+      "loss": 0.7254,
+      "step": 42176
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5409287393377749,
+      "learning_rate": 2.1064692988526093e-09,
+      "loss": 0.8514,
+      "step": 42177
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.822216478774411,
+      "learning_rate": 2.0908374724060156e-09,
+      "loss": 0.8077,
+      "step": 42178
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8850526638680796,
+      "learning_rate": 2.0752638569521144e-09,
+      "loss": 0.8753,
+      "step": 42179
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7460197853364143,
+      "learning_rate": 2.0597484525819446e-09,
+      "loss": 0.8134,
+      "step": 42180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5667553944581412,
+      "learning_rate": 2.0442912593843233e-09,
+      "loss": 0.7924,
+      "step": 42181
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1978676539043427,
+      "learning_rate": 2.0288922774502896e-09,
+      "loss": 0.7424,
+      "step": 42182
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6939014543349793,
+      "learning_rate": 2.013551506869771e-09,
+      "loss": 0.916,
+      "step": 42183
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6843601672946487,
+      "learning_rate": 1.9982689477315854e-09,
+      "loss": 0.8215,
+      "step": 42184
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7931350488999336,
+      "learning_rate": 1.9830446001245507e-09,
+      "loss": 0.6864,
+      "step": 42185
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.650443145608246,
+      "learning_rate": 1.967878464137485e-09,
+      "loss": 0.747,
+      "step": 42186
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.559196613479584,
+      "learning_rate": 1.9527705398592056e-09,
+      "loss": 0.8229,
+      "step": 42187
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8709111209527314,
+      "learning_rate": 1.9377208273763105e-09,
+      "loss": 0.8155,
+      "step": 42188
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7298597818401,
+      "learning_rate": 1.9227293267787272e-09,
+      "loss": 0.9268,
+      "step": 42189
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6001572021283632,
+      "learning_rate": 1.9077960381508333e-09,
+      "loss": 0.7337,
+      "step": 42190
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5986965800238577,
+      "learning_rate": 1.892920961582556e-09,
+      "loss": 0.8051,
+      "step": 42191
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6372080204414885,
+      "learning_rate": 1.878104097158273e-09,
+      "loss": 0.8761,
+      "step": 42192
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5787755525008982,
+      "learning_rate": 1.8633454449645814e-09,
+      "loss": 0.7712,
+      "step": 42193
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9517932818228103,
+      "learning_rate": 1.8486450050880788e-09,
+      "loss": 0.7647,
+      "step": 42194
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.784516249783182,
+      "learning_rate": 1.8340027776142521e-09,
+      "loss": 0.6917,
+      "step": 42195
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4938976545698655,
+      "learning_rate": 1.8194187626285887e-09,
+      "loss": 0.8944,
+      "step": 42196
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6856409757789463,
+      "learning_rate": 1.8048929602154652e-09,
+      "loss": 0.823,
+      "step": 42197
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.60931297550626,
+      "learning_rate": 1.790425370459259e-09,
+      "loss": 0.8991,
+      "step": 42198
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.573121428609393,
+      "learning_rate": 1.7760159934454568e-09,
+      "loss": 0.8473,
+      "step": 42199
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7440282285775388,
+      "learning_rate": 1.7616648292562154e-09,
+      "loss": 0.8926,
+      "step": 42200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6834977174711583,
+      "learning_rate": 1.7473718779759118e-09,
+      "loss": 0.9415,
+      "step": 42201
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5069536718520138,
+      "learning_rate": 1.733137139688923e-09,
+      "loss": 0.8116,
+      "step": 42202
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5859176017233483,
+      "learning_rate": 1.718960614476295e-09,
+      "loss": 0.8734,
+      "step": 42203
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5432410942177515,
+      "learning_rate": 1.7048423024201843e-09,
+      "loss": 0.7438,
+      "step": 42204
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3363354067030424,
+      "learning_rate": 1.6907822036049682e-09,
+      "loss": 0.7291,
+      "step": 42205
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7384978177897719,
+      "learning_rate": 1.6767803181116927e-09,
+      "loss": 0.6953,
+      "step": 42206
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3950069663381863,
+      "learning_rate": 1.6628366460214041e-09,
+      "loss": 0.8534,
+      "step": 42207
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.615596750951639,
+      "learning_rate": 1.6489511874151488e-09,
+      "loss": 0.6716,
+      "step": 42208
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.556367611635961,
+      "learning_rate": 1.6351239423739728e-09,
+      "loss": 0.744,
+      "step": 42209
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6459861180664621,
+      "learning_rate": 1.621354910978923e-09,
+      "loss": 0.8306,
+      "step": 42210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.448195277382428,
+      "learning_rate": 1.6076440933099347e-09,
+      "loss": 0.7786,
+      "step": 42211
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6520300869112916,
+      "learning_rate": 1.5939914894469444e-09,
+      "loss": 0.8989,
+      "step": 42212
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.456983615522293,
+      "learning_rate": 1.580397099468778e-09,
+      "loss": 0.876,
+      "step": 42213
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6897302656412625,
+      "learning_rate": 1.5668609234542608e-09,
+      "loss": 0.8896,
+      "step": 42214
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6028080666669793,
+      "learning_rate": 1.5533829614844397e-09,
+      "loss": 0.8234,
+      "step": 42215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.53510669948375,
+      "learning_rate": 1.5399632136348096e-09,
+      "loss": 0.8559,
+      "step": 42216
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7277982492932654,
+      "learning_rate": 1.526601679986417e-09,
+      "loss": 0.692,
+      "step": 42217
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5190545397473818,
+      "learning_rate": 1.5132983606147566e-09,
+      "loss": 0.793,
+      "step": 42218
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5206031621269702,
+      "learning_rate": 1.5000532555975444e-09,
+      "loss": 0.7581,
+      "step": 42219
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5663469522149425,
+      "learning_rate": 1.4868663650136062e-09,
+      "loss": 0.7551,
+      "step": 42220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5806638166641476,
+      "learning_rate": 1.473737688938437e-09,
+      "loss": 0.8491,
+      "step": 42221
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.545353836367418,
+      "learning_rate": 1.460667227447532e-09,
+      "loss": 0.8075,
+      "step": 42222
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6499616376109656,
+      "learning_rate": 1.447654980619717e-09,
+      "loss": 0.826,
+      "step": 42223
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7504109485767728,
+      "learning_rate": 1.434700948528267e-09,
+      "loss": 0.6704,
+      "step": 42224
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7989413337135927,
+      "learning_rate": 1.421805131248677e-09,
+      "loss": 0.8082,
+      "step": 42225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5662181118015797,
+      "learning_rate": 1.4089675288586623e-09,
+      "loss": 0.8097,
+      "step": 42226
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0637243217647288,
+      "learning_rate": 1.3961881414292776e-09,
+      "loss": 0.8247,
+      "step": 42227
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.429910254436577,
+      "learning_rate": 1.3834669690382385e-09,
+      "loss": 0.9412,
+      "step": 42228
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4858047580204126,
+      "learning_rate": 1.3708040117577093e-09,
+      "loss": 0.9545,
+      "step": 42229
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6708121284942288,
+      "learning_rate": 1.358199269662075e-09,
+      "loss": 0.9373,
+      "step": 42230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.700441639026598,
+      "learning_rate": 1.3456527428235e-09,
+      "loss": 0.8744,
+      "step": 42231
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6396660489094452,
+      "learning_rate": 1.3331644313174796e-09,
+      "loss": 0.7985,
+      "step": 42232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6038583153067674,
+      "learning_rate": 1.320734335213958e-09,
+      "loss": 0.9175,
+      "step": 42233
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7818891479886845,
+      "learning_rate": 1.3083624545873197e-09,
+      "loss": 0.9181,
+      "step": 42234
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5306121622686264,
+      "learning_rate": 1.2960487895097295e-09,
+      "loss": 0.9433,
+      "step": 42235
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7612041536232688,
+      "learning_rate": 1.2837933400511315e-09,
+      "loss": 0.7922,
+      "step": 42236
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8030333895482087,
+      "learning_rate": 1.2715961062836902e-09,
+      "loss": 0.7029,
+      "step": 42237
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6127173509332973,
+      "learning_rate": 1.2594570882784595e-09,
+      "loss": 0.847,
+      "step": 42238
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6884863380643635,
+      "learning_rate": 1.2473762861064943e-09,
+      "loss": 0.7256,
+      "step": 42239
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8991133747256839,
+      "learning_rate": 1.2353536998388483e-09,
+      "loss": 0.6644,
+      "step": 42240
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6504648766094798,
+      "learning_rate": 1.2233893295432453e-09,
+      "loss": 0.8854,
+      "step": 42241
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.616720699736026,
+      "learning_rate": 1.2114831752907397e-09,
+      "loss": 0.7648,
+      "step": 42242
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8463863935968141,
+      "learning_rate": 1.1996352371512755e-09,
+      "loss": 0.9085,
+      "step": 42243
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6466608750086302,
+      "learning_rate": 1.187845515192576e-09,
+      "loss": 0.7908,
+      "step": 42244
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8403234122938072,
+      "learning_rate": 1.1761140094834756e-09,
+      "loss": 0.8404,
+      "step": 42245
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.556225310680534,
+      "learning_rate": 1.1644407200939177e-09,
+      "loss": 0.9016,
+      "step": 42246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.583635362433815,
+      "learning_rate": 1.152825647089406e-09,
+      "loss": 0.91,
+      "step": 42247
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6118094523718134,
+      "learning_rate": 1.1412687905387743e-09,
+      "loss": 0.7034,
+      "step": 42248
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.528136374103904,
+      "learning_rate": 1.1297701505097459e-09,
+      "loss": 0.7141,
+      "step": 42249
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.872731341195446,
+      "learning_rate": 1.1183297270678239e-09,
+      "loss": 1.0002,
+      "step": 42250
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5375682743810994,
+      "learning_rate": 1.1069475202818425e-09,
+      "loss": 1.0606,
+      "step": 42251
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.687892947816652,
+      "learning_rate": 1.0956235302161944e-09,
+      "loss": 0.8855,
+      "step": 42252
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5825466991302086,
+      "learning_rate": 1.084357756937493e-09,
+      "loss": 0.9015,
+      "step": 42253
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5668107968014189,
+      "learning_rate": 1.0731502005112415e-09,
+      "loss": 0.8277,
+      "step": 42254
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7794203843938348,
+      "learning_rate": 1.062000861002943e-09,
+      "loss": 0.7416,
+      "step": 42255
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6195270888824105,
+      "learning_rate": 1.0509097384769906e-09,
+      "loss": 0.9355,
+      "step": 42256
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.407520417049264,
+      "learning_rate": 1.039876832997777e-09,
+      "loss": 0.7087,
+      "step": 42257
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6309065956679019,
+      "learning_rate": 1.0289021446308057e-09,
+      "loss": 1.0044,
+      "step": 42258
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.459135571482465,
+      "learning_rate": 1.017985673439359e-09,
+      "loss": 0.7196,
+      "step": 42259
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5447267997398826,
+      "learning_rate": 1.0071274194856096e-09,
+      "loss": 0.903,
+      "step": 42260
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6014100823077457,
+      "learning_rate": 9.963273828350606e-10,
+      "loss": 0.9738,
+      "step": 42261
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.720764188672103,
+      "learning_rate": 9.855855635498845e-10,
+      "loss": 0.8211,
+      "step": 42262
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5304025078359567,
+      "learning_rate": 9.749019616911437e-10,
+      "loss": 0.8706,
+      "step": 42263
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4149741674113137,
+      "learning_rate": 9.642765773232309e-10,
+      "loss": 0.8117,
+      "step": 42264
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8685324573485391,
+      "learning_rate": 9.537094105060985e-10,
+      "loss": 0.8382,
+      "step": 42265
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.658098783101975,
+      "learning_rate": 9.432004613019186e-10,
+      "loss": 0.8889,
+      "step": 42266
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4698905504818762,
+      "learning_rate": 9.327497297728638e-10,
+      "loss": 0.8617,
+      "step": 42267
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.561435612133822,
+      "learning_rate": 9.223572159788863e-10,
+      "loss": 0.8795,
+      "step": 42268
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4942406191978042,
+      "learning_rate": 9.12022919979938e-10,
+      "loss": 0.9209,
+      "step": 42269
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7358931358281551,
+      "learning_rate": 9.017468418370812e-10,
+      "loss": 0.6953,
+      "step": 42270
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6394580296837746,
+      "learning_rate": 8.915289816102679e-10,
+      "loss": 0.7341,
+      "step": 42271
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.641809558151472,
+      "learning_rate": 8.813693393583401e-10,
+      "loss": 0.7807,
+      "step": 42272
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7458863994980336,
+      "learning_rate": 8.712679151412495e-10,
+      "loss": 0.6914,
+      "step": 42273
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.581457239489462,
+      "learning_rate": 8.612247090167281e-10,
+      "loss": 0.7888,
+      "step": 42274
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6622614920500856,
+      "learning_rate": 8.512397210447276e-10,
+      "loss": 0.889,
+      "step": 42275
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6166720942146156,
+      "learning_rate": 8.413129512818697e-10,
+      "loss": 0.7681,
+      "step": 42276
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.589483607326514,
+      "learning_rate": 8.314443997858857e-10,
+      "loss": 0.8276,
+      "step": 42277
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7309164032048299,
+      "learning_rate": 8.216340666156175e-10,
+      "loss": 0.7062,
+      "step": 42278
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6242137097915939,
+      "learning_rate": 8.118819518276866e-10,
+      "loss": 0.808,
+      "step": 42279
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.050286579262783,
+      "learning_rate": 8.021880554787143e-10,
+      "loss": 0.8284,
+      "step": 42280
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5874953932421332,
+      "learning_rate": 7.925523776242116e-10,
+      "loss": 0.9091,
+      "step": 42281
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.617006544499708,
+      "learning_rate": 7.829749183219104e-10,
+      "loss": 0.8805,
+      "step": 42282
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6737721821327367,
+      "learning_rate": 7.734556776273217e-10,
+      "loss": 0.8895,
+      "step": 42283
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2077564683950848,
+      "learning_rate": 7.639946555937361e-10,
+      "loss": 0.8736,
+      "step": 42284
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7853933281167411,
+      "learning_rate": 7.545918522788853e-10,
+      "loss": 0.8181,
+      "step": 42285
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.556752577268482,
+      "learning_rate": 7.452472677371702e-10,
+      "loss": 0.9818,
+      "step": 42286
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6634121787367653,
+      "learning_rate": 7.359609020218816e-10,
+      "loss": 0.8782,
+      "step": 42287
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8490903030908792,
+      "learning_rate": 7.267327551863101e-10,
+      "loss": 0.7448,
+      "step": 42288
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6602348723271865,
+      "learning_rate": 7.175628272870772e-10,
+      "loss": 0.8001,
+      "step": 42289
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.576991422468891,
+      "learning_rate": 7.08451118375253e-10,
+      "loss": 0.7275,
+      "step": 42290
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7846716063582404,
+      "learning_rate": 6.993976285041282e-10,
+      "loss": 0.7641,
+      "step": 42291
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7453768372952588,
+      "learning_rate": 6.904023577269936e-10,
+      "loss": 0.78,
+      "step": 42292
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6320974001860407,
+      "learning_rate": 6.814653060971399e-10,
+      "loss": 0.8417,
+      "step": 42293
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3556884107485043,
+      "learning_rate": 6.725864736645271e-10,
+      "loss": 0.8855,
+      "step": 42294
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7324159694274599,
+      "learning_rate": 6.637658604813357e-10,
+      "loss": 0.6824,
+      "step": 42295
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.599218536294622,
+      "learning_rate": 6.550034666008564e-10,
+      "loss": 0.8722,
+      "step": 42296
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6284112338758971,
+      "learning_rate": 6.462992920719391e-10,
+      "loss": 0.6646,
+      "step": 42297
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8588420743819274,
+      "learning_rate": 6.376533369456539e-10,
+      "loss": 0.8939,
+      "step": 42298
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7689036483060439,
+      "learning_rate": 6.290656012730711e-10,
+      "loss": 0.7667,
+      "step": 42299
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4517499061216115,
+      "learning_rate": 6.205360851041509e-10,
+      "loss": 0.8854,
+      "step": 42300
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6171154175535973,
+      "learning_rate": 6.120647884877429e-10,
+      "loss": 0.9578,
+      "step": 42301
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5736671013010863,
+      "learning_rate": 6.036517114738072e-10,
+      "loss": 0.8503,
+      "step": 42302
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7601360942452673,
+      "learning_rate": 5.952968541111937e-10,
+      "loss": 0.6602,
+      "step": 42303
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1093186263050536,
+      "learning_rate": 5.87000216448752e-10,
+      "loss": 0.9219,
+      "step": 42304
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.307600342499896,
+      "learning_rate": 5.78761798534222e-10,
+      "loss": 0.8337,
+      "step": 42305
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5907129211369107,
+      "learning_rate": 5.70581600415343e-10,
+      "loss": 0.7965,
+      "step": 42306
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.690893779418313,
+      "learning_rate": 5.62459622140965e-10,
+      "loss": 0.9454,
+      "step": 42307
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7859540227926063,
+      "learning_rate": 5.543958637577174e-10,
+      "loss": 0.7016,
+      "step": 42308
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7335161054156536,
+      "learning_rate": 5.463903253122294e-10,
+      "loss": 0.8584,
+      "step": 42309
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.033168311924957,
+      "learning_rate": 5.384430068522406e-10,
+      "loss": 0.7658,
+      "step": 42310
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9727664325118242,
+      "learning_rate": 5.305539084221601e-10,
+      "loss": 0.8142,
+      "step": 42311
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5543304317599136,
+      "learning_rate": 5.227230300697273e-10,
+      "loss": 0.7846,
+      "step": 42312
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7676741904556148,
+      "learning_rate": 5.149503718393511e-10,
+      "loss": 0.8554,
+      "step": 42313
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4206666192803434,
+      "learning_rate": 5.072359337776611e-10,
+      "loss": 0.8317,
+      "step": 42314
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9850354264567294,
+      "learning_rate": 4.995797159279558e-10,
+      "loss": 0.8215,
+      "step": 42315
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5245094062735893,
+      "learning_rate": 4.919817183346443e-10,
+      "loss": 0.7401,
+      "step": 42316
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.904510801935316,
+      "learning_rate": 4.844419410443557e-10,
+      "loss": 0.8634,
+      "step": 42317
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6721150500826345,
+      "learning_rate": 4.769603840981684e-10,
+      "loss": 0.8873,
+      "step": 42318
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.887781265559562,
+      "learning_rate": 4.695370475416016e-10,
+      "loss": 0.7447,
+      "step": 42319
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.570040814057697,
+      "learning_rate": 4.621719314168438e-10,
+      "loss": 0.8307,
+      "step": 42320
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7431849610357777,
+      "learning_rate": 4.5486503576719354e-10,
+      "loss": 0.6606,
+      "step": 42321
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5000086896801124,
+      "learning_rate": 4.476163606348394e-10,
+      "loss": 0.8107,
+      "step": 42322
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7598010768700281,
+      "learning_rate": 4.4042590606308e-10,
+      "loss": 0.6639,
+      "step": 42323
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5281419894126889,
+      "learning_rate": 4.3329367209188344e-10,
+      "loss": 0.7378,
+      "step": 42324
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.530102193663849,
+      "learning_rate": 4.262196587645484e-10,
+      "loss": 0.8893,
+      "step": 42325
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8009162759554269,
+      "learning_rate": 4.1920386612215313e-10,
+      "loss": 0.6857,
+      "step": 42326
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5827992309832766,
+      "learning_rate": 4.1224629420466567e-10,
+      "loss": 0.812,
+      "step": 42327
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6130437045563712,
+      "learning_rate": 4.05346943052054e-10,
+      "loss": 0.7674,
+      "step": 42328
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7572872830577089,
+      "learning_rate": 3.9850581270539645e-10,
+      "loss": 0.8096,
+      "step": 42329
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7456067495461538,
+      "learning_rate": 3.917229032057712e-10,
+      "loss": 0.6819,
+      "step": 42330
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6103768962266374,
+      "learning_rate": 3.849982145898157e-10,
+      "loss": 0.7954,
+      "step": 42331
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.678809575434849,
+      "learning_rate": 3.7833174689860806e-10,
+      "loss": 0.7353,
+      "step": 42332
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5639480268501431,
+      "learning_rate": 3.717235001710062e-10,
+      "loss": 0.9178,
+      "step": 42333
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6673638802928474,
+      "learning_rate": 3.651734744447577e-10,
+      "loss": 0.9044,
+      "step": 42334
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6662385252412726,
+      "learning_rate": 3.586816697587203e-10,
+      "loss": 0.7697,
+      "step": 42335
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5016297580898017,
+      "learning_rate": 3.522480861506417e-10,
+      "loss": 0.8383,
+      "step": 42336
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.754537462342581,
+      "learning_rate": 3.458727236571591e-10,
+      "loss": 0.7803,
+      "step": 42337
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4129696153602096,
+      "learning_rate": 3.3955558231490994e-10,
+      "loss": 0.7087,
+      "step": 42338
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.54263393277406,
+      "learning_rate": 3.3329666216275205e-10,
+      "loss": 0.8123,
+      "step": 42339
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.929734894612325,
+      "learning_rate": 3.2709596323510227e-10,
+      "loss": 0.7339,
+      "step": 42340
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.903380903158645,
+      "learning_rate": 3.2095348556970827e-10,
+      "loss": 0.8473,
+      "step": 42341
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5583820869473068,
+      "learning_rate": 3.1486922920098696e-10,
+      "loss": 0.9118,
+      "step": 42342
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.76130199748163,
+      "learning_rate": 3.088431941655756e-10,
+      "loss": 0.8269,
+      "step": 42343
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7428066662309443,
+      "learning_rate": 3.028753804978912e-10,
+      "loss": 0.8805,
+      "step": 42344
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9835343917981951,
+      "learning_rate": 2.9696578823235067e-10,
+      "loss": 0.6295,
+      "step": 42345
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4699856380268777,
+      "learning_rate": 2.9111441740337085e-10,
+      "loss": 0.8557,
+      "step": 42346
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.754167816381501,
+      "learning_rate": 2.8532126804647895e-10,
+      "loss": 0.8181,
+      "step": 42347
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7562666280145057,
+      "learning_rate": 2.795863401927612e-10,
+      "loss": 0.6922,
+      "step": 42348
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6570736136817459,
+      "learning_rate": 2.739096338788549e-10,
+      "loss": 0.7897,
+      "step": 42349
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.557820296759749,
+      "learning_rate": 2.682911491347362e-10,
+      "loss": 0.9207,
+      "step": 42350
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7151487304376793,
+      "learning_rate": 2.627308859948219e-10,
+      "loss": 0.8413,
+      "step": 42351
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.620145260074418,
+      "learning_rate": 2.5722884449130845e-10,
+      "loss": 0.9372,
+      "step": 42352
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.672510711756633,
+      "learning_rate": 2.517850246563924e-10,
+      "loss": 0.6785,
+      "step": 42353
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5487065206626083,
+      "learning_rate": 2.4639942652116e-10,
+      "loss": 0.894,
+      "step": 42354
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2441762691374767,
+      "learning_rate": 2.4107205011780764e-10,
+      "loss": 0.9329,
+      "step": 42355
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8377528113923243,
+      "learning_rate": 2.358028954763114e-10,
+      "loss": 0.9403,
+      "step": 42356
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.368813522864278,
+      "learning_rate": 2.305919626277575e-10,
+      "loss": 0.8236,
+      "step": 42357
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7267870910597,
+      "learning_rate": 2.25439251602122e-10,
+      "loss": 0.6819,
+      "step": 42358
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4021410201983482,
+      "learning_rate": 2.203447624304911e-10,
+      "loss": 0.8096,
+      "step": 42359
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6283404559527797,
+      "learning_rate": 2.1530849514173057e-10,
+      "loss": 0.8929,
+      "step": 42360
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8590806005131697,
+      "learning_rate": 2.103304497658165e-10,
+      "loss": 0.7862,
+      "step": 42361
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.491821665233555,
+      "learning_rate": 2.0541062633050446e-10,
+      "loss": 0.7952,
+      "step": 42362
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4912443547604637,
+      "learning_rate": 2.0054902486466022e-10,
+      "loss": 0.8533,
+      "step": 42363
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9175440128812808,
+      "learning_rate": 1.957456453982598e-10,
+      "loss": 0.7785,
+      "step": 42364
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.656828818899101,
+      "learning_rate": 1.9100048795794858e-10,
+      "loss": 0.6997,
+      "step": 42365
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3646361269618206,
+      "learning_rate": 1.8631355257148209e-10,
+      "loss": 0.5688,
+      "step": 42366
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.554456251505699,
+      "learning_rate": 1.8168483926550574e-10,
+      "loss": 0.7508,
+      "step": 42367
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6163076357423498,
+      "learning_rate": 1.7711434806888527e-10,
+      "loss": 0.8134,
+      "step": 42368
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6567461972478963,
+      "learning_rate": 1.7260207900604565e-10,
+      "loss": 0.8399,
+      "step": 42369
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.618767125981532,
+      "learning_rate": 1.6814803210363217e-10,
+      "loss": 0.793,
+      "step": 42370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4749044005377998,
+      "learning_rate": 1.6375220738940046e-10,
+      "loss": 0.8276,
+      "step": 42371
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5494425664509803,
+      "learning_rate": 1.5941460488777538e-10,
+      "loss": 0.8955,
+      "step": 42372
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4794741559031412,
+      "learning_rate": 1.5513522462318188e-10,
+      "loss": 0.8381,
+      "step": 42373
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.517597682791875,
+      "learning_rate": 1.5091406662115505e-10,
+      "loss": 0.8073,
+      "step": 42374
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.737958664827841,
+      "learning_rate": 1.4675113090723004e-10,
+      "loss": 0.8475,
+      "step": 42375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.743797678119651,
+      "learning_rate": 1.426464175036113e-10,
+      "loss": 0.791,
+      "step": 42376
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4951157662146957,
+      "learning_rate": 1.385999264369442e-10,
+      "loss": 0.7639,
+      "step": 42377
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.655906742083098,
+      "learning_rate": 1.3461165772832297e-10,
+      "loss": 0.9133,
+      "step": 42378
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.711481373369709,
+      "learning_rate": 1.3068161140217251e-10,
+      "loss": 0.746,
+      "step": 42379
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6205998008034985,
+      "learning_rate": 1.2680978748069728e-10,
+      "loss": 0.7897,
+      "step": 42380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4254557830862515,
+      "learning_rate": 1.2299618598721197e-10,
+      "loss": 0.9142,
+      "step": 42381
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8591513663749548,
+      "learning_rate": 1.1924080694281083e-10,
+      "loss": 0.8627,
+      "step": 42382
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4162508723703597,
+      "learning_rate": 1.1554365037080851e-10,
+      "loss": 0.8123,
+      "step": 42383
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4835954076530324,
+      "learning_rate": 1.1190471629229927e-10,
+      "loss": 0.6787,
+      "step": 42384
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0974411100014017,
+      "learning_rate": 1.0832400472726711e-10,
+      "loss": 0.8004,
+      "step": 42385
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5188324452325102,
+      "learning_rate": 1.0480151569791652e-10,
+      "loss": 0.8004,
+      "step": 42386
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5881754973867008,
+      "learning_rate": 1.0133724922534171e-10,
+      "loss": 0.7761,
+      "step": 42387
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.530300597311648,
+      "learning_rate": 9.793120532730627e-11,
+      "loss": 0.8445,
+      "step": 42388
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.752864017170777,
+      "learning_rate": 9.458338402601464e-11,
+      "loss": 0.7353,
+      "step": 42389
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.972268878588096,
+      "learning_rate": 9.129378533923038e-11,
+      "loss": 0.7017,
+      "step": 42390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5670237928647317,
+      "learning_rate": 8.806240928693755e-11,
+      "loss": 0.7622,
+      "step": 42391
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.523320874941075,
+      "learning_rate": 8.488925588800989e-11,
+      "loss": 0.7929,
+      "step": 42392
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1553693073998654,
+      "learning_rate": 8.177432516132122e-11,
+      "loss": 0.8052,
+      "step": 42393
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5233110572155484,
+      "learning_rate": 7.871761712352489e-11,
+      "loss": 0.8672,
+      "step": 42394
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6344758478578412,
+      "learning_rate": 7.571913179349466e-11,
+      "loss": 0.9136,
+      "step": 42395
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5261513944129594,
+      "learning_rate": 7.277886918899413e-11,
+      "loss": 0.8005,
+      "step": 42396
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2657167507705034,
+      "learning_rate": 6.989682932667663e-11,
+      "loss": 0.7965,
+      "step": 42397
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.506437670504428,
+      "learning_rate": 6.707301222319551e-11,
+      "loss": 0.7731,
+      "step": 42398
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6247860403611012,
+      "learning_rate": 6.430741789520412e-11,
+      "loss": 0.8817,
+      "step": 42399
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8173171461774766,
+      "learning_rate": 6.160004635824556e-11,
+      "loss": 0.7117,
+      "step": 42400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6906380678175226,
+      "learning_rate": 5.895089762897322e-11,
+      "loss": 0.8592,
+      "step": 42401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3813259537312768,
+      "learning_rate": 5.6359971722930176e-11,
+      "loss": 0.7661,
+      "step": 42402
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.532038578128869,
+      "learning_rate": 5.3827268654549346e-11,
+      "loss": 0.7459,
+      "step": 42403
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.457747653819832,
+      "learning_rate": 5.135278843826363e-11,
+      "loss": 0.8061,
+      "step": 42404
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7749156422696826,
+      "learning_rate": 4.8936531088505936e-11,
+      "loss": 0.7089,
+      "step": 42405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0874771757125266,
+      "learning_rate": 4.657849662081937e-11,
+      "loss": 0.8608,
+      "step": 42406
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6188922329249824,
+      "learning_rate": 4.4278685047416395e-11,
+      "loss": 0.7246,
+      "step": 42407
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6848098459154524,
+      "learning_rate": 4.203709638161968e-11,
+      "loss": 0.8646,
+      "step": 42408
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1026916178637434,
+      "learning_rate": 3.985373063786213e-11,
+      "loss": 0.7923,
+      "step": 42409
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6494247039239345,
+      "learning_rate": 3.772858782724598e-11,
+      "loss": 0.7838,
+      "step": 42410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6365596778509703,
+      "learning_rate": 3.566166796420412e-11,
+      "loss": 0.8066,
+      "step": 42411
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5408723679261327,
+      "learning_rate": 3.3652971057618336e-11,
+      "loss": 0.7382,
+      "step": 42412
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7499157608675954,
+      "learning_rate": 3.170249712303175e-11,
+      "loss": 0.706,
+      "step": 42413
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4740685944299978,
+      "learning_rate": 2.9810246168215926e-11,
+      "loss": 0.7186,
+      "step": 42414
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6854286601845205,
+      "learning_rate": 2.7976218206493543e-11,
+      "loss": 0.8248,
+      "step": 42415
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7837237791919032,
+      "learning_rate": 2.6200413247856605e-11,
+      "loss": 0.9775,
+      "step": 42416
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4275282060553205,
+      "learning_rate": 2.4482831302297118e-11,
+      "loss": 0.9373,
+      "step": 42417
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5726815320551197,
+      "learning_rate": 2.2823472379807086e-11,
+      "loss": 0.9319,
+      "step": 42418
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8312925336539139,
+      "learning_rate": 2.1222336491488748e-11,
+      "loss": 1.0068,
+      "step": 42419
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6000510701129718,
+      "learning_rate": 1.967942364511366e-11,
+      "loss": 0.8118,
+      "step": 42420
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.536595500207733,
+      "learning_rate": 1.8194733849563605e-11,
+      "loss": 0.7569,
+      "step": 42421
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5529557373756895,
+      "learning_rate": 1.6768267114830594e-11,
+      "loss": 0.8572,
+      "step": 42422
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.592301714447548,
+      "learning_rate": 1.5400023447575962e-11,
+      "loss": 0.8736,
+      "step": 42423
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5182829705238743,
+      "learning_rate": 1.4090002856681495e-11,
+      "loss": 0.8393,
+      "step": 42424
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6652589112893745,
+      "learning_rate": 1.2838205351028976e-11,
+      "loss": 0.8292,
+      "step": 42425
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6035177651211294,
+      "learning_rate": 1.1644630935059298e-11,
+      "loss": 0.8648,
+      "step": 42426
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4846239784637403,
+      "learning_rate": 1.0509279617654245e-11,
+      "loss": 0.7653,
+      "step": 42427
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7372140680678336,
+      "learning_rate": 9.432151405475153e-12,
+      "loss": 0.8524,
+      "step": 42428
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.524340191444371,
+      "learning_rate": 8.413246304073142e-12,
+      "loss": 0.7568,
+      "step": 42429
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7091729237564082,
+      "learning_rate": 7.452564318999322e-12,
+      "loss": 0.8424,
+      "step": 42430
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4610877225365777,
+      "learning_rate": 6.550105456915035e-12,
+      "loss": 0.7819,
+      "step": 42431
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0087337206794924,
+      "learning_rate": 5.705869723371394e-12,
+      "loss": 0.8306,
+      "step": 42432
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6080479046297582,
+      "learning_rate": 4.919857121699068e-12,
+      "loss": 0.8515,
+      "step": 42433
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6577693568579919,
+      "learning_rate": 4.192067657449173e-12,
+      "loss": 0.8123,
+      "step": 42434
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.074026779924166,
+      "learning_rate": 3.5225013350626004e-12,
+      "loss": 0.785,
+      "step": 42435
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7912241012690377,
+      "learning_rate": 2.9111581578700197e-12,
+      "loss": 0.8867,
+      "step": 42436
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6805713546674979,
+      "learning_rate": 2.3580381292021006e-12,
+      "loss": 0.7668,
+      "step": 42437
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7593399916820958,
+      "learning_rate": 1.863141253499734e-12,
+      "loss": 0.8347,
+      "step": 42438
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3698041207772478,
+      "learning_rate": 1.4264675329833665e-12,
+      "loss": 0.8177,
+      "step": 42439
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.731730163913853,
+      "learning_rate": 1.0480169687632214e-12,
+      "loss": 0.6858,
+      "step": 42440
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.453824357005901,
+      "learning_rate": 7.277895663904133e-13,
+      "loss": 0.8498,
+      "step": 42441
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6918765697035005,
+      "learning_rate": 4.657853236444964e-13,
+      "loss": 0.7151,
+      "step": 42442
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4599947086625775,
+      "learning_rate": 2.620042460765859e-13,
+      "loss": 0.8809,
+      "step": 42443
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6481103478531685,
+      "learning_rate": 1.1644633146623564e-13,
+      "loss": 0.8487,
+      "step": 42444
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6794883078153966,
+      "learning_rate": 2.9111583144114664e-14,
+      "loss": 0.7506,
+      "step": 42445
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3701915655778307,
+      "learning_rate": 0.0,
+      "loss": 0.6517,
+      "step": 42446
+    },
+    {
+      "epoch": 1.0,
+      "step": 42446,
+      "total_flos": 1.6390346793348497e+19,
+      "train_loss": 0.8575404643863601,
+      "train_runtime": 148237.3426,
+      "train_samples_per_second": 36.651,
+      "train_steps_per_second": 0.286
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 42446,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 10000,
+  "total_flos": 1.6390346793348497e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}