diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,15764 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 382,
+  "global_step": 2242,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 115.0,
+      "learning_rate": 2.9411764705882356e-07,
+      "loss": 4.0725,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 112.0,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 4.3488,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 111.0,
+      "learning_rate": 8.823529411764707e-07,
+      "loss": 4.4319,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 107.0,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 4.2862,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 112.5,
+      "learning_rate": 1.4705882352941177e-06,
+      "loss": 4.0086,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 102.5,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 4.1965,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 98.5,
+      "learning_rate": 2.058823529411765e-06,
+      "loss": 3.7466,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 93.5,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 3.6961,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 103.0,
+      "learning_rate": 2.647058823529412e-06,
+      "loss": 4.0201,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 89.0,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 3.6967,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 86.5,
+      "learning_rate": 3.2352941176470594e-06,
+      "loss": 3.4749,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 90.0,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 3.6254,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 71.5,
+      "learning_rate": 3.8235294117647055e-06,
+      "loss": 3.0627,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 74.0,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 3.3296,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 75.5,
+      "learning_rate": 4.411764705882353e-06,
+      "loss": 3.2013,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 63.75,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 3.0594,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 65.5,
+      "learning_rate": 5e-06,
+      "loss": 3.0923,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 61.5,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 2.8753,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 53.25,
+      "learning_rate": 5.588235294117647e-06,
+      "loss": 2.6746,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 48.25,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 2.4675,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 44.25,
+      "learning_rate": 6.176470588235295e-06,
+      "loss": 2.3024,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 39.75,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 2.0897,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 32.25,
+      "learning_rate": 6.764705882352942e-06,
+      "loss": 1.9655,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 32.25,
+      "learning_rate": 7.058823529411766e-06,
+      "loss": 1.7342,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 35.5,
+      "learning_rate": 7.352941176470589e-06,
+      "loss": 1.9268,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 32.75,
+      "learning_rate": 7.647058823529411e-06,
+      "loss": 1.7544,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 27.5,
+      "learning_rate": 7.941176470588236e-06,
+      "loss": 1.5619,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 26.625,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 1.7214,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.375,
+      "learning_rate": 8.529411764705883e-06,
+      "loss": 1.5131,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 23.5,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 1.2882,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.375,
+      "learning_rate": 9.11764705882353e-06,
+      "loss": 1.1836,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.5,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 1.1409,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.875,
+      "learning_rate": 9.705882352941177e-06,
+      "loss": 1.0394,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.625,
+      "learning_rate": 1e-05,
+      "loss": 1.102,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.5,
+      "learning_rate": 1.0294117647058823e-05,
+      "loss": 0.8235,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.0588235294117648e-05,
+      "loss": 0.7823,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.0882352941176471e-05,
+      "loss": 0.8871,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.1176470588235295e-05,
+      "loss": 0.7424,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.25,
+      "learning_rate": 1.1470588235294118e-05,
+      "loss": 0.7749,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.0,
+      "learning_rate": 1.1764705882352942e-05,
+      "loss": 0.7763,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.2058823529411765e-05,
+      "loss": 0.6946,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.235294117647059e-05,
+      "loss": 0.5177,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.2647058823529412e-05,
+      "loss": 0.6325,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2941176470588238e-05,
+      "loss": 0.6239,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.59375,
+      "learning_rate": 1.323529411764706e-05,
+      "loss": 0.5248,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.3529411764705885e-05,
+      "loss": 0.5823,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.3823529411764706e-05,
+      "loss": 0.5379,
+      "step": 47
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.4117647058823532e-05,
+      "loss": 0.5101,
+      "step": 48
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.4411764705882353e-05,
+      "loss": 0.4993,
+      "step": 49
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.96875,
+      "learning_rate": 1.4705882352941179e-05,
+      "loss": 0.4116,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.375,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.4443,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.5,
+      "learning_rate": 1.5294117647058822e-05,
+      "loss": 0.3809,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.40625,
+      "learning_rate": 1.558823529411765e-05,
+      "loss": 0.3316,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.375,
+      "learning_rate": 1.5882352941176473e-05,
+      "loss": 0.3766,
+      "step": 54
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.6176470588235296e-05,
+      "loss": 0.3426,
+      "step": 55
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.34375,
+      "learning_rate": 1.647058823529412e-05,
+      "loss": 0.3475,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.6764705882352943e-05,
+      "loss": 0.3355,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.7058823529411767e-05,
+      "loss": 0.2909,
+      "step": 58
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.735294117647059e-05,
+      "loss": 0.3247,
+      "step": 59
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 0.3539,
+      "step": 60
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.7941176470588237e-05,
+      "loss": 0.3005,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.823529411764706e-05,
+      "loss": 0.2696,
+      "step": 62
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.8529411764705884e-05,
+      "loss": 0.3055,
+      "step": 63
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.8823529411764708e-05,
+      "loss": 0.2734,
+      "step": 64
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.911764705882353e-05,
+      "loss": 0.2543,
+      "step": 65
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.9411764705882355e-05,
+      "loss": 0.2311,
+      "step": 66
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.9705882352941178e-05,
+      "loss": 0.264,
+      "step": 67
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.40625,
+      "learning_rate": 2e-05,
+      "loss": 0.2379,
+      "step": 68
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9999989558795673e-05,
+      "loss": 0.2357,
+      "step": 69
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.75,
+      "learning_rate": 1.99999582352045e-05,
+      "loss": 0.2628,
+      "step": 70
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.9999906029291888e-05,
+      "loss": 0.2319,
+      "step": 71
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.9999832941166858e-05,
+      "loss": 0.2541,
+      "step": 72
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.999973897098203e-05,
+      "loss": 0.2577,
+      "step": 73
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.999962411893365e-05,
+      "loss": 0.2316,
+      "step": 74
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.999948838526154e-05,
+      "loss": 0.2463,
+      "step": 75
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.9999331770249155e-05,
+      "loss": 0.2296,
+      "step": 76
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.9999154274223543e-05,
+      "loss": 0.2696,
+      "step": 77
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9998955897555355e-05,
+      "loss": 0.1864,
+      "step": 78
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.9998736640658856e-05,
+      "loss": 0.2339,
+      "step": 79
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9998496503991902e-05,
+      "loss": 0.2223,
+      "step": 80
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.9998235488055956e-05,
+      "loss": 0.24,
+      "step": 81
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.9997953593396086e-05,
+      "loss": 0.1875,
+      "step": 82
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9997650820600952e-05,
+      "loss": 0.1856,
+      "step": 83
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9997327170302818e-05,
+      "loss": 0.1628,
+      "step": 84
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9996982643177544e-05,
+      "loss": 0.2195,
+      "step": 85
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9996617239944584e-05,
+      "loss": 0.2017,
+      "step": 86
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9996230961366995e-05,
+      "loss": 0.1916,
+      "step": 87
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.9995823808251407e-05,
+      "loss": 0.2233,
+      "step": 88
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9995395781448065e-05,
+      "loss": 0.1969,
+      "step": 89
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.875,
+      "learning_rate": 1.999494688185079e-05,
+      "loss": 0.1936,
+      "step": 90
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.9994477110396982e-05,
+      "loss": 0.1794,
+      "step": 91
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.9993986468067652e-05,
+      "loss": 0.1885,
+      "step": 92
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.999347495588737e-05,
+      "loss": 0.1571,
+      "step": 93
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9992942574924302e-05,
+      "loss": 0.146,
+      "step": 94
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9992389326290185e-05,
+      "loss": 0.1623,
+      "step": 95
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9991815211140335e-05,
+      "loss": 0.1853,
+      "step": 96
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9991220230673642e-05,
+      "loss": 0.1668,
+      "step": 97
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9990604386132572e-05,
+      "loss": 0.1408,
+      "step": 98
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.998996767880315e-05,
+      "loss": 0.1515,
+      "step": 99
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9989310110014983e-05,
+      "loss": 0.1509,
+      "step": 100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9988631681141228e-05,
+      "loss": 0.1708,
+      "step": 101
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.998793239359861e-05,
+      "loss": 0.1496,
+      "step": 102
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.9987212248847404e-05,
+      "loss": 0.119,
+      "step": 103
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9986471248391456e-05,
+      "loss": 0.1377,
+      "step": 104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.998570939377814e-05,
+      "loss": 0.1503,
+      "step": 105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.375,
+      "learning_rate": 1.9984926686598405e-05,
+      "loss": 0.1602,
+      "step": 106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9984123128486726e-05,
+      "loss": 0.1448,
+      "step": 107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.998329872112113e-05,
+      "loss": 0.1522,
+      "step": 108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.998245346622317e-05,
+      "loss": 0.1758,
+      "step": 109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.9981587365557946e-05,
+      "loss": 0.1299,
+      "step": 110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9980700420934087e-05,
+      "loss": 0.1531,
+      "step": 111
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9979792634203745e-05,
+      "loss": 0.128,
+      "step": 112
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.9978864007262597e-05,
+      "loss": 0.104,
+      "step": 113
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9977914542049843e-05,
+      "loss": 0.1078,
+      "step": 114
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.997694424054819e-05,
+      "loss": 0.1255,
+      "step": 115
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.9975953104783865e-05,
+      "loss": 0.135,
+      "step": 116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.9974941136826595e-05,
+      "loss": 0.0979,
+      "step": 117
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.997390833878962e-05,
+      "loss": 0.1054,
+      "step": 118
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9972854712829658e-05,
+      "loss": 0.089,
+      "step": 119
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.997178026114695e-05,
+      "loss": 0.115,
+      "step": 120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9970684985985193e-05,
+      "loss": 0.121,
+      "step": 121
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.9969568889631598e-05,
+      "loss": 0.1117,
+      "step": 122
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.996843197441684e-05,
+      "loss": 0.0928,
+      "step": 123
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.996727424271507e-05,
+      "loss": 0.0935,
+      "step": 124
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.996609569694391e-05,
+      "loss": 0.1038,
+      "step": 125
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.996489633956445e-05,
+      "loss": 0.1019,
+      "step": 126
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.9963676173081236e-05,
+      "loss": 0.0908,
+      "step": 127
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.9962435200042266e-05,
+      "loss": 0.0894,
+      "step": 128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9961173423039e-05,
+      "loss": 0.1005,
+      "step": 129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.995989084470632e-05,
+      "loss": 0.1123,
+      "step": 130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.995858746772257e-05,
+      "loss": 0.1083,
+      "step": 131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.75,
+      "learning_rate": 1.995726329480951e-05,
+      "loss": 0.0965,
+      "step": 132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.995591832873233e-05,
+      "loss": 0.1124,
+      "step": 133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9954552572299643e-05,
+      "loss": 0.0974,
+      "step": 134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.9953166028363482e-05,
+      "loss": 0.1108,
+      "step": 135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.995175869981928e-05,
+      "loss": 0.1039,
+      "step": 136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.9950330589605885e-05,
+      "loss": 0.1037,
+      "step": 137
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.9948881700705525e-05,
+      "loss": 0.0834,
+      "step": 138
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9947412036143835e-05,
+      "loss": 0.0822,
+      "step": 139
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0,
+      "learning_rate": 1.994592159898983e-05,
+      "loss": 0.0833,
+      "step": 140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.9944410392355896e-05,
+      "loss": 0.1039,
+      "step": 141
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9942878419397805e-05,
+      "loss": 0.0593,
+      "step": 142
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.9941325683314675e-05,
+      "loss": 0.0899,
+      "step": 143
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.125,
+      "learning_rate": 1.9939752187349004e-05,
+      "loss": 0.1014,
+      "step": 144
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9938157934786622e-05,
+      "loss": 0.0908,
+      "step": 145
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.9936542928956713e-05,
+      "loss": 0.0956,
+      "step": 146
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.9934907173231804e-05,
+      "loss": 0.0862,
+      "step": 147
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.9933250671027745e-05,
+      "loss": 0.0763,
+      "step": 148
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9931573425803707e-05,
+      "loss": 0.0797,
+      "step": 149
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.9929875441062187e-05,
+      "loss": 0.1047,
+      "step": 150
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.9928156720348985e-05,
+      "loss": 0.0759,
+      "step": 151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.99264172672532e-05,
+      "loss": 0.1031,
+      "step": 152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.992465708540723e-05,
+      "loss": 0.0811,
+      "step": 153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.992287617848677e-05,
+      "loss": 0.0798,
+      "step": 154
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0,
+      "learning_rate": 1.9921074550210766e-05,
+      "loss": 0.0917,
+      "step": 155
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.991925220434146e-05,
+      "loss": 0.0789,
+      "step": 156
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.991740914468435e-05,
+      "loss": 0.0831,
+      "step": 157
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.9915545375088187e-05,
+      "loss": 0.0691,
+      "step": 158
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.991366089944497e-05,
+      "loss": 0.0793,
+      "step": 159
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.9911755721689938e-05,
+      "loss": 0.0858,
+      "step": 160
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.9909829845801564e-05,
+      "loss": 0.0868,
+      "step": 161
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.9907883275801537e-05,
+      "loss": 0.0915,
+      "step": 162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.9905916015754764e-05,
+      "loss": 0.085,
+      "step": 163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.990392806976936e-05,
+      "loss": 0.0624,
+      "step": 164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.9901919441996634e-05,
+      "loss": 0.0767,
+      "step": 165
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.9899890136631088e-05,
+      "loss": 0.0786,
+      "step": 166
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.9897840157910395e-05,
+      "loss": 0.0778,
+      "step": 167
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9895769510115406e-05,
+      "loss": 0.0576,
+      "step": 168
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.9893678197570136e-05,
+      "loss": 0.0751,
+      "step": 169
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.125,
+      "learning_rate": 1.989156622464174e-05,
+      "loss": 0.0998,
+      "step": 170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.9889433595740542e-05,
+      "loss": 0.0803,
+      "step": 171
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.988728031531997e-05,
+      "loss": 0.0601,
+      "step": 172
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9885106387876597e-05,
+      "loss": 0.0779,
+      "step": 173
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.9882911817950107e-05,
+      "loss": 0.083,
+      "step": 174
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.9880696610123294e-05,
+      "loss": 0.0861,
+      "step": 175
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.987846076902204e-05,
+      "loss": 0.0623,
+      "step": 176
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.9876204299315323e-05,
+      "loss": 0.0901,
+      "step": 177
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.98739272057152e-05,
+      "loss": 0.0915,
+      "step": 178
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.9871629492976778e-05,
+      "loss": 0.0728,
+      "step": 179
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9869311165898246e-05,
+      "loss": 0.0609,
+      "step": 180
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.9866972229320826e-05,
+      "loss": 0.0656,
+      "step": 181
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0,
+      "learning_rate": 1.986461268812878e-05,
+      "loss": 0.0964,
+      "step": 182
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.98622325472494e-05,
+      "loss": 0.0728,
+      "step": 183
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.9859831811652988e-05,
+      "loss": 0.0878,
+      "step": 184
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.9857410486352865e-05,
+      "loss": 0.0695,
+      "step": 185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9854968576405343e-05,
+      "loss": 0.0764,
+      "step": 186
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.9852506086909707e-05,
+      "loss": 0.0765,
+      "step": 187
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.985002302300824e-05,
+      "loss": 0.0709,
+      "step": 188
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.9847519389886174e-05,
+      "loss": 0.0688,
+      "step": 189
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.9844995192771695e-05,
+      "loss": 0.0844,
+      "step": 190
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.9842450436935937e-05,
+      "loss": 0.0683,
+      "step": 191
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.9839885127692963e-05,
+      "loss": 0.0832,
+      "step": 192
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9837299270399755e-05,
+      "loss": 0.0663,
+      "step": 193
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.9834692870456203e-05,
+      "loss": 0.0752,
+      "step": 194
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9832065933305107e-05,
+      "loss": 0.0647,
+      "step": 195
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.982941846443214e-05,
+      "loss": 0.0625,
+      "step": 196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9826750469365853e-05,
+      "loss": 0.0933,
+      "step": 197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.9824061953677664e-05,
+      "loss": 0.0967,
+      "step": 198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.9821352922981836e-05,
+      "loss": 0.0704,
+      "step": 199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.9818623382935488e-05,
+      "loss": 0.0688,
+      "step": 200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9815873339238546e-05,
+      "loss": 0.054,
+      "step": 201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.9813102797633773e-05,
+      "loss": 0.0723,
+      "step": 202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.9810311763906722e-05,
+      "loss": 0.0915,
+      "step": 203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.9807500243885746e-05,
+      "loss": 0.0685,
+      "step": 204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.9804668243441975e-05,
+      "loss": 0.0609,
+      "step": 205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.980181576848931e-05,
+      "loss": 0.0851,
+      "step": 206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.9798942824984398e-05,
+      "loss": 0.0838,
+      "step": 207
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.9796049418926645e-05,
+      "loss": 0.0751,
+      "step": 208
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.979313555635818e-05,
+      "loss": 0.0561,
+      "step": 209
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.9790201243363847e-05,
+      "loss": 0.0717,
+      "step": 210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9787246486071197e-05,
+      "loss": 0.059,
+      "step": 211
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9784271290650476e-05,
+      "loss": 0.0767,
+      "step": 212
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.9781275663314613e-05,
+      "loss": 0.0876,
+      "step": 213
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.977825961031919e-05,
+      "loss": 0.0675,
+      "step": 214
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.977522313796246e-05,
+      "loss": 0.1007,
+      "step": 215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.977216625258531e-05,
+      "loss": 0.0661,
+      "step": 216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.9769088960571247e-05,
+      "loss": 0.0754,
+      "step": 217
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.97659912683464e-05,
+      "loss": 0.0708,
+      "step": 218
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.9762873182379495e-05,
+      "loss": 0.1077,
+      "step": 219
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.975973470918185e-05,
+      "loss": 0.0672,
+      "step": 220
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9756575855307356e-05,
+      "loss": 0.0654,
+      "step": 221
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.975339662735245e-05,
+      "loss": 0.0616,
+      "step": 222
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.9750197031956135e-05,
+      "loss": 0.0812,
+      "step": 223
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.875,
+      "learning_rate": 1.9746977075799934e-05,
+      "loss": 0.0835,
+      "step": 224
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.974373676560789e-05,
+      "loss": 0.0751,
+      "step": 225
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.974047610814655e-05,
+      "loss": 0.0579,
+      "step": 226
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.9737195110224957e-05,
+      "loss": 0.0618,
+      "step": 227
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9733893778694622e-05,
+      "loss": 0.0608,
+      "step": 228
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.973057212044952e-05,
+      "loss": 0.08,
+      "step": 229
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.972723014242607e-05,
+      "loss": 0.0597,
+      "step": 230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.9723867851603136e-05,
+      "loss": 0.093,
+      "step": 231
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.9720485255001987e-05,
+      "loss": 0.0649,
+      "step": 232
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9717082359686295e-05,
+      "loss": 0.0559,
+      "step": 233
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.971365917276213e-05,
+      "loss": 0.0577,
+      "step": 234
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.971021570137793e-05,
+      "loss": 0.1001,
+      "step": 235
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.970675195272449e-05,
+      "loss": 0.0759,
+      "step": 236
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.9703267934034955e-05,
+      "loss": 0.0707,
+      "step": 237
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.9699763652584797e-05,
+      "loss": 0.0588,
+      "step": 238
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9696239115691795e-05,
+      "loss": 0.0713,
+      "step": 239
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9692694330716027e-05,
+      "loss": 0.0555,
+      "step": 240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.9689129305059868e-05,
+      "loss": 0.0772,
+      "step": 241
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.9685544046167944e-05,
+      "loss": 0.0734,
+      "step": 242
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9681938561527142e-05,
+      "loss": 0.063,
+      "step": 243
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9678312858666583e-05,
+      "loss": 0.0719,
+      "step": 244
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.9674666945157605e-05,
+      "loss": 0.0847,
+      "step": 245
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.967100082861375e-05,
+      "loss": 0.0644,
+      "step": 246
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9667314516690764e-05,
+      "loss": 0.0643,
+      "step": 247
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.9663608017086543e-05,
+      "loss": 0.0836,
+      "step": 248
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9659881337541155e-05,
+      "loss": 0.0836,
+      "step": 249
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.9656134485836804e-05,
+      "loss": 0.0692,
+      "step": 250
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.965236746979782e-05,
+      "loss": 0.0656,
+      "step": 251
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9648580297290643e-05,
+      "loss": 0.0715,
+      "step": 252
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.875,
+      "learning_rate": 1.9644772976223793e-05,
+      "loss": 0.0811,
+      "step": 253
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.964094551454788e-05,
+      "loss": 0.0666,
+      "step": 254
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.9637097920255565e-05,
+      "loss": 0.0664,
+      "step": 255
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.9633230201381547e-05,
+      "loss": 0.0818,
+      "step": 256
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.962934236600256e-05,
+      "loss": 0.0867,
+      "step": 257
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9625434422237345e-05,
+      "loss": 0.0877,
+      "step": 258
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0,
+      "learning_rate": 1.9621506378246616e-05,
+      "loss": 0.0961,
+      "step": 259
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.9617558242233085e-05,
+      "loss": 0.0876,
+      "step": 260
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.961359002244141e-05,
+      "loss": 0.058,
+      "step": 261
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9609601727158187e-05,
+      "loss": 0.0821,
+      "step": 262
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.960559336471194e-05,
+      "loss": 0.0678,
+      "step": 263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.9601564943473093e-05,
+      "loss": 0.0757,
+      "step": 264
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9597516471853964e-05,
+      "loss": 0.0587,
+      "step": 265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.9593447958308732e-05,
+      "loss": 0.0725,
+      "step": 266
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9589359411333433e-05,
+      "loss": 0.1073,
+      "step": 267
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.9585250839465938e-05,
+      "loss": 0.0698,
+      "step": 268
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.958112225128594e-05,
+      "loss": 0.0672,
+      "step": 269
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.9576973655414923e-05,
+      "loss": 0.099,
+      "step": 270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.957280506051615e-05,
+      "loss": 0.0689,
+      "step": 271
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.9568616475294657e-05,
+      "loss": 0.0769,
+      "step": 272
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.9564407908497217e-05,
+      "loss": 0.0733,
+      "step": 273
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0,
+      "learning_rate": 1.9560179368912327e-05,
+      "loss": 0.0711,
+      "step": 274
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9555930865370207e-05,
+      "loss": 0.0866,
+      "step": 275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.9551662406742742e-05,
+      "loss": 0.0687,
+      "step": 276
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.954737400194351e-05,
+      "loss": 0.0794,
+      "step": 277
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.954306565992773e-05,
+      "loss": 0.0777,
+      "step": 278
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.9538737389692264e-05,
+      "loss": 0.0925,
+      "step": 279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.9534389200275577e-05,
+      "loss": 0.0829,
+      "step": 280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9530021100757732e-05,
+      "loss": 0.0613,
+      "step": 281
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9525633100260385e-05,
+      "loss": 0.0701,
+      "step": 282
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9521225207946725e-05,
+      "loss": 0.0717,
+      "step": 283
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9516797433021502e-05,
+      "loss": 0.054,
+      "step": 284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.951234978473097e-05,
+      "loss": 0.0615,
+      "step": 285
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.9507882272362896e-05,
+      "loss": 0.0846,
+      "step": 286
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9503394905246517e-05,
+      "loss": 0.0621,
+      "step": 287
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.949888769275254e-05,
+      "loss": 0.0762,
+      "step": 288
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.949436064429311e-05,
+      "loss": 0.0723,
+      "step": 289
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.9489813769321787e-05,
+      "loss": 0.0765,
+      "step": 290
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.9485247077333555e-05,
+      "loss": 0.0743,
+      "step": 291
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9480660577864757e-05,
+      "loss": 0.078,
+      "step": 292
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.947605428049311e-05,
+      "loss": 0.0688,
+      "step": 293
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.947142819483767e-05,
+      "loss": 0.0811,
+      "step": 294
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.9466782330558828e-05,
+      "loss": 0.0858,
+      "step": 295
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9462116697358256e-05,
+      "loss": 0.0667,
+      "step": 296
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.9457431304978928e-05,
+      "loss": 0.0771,
+      "step": 297
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.9452726163205064e-05,
+      "loss": 0.0623,
+      "step": 298
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.9448001281862143e-05,
+      "loss": 0.0717,
+      "step": 299
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9443256670816847e-05,
+      "loss": 0.0646,
+      "step": 300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.9438492339977072e-05,
+      "loss": 0.0783,
+      "step": 301
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.943370829929189e-05,
+      "loss": 0.0699,
+      "step": 302
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9428904558751522e-05,
+      "loss": 0.0912,
+      "step": 303
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.942408112838734e-05,
+      "loss": 0.0714,
+      "step": 304
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.941923801827183e-05,
+      "loss": 0.0743,
+      "step": 305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9414375238518572e-05,
+      "loss": 0.0774,
+      "step": 306
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.940949279928222e-05,
+      "loss": 0.0914,
+      "step": 307
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.940459071075848e-05,
+      "loss": 0.0658,
+      "step": 308
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.93996689831841e-05,
+      "loss": 0.0742,
+      "step": 309
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.939472762683683e-05,
+      "loss": 0.0958,
+      "step": 310
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.9389766652035408e-05,
+      "loss": 0.0761,
+      "step": 311
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9384786069139552e-05,
+      "loss": 0.0587,
+      "step": 312
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.9379785888549914e-05,
+      "loss": 0.0653,
+      "step": 313
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.9374766120708077e-05,
+      "loss": 0.0997,
+      "step": 314
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.9369726776096526e-05,
+      "loss": 0.0875,
+      "step": 315
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.936466786523862e-05,
+      "loss": 0.0747,
+      "step": 316
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.75,
+      "learning_rate": 1.935958939869859e-05,
+      "loss": 0.0684,
+      "step": 317
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.9354491387081498e-05,
+      "loss": 0.0822,
+      "step": 318
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9349373841033217e-05,
+      "loss": 0.0789,
+      "step": 319
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9344236771240417e-05,
+      "loss": 0.0666,
+      "step": 320
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9339080188430537e-05,
+      "loss": 0.069,
+      "step": 321
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9333904103371764e-05,
+      "loss": 0.0529,
+      "step": 322
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.932870852687301e-05,
+      "loss": 0.0735,
+      "step": 323
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9323493469783893e-05,
+      "loss": 0.0629,
+      "step": 324
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9318258942994702e-05,
+      "loss": 0.0637,
+      "step": 325
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.9313004957436394e-05,
+      "loss": 0.0705,
+      "step": 326
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9307731524080555e-05,
+      "loss": 0.0829,
+      "step": 327
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.9302438653939387e-05,
+      "loss": 0.0718,
+      "step": 328
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.9297126358065675e-05,
+      "loss": 0.0956,
+      "step": 329
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9291794647552772e-05,
+      "loss": 0.086,
+      "step": 330
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.928644353353457e-05,
+      "loss": 0.09,
+      "step": 331
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.92810730271855e-05,
+      "loss": 0.0694,
+      "step": 332
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.9275683139720452e-05,
+      "loss": 0.0724,
+      "step": 333
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9270273882394818e-05,
+      "loss": 0.0736,
+      "step": 334
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9264845266504438e-05,
+      "loss": 0.0816,
+      "step": 335
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9259397303385557e-05,
+      "loss": 0.0641,
+      "step": 336
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.925393000441484e-05,
+      "loss": 0.0929,
+      "step": 337
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9248443381009326e-05,
+      "loss": 0.0718,
+      "step": 338
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.92429374446264e-05,
+      "loss": 0.0991,
+      "step": 339
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.923741220676379e-05,
+      "loss": 0.0715,
+      "step": 340
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9231867678959524e-05,
+      "loss": 0.078,
+      "step": 341
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9226303872791904e-05,
+      "loss": 0.0761,
+      "step": 342
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.9220720799879506e-05,
+      "loss": 0.0969,
+      "step": 343
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.9215118471881124e-05,
+      "loss": 0.0874,
+      "step": 344
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9209496900495768e-05,
+      "loss": 0.0694,
+      "step": 345
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.920385609746264e-05,
+      "loss": 0.1018,
+      "step": 346
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.919819607456109e-05,
+      "loss": 0.0883,
+      "step": 347
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.9192516843610612e-05,
+      "loss": 0.1028,
+      "step": 348
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0,
+      "learning_rate": 1.9186818416470806e-05,
+      "loss": 0.0923,
+      "step": 349
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.9181100805041365e-05,
+      "loss": 0.0945,
+      "step": 350
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.917536402126203e-05,
+      "loss": 0.0884,
+      "step": 351
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.9169608077112594e-05,
+      "loss": 0.0912,
+      "step": 352
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.9163832984612855e-05,
+      "loss": 0.0686,
+      "step": 353
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.9158038755822592e-05,
+      "loss": 0.067,
+      "step": 354
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.9152225402841552e-05,
+      "loss": 0.0942,
+      "step": 355
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9146392937809422e-05,
+      "loss": 0.071,
+      "step": 356
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.9140541372905787e-05,
+      "loss": 0.0833,
+      "step": 357
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.913467072035013e-05,
+      "loss": 0.0823,
+      "step": 358
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9128780992401783e-05,
+      "loss": 0.0814,
+      "step": 359
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.9122872201359918e-05,
+      "loss": 0.0828,
+      "step": 360
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.9116944359563514e-05,
+      "loss": 0.0918,
+      "step": 361
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.911099747939133e-05,
+      "loss": 0.0887,
+      "step": 362
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.910503157326189e-05,
+      "loss": 0.0808,
+      "step": 363
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.909904665363344e-05,
+      "loss": 0.0697,
+      "step": 364
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.909304273300393e-05,
+      "loss": 0.0749,
+      "step": 365
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9087019823910995e-05,
+      "loss": 0.0837,
+      "step": 366
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.9080977938931922e-05,
+      "loss": 0.0862,
+      "step": 367
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.907491709068362e-05,
+      "loss": 0.1013,
+      "step": 368
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.90688372918226e-05,
+      "loss": 0.0838,
+      "step": 369
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.906273855504495e-05,
+      "loss": 0.0987,
+      "step": 370
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.9056620893086296e-05,
+      "loss": 0.0752,
+      "step": 371
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.9050484318721794e-05,
+      "loss": 0.0811,
+      "step": 372
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.9044328844766084e-05,
+      "loss": 0.085,
+      "step": 373
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9038154484073286e-05,
+      "loss": 0.0702,
+      "step": 374
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.9031961249536946e-05,
+      "loss": 0.0724,
+      "step": 375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9025749154090032e-05,
+      "loss": 0.0624,
+      "step": 376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9019518210704896e-05,
+      "loss": 0.066,
+      "step": 377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.9013268432393242e-05,
+      "loss": 0.0843,
+      "step": 378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9006999832206127e-05,
+      "loss": 0.0663,
+      "step": 379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.9000712423233882e-05,
+      "loss": 0.0803,
+      "step": 380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.899440621860614e-05,
+      "loss": 0.0846,
+      "step": 381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8988081231491774e-05,
+      "loss": 0.0808,
+      "step": 382
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 0.09984159469604492,
+      "eval_runtime": 229.9749,
+      "eval_samples_per_second": 34.665,
+      "eval_steps_per_second": 4.335,
+      "step": 382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8981737475098882e-05,
+      "loss": 0.1313,
+      "step": 383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8975374962674754e-05,
+      "loss": 0.1294,
+      "step": 384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.896899370750585e-05,
+      "loss": 0.1134,
+      "step": 385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.8962593722917763e-05,
+      "loss": 0.1158,
+      "step": 386
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.8956175022275203e-05,
+      "loss": 0.1237,
+      "step": 387
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.8949737618981968e-05,
+      "loss": 0.1275,
+      "step": 388
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.8943281526480904e-05,
+      "loss": 0.127,
+      "step": 389
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.8936806758253887e-05,
+      "loss": 0.1202,
+      "step": 390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.893031332782179e-05,
+      "loss": 0.125,
+      "step": 391
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.8923801248744465e-05,
+      "loss": 0.1428,
+      "step": 392
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.89172705346207e-05,
+      "loss": 0.143,
+      "step": 393
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.89107211990882e-05,
+      "loss": 0.1296,
+      "step": 394
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.890415325582355e-05,
+      "loss": 0.1099,
+      "step": 395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.8897566718542203e-05,
+      "loss": 0.1294,
+      "step": 396
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.889096160099843e-05,
+      "loss": 0.1181,
+      "step": 397
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.8884337916985315e-05,
+      "loss": 0.1229,
+      "step": 398
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8877695680334695e-05,
+      "loss": 0.1311,
+      "step": 399
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.887103490491717e-05,
+      "loss": 0.131,
+      "step": 400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.886435560464204e-05,
+      "loss": 0.1148,
+      "step": 401
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.885765779345729e-05,
+      "loss": 0.1054,
+      "step": 402
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.885094148534957e-05,
+      "loss": 0.1192,
+      "step": 403
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.884420669434414e-05,
+      "loss": 0.1199,
+      "step": 404
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.883745343450487e-05,
+      "loss": 0.1441,
+      "step": 405
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.8830681719934197e-05,
+      "loss": 0.1133,
+      "step": 406
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.882389156477309e-05,
+      "loss": 0.1162,
+      "step": 407
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.8817082983201023e-05,
+      "loss": 0.113,
+      "step": 408
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.881025598943596e-05,
+      "loss": 0.1121,
+      "step": 409
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.8803410597734303e-05,
+      "loss": 0.1089,
+      "step": 410
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.8796546822390886e-05,
+      "loss": 0.1032,
+      "step": 411
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.878966467773892e-05,
+      "loss": 0.1192,
+      "step": 412
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.8782764178149983e-05,
+      "loss": 0.1154,
+      "step": 413
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.8775845338033984e-05,
+      "loss": 0.1193,
+      "step": 414
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.8768908171839117e-05,
+      "loss": 0.1239,
+      "step": 415
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.8761952694051864e-05,
+      "loss": 0.1283,
+      "step": 416
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.8754978919196935e-05,
+      "loss": 0.1281,
+      "step": 417
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.8747986861837253e-05,
+      "loss": 0.1251,
+      "step": 418
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.874097653657392e-05,
+      "loss": 0.1129,
+      "step": 419
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.873394795804618e-05,
+      "loss": 0.0953,
+      "step": 420
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.8726901140931396e-05,
+      "loss": 0.1025,
+      "step": 421
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.8719836099945024e-05,
+      "loss": 0.0951,
+      "step": 422
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.871275284984057e-05,
+      "loss": 0.0938,
+      "step": 423
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.875,
+      "learning_rate": 1.8705651405409568e-05,
+      "loss": 0.1112,
+      "step": 424
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.869853178148154e-05,
+      "loss": 0.0919,
+      "step": 425
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.8691393992923984e-05,
+      "loss": 0.121,
+      "step": 426
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.8684238054642313e-05,
+      "loss": 0.1035,
+      "step": 427
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.867706398157985e-05,
+      "loss": 0.1105,
+      "step": 428
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.866987178871779e-05,
+      "loss": 0.1052,
+      "step": 429
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.866266149107517e-05,
+      "loss": 0.1211,
+      "step": 430
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.865543310370882e-05,
+      "loss": 0.1094,
+      "step": 431
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.8648186641713357e-05,
+      "loss": 0.1309,
+      "step": 432
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.8640922120221138e-05,
+      "loss": 0.0963,
+      "step": 433
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.8633639554402238e-05,
+      "loss": 0.1083,
+      "step": 434
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.8626338959464396e-05,
+      "loss": 0.1029,
+      "step": 435
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.861902035065303e-05,
+      "loss": 0.0953,
+      "step": 436
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8611683743251144e-05,
+      "loss": 0.1017,
+      "step": 437
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.8604329152579348e-05,
+      "loss": 0.1303,
+      "step": 438
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.8596956593995798e-05,
+      "loss": 0.1124,
+      "step": 439
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.8589566082896176e-05,
+      "loss": 0.117,
+      "step": 440
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.858215763471364e-05,
+      "loss": 0.1157,
+      "step": 441
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.857473126491882e-05,
+      "loss": 0.1076,
+      "step": 442
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.8567286989019765e-05,
+      "loss": 0.1047,
+      "step": 443
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.8559824822561915e-05,
+      "loss": 0.0992,
+      "step": 444
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.855234478112807e-05,
+      "loss": 0.1112,
+      "step": 445
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.8544846880338362e-05,
+      "loss": 0.1085,
+      "step": 446
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.853733113585021e-05,
+      "loss": 0.0973,
+      "step": 447
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.85297975633583e-05,
+      "loss": 0.141,
+      "step": 448
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.852224617859455e-05,
+      "loss": 0.1052,
+      "step": 449
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.851467699732806e-05,
+      "loss": 0.0935,
+      "step": 450
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.8507090035365113e-05,
+      "loss": 0.1078,
+      "step": 451
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.849948530854911e-05,
+      "loss": 0.1033,
+      "step": 452
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.849186283276055e-05,
+      "loss": 0.099,
+      "step": 453
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.8484222623917003e-05,
+      "loss": 0.1005,
+      "step": 454
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8476564697973062e-05,
+      "loss": 0.0964,
+      "step": 455
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.8468889070920322e-05,
+      "loss": 0.1054,
+      "step": 456
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.846119575878734e-05,
+      "loss": 0.1047,
+      "step": 457
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.845348477763961e-05,
+      "loss": 0.0995,
+      "step": 458
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.8445756143579506e-05,
+      "loss": 0.1062,
+      "step": 459
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.8438009872746292e-05,
+      "loss": 0.105,
+      "step": 460
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.8430245981316036e-05,
+      "loss": 0.1187,
+      "step": 461
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.842246448550162e-05,
+      "loss": 0.0894,
+      "step": 462
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.841466540155268e-05,
+      "loss": 0.0975,
+      "step": 463
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.8406848745755582e-05,
+      "loss": 0.1314,
+      "step": 464
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.8399014534433384e-05,
+      "loss": 0.0983,
+      "step": 465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.8391162783945805e-05,
+      "loss": 0.1178,
+      "step": 466
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.8383293510689196e-05,
+      "loss": 0.0988,
+      "step": 467
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.8375406731096495e-05,
+      "loss": 0.1035,
+      "step": 468
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.8367502461637193e-05,
+      "loss": 0.1014,
+      "step": 469
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.835958071881731e-05,
+      "loss": 0.1177,
+      "step": 470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.8351641519179356e-05,
+      "loss": 0.1154,
+      "step": 471
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.8343684879302285e-05,
+      "loss": 0.1076,
+      "step": 472
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.8335710815801486e-05,
+      "loss": 0.0854,
+      "step": 473
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.832771934532872e-05,
+      "loss": 0.0975,
+      "step": 474
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8319710484572103e-05,
+      "loss": 0.0984,
+      "step": 475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.8311684250256066e-05,
+      "loss": 0.0867,
+      "step": 476
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.830364065914132e-05,
+      "loss": 0.0949,
+      "step": 477
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.8295579728024815e-05,
+      "loss": 0.0971,
+      "step": 478
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.8287501473739724e-05,
+      "loss": 0.0953,
+      "step": 479
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.8279405913155385e-05,
+      "loss": 0.0919,
+      "step": 480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.8271293063177277e-05,
+      "loss": 0.1067,
+      "step": 481
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.826316294074699e-05,
+      "loss": 0.1172,
+      "step": 482
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.8255015562842167e-05,
+      "loss": 0.1093,
+      "step": 483
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.8246850946476505e-05,
+      "loss": 0.1033,
+      "step": 484
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.8238669108699688e-05,
+      "loss": 0.1165,
+      "step": 485
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.8230470066597365e-05,
+      "loss": 0.1111,
+      "step": 486
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.8222253837291108e-05,
+      "loss": 0.0908,
+      "step": 487
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.875,
+      "learning_rate": 1.8214020437938382e-05,
+      "loss": 0.0957,
+      "step": 488
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.820576988573251e-05,
+      "loss": 0.0937,
+      "step": 489
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.8197502197902633e-05,
+      "loss": 0.0944,
+      "step": 490
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.8189217391713675e-05,
+      "loss": 0.0981,
+      "step": 491
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.8180915484466303e-05,
+      "loss": 0.0942,
+      "step": 492
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.81725964934969e-05,
+      "loss": 0.1046,
+      "step": 493
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.8164260436177524e-05,
+      "loss": 0.097,
+      "step": 494
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.8155907329915873e-05,
+      "loss": 0.0983,
+      "step": 495
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.814753719215524e-05,
+      "loss": 0.1029,
+      "step": 496
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.8139150040374497e-05,
+      "loss": 0.0907,
+      "step": 497
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8130745892088028e-05,
+      "loss": 0.1036,
+      "step": 498
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.8122324764845723e-05,
+      "loss": 0.0891,
+      "step": 499
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.8113886676232922e-05,
+      "loss": 0.1002,
+      "step": 500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.8105431643870387e-05,
+      "loss": 0.107,
+      "step": 501
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.8096959685414264e-05,
+      "loss": 0.0981,
+      "step": 502
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8088470818556044e-05,
+      "loss": 0.0972,
+      "step": 503
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.807996506102252e-05,
+      "loss": 0.0951,
+      "step": 504
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.8071442430575764e-05,
+      "loss": 0.1206,
+      "step": 505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.806290294501309e-05,
+      "loss": 0.0941,
+      "step": 506
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.805434662216699e-05,
+      "loss": 0.0799,
+      "step": 507
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.804577347990513e-05,
+      "loss": 0.104,
+      "step": 508
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0,
+      "learning_rate": 1.8037183536130298e-05,
+      "loss": 0.1018,
+      "step": 509
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0,
+      "learning_rate": 1.8028576808780363e-05,
+      "loss": 0.1132,
+      "step": 510
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8019953315828247e-05,
+      "loss": 0.1114,
+      "step": 511
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.8011313075281883e-05,
+      "loss": 0.1135,
+      "step": 512
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.8002656105184164e-05,
+      "loss": 0.099,
+      "step": 513
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.7993982423612942e-05,
+      "loss": 0.0983,
+      "step": 514
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.7985292048680945e-05,
+      "loss": 0.0942,
+      "step": 515
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.797658499853577e-05,
+      "loss": 0.0821,
+      "step": 516
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0,
+      "learning_rate": 1.796786129135984e-05,
+      "loss": 0.1067,
+      "step": 517
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.795912094537035e-05,
+      "loss": 0.1065,
+      "step": 518
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.7950363978819248e-05,
+      "loss": 0.0944,
+      "step": 519
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.875,
+      "learning_rate": 1.7941590409993193e-05,
+      "loss": 0.0913,
+      "step": 520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.7932800257213513e-05,
+      "loss": 0.0821,
+      "step": 521
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.7923993538836154e-05,
+      "loss": 0.1171,
+      "step": 522
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.7915170273251676e-05,
+      "loss": 0.0849,
+      "step": 523
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.7906330478885178e-05,
+      "loss": 0.0988,
+      "step": 524
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.7897474174196278e-05,
+      "loss": 0.1047,
+      "step": 525
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.75,
+      "learning_rate": 1.7888601377679075e-05,
+      "loss": 0.0845,
+      "step": 526
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.7879712107862108e-05,
+      "loss": 0.1106,
+      "step": 527
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.787080638330831e-05,
+      "loss": 0.099,
+      "step": 528
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.875,
+      "learning_rate": 1.786188422261498e-05,
+      "loss": 0.1002,
+      "step": 529
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.7852945644413738e-05,
+      "loss": 0.1205,
+      "step": 530
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.7843990667370486e-05,
+      "loss": 0.0862,
+      "step": 531
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.7835019310185383e-05,
+      "loss": 0.1132,
+      "step": 532
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.782603159159277e-05,
+      "loss": 0.0958,
+      "step": 533
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.781702753036118e-05,
+      "loss": 0.0917,
+      "step": 534
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.780800714529325e-05,
+      "loss": 0.089,
+      "step": 535
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.779897045522573e-05,
+      "loss": 0.0721,
+      "step": 536
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.7789917479029394e-05,
+      "loss": 0.086,
+      "step": 537
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.7780848235609044e-05,
+      "loss": 0.08,
+      "step": 538
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.777176274390344e-05,
+      "loss": 0.0947,
+      "step": 539
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.776266102288528e-05,
+      "loss": 0.0897,
+      "step": 540
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.875,
+      "learning_rate": 1.775354309156115e-05,
+      "loss": 0.1017,
+      "step": 541
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.7744408968971486e-05,
+      "loss": 0.1228,
+      "step": 542
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.7735258674190533e-05,
+      "loss": 0.0931,
+      "step": 543
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.7726092226326317e-05,
+      "loss": 0.1015,
+      "step": 544
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.7716909644520584e-05,
+      "loss": 0.0957,
+      "step": 545
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.770771094794878e-05,
+      "loss": 0.1029,
+      "step": 546
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7698496155819994e-05,
+      "loss": 0.0784,
+      "step": 547
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.7689265287376935e-05,
+      "loss": 0.0888,
+      "step": 548
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.768001836189588e-05,
+      "loss": 0.0935,
+      "step": 549
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.7670755398686635e-05,
+      "loss": 0.0834,
+      "step": 550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.7661476417092502e-05,
+      "loss": 0.1196,
+      "step": 551
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.7652181436490225e-05,
+      "loss": 0.0955,
+      "step": 552
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.7642870476289965e-05,
+      "loss": 0.0845,
+      "step": 553
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.7633543555935246e-05,
+      "loss": 0.0823,
+      "step": 554
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.762420069490293e-05,
+      "loss": 0.1031,
+      "step": 555
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.761484191270316e-05,
+      "loss": 0.0686,
+      "step": 556
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.760546722887932e-05,
+      "loss": 0.0917,
+      "step": 557
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.7596076663008016e-05,
+      "loss": 0.0993,
+      "step": 558
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.758667023469901e-05,
+      "loss": 0.101,
+      "step": 559
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.7577247963595184e-05,
+      "loss": 0.0922,
+      "step": 560
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.7567809869372522e-05,
+      "loss": 0.0875,
+      "step": 561
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.7558355971740024e-05,
+      "loss": 0.1121,
+      "step": 562
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.7548886290439716e-05,
+      "loss": 0.0902,
+      "step": 563
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.7539400845246565e-05,
+      "loss": 0.0741,
+      "step": 564
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.7529899655968474e-05,
+      "loss": 0.0831,
+      "step": 565
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.7520382742446208e-05,
+      "loss": 0.0889,
+      "step": 566
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.7510850124553378e-05,
+      "loss": 0.0956,
+      "step": 567
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.7501301822196387e-05,
+      "loss": 0.101,
+      "step": 568
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.7491737855314383e-05,
+      "loss": 0.0949,
+      "step": 569
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.7482158243879243e-05,
+      "loss": 0.1013,
+      "step": 570
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.7472563007895497e-05,
+      "loss": 0.1046,
+      "step": 571
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.7462952167400308e-05,
+      "loss": 0.1069,
+      "step": 572
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.7453325742463426e-05,
+      "loss": 0.0747,
+      "step": 573
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.7443683753187152e-05,
+      "loss": 0.0869,
+      "step": 574
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.7434026219706273e-05,
+      "loss": 0.1059,
+      "step": 575
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.742435316218805e-05,
+      "loss": 0.1011,
+      "step": 576
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.7414664600832155e-05,
+      "loss": 0.091,
+      "step": 577
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.7404960555870638e-05,
+      "loss": 0.0779,
+      "step": 578
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.7395241047567884e-05,
+      "loss": 0.0855,
+      "step": 579
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.7385506096220566e-05,
+      "loss": 0.1022,
+      "step": 580
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.125,
+      "learning_rate": 1.7375755722157608e-05,
+      "loss": 0.1207,
+      "step": 581
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.736598994574014e-05,
+      "loss": 0.0996,
+      "step": 582
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.735620878736145e-05,
+      "loss": 0.0852,
+      "step": 583
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.734641226744696e-05,
+      "loss": 0.0935,
+      "step": 584
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.7336600406454163e-05,
+      "loss": 0.098,
+      "step": 585
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.732677322487258e-05,
+      "loss": 0.0948,
+      "step": 586
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.7316930743223742e-05,
+      "loss": 0.0806,
+      "step": 587
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.7307072982061118e-05,
+      "loss": 0.0797,
+      "step": 588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.729719996197009e-05,
+      "loss": 0.0809,
+      "step": 589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.7287311703567894e-05,
+      "loss": 0.0941,
+      "step": 590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.7277408227503604e-05,
+      "loss": 0.0869,
+      "step": 591
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.7267489554458057e-05,
+      "loss": 0.0823,
+      "step": 592
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.725755570514384e-05,
+      "loss": 0.1041,
+      "step": 593
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.724760670030521e-05,
+      "loss": 0.1031,
+      "step": 594
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.7237642560718102e-05,
+      "loss": 0.0978,
+      "step": 595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0,
+      "learning_rate": 1.7227663307190027e-05,
+      "loss": 0.0996,
+      "step": 596
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.7217668960560074e-05,
+      "loss": 0.0771,
+      "step": 597
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.7207659541698843e-05,
+      "loss": 0.097,
+      "step": 598
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.7197635071508417e-05,
+      "loss": 0.0807,
+      "step": 599
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.71875955709223e-05,
+      "loss": 0.1252,
+      "step": 600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.7177541060905394e-05,
+      "loss": 0.0908,
+      "step": 601
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.7167471562453928e-05,
+      "loss": 0.0765,
+      "step": 602
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.7157387096595445e-05,
+      "loss": 0.0775,
+      "step": 603
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.714728768438874e-05,
+      "loss": 0.1118,
+      "step": 604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.7137173346923816e-05,
+      "loss": 0.1091,
+      "step": 605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.712704410532185e-05,
+      "loss": 0.0819,
+      "step": 606
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.7116899980735132e-05,
+      "loss": 0.1062,
+      "step": 607
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.7106740994347036e-05,
+      "loss": 0.1005,
+      "step": 608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7096567167371984e-05,
+      "loss": 0.0748,
+      "step": 609
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.708637852105537e-05,
+      "loss": 0.1036,
+      "step": 610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.875,
+      "learning_rate": 1.7076175076673537e-05,
+      "loss": 0.0891,
+      "step": 611
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.706595685553374e-05,
+      "loss": 0.1007,
+      "step": 612
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.7055723878974092e-05,
+      "loss": 0.0783,
+      "step": 613
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.7045476168363498e-05,
+      "loss": 0.0779,
+      "step": 614
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.875,
+      "learning_rate": 1.703521374510166e-05,
+      "loss": 0.0806,
+      "step": 615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.702493663061898e-05,
+      "loss": 0.0895,
+      "step": 616
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.7014644846376553e-05,
+      "loss": 0.0911,
+      "step": 617
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.7004338413866105e-05,
+      "loss": 0.0773,
+      "step": 618
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.6994017354609944e-05,
+      "loss": 0.0979,
+      "step": 619
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.6983681690160934e-05,
+      "loss": 0.0787,
+      "step": 620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.6973331442102427e-05,
+      "loss": 0.0855,
+      "step": 621
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.6962966632048234e-05,
+      "loss": 0.0819,
+      "step": 622
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6952587281642577e-05,
+      "loss": 0.0878,
+      "step": 623
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.6942193412560044e-05,
+      "loss": 0.0838,
+      "step": 624
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.6931785046505533e-05,
+      "loss": 0.0896,
+      "step": 625
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.6921362205214212e-05,
+      "loss": 0.0906,
+      "step": 626
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.6910924910451493e-05,
+      "loss": 0.101,
+      "step": 627
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.6900473184012963e-05,
+      "loss": 0.0834,
+      "step": 628
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.689000704772434e-05,
+      "loss": 0.0924,
+      "step": 629
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.875,
+      "learning_rate": 1.6879526523441433e-05,
+      "loss": 0.0901,
+      "step": 630
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.686903163305011e-05,
+      "loss": 0.107,
+      "step": 631
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6858522398466228e-05,
+      "loss": 0.0945,
+      "step": 632
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.68479988416356e-05,
+      "loss": 0.0682,
+      "step": 633
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.6837460984533937e-05,
+      "loss": 0.1066,
+      "step": 634
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.682690884916684e-05,
+      "loss": 0.1051,
+      "step": 635
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.6816342457569694e-05,
+      "loss": 0.0864,
+      "step": 636
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.6805761831807683e-05,
+      "loss": 0.0874,
+      "step": 637
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6795166993975693e-05,
+      "loss": 0.0834,
+      "step": 638
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.6784557966198304e-05,
+      "loss": 0.0981,
+      "step": 639
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.6773934770629713e-05,
+      "loss": 0.0859,
+      "step": 640
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.6763297429453715e-05,
+      "loss": 0.0962,
+      "step": 641
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.6752645964883642e-05,
+      "loss": 0.0958,
+      "step": 642
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.674198039916232e-05,
+      "loss": 0.0832,
+      "step": 643
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.6731300754562008e-05,
+      "loss": 0.0799,
+      "step": 644
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.6720607053384393e-05,
+      "loss": 0.1109,
+      "step": 645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.6709899317960486e-05,
+      "loss": 0.0945,
+      "step": 646
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.669917757065062e-05,
+      "loss": 0.0775,
+      "step": 647
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.668844183384439e-05,
+      "loss": 0.0835,
+      "step": 648
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.75,
+      "learning_rate": 1.66776921299606e-05,
+      "loss": 0.0827,
+      "step": 649
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.6666928481447216e-05,
+      "loss": 0.1027,
+      "step": 650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.665615091078133e-05,
+      "loss": 0.0833,
+      "step": 651
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.6645359440469106e-05,
+      "loss": 0.0858,
+      "step": 652
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.6634554093045738e-05,
+      "loss": 0.1068,
+      "step": 653
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.662373489107539e-05,
+      "loss": 0.1033,
+      "step": 654
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.661290185715116e-05,
+      "loss": 0.0952,
+      "step": 655
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.6602055013895032e-05,
+      "loss": 0.0829,
+      "step": 656
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.6591194383957834e-05,
+      "loss": 0.0916,
+      "step": 657
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.6580319990019173e-05,
+      "loss": 0.1042,
+      "step": 658
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.6569431854787397e-05,
+      "loss": 0.0885,
+      "step": 659
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.6558530000999566e-05,
+      "loss": 0.0829,
+      "step": 660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6547614451421366e-05,
+      "loss": 0.0717,
+      "step": 661
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.6536685228847106e-05,
+      "loss": 0.0783,
+      "step": 662
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.6525742356099622e-05,
+      "loss": 0.0911,
+      "step": 663
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.6514785856030276e-05,
+      "loss": 0.0876,
+      "step": 664
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.650381575151888e-05,
+      "loss": 0.0786,
+      "step": 665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.6492832065473648e-05,
+      "loss": 0.0907,
+      "step": 666
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.6481834820831164e-05,
+      "loss": 0.0994,
+      "step": 667
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.6470824040556328e-05,
+      "loss": 0.0876,
+      "step": 668
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.64597997476423e-05,
+      "loss": 0.0917,
+      "step": 669
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.6448761965110452e-05,
+      "loss": 0.0976,
+      "step": 670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.6437710716010348e-05,
+      "loss": 0.0883,
+      "step": 671
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.642664602341964e-05,
+      "loss": 0.09,
+      "step": 672
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.6415567910444082e-05,
+      "loss": 0.0973,
+      "step": 673
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.6404476400217443e-05,
+      "loss": 0.0859,
+      "step": 674
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.6393371515901458e-05,
+      "loss": 0.1135,
+      "step": 675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.6382253280685813e-05,
+      "loss": 0.0947,
+      "step": 676
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.6371121717788053e-05,
+      "loss": 0.0823,
+      "step": 677
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.635997685045356e-05,
+      "loss": 0.0861,
+      "step": 678
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.6348818701955513e-05,
+      "loss": 0.0789,
+      "step": 679
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.6337647295594803e-05,
+      "loss": 0.0762,
+      "step": 680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.632646265470002e-05,
+      "loss": 0.0748,
+      "step": 681
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.631526480262739e-05,
+      "loss": 0.1057,
+      "step": 682
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.6304053762760728e-05,
+      "loss": 0.0829,
+      "step": 683
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.6292829558511377e-05,
+      "loss": 0.0877,
+      "step": 684
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.628159221331819e-05,
+      "loss": 0.0899,
+      "step": 685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.627034175064744e-05,
+      "loss": 0.0884,
+      "step": 686
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.6259078193992806e-05,
+      "loss": 0.0951,
+      "step": 687
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.624780156687531e-05,
+      "loss": 0.0688,
+      "step": 688
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.6236511892843267e-05,
+      "loss": 0.0816,
+      "step": 689
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.6225209195472226e-05,
+      "loss": 0.0868,
+      "step": 690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.621389349836495e-05,
+      "loss": 0.1009,
+      "step": 691
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.6202564825151346e-05,
+      "loss": 0.0796,
+      "step": 692
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.61912231994884e-05,
+      "loss": 0.0684,
+      "step": 693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.6179868645060163e-05,
+      "loss": 0.0745,
+      "step": 694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.6168501185577677e-05,
+      "loss": 0.0844,
+      "step": 695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.6157120844778945e-05,
+      "loss": 0.0843,
+      "step": 696
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6145727646428844e-05,
+      "loss": 0.0676,
+      "step": 697
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.6134321614319132e-05,
+      "loss": 0.0912,
+      "step": 698
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.6122902772268344e-05,
+      "loss": 0.086,
+      "step": 699
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.6111471144121765e-05,
+      "loss": 0.1003,
+      "step": 700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.6100026753751407e-05,
+      "loss": 0.0987,
+      "step": 701
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.6088569625055897e-05,
+      "loss": 0.0821,
+      "step": 702
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.6077099781960484e-05,
+      "loss": 0.0839,
+      "step": 703
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.606561724841697e-05,
+      "loss": 0.0769,
+      "step": 704
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.6054122048403636e-05,
+      "loss": 0.0856,
+      "step": 705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.6042614205925246e-05,
+      "loss": 0.0889,
+      "step": 706
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.6031093745012935e-05,
+      "loss": 0.0824,
+      "step": 707
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.6019560689724203e-05,
+      "loss": 0.0972,
+      "step": 708
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.600801506414285e-05,
+      "loss": 0.0817,
+      "step": 709
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.599645689237892e-05,
+      "loss": 0.0814,
+      "step": 710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5984886198568658e-05,
+      "loss": 0.0737,
+      "step": 711
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.5973303006874465e-05,
+      "loss": 0.071,
+      "step": 712
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.5961707341484832e-05,
+      "loss": 0.0899,
+      "step": 713
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.5950099226614303e-05,
+      "loss": 0.0737,
+      "step": 714
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.593847868650342e-05,
+      "loss": 0.0803,
+      "step": 715
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.125,
+      "learning_rate": 1.592684574541866e-05,
+      "loss": 0.088,
+      "step": 716
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.5915200427652417e-05,
+      "loss": 0.1083,
+      "step": 717
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.590354275752291e-05,
+      "loss": 0.0642,
+      "step": 718
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.589187275937417e-05,
+      "loss": 0.0791,
+      "step": 719
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.25,
+      "learning_rate": 1.588019045757596e-05,
+      "loss": 0.0832,
+      "step": 720
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.5868495876523743e-05,
+      "loss": 0.0872,
+      "step": 721
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.5856789040638622e-05,
+      "loss": 0.1022,
+      "step": 722
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.5845069974367282e-05,
+      "loss": 0.0975,
+      "step": 723
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.5833338702181962e-05,
+      "loss": 0.0879,
+      "step": 724
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.5821595248580383e-05,
+      "loss": 0.0828,
+      "step": 725
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0,
+      "learning_rate": 1.580983963808571e-05,
+      "loss": 0.0799,
+      "step": 726
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.579807189524648e-05,
+      "loss": 0.0885,
+      "step": 727
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.578629204463658e-05,
+      "loss": 0.0757,
+      "step": 728
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.5774500110855172e-05,
+      "loss": 0.0775,
+      "step": 729
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.576269611852666e-05,
+      "loss": 0.0792,
+      "step": 730
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.575088009230062e-05,
+      "loss": 0.0747,
+      "step": 731
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.5739052056851756e-05,
+      "loss": 0.0982,
+      "step": 732
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.572721203687986e-05,
+      "loss": 0.0752,
+      "step": 733
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.5715360057109746e-05,
+      "loss": 0.0917,
+      "step": 734
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.57034961422912e-05,
+      "loss": 0.0767,
+      "step": 735
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.5691620317198933e-05,
+      "loss": 0.0856,
+      "step": 736
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5679732606632538e-05,
+      "loss": 0.0887,
+      "step": 737
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5667833035416405e-05,
+      "loss": 0.0805,
+      "step": 738
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.5655921628399708e-05,
+      "loss": 0.068,
+      "step": 739
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.5643998410456336e-05,
+      "loss": 0.0795,
+      "step": 740
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.5632063406484844e-05,
+      "loss": 0.0851,
+      "step": 741
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.5620116641408388e-05,
+      "loss": 0.0905,
+      "step": 742
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5608158140174696e-05,
+      "loss": 0.0711,
+      "step": 743
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.5596187927755996e-05,
+      "loss": 0.0766,
+      "step": 744
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.5584206029148977e-05,
+      "loss": 0.0922,
+      "step": 745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.5572212469374725e-05,
+      "loss": 0.0846,
+      "step": 746
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.5560207273478688e-05,
+      "loss": 0.074,
+      "step": 747
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.55481904665306e-05,
+      "loss": 0.0908,
+      "step": 748
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.5536162073624454e-05,
+      "loss": 0.0877,
+      "step": 749
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.5524122119878428e-05,
+      "loss": 0.0835,
+      "step": 750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.551207063043485e-05,
+      "loss": 0.0871,
+      "step": 751
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.5500007630460124e-05,
+      "loss": 0.068,
+      "step": 752
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.5487933145144708e-05,
+      "loss": 0.0913,
+      "step": 753
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.5475847199703037e-05,
+      "loss": 0.0921,
+      "step": 754
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.5463749819373466e-05,
+      "loss": 0.0837,
+      "step": 755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.75,
+      "learning_rate": 1.545164102941825e-05,
+      "loss": 0.0678,
+      "step": 756
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.5439520855123445e-05,
+      "loss": 0.0686,
+      "step": 757
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.542738932179891e-05,
+      "loss": 0.091,
+      "step": 758
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.54152464547782e-05,
+      "loss": 0.0804,
+      "step": 759
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.5403092279418544e-05,
+      "loss": 0.0755,
+      "step": 760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.5390926821100796e-05,
+      "loss": 0.0783,
+      "step": 761
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.5378750105229362e-05,
+      "loss": 0.0751,
+      "step": 762
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.5366562157232147e-05,
+      "loss": 0.0767,
+      "step": 763
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.5354363002560532e-05,
+      "loss": 0.091,
+      "step": 764
+    },
+    {
+      "epoch": 0.68,
+      "eval_loss": 0.08239088952541351,
+      "eval_runtime": 229.4221,
+      "eval_samples_per_second": 34.748,
+      "eval_steps_per_second": 4.346,
+      "step": 764
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.534215266668929e-05,
+      "loss": 0.0831,
+      "step": 765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5329931175116542e-05,
+      "loss": 0.0702,
+      "step": 766
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.53176985533637e-05,
+      "loss": 0.0846,
+      "step": 767
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5305454826975433e-05,
+      "loss": 0.0868,
+      "step": 768
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.5293200021519585e-05,
+      "loss": 0.0803,
+      "step": 769
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.5280934162587145e-05,
+      "loss": 0.0665,
+      "step": 770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5268657275792183e-05,
+      "loss": 0.0746,
+      "step": 771
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.5256369386771792e-05,
+      "loss": 0.081,
+      "step": 772
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5244070521186044e-05,
+      "loss": 0.0673,
+      "step": 773
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5231760704717933e-05,
+      "loss": 0.0732,
+      "step": 774
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.521943996307332e-05,
+      "loss": 0.068,
+      "step": 775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.5207108321980885e-05,
+      "loss": 0.0777,
+      "step": 776
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5194765807192063e-05,
+      "loss": 0.1049,
+      "step": 777
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.5182412444480993e-05,
+      "loss": 0.0916,
+      "step": 778
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.517004825964448e-05,
+      "loss": 0.084,
+      "step": 779
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.5157673278501916e-05,
+      "loss": 0.0716,
+      "step": 780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.514528752689524e-05,
+      "loss": 0.0823,
+      "step": 781
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.513289103068889e-05,
+      "loss": 0.0775,
+      "step": 782
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.5120483815769729e-05,
+      "loss": 0.0904,
+      "step": 783
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5108065908047015e-05,
+      "loss": 0.0597,
+      "step": 784
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.509563733345233e-05,
+      "loss": 0.0874,
+      "step": 785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.5083198117939532e-05,
+      "loss": 0.0809,
+      "step": 786
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.5070748287484698e-05,
+      "loss": 0.0758,
+      "step": 787
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.5058287868086072e-05,
+      "loss": 0.0808,
+      "step": 788
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.5045816885764011e-05,
+      "loss": 0.074,
+      "step": 789
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.5033335366560931e-05,
+      "loss": 0.0775,
+      "step": 790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.5020843336541253e-05,
+      "loss": 0.0962,
+      "step": 791
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.5008340821791336e-05,
+      "loss": 0.0985,
+      "step": 792
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.4995827848419452e-05,
+      "loss": 0.0827,
+      "step": 793
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.49833044425557e-05,
+      "loss": 0.0896,
+      "step": 794
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.4970770630351965e-05,
+      "loss": 0.0828,
+      "step": 795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.4958226437981868e-05,
+      "loss": 0.0766,
+      "step": 796
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4945671891640705e-05,
+      "loss": 0.0643,
+      "step": 797
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.493310701754539e-05,
+      "loss": 0.0861,
+      "step": 798
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.75,
+      "learning_rate": 1.492053184193441e-05,
+      "loss": 0.0752,
+      "step": 799
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.490794639106776e-05,
+      "loss": 0.0681,
+      "step": 800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.4895350691226891e-05,
+      "loss": 0.0984,
+      "step": 801
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.4882744768714656e-05,
+      "loss": 0.0792,
+      "step": 802
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.4870128649855265e-05,
+      "loss": 0.0638,
+      "step": 803
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4857502360994207e-05,
+      "loss": 0.0615,
+      "step": 804
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4844865928498213e-05,
+      "loss": 0.0534,
+      "step": 805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4832219378755206e-05,
+      "loss": 0.0721,
+      "step": 806
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.4819562738174217e-05,
+      "loss": 0.0827,
+      "step": 807
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.480689603318537e-05,
+      "loss": 0.0677,
+      "step": 808
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.4794219290239787e-05,
+      "loss": 0.0686,
+      "step": 809
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4781532535809566e-05,
+      "loss": 0.0748,
+      "step": 810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.4768835796387708e-05,
+      "loss": 0.0895,
+      "step": 811
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4756129098488062e-05,
+      "loss": 0.0752,
+      "step": 812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0,
+      "learning_rate": 1.4743412468645266e-05,
+      "loss": 0.0972,
+      "step": 813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.4730685933414717e-05,
+      "loss": 0.0675,
+      "step": 814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.875,
+      "learning_rate": 1.4717949519372481e-05,
+      "loss": 0.0837,
+      "step": 815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.4705203253115257e-05,
+      "loss": 0.088,
+      "step": 816
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.4692447161260322e-05,
+      "loss": 0.0746,
+      "step": 817
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.4679681270445464e-05,
+      "loss": 0.0746,
+      "step": 818
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4666905607328946e-05,
+      "loss": 0.0632,
+      "step": 819
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.465412019858942e-05,
+      "loss": 0.0907,
+      "step": 820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.4641325070925904e-05,
+      "loss": 0.084,
+      "step": 821
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4628520251057708e-05,
+      "loss": 0.0697,
+      "step": 822
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.4615705765724374e-05,
+      "loss": 0.0757,
+      "step": 823
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.4602881641685644e-05,
+      "loss": 0.086,
+      "step": 824
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.459004790572137e-05,
+      "loss": 0.0795,
+      "step": 825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.4577204584631484e-05,
+      "loss": 0.0956,
+      "step": 826
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4564351705235935e-05,
+      "loss": 0.0611,
+      "step": 827
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.4551489294374637e-05,
+      "loss": 0.089,
+      "step": 828
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.4538617378907392e-05,
+      "loss": 0.0985,
+      "step": 829
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.4525735985713868e-05,
+      "loss": 0.0768,
+      "step": 830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.125,
+      "learning_rate": 1.451284514169351e-05,
+      "loss": 0.105,
+      "step": 831
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.4499944873765512e-05,
+      "loss": 0.0654,
+      "step": 832
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.4487035208868734e-05,
+      "loss": 0.0703,
+      "step": 833
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.447411617396167e-05,
+      "loss": 0.0733,
+      "step": 834
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.4461187796022378e-05,
+      "loss": 0.0709,
+      "step": 835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.444825010204842e-05,
+      "loss": 0.0861,
+      "step": 836
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4435303119056822e-05,
+      "loss": 0.0771,
+      "step": 837
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4422346874084001e-05,
+      "loss": 0.0679,
+      "step": 838
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4409381394185718e-05,
+      "loss": 0.0603,
+      "step": 839
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.4396406706437016e-05,
+      "loss": 0.0911,
+      "step": 840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.4383422837932168e-05,
+      "loss": 0.0684,
+      "step": 841
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.4370429815784621e-05,
+      "loss": 0.0783,
+      "step": 842
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.4357427667126935e-05,
+      "loss": 0.0859,
+      "step": 843
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.434441641911073e-05,
+      "loss": 0.0683,
+      "step": 844
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4331396098906618e-05,
+      "loss": 0.0786,
+      "step": 845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.875,
+      "learning_rate": 1.431836673370417e-05,
+      "loss": 0.0825,
+      "step": 846
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.4305328350711839e-05,
+      "loss": 0.0915,
+      "step": 847
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.4292280977156909e-05,
+      "loss": 0.0963,
+      "step": 848
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.4279224640285433e-05,
+      "loss": 0.074,
+      "step": 849
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4266159367362198e-05,
+      "loss": 0.0689,
+      "step": 850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4253085185670628e-05,
+      "loss": 0.0659,
+      "step": 851
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.4240002122512773e-05,
+      "loss": 0.0811,
+      "step": 852
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.4226910205209217e-05,
+      "loss": 0.0845,
+      "step": 853
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.4213809461099035e-05,
+      "loss": 0.0841,
+      "step": 854
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.4200699917539743e-05,
+      "loss": 0.0828,
+      "step": 855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4187581601907212e-05,
+      "loss": 0.0672,
+      "step": 856
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.4174454541595662e-05,
+      "loss": 0.0806,
+      "step": 857
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.4161318764017544e-05,
+      "loss": 0.0853,
+      "step": 858
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.4148174296603532e-05,
+      "loss": 0.0777,
+      "step": 859
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.4135021166802434e-05,
+      "loss": 0.0841,
+      "step": 860
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.4121859402081159e-05,
+      "loss": 0.0792,
+      "step": 861
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.4108689029924638e-05,
+      "loss": 0.078,
+      "step": 862
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.4095510077835789e-05,
+      "loss": 0.0682,
+      "step": 863
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4082322573335425e-05,
+      "loss": 0.0676,
+      "step": 864
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4069126543962238e-05,
+      "loss": 0.0769,
+      "step": 865
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4055922017272718e-05,
+      "loss": 0.0755,
+      "step": 866
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.75,
+      "learning_rate": 1.4042709020841095e-05,
+      "loss": 0.0816,
+      "step": 867
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.4029487582259289e-05,
+      "loss": 0.0865,
+      "step": 868
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4016257729136848e-05,
+      "loss": 0.0601,
+      "step": 869
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.4003019489100891e-05,
+      "loss": 0.0803,
+      "step": 870
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3989772889796052e-05,
+      "loss": 0.0681,
+      "step": 871
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.3976517958884421e-05,
+      "loss": 0.0587,
+      "step": 872
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.396325472404549e-05,
+      "loss": 0.0718,
+      "step": 873
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.3949983212976081e-05,
+      "loss": 0.1046,
+      "step": 874
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.393670345339031e-05,
+      "loss": 0.0856,
+      "step": 875
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3923415473019517e-05,
+      "loss": 0.0746,
+      "step": 876
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0,
+      "learning_rate": 1.3910119299612199e-05,
+      "loss": 0.0849,
+      "step": 877
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.3896814960933973e-05,
+      "loss": 0.0573,
+      "step": 878
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.3883502484767504e-05,
+      "loss": 0.0865,
+      "step": 879
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.3870181898912437e-05,
+      "loss": 0.0814,
+      "step": 880
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3856853231185384e-05,
+      "loss": 0.08,
+      "step": 881
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.3843516509419795e-05,
+      "loss": 0.0711,
+      "step": 882
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.3830171761465968e-05,
+      "loss": 0.0808,
+      "step": 883
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3816819015190945e-05,
+      "loss": 0.0748,
+      "step": 884
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.380345829847848e-05,
+      "loss": 0.0764,
+      "step": 885
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.3790089639228969e-05,
+      "loss": 0.0802,
+      "step": 886
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.3776713065359388e-05,
+      "loss": 0.0713,
+      "step": 887
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3763328604803252e-05,
+      "loss": 0.0844,
+      "step": 888
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.374993628551053e-05,
+      "loss": 0.0692,
+      "step": 889
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.3736536135447615e-05,
+      "loss": 0.07,
+      "step": 890
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.3723128182597248e-05,
+      "loss": 0.0649,
+      "step": 891
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.3709712454958465e-05,
+      "loss": 0.0741,
+      "step": 892
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.3696288980546534e-05,
+      "loss": 0.0706,
+      "step": 893
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.3682857787392907e-05,
+      "loss": 0.0918,
+      "step": 894
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.366941890354514e-05,
+      "loss": 0.0687,
+      "step": 895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3655972357066874e-05,
+      "loss": 0.0813,
+      "step": 896
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.3642518176037727e-05,
+      "loss": 0.0712,
+      "step": 897
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.3629056388553272e-05,
+      "loss": 0.0631,
+      "step": 898
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.3615587022724967e-05,
+      "loss": 0.072,
+      "step": 899
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.3602110106680085e-05,
+      "loss": 0.1011,
+      "step": 900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3588625668561681e-05,
+      "loss": 0.0786,
+      "step": 901
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.3575133736528506e-05,
+      "loss": 0.0706,
+      "step": 902
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.3561634338754961e-05,
+      "loss": 0.0738,
+      "step": 903
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.3548127503431041e-05,
+      "loss": 0.0886,
+      "step": 904
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.3534613258762278e-05,
+      "loss": 0.0844,
+      "step": 905
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3521091632969663e-05,
+      "loss": 0.0594,
+      "step": 906
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.3507562654289613e-05,
+      "loss": 0.0671,
+      "step": 907
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.3494026350973888e-05,
+      "loss": 0.0895,
+      "step": 908
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.3480482751289552e-05,
+      "loss": 0.0657,
+      "step": 909
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.3466931883518903e-05,
+      "loss": 0.0824,
+      "step": 910
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.3453373775959419e-05,
+      "loss": 0.0852,
+      "step": 911
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3439808456923697e-05,
+      "loss": 0.0598,
+      "step": 912
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3426235954739384e-05,
+      "loss": 0.079,
+      "step": 913
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.3412656297749136e-05,
+      "loss": 0.0518,
+      "step": 914
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.3399069514310545e-05,
+      "loss": 0.0755,
+      "step": 915
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.3385475632796096e-05,
+      "loss": 0.1082,
+      "step": 916
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.337187468159308e-05,
+      "loss": 0.0699,
+      "step": 917
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.335826668910356e-05,
+      "loss": 0.0631,
+      "step": 918
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.3344651683744303e-05,
+      "loss": 0.0873,
+      "step": 919
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.3331029693946722e-05,
+      "loss": 0.0767,
+      "step": 920
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.3317400748156812e-05,
+      "loss": 0.0695,
+      "step": 921
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.875,
+      "learning_rate": 1.3303764874835092e-05,
+      "loss": 0.0774,
+      "step": 922
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.3290122102456553e-05,
+      "loss": 0.0768,
+      "step": 923
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.3276472459510584e-05,
+      "loss": 0.0678,
+      "step": 924
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.3262815974500934e-05,
+      "loss": 0.0825,
+      "step": 925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.324915267594563e-05,
+      "loss": 0.0852,
+      "step": 926
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.3235482592376931e-05,
+      "loss": 0.0936,
+      "step": 927
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3221805752341262e-05,
+      "loss": 0.0713,
+      "step": 928
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.320812218439916e-05,
+      "loss": 0.0955,
+      "step": 929
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.3194431917125217e-05,
+      "loss": 0.0792,
+      "step": 930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.3180734979107998e-05,
+      "loss": 0.0769,
+      "step": 931
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.3167031398950019e-05,
+      "loss": 0.0795,
+      "step": 932
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.3153321205267648e-05,
+      "loss": 0.0838,
+      "step": 933
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.3139604426691075e-05,
+      "loss": 0.0756,
+      "step": 934
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.3125881091864239e-05,
+      "loss": 0.071,
+      "step": 935
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.3112151229444766e-05,
+      "loss": 0.0833,
+      "step": 936
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.309841486810392e-05,
+      "loss": 0.0869,
+      "step": 937
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.3084672036526524e-05,
+      "loss": 0.0794,
+      "step": 938
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.3070922763410926e-05,
+      "loss": 0.0945,
+      "step": 939
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.3057167077468926e-05,
+      "loss": 0.0817,
+      "step": 940
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3043405007425695e-05,
+      "loss": 0.0818,
+      "step": 941
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.3029636582019762e-05,
+      "loss": 0.0805,
+      "step": 942
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.3015861830002911e-05,
+      "loss": 0.0877,
+      "step": 943
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.3002080780140142e-05,
+      "loss": 0.0627,
+      "step": 944
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.2988293461209609e-05,
+      "loss": 0.0859,
+      "step": 945
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.2974499902002551e-05,
+      "loss": 0.0655,
+      "step": 946
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.2960700131323247e-05,
+      "loss": 0.0878,
+      "step": 947
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.2946894177988934e-05,
+      "loss": 0.0701,
+      "step": 948
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2933082070829777e-05,
+      "loss": 0.0657,
+      "step": 949
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2919263838688779e-05,
+      "loss": 0.0637,
+      "step": 950
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.2905439510421734e-05,
+      "loss": 0.0565,
+      "step": 951
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.2891609114897175e-05,
+      "loss": 0.093,
+      "step": 952
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.287777268099629e-05,
+      "loss": 0.0574,
+      "step": 953
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2863930237612899e-05,
+      "loss": 0.0674,
+      "step": 954
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.2850081813653346e-05,
+      "loss": 0.0535,
+      "step": 955
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.2836227438036483e-05,
+      "loss": 0.075,
+      "step": 956
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.2822367139693579e-05,
+      "loss": 0.0755,
+      "step": 957
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.2808500947568277e-05,
+      "loss": 0.0751,
+      "step": 958
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.2794628890616525e-05,
+      "loss": 0.0605,
+      "step": 959
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.875,
+      "learning_rate": 1.2780750997806526e-05,
+      "loss": 0.0953,
+      "step": 960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.2766867298118653e-05,
+      "loss": 0.0874,
+      "step": 961
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.2752977820545417e-05,
+      "loss": 0.0762,
+      "step": 962
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2739082594091402e-05,
+      "loss": 0.067,
+      "step": 963
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.2725181647773176e-05,
+      "loss": 0.0812,
+      "step": 964
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2711275010619269e-05,
+      "loss": 0.0698,
+      "step": 965
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2697362711670087e-05,
+      "loss": 0.0718,
+      "step": 966
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.2683444779977865e-05,
+      "loss": 0.0871,
+      "step": 967
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.2669521244606588e-05,
+      "loss": 0.0749,
+      "step": 968
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.2655592134631965e-05,
+      "loss": 0.071,
+      "step": 969
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2641657479141319e-05,
+      "loss": 0.0602,
+      "step": 970
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2627717307233576e-05,
+      "loss": 0.0692,
+      "step": 971
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.261377164801917e-05,
+      "loss": 0.0807,
+      "step": 972
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2599820530619996e-05,
+      "loss": 0.0681,
+      "step": 973
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.2585863984169345e-05,
+      "loss": 0.069,
+      "step": 974
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.2571902037811852e-05,
+      "loss": 0.0776,
+      "step": 975
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.2557934720703425e-05,
+      "loss": 0.0814,
+      "step": 976
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.2543962062011179e-05,
+      "loss": 0.0615,
+      "step": 977
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.2529984090913398e-05,
+      "loss": 0.0869,
+      "step": 978
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.251600083659945e-05,
+      "loss": 0.0881,
+      "step": 979
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.2502012328269739e-05,
+      "loss": 0.0798,
+      "step": 980
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.2488018595135636e-05,
+      "loss": 0.0774,
+      "step": 981
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.247401966641943e-05,
+      "loss": 0.0683,
+      "step": 982
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.2460015571354255e-05,
+      "loss": 0.056,
+      "step": 983
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.2446006339184036e-05,
+      "loss": 0.0823,
+      "step": 984
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2431991999163425e-05,
+      "loss": 0.0639,
+      "step": 985
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2417972580557733e-05,
+      "loss": 0.0754,
+      "step": 986
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.240394811264289e-05,
+      "loss": 0.064,
+      "step": 987
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.238991862470536e-05,
+      "loss": 0.0655,
+      "step": 988
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.2375884146042093e-05,
+      "loss": 0.082,
+      "step": 989
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.236184470596046e-05,
+      "loss": 0.0809,
+      "step": 990
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.2347800333778197e-05,
+      "loss": 0.0717,
+      "step": 991
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.2333751058823331e-05,
+      "loss": 0.0848,
+      "step": 992
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2319696910434137e-05,
+      "loss": 0.0817,
+      "step": 993
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.2305637917959059e-05,
+      "loss": 0.0797,
+      "step": 994
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.229157411075666e-05,
+      "loss": 0.0772,
+      "step": 995
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2277505518195556e-05,
+      "loss": 0.0671,
+      "step": 996
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2263432169654356e-05,
+      "loss": 0.0721,
+      "step": 997
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.224935409452161e-05,
+      "loss": 0.0656,
+      "step": 998
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.223527132219572e-05,
+      "loss": 0.0709,
+      "step": 999
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.2221183882084913e-05,
+      "loss": 0.0754,
+      "step": 1000
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.220709180360715e-05,
+      "loss": 0.0645,
+      "step": 1001
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.2192995116190089e-05,
+      "loss": 0.0812,
+      "step": 1002
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.2178893849271012e-05,
+      "loss": 0.0658,
+      "step": 1003
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2164788032296758e-05,
+      "loss": 0.0702,
+      "step": 1004
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.215067769472367e-05,
+      "loss": 0.0749,
+      "step": 1005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.213656286601753e-05,
+      "loss": 0.0664,
+      "step": 1006
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.875,
+      "learning_rate": 1.2122443575653502e-05,
+      "loss": 0.089,
+      "step": 1007
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.2108319853116066e-05,
+      "loss": 0.0655,
+      "step": 1008
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2094191727898958e-05,
+      "loss": 0.0743,
+      "step": 1009
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.2080059229505103e-05,
+      "loss": 0.0858,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2065922387446563e-05,
+      "loss": 0.077,
+      "step": 1011
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.2051781231244465e-05,
+      "loss": 0.0787,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2037635790428958e-05,
+      "loss": 0.0711,
+      "step": 1013
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2023486094539125e-05,
+      "loss": 0.0667,
+      "step": 1014
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2009332173122936e-05,
+      "loss": 0.0608,
+      "step": 1015
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.1995174055737196e-05,
+      "loss": 0.0609,
+      "step": 1016
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1981011771947455e-05,
+      "loss": 0.082,
+      "step": 1017
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.196684535132798e-05,
+      "loss": 0.0888,
+      "step": 1018
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.1952674823461666e-05,
+      "loss": 0.0591,
+      "step": 1019
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.193850021793999e-05,
+      "loss": 0.0833,
+      "step": 1020
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.1924321564362938e-05,
+      "loss": 0.0764,
+      "step": 1021
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.1910138892338961e-05,
+      "loss": 0.0619,
+      "step": 1022
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.1895952231484888e-05,
+      "loss": 0.0615,
+      "step": 1023
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.188176161142589e-05,
+      "loss": 0.0781,
+      "step": 1024
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.1867567061795395e-05,
+      "loss": 0.08,
+      "step": 1025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.75,
+      "learning_rate": 1.1853368612235043e-05,
+      "loss": 0.07,
+      "step": 1026
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.1839166292394615e-05,
+      "loss": 0.076,
+      "step": 1027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.75,
+      "learning_rate": 1.182496013193198e-05,
+      "loss": 0.0716,
+      "step": 1028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.181075016051302e-05,
+      "loss": 0.068,
+      "step": 1029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.1796536407811576e-05,
+      "loss": 0.062,
+      "step": 1030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.1782318903509389e-05,
+      "loss": 0.0975,
+      "step": 1031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.1768097677296034e-05,
+      "loss": 0.0838,
+      "step": 1032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.1753872758868857e-05,
+      "loss": 0.0694,
+      "step": 1033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.173964417793291e-05,
+      "loss": 0.0937,
+      "step": 1034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.1725411964200902e-05,
+      "loss": 0.0788,
+      "step": 1035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.1711176147393118e-05,
+      "loss": 0.0902,
+      "step": 1036
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.1696936757237378e-05,
+      "loss": 0.0679,
+      "step": 1037
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.1682693823468953e-05,
+      "loss": 0.066,
+      "step": 1038
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.1668447375830525e-05,
+      "loss": 0.0858,
+      "step": 1039
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.1654197444072106e-05,
+      "loss": 0.0702,
+      "step": 1040
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1639944057950981e-05,
+      "loss": 0.0745,
+      "step": 1041
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.162568724723166e-05,
+      "loss": 0.0742,
+      "step": 1042
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.1611427041685797e-05,
+      "loss": 0.0572,
+      "step": 1043
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.1597163471092132e-05,
+      "loss": 0.0756,
+      "step": 1044
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.1582896565236439e-05,
+      "loss": 0.0817,
+      "step": 1045
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.1568626353911452e-05,
+      "loss": 0.0749,
+      "step": 1046
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.1554352866916811e-05,
+      "loss": 0.0855,
+      "step": 1047
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.1540076134058992e-05,
+      "loss": 0.0717,
+      "step": 1048
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.152579618515126e-05,
+      "loss": 0.0704,
+      "step": 1049
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.1511513050013579e-05,
+      "loss": 0.0634,
+      "step": 1050
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.1497226758472578e-05,
+      "loss": 0.0815,
+      "step": 1051
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.1482937340361475e-05,
+      "loss": 0.0626,
+      "step": 1052
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1468644825520019e-05,
+      "loss": 0.0661,
+      "step": 1053
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.1454349243794421e-05,
+      "loss": 0.0743,
+      "step": 1054
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.14400506250373e-05,
+      "loss": 0.0725,
+      "step": 1055
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.1425748999107614e-05,
+      "loss": 0.0579,
+      "step": 1056
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.1411444395870608e-05,
+      "loss": 0.0706,
+      "step": 1057
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.1397136845197733e-05,
+      "loss": 0.0673,
+      "step": 1058
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1382826376966602e-05,
+      "loss": 0.0687,
+      "step": 1059
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.1368513021060922e-05,
+      "loss": 0.0692,
+      "step": 1060
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.135419680737043e-05,
+      "loss": 0.0699,
+      "step": 1061
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1339877765790824e-05,
+      "loss": 0.073,
+      "step": 1062
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.1325555926223713e-05,
+      "loss": 0.0893,
+      "step": 1063
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.1311231318576548e-05,
+      "loss": 0.0567,
+      "step": 1064
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.1296903972762556e-05,
+      "loss": 0.0894,
+      "step": 1065
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.1282573918700695e-05,
+      "loss": 0.0849,
+      "step": 1066
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1268241186315561e-05,
+      "loss": 0.0721,
+      "step": 1067
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1253905805537357e-05,
+      "loss": 0.0785,
+      "step": 1068
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1239567806301807e-05,
+      "loss": 0.0774,
+      "step": 1069
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.1225227218550107e-05,
+      "loss": 0.0776,
+      "step": 1070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.1210884072228862e-05,
+      "loss": 0.0649,
+      "step": 1071
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.1196538397290015e-05,
+      "loss": 0.0635,
+      "step": 1072
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.118219022369079e-05,
+      "loss": 0.0715,
+      "step": 1073
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.1167839581393628e-05,
+      "loss": 0.0713,
+      "step": 1074
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.1153486500366129e-05,
+      "loss": 0.0799,
+      "step": 1075
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.1139131010580981e-05,
+      "loss": 0.06,
+      "step": 1076
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.112477314201591e-05,
+      "loss": 0.0821,
+      "step": 1077
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1110412924653597e-05,
+      "loss": 0.0744,
+      "step": 1078
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.1096050388481636e-05,
+      "loss": 0.0843,
+      "step": 1079
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1081685563492463e-05,
+      "loss": 0.0823,
+      "step": 1080
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1067318479683295e-05,
+      "loss": 0.071,
+      "step": 1081
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.1052949167056064e-05,
+      "loss": 0.0756,
+      "step": 1082
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.1038577655617349e-05,
+      "loss": 0.078,
+      "step": 1083
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1024203975378336e-05,
+      "loss": 0.0735,
+      "step": 1084
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.100982815635472e-05,
+      "loss": 0.0731,
+      "step": 1085
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.0995450228566689e-05,
+      "loss": 0.0742,
+      "step": 1086
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.0981070222038808e-05,
+      "loss": 0.0635,
+      "step": 1087
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.09666881668e-05,
+      "loss": 0.0713,
+      "step": 1088
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0952304092883454e-05,
+      "loss": 0.0812,
+      "step": 1089
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.0937918030326587e-05,
+      "loss": 0.0591,
+      "step": 1090
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.092353000917096e-05,
+      "loss": 0.0573,
+      "step": 1091
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.090914005946223e-05,
+      "loss": 0.0576,
+      "step": 1092
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0894748211250072e-05,
+      "loss": 0.0756,
+      "step": 1093
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.088035449458814e-05,
+      "loss": 0.065,
+      "step": 1094
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.0865958939533972e-05,
+      "loss": 0.0549,
+      "step": 1095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.085156157614896e-05,
+      "loss": 0.0593,
+      "step": 1096
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.0837162434498267e-05,
+      "loss": 0.0719,
+      "step": 1097
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0822761544650766e-05,
+      "loss": 0.0781,
+      "step": 1098
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.0808358936678984e-05,
+      "loss": 0.0677,
+      "step": 1099
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.079395464065904e-05,
+      "loss": 0.0704,
+      "step": 1100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.077954868667057e-05,
+      "loss": 0.0938,
+      "step": 1101
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.75,
+      "learning_rate": 1.0765141104796673e-05,
+      "loss": 0.068,
+      "step": 1102
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0750731925123854e-05,
+      "loss": 0.0721,
+      "step": 1103
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.0736321177741951e-05,
+      "loss": 0.1041,
+      "step": 1104
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.0721908892744073e-05,
+      "loss": 0.0523,
+      "step": 1105
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.0707495100226547e-05,
+      "loss": 0.0625,
+      "step": 1106
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.0693079830288839e-05,
+      "loss": 0.0555,
+      "step": 1107
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.067866311303351e-05,
+      "loss": 0.0696,
+      "step": 1108
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.066424497856613e-05,
+      "loss": 0.0573,
+      "step": 1109
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0,
+      "learning_rate": 1.0649825456995242e-05,
+      "loss": 0.0807,
+      "step": 1110
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.063540457843228e-05,
+      "loss": 0.057,
+      "step": 1111
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.0620982372991514e-05,
+      "loss": 0.062,
+      "step": 1112
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.0606558870789978e-05,
+      "loss": 0.0842,
+      "step": 1113
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.125,
+      "learning_rate": 1.0592134101947418e-05,
+      "loss": 0.0939,
+      "step": 1114
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.25,
+      "learning_rate": 1.0577708096586232e-05,
+      "loss": 0.0667,
+      "step": 1115
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.056328088483139e-05,
+      "loss": 0.0734,
+      "step": 1116
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.0548852496810387e-05,
+      "loss": 0.0666,
+      "step": 1117
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.0534422962653169e-05,
+      "loss": 0.0705,
+      "step": 1118
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.0519992312492081e-05,
+      "loss": 0.0686,
+      "step": 1119
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.0505560576461798e-05,
+      "loss": 0.0834,
+      "step": 1120
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.049112778469926e-05,
+      "loss": 0.0621,
+      "step": 1121
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.047669396734361e-05,
+      "loss": 0.0937,
+      "step": 1122
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.0462259154536136e-05,
+      "loss": 0.0647,
+      "step": 1123
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.0447823376420208e-05,
+      "loss": 0.0676,
+      "step": 1124
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.0433386663141202e-05,
+      "loss": 0.0624,
+      "step": 1125
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.0418949044846461e-05,
+      "loss": 0.0688,
+      "step": 1126
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.0404510551685199e-05,
+      "loss": 0.0671,
+      "step": 1127
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0390071213808475e-05,
+      "loss": 0.0662,
+      "step": 1128
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.0375631061369098e-05,
+      "loss": 0.0694,
+      "step": 1129
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.875,
+      "learning_rate": 1.0361190124521595e-05,
+      "loss": 0.0732,
+      "step": 1130
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.0346748433422109e-05,
+      "loss": 0.0755,
+      "step": 1131
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.0332306018228377e-05,
+      "loss": 0.0842,
+      "step": 1132
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.0,
+      "learning_rate": 1.0317862909099636e-05,
+      "loss": 0.0697,
+      "step": 1133
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.0303419136196578e-05,
+      "loss": 0.064,
+      "step": 1134
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.0288974729681282e-05,
+      "loss": 0.0641,
+      "step": 1135
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0274529719717145e-05,
+      "loss": 0.0705,
+      "step": 1136
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.026008413646883e-05,
+      "loss": 0.102,
+      "step": 1137
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.0245638010102192e-05,
+      "loss": 0.0725,
+      "step": 1138
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.0231191370784226e-05,
+      "loss": 0.0845,
+      "step": 1139
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.0216744248682988e-05,
+      "loss": 0.0923,
+      "step": 1140
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0202296673967558e-05,
+      "loss": 0.0724,
+      "step": 1141
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0187848676807942e-05,
+      "loss": 0.0468,
+      "step": 1142
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.0173400287375041e-05,
+      "loss": 0.0774,
+      "step": 1143
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.0158951535840577e-05,
+      "loss": 0.073,
+      "step": 1144
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0144502452377019e-05,
+      "loss": 0.0893,
+      "step": 1145
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.0130053067157537e-05,
+      "loss": 0.0868,
+      "step": 1146
+    },
+    {
+      "epoch": 1.02,
+      "eval_loss": 0.07227291166782379,
+      "eval_runtime": 228.7587,
+      "eval_samples_per_second": 34.849,
+      "eval_steps_per_second": 4.358,
+      "step": 1146
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.0115603410355923e-05,
+      "loss": 0.0642,
+      "step": 1147
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.0101153512146544e-05,
+      "loss": 0.0673,
+      "step": 1148
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0086703402704267e-05,
+      "loss": 0.0768,
+      "step": 1149
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.0072253112204405e-05,
+      "loss": 0.1185,
+      "step": 1150
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.0057802670822636e-05,
+      "loss": 0.0771,
+      "step": 1151
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.0043352108734971e-05,
+      "loss": 0.0728,
+      "step": 1152
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.002890145611766e-05,
+      "loss": 0.0732,
+      "step": 1153
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.0014450743147147e-05,
+      "loss": 0.0612,
+      "step": 1154
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.765625,
+      "learning_rate": 1e-05,
+      "loss": 0.0662,
+      "step": 1155
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.985549256852855e-06,
+      "loss": 0.0675,
+      "step": 1156
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.971098543882345e-06,
+      "loss": 0.0539,
+      "step": 1157
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.95664789126503e-06,
+      "loss": 0.0747,
+      "step": 1158
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.942197329177369e-06,
+      "loss": 0.0701,
+      "step": 1159
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.927746887795599e-06,
+      "loss": 0.0745,
+      "step": 1160
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.913296597295735e-06,
+      "loss": 0.0709,
+      "step": 1161
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.898846487853459e-06,
+      "loss": 0.0952,
+      "step": 1162
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.88439658964408e-06,
+      "loss": 0.0554,
+      "step": 1163
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.869946932842468e-06,
+      "loss": 0.0838,
+      "step": 1164
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.855497547622985e-06,
+      "loss": 0.0655,
+      "step": 1165
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.841048464159424e-06,
+      "loss": 0.0774,
+      "step": 1166
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.826599712624962e-06,
+      "loss": 0.063,
+      "step": 1167
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.812151323192062e-06,
+      "loss": 0.0686,
+      "step": 1168
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.797703326032449e-06,
+      "loss": 0.0791,
+      "step": 1169
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.783255751317015e-06,
+      "loss": 0.0737,
+      "step": 1170
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.768808629215776e-06,
+      "loss": 0.061,
+      "step": 1171
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.754361989897811e-06,
+      "loss": 0.098,
+      "step": 1172
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.739915863531171e-06,
+      "loss": 0.0645,
+      "step": 1173
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.725470280282857e-06,
+      "loss": 0.0489,
+      "step": 1174
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.711025270318721e-06,
+      "loss": 0.075,
+      "step": 1175
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.696580863803424e-06,
+      "loss": 0.0826,
+      "step": 1176
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.68213709090037e-06,
+      "loss": 0.0718,
+      "step": 1177
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.667693981771627e-06,
+      "loss": 0.0641,
+      "step": 1178
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.653251566577898e-06,
+      "loss": 0.0831,
+      "step": 1179
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.63880987547841e-06,
+      "loss": 0.061,
+      "step": 1180
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.624368938630902e-06,
+      "loss": 0.073,
+      "step": 1181
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.60992878619153e-06,
+      "loss": 0.0928,
+      "step": 1182
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.875,
+      "learning_rate": 9.595489448314805e-06,
+      "loss": 0.0682,
+      "step": 1183
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.581050955153546e-06,
+      "loss": 0.0921,
+      "step": 1184
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.566613336858801e-06,
+      "loss": 0.0622,
+      "step": 1185
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.552176623579793e-06,
+      "loss": 0.0723,
+      "step": 1186
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.537740845463866e-06,
+      "loss": 0.0725,
+      "step": 1187
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.523306032656394e-06,
+      "loss": 0.0647,
+      "step": 1188
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.508872215300747e-06,
+      "loss": 0.072,
+      "step": 1189
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.494439423538205e-06,
+      "loss": 0.0805,
+      "step": 1190
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.48000768750792e-06,
+      "loss": 0.0586,
+      "step": 1191
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.465577037346836e-06,
+      "loss": 0.0613,
+      "step": 1192
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.451147503189616e-06,
+      "loss": 0.0603,
+      "step": 1193
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.436719115168611e-06,
+      "loss": 0.0834,
+      "step": 1194
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.422291903413771e-06,
+      "loss": 0.078,
+      "step": 1195
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.407865898052582e-06,
+      "loss": 0.0609,
+      "step": 1196
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.393441129210029e-06,
+      "loss": 0.0721,
+      "step": 1197
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.379017627008492e-06,
+      "loss": 0.072,
+      "step": 1198
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.36459542156772e-06,
+      "loss": 0.0609,
+      "step": 1199
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.35017454300476e-06,
+      "loss": 0.0674,
+      "step": 1200
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.75,
+      "learning_rate": 9.335755021433872e-06,
+      "loss": 0.0624,
+      "step": 1201
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.321336886966495e-06,
+      "loss": 0.0741,
+      "step": 1202
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.306920169711163e-06,
+      "loss": 0.0466,
+      "step": 1203
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.292504899773454e-06,
+      "loss": 0.0585,
+      "step": 1204
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.27809110725593e-06,
+      "loss": 0.0547,
+      "step": 1205
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.26367882225805e-06,
+      "loss": 0.0804,
+      "step": 1206
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.24926807487615e-06,
+      "loss": 0.0735,
+      "step": 1207
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.23485889520333e-06,
+      "loss": 0.0744,
+      "step": 1208
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.220451313329434e-06,
+      "loss": 0.0649,
+      "step": 1209
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.206045359340964e-06,
+      "loss": 0.0699,
+      "step": 1210
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.191641063321018e-06,
+      "loss": 0.0756,
+      "step": 1211
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.0,
+      "learning_rate": 9.177238455349239e-06,
+      "loss": 0.0811,
+      "step": 1212
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.162837565501736e-06,
+      "loss": 0.056,
+      "step": 1213
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.14843842385104e-06,
+      "loss": 0.0706,
+      "step": 1214
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.134041060466031e-06,
+      "loss": 0.0656,
+      "step": 1215
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.119645505411866e-06,
+      "loss": 0.064,
+      "step": 1216
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.105251788749932e-06,
+      "loss": 0.077,
+      "step": 1217
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.090859940537776e-06,
+      "loss": 0.0639,
+      "step": 1218
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.07646999082904e-06,
+      "loss": 0.0694,
+      "step": 1219
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.28125,
+      "learning_rate": 9.062081969673417e-06,
+      "loss": 0.0761,
+      "step": 1220
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.04769590711655e-06,
+      "loss": 0.0815,
+      "step": 1221
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.033311833200007e-06,
+      "loss": 0.078,
+      "step": 1222
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.018929777961197e-06,
+      "loss": 0.0632,
+      "step": 1223
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.004549771433313e-06,
+      "loss": 0.0719,
+      "step": 1224
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.990171843645281e-06,
+      "loss": 0.0602,
+      "step": 1225
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.975796024621668e-06,
+      "loss": 0.0692,
+      "step": 1226
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.961422344382653e-06,
+      "loss": 0.0528,
+      "step": 1227
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.94705083294394e-06,
+      "loss": 0.0683,
+      "step": 1228
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.932681520316705e-06,
+      "loss": 0.0616,
+      "step": 1229
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.918314436507539e-06,
+      "loss": 0.0777,
+      "step": 1230
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.903949611518367e-06,
+      "loss": 0.0832,
+      "step": 1231
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.889587075346408e-06,
+      "loss": 0.0783,
+      "step": 1232
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.875226857984093e-06,
+      "loss": 0.0552,
+      "step": 1233
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.860868989419017e-06,
+      "loss": 0.0726,
+      "step": 1234
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.846513499633873e-06,
+      "loss": 0.0632,
+      "step": 1235
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.832160418606373e-06,
+      "loss": 0.0589,
+      "step": 1236
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.817809776309215e-06,
+      "loss": 0.0774,
+      "step": 1237
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.803461602709988e-06,
+      "loss": 0.0639,
+      "step": 1238
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.789115927771138e-06,
+      "loss": 0.0529,
+      "step": 1239
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.774772781449894e-06,
+      "loss": 0.0805,
+      "step": 1240
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.760432193698196e-06,
+      "loss": 0.0687,
+      "step": 1241
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.746094194462648e-06,
+      "loss": 0.054,
+      "step": 1242
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.731758813684442e-06,
+      "loss": 0.0713,
+      "step": 1243
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.717426081299308e-06,
+      "loss": 0.0511,
+      "step": 1244
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.703096027237446e-06,
+      "loss": 0.0664,
+      "step": 1245
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.688768681423455e-06,
+      "loss": 0.0527,
+      "step": 1246
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.674444073776292e-06,
+      "loss": 0.0652,
+      "step": 1247
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.660122234209178e-06,
+      "loss": 0.0737,
+      "step": 1248
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.64580319262957e-06,
+      "loss": 0.0573,
+      "step": 1249
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.63148697893908e-06,
+      "loss": 0.0588,
+      "step": 1250
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.6171736230334e-06,
+      "loss": 0.0605,
+      "step": 1251
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.602863154802272e-06,
+      "loss": 0.0658,
+      "step": 1252
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.588555604129395e-06,
+      "loss": 0.0696,
+      "step": 1253
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.574251000892386e-06,
+      "loss": 0.0755,
+      "step": 1254
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.559949374962703e-06,
+      "loss": 0.0595,
+      "step": 1255
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.545650756205582e-06,
+      "loss": 0.055,
+      "step": 1256
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.531355174479986e-06,
+      "loss": 0.0652,
+      "step": 1257
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.517062659638528e-06,
+      "loss": 0.0695,
+      "step": 1258
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.502773241527424e-06,
+      "loss": 0.0759,
+      "step": 1259
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.4140625,
+      "learning_rate": 8.488486949986428e-06,
+      "loss": 0.0924,
+      "step": 1260
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.474203814848742e-06,
+      "loss": 0.0632,
+      "step": 1261
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.875,
+      "learning_rate": 8.45992386594101e-06,
+      "loss": 0.0673,
+      "step": 1262
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.445647133083192e-06,
+      "loss": 0.0819,
+      "step": 1263
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.431373646088551e-06,
+      "loss": 0.0635,
+      "step": 1264
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.417103434763566e-06,
+      "loss": 0.0735,
+      "step": 1265
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.40283652890787e-06,
+      "loss": 0.0585,
+      "step": 1266
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.38857295831421e-06,
+      "loss": 0.0679,
+      "step": 1267
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.374312752768341e-06,
+      "loss": 0.07,
+      "step": 1268
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.36005594204902e-06,
+      "loss": 0.0548,
+      "step": 1269
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.345802555927899e-06,
+      "loss": 0.0708,
+      "step": 1270
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.331552624169477e-06,
+      "loss": 0.0753,
+      "step": 1271
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.317306176531045e-06,
+      "loss": 0.0561,
+      "step": 1272
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.303063242762625e-06,
+      "loss": 0.0719,
+      "step": 1273
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.288823852606884e-06,
+      "loss": 0.0789,
+      "step": 1274
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.274588035799103e-06,
+      "loss": 0.0582,
+      "step": 1275
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.260355822067092e-06,
+      "loss": 0.0632,
+      "step": 1276
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.246127241131145e-06,
+      "loss": 0.0827,
+      "step": 1277
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.23190232270397e-06,
+      "loss": 0.0853,
+      "step": 1278
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.217681096490611e-06,
+      "loss": 0.0707,
+      "step": 1279
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.20346359218843e-06,
+      "loss": 0.0629,
+      "step": 1280
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.189249839486983e-06,
+      "loss": 0.0902,
+      "step": 1281
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.175039868068022e-06,
+      "loss": 0.077,
+      "step": 1282
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.160833707605387e-06,
+      "loss": 0.0713,
+      "step": 1283
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.875,
+      "learning_rate": 8.14663138776496e-06,
+      "loss": 0.0715,
+      "step": 1284
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.13243293820461e-06,
+      "loss": 0.0716,
+      "step": 1285
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.118238388574112e-06,
+      "loss": 0.0499,
+      "step": 1286
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.104047768515112e-06,
+      "loss": 0.0764,
+      "step": 1287
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.08986110766104e-06,
+      "loss": 0.0862,
+      "step": 1288
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.075678435637064e-06,
+      "loss": 0.0781,
+      "step": 1289
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.061499782060016e-06,
+      "loss": 0.0675,
+      "step": 1290
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.047325176538338e-06,
+      "loss": 0.0552,
+      "step": 1291
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.033154648672021e-06,
+      "loss": 0.0585,
+      "step": 1292
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.018988228052548e-06,
+      "loss": 0.0739,
+      "step": 1293
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.004825944262806e-06,
+      "loss": 0.0623,
+      "step": 1294
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.990667826877066e-06,
+      "loss": 0.0791,
+      "step": 1295
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.976513905460878e-06,
+      "loss": 0.0702,
+      "step": 1296
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.962364209571044e-06,
+      "loss": 0.0759,
+      "step": 1297
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.948218768755536e-06,
+      "loss": 0.0805,
+      "step": 1298
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.934077612553439e-06,
+      "loss": 0.0649,
+      "step": 1299
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.919940770494902e-06,
+      "loss": 0.0696,
+      "step": 1300
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.905808272101044e-06,
+      "loss": 0.0598,
+      "step": 1301
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.891680146883934e-06,
+      "loss": 0.0909,
+      "step": 1302
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.8775564243465e-06,
+      "loss": 0.0688,
+      "step": 1303
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.863437133982472e-06,
+      "loss": 0.0614,
+      "step": 1304
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.849322305276335e-06,
+      "loss": 0.0686,
+      "step": 1305
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.835211967703244e-06,
+      "loss": 0.0617,
+      "step": 1306
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.75,
+      "learning_rate": 7.821106150728988e-06,
+      "loss": 0.0652,
+      "step": 1307
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.807004883809913e-06,
+      "loss": 0.0613,
+      "step": 1308
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.792908196392854e-06,
+      "loss": 0.0772,
+      "step": 1309
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.778816117915094e-06,
+      "loss": 0.072,
+      "step": 1310
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.764728677804282e-06,
+      "loss": 0.0751,
+      "step": 1311
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.0,
+      "learning_rate": 7.750645905478391e-06,
+      "loss": 0.0804,
+      "step": 1312
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.736567830345646e-06,
+      "loss": 0.0805,
+      "step": 1313
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.722494481804446e-06,
+      "loss": 0.0755,
+      "step": 1314
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.708425889243346e-06,
+      "loss": 0.0652,
+      "step": 1315
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.694362082040945e-06,
+      "loss": 0.0695,
+      "step": 1316
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.680303089565865e-06,
+      "loss": 0.052,
+      "step": 1317
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.666248941176672e-06,
+      "loss": 0.0736,
+      "step": 1318
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.652199666221805e-06,
+      "loss": 0.0711,
+      "step": 1319
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.638155294039544e-06,
+      "loss": 0.0653,
+      "step": 1320
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.62411585395791e-06,
+      "loss": 0.0647,
+      "step": 1321
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.610081375294642e-06,
+      "loss": 0.0761,
+      "step": 1322
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.596051887357113e-06,
+      "loss": 0.0624,
+      "step": 1323
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.582027419442269e-06,
+      "loss": 0.062,
+      "step": 1324
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.568008000836581e-06,
+      "loss": 0.0689,
+      "step": 1325
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.553993660815967e-06,
+      "loss": 0.0643,
+      "step": 1326
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.5399844286457445e-06,
+      "loss": 0.0565,
+      "step": 1327
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.525980333580572e-06,
+      "loss": 0.049,
+      "step": 1328
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.511981404864366e-06,
+      "loss": 0.0783,
+      "step": 1329
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.497987671730267e-06,
+      "loss": 0.0793,
+      "step": 1330
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.483999163400554e-06,
+      "loss": 0.0585,
+      "step": 1331
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.470015909086603e-06,
+      "loss": 0.0803,
+      "step": 1332
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.456037937988825e-06,
+      "loss": 0.0776,
+      "step": 1333
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.442065279296579e-06,
+      "loss": 0.0621,
+      "step": 1334
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.42809796218815e-06,
+      "loss": 0.0675,
+      "step": 1335
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.414136015830658e-06,
+      "loss": 0.0721,
+      "step": 1336
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.4001794693800085e-06,
+      "loss": 0.08,
+      "step": 1337
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.386228351980835e-06,
+      "loss": 0.0599,
+      "step": 1338
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.3722826927664285e-06,
+      "loss": 0.0582,
+      "step": 1339
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.358342520858686e-06,
+      "loss": 0.0816,
+      "step": 1340
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.34440786536804e-06,
+      "loss": 0.0527,
+      "step": 1341
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.330478755393412e-06,
+      "loss": 0.0794,
+      "step": 1342
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.125,
+      "learning_rate": 7.31655522002214e-06,
+      "loss": 0.0738,
+      "step": 1343
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.302637288329916e-06,
+      "loss": 0.0762,
+      "step": 1344
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.2887249893807355e-06,
+      "loss": 0.0754,
+      "step": 1345
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.274818352226829e-06,
+      "loss": 0.0639,
+      "step": 1346
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.260917405908601e-06,
+      "loss": 0.07,
+      "step": 1347
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.2470221794545834e-06,
+      "loss": 0.0577,
+      "step": 1348
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.23313270188135e-06,
+      "loss": 0.0777,
+      "step": 1349
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.219249002193478e-06,
+      "loss": 0.065,
+      "step": 1350
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.205371109383477e-06,
+      "loss": 0.0598,
+      "step": 1351
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.191499052431724e-06,
+      "loss": 0.0606,
+      "step": 1352
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.177632860306426e-06,
+      "loss": 0.0667,
+      "step": 1353
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.16377256196352e-06,
+      "loss": 0.0618,
+      "step": 1354
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.149918186346656e-06,
+      "loss": 0.0665,
+      "step": 1355
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.1360697623871055e-06,
+      "loss": 0.0685,
+      "step": 1356
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.122227319003711e-06,
+      "loss": 0.083,
+      "step": 1357
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.108390885102832e-06,
+      "loss": 0.065,
+      "step": 1358
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.094560489578269e-06,
+      "loss": 0.0715,
+      "step": 1359
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.080736161311224e-06,
+      "loss": 0.0626,
+      "step": 1360
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.066917929170225e-06,
+      "loss": 0.0587,
+      "step": 1361
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.053105822011066e-06,
+      "loss": 0.0709,
+      "step": 1362
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.039299868676758e-06,
+      "loss": 0.0627,
+      "step": 1363
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.0255000979974505e-06,
+      "loss": 0.0887,
+      "step": 1364
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.01170653879039e-06,
+      "loss": 0.0586,
+      "step": 1365
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.997919219859861e-06,
+      "loss": 0.0724,
+      "step": 1366
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.98413816999709e-06,
+      "loss": 0.0665,
+      "step": 1367
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.970363417980241e-06,
+      "loss": 0.0728,
+      "step": 1368
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.956594992574306e-06,
+      "loss": 0.0558,
+      "step": 1369
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.942832922531077e-06,
+      "loss": 0.061,
+      "step": 1370
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.9290772365890744e-06,
+      "loss": 0.0784,
+      "step": 1371
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.915327963473478e-06,
+      "loss": 0.073,
+      "step": 1372
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.901585131896086e-06,
+      "loss": 0.0658,
+      "step": 1373
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.887848770555234e-06,
+      "loss": 0.0613,
+      "step": 1374
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.874118908135762e-06,
+      "loss": 0.0742,
+      "step": 1375
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.860395573308927e-06,
+      "loss": 0.0525,
+      "step": 1376
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.846678794732354e-06,
+      "loss": 0.0727,
+      "step": 1377
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.84375,
+      "learning_rate": 6.8329686010499855e-06,
+      "loss": 0.0747,
+      "step": 1378
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.819265020892004e-06,
+      "loss": 0.0652,
+      "step": 1379
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.805568082874785e-06,
+      "loss": 0.0551,
+      "step": 1380
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.79187781560084e-06,
+      "loss": 0.0584,
+      "step": 1381
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.778194247658741e-06,
+      "loss": 0.0544,
+      "step": 1382
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.764517407623074e-06,
+      "loss": 0.0565,
+      "step": 1383
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.85546875,
+      "learning_rate": 6.750847324054375e-06,
+      "loss": 0.06,
+      "step": 1384
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.737184025499067e-06,
+      "loss": 0.0608,
+      "step": 1385
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8671875,
+      "learning_rate": 6.72352754048942e-06,
+      "loss": 0.0655,
+      "step": 1386
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.709877897543451e-06,
+      "loss": 0.0514,
+      "step": 1387
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.696235125164914e-06,
+      "loss": 0.058,
+      "step": 1388
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.68259925184319e-06,
+      "loss": 0.0616,
+      "step": 1389
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.668970306053278e-06,
+      "loss": 0.0685,
+      "step": 1390
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8203125,
+      "learning_rate": 6.655348316255699e-06,
+      "loss": 0.0528,
+      "step": 1391
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.641733310896443e-06,
+      "loss": 0.0636,
+      "step": 1392
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.628125318406926e-06,
+      "loss": 0.065,
+      "step": 1393
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.614524367203907e-06,
+      "loss": 0.0628,
+      "step": 1394
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.6009304856894545e-06,
+      "loss": 0.0722,
+      "step": 1395
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.587343702250868e-06,
+      "loss": 0.0565,
+      "step": 1396
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.57376404526062e-06,
+      "loss": 0.062,
+      "step": 1397
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.560191543076307e-06,
+      "loss": 0.0709,
+      "step": 1398
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.546626224040582e-06,
+      "loss": 0.0665,
+      "step": 1399
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.533068116481096e-06,
+      "loss": 0.0679,
+      "step": 1400
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.5195172487104505e-06,
+      "loss": 0.0632,
+      "step": 1401
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.505973649026115e-06,
+      "loss": 0.0673,
+      "step": 1402
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.492437345710391e-06,
+      "loss": 0.0607,
+      "step": 1403
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.4789083670303385e-06,
+      "loss": 0.0641,
+      "step": 1404
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.465386741237723e-06,
+      "loss": 0.0808,
+      "step": 1405
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.451872496568961e-06,
+      "loss": 0.0577,
+      "step": 1406
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.438365661245041e-06,
+      "loss": 0.066,
+      "step": 1407
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.424866263471501e-06,
+      "loss": 0.0704,
+      "step": 1408
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.411374331438321e-06,
+      "loss": 0.0622,
+      "step": 1409
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.397889893319915e-06,
+      "loss": 0.0677,
+      "step": 1410
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.384412977275038e-06,
+      "loss": 0.0539,
+      "step": 1411
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.37094361144673e-06,
+      "loss": 0.0847,
+      "step": 1412
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.357481823962277e-06,
+      "loss": 0.0473,
+      "step": 1413
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.85546875,
+      "learning_rate": 6.344027642933129e-06,
+      "loss": 0.0594,
+      "step": 1414
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.33058109645486e-06,
+      "loss": 0.0503,
+      "step": 1415
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.317142212607099e-06,
+      "loss": 0.049,
+      "step": 1416
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.303711019453469e-06,
+      "loss": 0.0672,
+      "step": 1417
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.875,
+      "learning_rate": 6.29028754504154e-06,
+      "loss": 0.0668,
+      "step": 1418
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.83984375,
+      "learning_rate": 6.2768718174027546e-06,
+      "loss": 0.0571,
+      "step": 1419
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.263463864552386e-06,
+      "loss": 0.0637,
+      "step": 1420
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.250063714489475e-06,
+      "loss": 0.0701,
+      "step": 1421
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.236671395196752e-06,
+      "loss": 0.0453,
+      "step": 1422
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.223286934640613e-06,
+      "loss": 0.0489,
+      "step": 1423
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.2099103607710345e-06,
+      "loss": 0.052,
+      "step": 1424
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.196541701521521e-06,
+      "loss": 0.0638,
+      "step": 1425
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.183180984809059e-06,
+      "loss": 0.0556,
+      "step": 1426
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.85546875,
+      "learning_rate": 6.169828238534035e-06,
+      "loss": 0.0632,
+      "step": 1427
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.156483490580206e-06,
+      "loss": 0.045,
+      "step": 1428
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.14314676881462e-06,
+      "loss": 0.0609,
+      "step": 1429
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.129818101087561e-06,
+      "loss": 0.0587,
+      "step": 1430
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.116497515232502e-06,
+      "loss": 0.0596,
+      "step": 1431
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.1031850390660295e-06,
+      "loss": 0.0647,
+      "step": 1432
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.0898807003878e-06,
+      "loss": 0.0572,
+      "step": 1433
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.076584526980484e-06,
+      "loss": 0.0747,
+      "step": 1434
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.063296546609689e-06,
+      "loss": 0.0663,
+      "step": 1435
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.050016787023922e-06,
+      "loss": 0.0532,
+      "step": 1436
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.036745275954514e-06,
+      "loss": 0.0705,
+      "step": 1437
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.023482041115581e-06,
+      "loss": 0.0558,
+      "step": 1438
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.010227110203952e-06,
+      "loss": 0.0617,
+      "step": 1439
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.87109375,
+      "learning_rate": 5.996980510899111e-06,
+      "loss": 0.0658,
+      "step": 1440
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.82421875,
+      "learning_rate": 5.983742270863157e-06,
+      "loss": 0.057,
+      "step": 1441
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.970512417740714e-06,
+      "loss": 0.0541,
+      "step": 1442
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.957290979158907e-06,
+      "loss": 0.073,
+      "step": 1443
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.944077982727286e-06,
+      "loss": 0.0766,
+      "step": 1444
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8046875,
+      "learning_rate": 5.930873456037765e-06,
+      "loss": 0.0755,
+      "step": 1445
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.81640625,
+      "learning_rate": 5.9176774266645805e-06,
+      "loss": 0.061,
+      "step": 1446
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7890625,
+      "learning_rate": 5.9044899221642164e-06,
+      "loss": 0.0457,
+      "step": 1447
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.77734375,
+      "learning_rate": 5.891310970075361e-06,
+      "loss": 0.0631,
+      "step": 1448
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8359375,
+      "learning_rate": 5.878140597918843e-06,
+      "loss": 0.0711,
+      "step": 1449
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8359375,
+      "learning_rate": 5.864978833197567e-06,
+      "loss": 0.0528,
+      "step": 1450
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.851825703396474e-06,
+      "loss": 0.0612,
+      "step": 1451
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.83984375,
+      "learning_rate": 5.8386812359824594e-06,
+      "loss": 0.058,
+      "step": 1452
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.828125,
+      "learning_rate": 5.825545458404342e-06,
+      "loss": 0.057,
+      "step": 1453
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.81640625,
+      "learning_rate": 5.812418398092787e-06,
+      "loss": 0.0626,
+      "step": 1454
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.799300082460262e-06,
+      "loss": 0.0805,
+      "step": 1455
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7890625,
+      "learning_rate": 5.7861905389009686e-06,
+      "loss": 0.0637,
+      "step": 1456
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7890625,
+      "learning_rate": 5.773089794790785e-06,
+      "loss": 0.0649,
+      "step": 1457
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.7599978774872275e-06,
+      "loss": 0.0549,
+      "step": 1458
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.746914814329376e-06,
+      "loss": 0.0569,
+      "step": 1459
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.733840632637808e-06,
+      "loss": 0.0646,
+      "step": 1460
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.80859375,
+      "learning_rate": 5.7207753597145684e-06,
+      "loss": 0.0535,
+      "step": 1461
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.7077190228430944e-06,
+      "loss": 0.0539,
+      "step": 1462
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.694671649288161e-06,
+      "loss": 0.0526,
+      "step": 1463
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.78125,
+      "learning_rate": 5.681633266295834e-06,
+      "loss": 0.0535,
+      "step": 1464
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.668603901093383e-06,
+      "loss": 0.0518,
+      "step": 1465
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.655583580889276e-06,
+      "loss": 0.0643,
+      "step": 1466
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.642572332873067e-06,
+      "loss": 0.0669,
+      "step": 1467
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.629570184215381e-06,
+      "loss": 0.0528,
+      "step": 1468
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.6165771620678335e-06,
+      "loss": 0.0714,
+      "step": 1469
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.6035932935629865e-06,
+      "loss": 0.0528,
+      "step": 1470
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.590618605814288e-06,
+      "loss": 0.0706,
+      "step": 1471
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.577653125916e-06,
+      "loss": 0.0714,
+      "step": 1472
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.564696880943178e-06,
+      "loss": 0.0553,
+      "step": 1473
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.551749897951583e-06,
+      "loss": 0.055,
+      "step": 1474
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.538812203977626e-06,
+      "loss": 0.0767,
+      "step": 1475
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.81640625,
+      "learning_rate": 5.525883826038332e-06,
+      "loss": 0.0738,
+      "step": 1476
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8125,
+      "learning_rate": 5.5129647911312675e-06,
+      "loss": 0.0597,
+      "step": 1477
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.875,
+      "learning_rate": 5.500055126234491e-06,
+      "loss": 0.0648,
+      "step": 1478
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.487154858306491e-06,
+      "loss": 0.0627,
+      "step": 1479
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.474264014286135e-06,
+      "loss": 0.0492,
+      "step": 1480
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7578125,
+      "learning_rate": 5.461382621092612e-06,
+      "loss": 0.0566,
+      "step": 1481
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.4485107056253684e-06,
+      "loss": 0.0532,
+      "step": 1482
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.89453125,
+      "learning_rate": 5.435648294764062e-06,
+      "loss": 0.048,
+      "step": 1483
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.4227954153685194e-06,
+      "loss": 0.0509,
+      "step": 1484
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.409952094278634e-06,
+      "loss": 0.0757,
+      "step": 1485
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.39711835831436e-06,
+      "loss": 0.0433,
+      "step": 1486
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.384294234275626e-06,
+      "loss": 0.0673,
+      "step": 1487
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.3714797489422945e-06,
+      "loss": 0.0558,
+      "step": 1488
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.80078125,
+      "learning_rate": 5.358674929074101e-06,
+      "loss": 0.0667,
+      "step": 1489
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.75,
+      "learning_rate": 5.34587980141058e-06,
+      "loss": 0.0489,
+      "step": 1490
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.33309439267106e-06,
+      "loss": 0.0569,
+      "step": 1491
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7734375,
+      "learning_rate": 5.320318729554539e-06,
+      "loss": 0.053,
+      "step": 1492
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8125,
+      "learning_rate": 5.3075528387396826e-06,
+      "loss": 0.0547,
+      "step": 1493
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.294796746884745e-06,
+      "loss": 0.0552,
+      "step": 1494
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.2820504806275215e-06,
+      "loss": 0.0539,
+      "step": 1495
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.2578125,
+      "learning_rate": 5.269314066585288e-06,
+      "loss": 0.0616,
+      "step": 1496
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.82421875,
+      "learning_rate": 5.256587531354738e-06,
+      "loss": 0.0703,
+      "step": 1497
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.24387090151194e-06,
+      "loss": 0.0701,
+      "step": 1498
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.79296875,
+      "learning_rate": 5.231164203612295e-06,
+      "loss": 0.0579,
+      "step": 1499
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.79296875,
+      "learning_rate": 5.218467464190436e-06,
+      "loss": 0.0521,
+      "step": 1500
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.84375,
+      "learning_rate": 5.205780709760215e-06,
+      "loss": 0.0579,
+      "step": 1501
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.859375,
+      "learning_rate": 5.1931039668146344e-06,
+      "loss": 0.0587,
+      "step": 1502
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.180437261825785e-06,
+      "loss": 0.0558,
+      "step": 1503
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.167780621244801e-06,
+      "loss": 0.045,
+      "step": 1504
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.155134071501786e-06,
+      "loss": 0.0598,
+      "step": 1505
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.1424976390057945e-06,
+      "loss": 0.0534,
+      "step": 1506
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.1298713501447396e-06,
+      "loss": 0.0555,
+      "step": 1507
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.82421875,
+      "learning_rate": 5.117255231285345e-06,
+      "loss": 0.0585,
+      "step": 1508
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.104649308773113e-06,
+      "loss": 0.0507,
+      "step": 1509
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.092053608932243e-06,
+      "loss": 0.072,
+      "step": 1510
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.079468158065591e-06,
+      "loss": 0.0481,
+      "step": 1511
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.06689298245461e-06,
+      "loss": 0.0805,
+      "step": 1512
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.875,
+      "learning_rate": 5.054328108359297e-06,
+      "loss": 0.0559,
+      "step": 1513
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.041773562018136e-06,
+      "loss": 0.0499,
+      "step": 1514
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.029229369648039e-06,
+      "loss": 0.0549,
+      "step": 1515
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.0166955574443e-06,
+      "loss": 0.0544,
+      "step": 1516
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.75,
+      "learning_rate": 5.004172151580551e-06,
+      "loss": 0.0509,
+      "step": 1517
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.991659178208666e-06,
+      "loss": 0.0495,
+      "step": 1518
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.979156663458751e-06,
+      "loss": 0.0418,
+      "step": 1519
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.96666463343907e-06,
+      "loss": 0.0558,
+      "step": 1520
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.95418311423599e-06,
+      "loss": 0.0502,
+      "step": 1521
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.941712131913933e-06,
+      "loss": 0.0573,
+      "step": 1522
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9375,
+      "learning_rate": 4.929251712515306e-06,
+      "loss": 0.0637,
+      "step": 1523
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.916801882060471e-06,
+      "loss": 0.0594,
+      "step": 1524
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.904362666547672e-06,
+      "loss": 0.0693,
+      "step": 1525
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.891934091952986e-06,
+      "loss": 0.0558,
+      "step": 1526
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.8795161842302726e-06,
+      "loss": 0.0386,
+      "step": 1527
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.867108969311113e-06,
+      "loss": 0.057,
+      "step": 1528
+    },
+    {
+      "epoch": 1.36,
+      "eval_loss": 0.05325327813625336,
+      "eval_runtime": 228.5302,
+      "eval_samples_per_second": 34.884,
+      "eval_steps_per_second": 4.363,
+      "step": 1528
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.8547124731047645e-06,
+      "loss": 0.0565,
+      "step": 1529
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.8423267214980885e-06,
+      "loss": 0.0436,
+      "step": 1530
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.875,
+      "learning_rate": 4.8299517403555186e-06,
+      "loss": 0.0583,
+      "step": 1531
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.817587555519008e-06,
+      "loss": 0.0554,
+      "step": 1532
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.8052341928079395e-06,
+      "loss": 0.0449,
+      "step": 1533
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.792891678019116e-06,
+      "loss": 0.0525,
+      "step": 1534
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.780560036926679e-06,
+      "loss": 0.0395,
+      "step": 1535
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.768239295282067e-06,
+      "loss": 0.0397,
+      "step": 1536
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.755929478813961e-06,
+      "loss": 0.0469,
+      "step": 1537
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.743630613228208e-06,
+      "loss": 0.0521,
+      "step": 1538
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.73134272420782e-06,
+      "loss": 0.058,
+      "step": 1539
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.7190658374128565e-06,
+      "loss": 0.0576,
+      "step": 1540
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.706799978480416e-06,
+      "loss": 0.048,
+      "step": 1541
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.694545173024571e-06,
+      "loss": 0.0556,
+      "step": 1542
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.682301446636303e-06,
+      "loss": 0.0451,
+      "step": 1543
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.670068824883467e-06,
+      "loss": 0.0482,
+      "step": 1544
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.657847333310711e-06,
+      "loss": 0.0456,
+      "step": 1545
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.645636997439468e-06,
+      "loss": 0.0461,
+      "step": 1546
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.633437842767856e-06,
+      "loss": 0.0523,
+      "step": 1547
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.621249894770644e-06,
+      "loss": 0.0523,
+      "step": 1548
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.609073178899205e-06,
+      "loss": 0.0405,
+      "step": 1549
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.596907720581456e-06,
+      "loss": 0.0494,
+      "step": 1550
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.5847535452218024e-06,
+      "loss": 0.0475,
+      "step": 1551
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.5726106782010924e-06,
+      "loss": 0.0518,
+      "step": 1552
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.560479144876555e-06,
+      "loss": 0.0481,
+      "step": 1553
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.548358970581758e-06,
+      "loss": 0.0536,
+      "step": 1554
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.536250180626538e-06,
+      "loss": 0.0421,
+      "step": 1555
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.524152800296968e-06,
+      "loss": 0.0489,
+      "step": 1556
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.512066854855293e-06,
+      "loss": 0.0446,
+      "step": 1557
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.499992369539876e-06,
+      "loss": 0.0366,
+      "step": 1558
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.487929369565156e-06,
+      "loss": 0.0429,
+      "step": 1559
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.475877880121572e-06,
+      "loss": 0.0433,
+      "step": 1560
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.463837926375546e-06,
+      "loss": 0.0512,
+      "step": 1561
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.451809533469404e-06,
+      "loss": 0.0597,
+      "step": 1562
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.439792726521317e-06,
+      "loss": 0.0409,
+      "step": 1563
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.4277875306252785e-06,
+      "loss": 0.0477,
+      "step": 1564
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.415793970851027e-06,
+      "loss": 0.0461,
+      "step": 1565
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.403812072244007e-06,
+      "loss": 0.0395,
+      "step": 1566
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.391841859825307e-06,
+      "loss": 0.0395,
+      "step": 1567
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.379883358591614e-06,
+      "loss": 0.049,
+      "step": 1568
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.367936593515162e-06,
+      "loss": 0.0371,
+      "step": 1569
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.356001589543667e-06,
+      "loss": 0.0555,
+      "step": 1570
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.344078371600293e-06,
+      "loss": 0.048,
+      "step": 1571
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.332166964583602e-06,
+      "loss": 0.053,
+      "step": 1572
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.320267393367469e-06,
+      "loss": 0.0431,
+      "step": 1573
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.308379682801069e-06,
+      "loss": 0.0493,
+      "step": 1574
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.2965038577088024e-06,
+      "loss": 0.0585,
+      "step": 1575
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.284639942890258e-06,
+      "loss": 0.0498,
+      "step": 1576
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.272787963120146e-06,
+      "loss": 0.0607,
+      "step": 1577
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.260947943148247e-06,
+      "loss": 0.0541,
+      "step": 1578
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.2491199076993875e-06,
+      "loss": 0.0508,
+      "step": 1579
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.237303881473345e-06,
+      "loss": 0.0509,
+      "step": 1580
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.2254998891448295e-06,
+      "loss": 0.0451,
+      "step": 1581
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.213707955363423e-06,
+      "loss": 0.0429,
+      "step": 1582
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.201928104753522e-06,
+      "loss": 0.052,
+      "step": 1583
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.190160361914293e-06,
+      "loss": 0.0482,
+      "step": 1584
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.1784047514196155e-06,
+      "loss": 0.0537,
+      "step": 1585
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.166661297818037e-06,
+      "loss": 0.0405,
+      "step": 1586
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.1549300256327215e-06,
+      "loss": 0.0497,
+      "step": 1587
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.143210959361383e-06,
+      "loss": 0.0419,
+      "step": 1588
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.131504123476259e-06,
+      "loss": 0.042,
+      "step": 1589
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.11980954242404e-06,
+      "loss": 0.0443,
+      "step": 1590
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.108127240625831e-06,
+      "loss": 0.0441,
+      "step": 1591
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.096457242477094e-06,
+      "loss": 0.054,
+      "step": 1592
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.084799572347587e-06,
+      "loss": 0.0473,
+      "step": 1593
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.07315425458134e-06,
+      "loss": 0.0451,
+      "step": 1594
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.061521313496587e-06,
+      "loss": 0.0379,
+      "step": 1595
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.0499007733857e-06,
+      "loss": 0.0409,
+      "step": 1596
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.038292658515169e-06,
+      "loss": 0.0459,
+      "step": 1597
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.0266969931255366e-06,
+      "loss": 0.0489,
+      "step": 1598
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.0151138014313425e-06,
+      "loss": 0.0528,
+      "step": 1599
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.003543107621083e-06,
+      "loss": 0.0406,
+      "step": 1600
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.991984935857152e-06,
+      "loss": 0.0356,
+      "step": 1601
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.625,
+      "learning_rate": 3.980439310275801e-06,
+      "loss": 0.0498,
+      "step": 1602
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.96890625498707e-06,
+      "loss": 0.0655,
+      "step": 1603
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.957385794074754e-06,
+      "loss": 0.0363,
+      "step": 1604
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.9458779515963645e-06,
+      "loss": 0.038,
+      "step": 1605
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.9343827515830346e-06,
+      "loss": 0.0407,
+      "step": 1606
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.922900218039516e-06,
+      "loss": 0.0436,
+      "step": 1607
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.911430374944105e-06,
+      "loss": 0.0537,
+      "step": 1608
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.899973246248596e-06,
+      "loss": 0.0587,
+      "step": 1609
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.888528855878236e-06,
+      "loss": 0.0483,
+      "step": 1610
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.8770972277316585e-06,
+      "loss": 0.0696,
+      "step": 1611
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.865678385680872e-06,
+      "loss": 0.0572,
+      "step": 1612
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.8542723535711565e-06,
+      "loss": 0.042,
+      "step": 1613
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.8428791552210595e-06,
+      "loss": 0.05,
+      "step": 1614
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.831498814422323e-06,
+      "loss": 0.0358,
+      "step": 1615
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.82013135493984e-06,
+      "loss": 0.0472,
+      "step": 1616
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.8087768005116053e-06,
+      "loss": 0.047,
+      "step": 1617
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.797435174848656e-06,
+      "loss": 0.0533,
+      "step": 1618
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.7861065016350462e-06,
+      "loss": 0.0515,
+      "step": 1619
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.774790804527776e-06,
+      "loss": 0.0377,
+      "step": 1620
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.7634881071567386e-06,
+      "loss": 0.0371,
+      "step": 1621
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.752198433124693e-06,
+      "loss": 0.041,
+      "step": 1622
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.740921806007196e-06,
+      "loss": 0.0486,
+      "step": 1623
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.7296582493525636e-06,
+      "loss": 0.0635,
+      "step": 1624
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.718407786681817e-06,
+      "loss": 0.0404,
+      "step": 1625
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.707170441488622e-06,
+      "loss": 0.0531,
+      "step": 1626
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.695946237239276e-06,
+      "loss": 0.0465,
+      "step": 1627
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.684735197372611e-06,
+      "loss": 0.0495,
+      "step": 1628
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.6735373452999813e-06,
+      "loss": 0.0499,
+      "step": 1629
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.6623527044051988e-06,
+      "loss": 0.0534,
+      "step": 1630
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.875,
+      "learning_rate": 3.6511812980444896e-06,
+      "loss": 0.0602,
+      "step": 1631
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.640023149546442e-06,
+      "loss": 0.0418,
+      "step": 1632
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.6288782822119495e-06,
+      "loss": 0.0515,
+      "step": 1633
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.617746719314189e-06,
+      "loss": 0.0414,
+      "step": 1634
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.6066284840985446e-06,
+      "loss": 0.0465,
+      "step": 1635
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.595523599782562e-06,
+      "loss": 0.0422,
+      "step": 1636
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.58443208955592e-06,
+      "loss": 0.045,
+      "step": 1637
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.5733539765803627e-06,
+      "loss": 0.064,
+      "step": 1638
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.5622892839896563e-06,
+      "loss": 0.0486,
+      "step": 1639
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.551238034889547e-06,
+      "loss": 0.0371,
+      "step": 1640
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.540200252357703e-06,
+      "loss": 0.0443,
+      "step": 1641
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.529175959443676e-06,
+      "loss": 0.0395,
+      "step": 1642
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.51816517916884e-06,
+      "loss": 0.038,
+      "step": 1643
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.5071679345263543e-06,
+      "loss": 0.0394,
+      "step": 1644
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.496184248481126e-06,
+      "loss": 0.0451,
+      "step": 1645
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.4852141439697264e-06,
+      "loss": 0.0485,
+      "step": 1646
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.4742576439003796e-06,
+      "loss": 0.0361,
+      "step": 1647
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.463314771152897e-06,
+      "loss": 0.0609,
+      "step": 1648
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.452385548578633e-06,
+      "loss": 0.0551,
+      "step": 1649
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.4414699990004385e-06,
+      "loss": 0.0473,
+      "step": 1650
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.430568145212605e-06,
+      "loss": 0.0735,
+      "step": 1651
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.4196800099808327e-06,
+      "loss": 0.0427,
+      "step": 1652
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.408805616042169e-06,
+      "loss": 0.0572,
+      "step": 1653
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.3979449861049686e-06,
+      "loss": 0.0539,
+      "step": 1654
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.3870981428488426e-06,
+      "loss": 0.0573,
+      "step": 1655
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.3762651089246134e-06,
+      "loss": 0.0473,
+      "step": 1656
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.3654459069542668e-06,
+      "loss": 0.0306,
+      "step": 1657
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.3546405595308973e-06,
+      "loss": 0.0507,
+      "step": 1658
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.3438490892186715e-06,
+      "loss": 0.0577,
+      "step": 1659
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.3330715185527895e-06,
+      "loss": 0.059,
+      "step": 1660
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.322307870039405e-06,
+      "loss": 0.0454,
+      "step": 1661
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.3115581661556117e-06,
+      "loss": 0.0439,
+      "step": 1662
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.3008224293493805e-06,
+      "loss": 0.0542,
+      "step": 1663
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.2901006820395164e-06,
+      "loss": 0.0536,
+      "step": 1664
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.279392946615613e-06,
+      "loss": 0.047,
+      "step": 1665
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.2686992454379896e-06,
+      "loss": 0.0424,
+      "step": 1666
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.258019600837683e-06,
+      "loss": 0.0485,
+      "step": 1667
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.2473540351163614e-06,
+      "loss": 0.0516,
+      "step": 1668
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.2367025705462885e-06,
+      "loss": 0.0527,
+      "step": 1669
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.226065229370291e-06,
+      "loss": 0.0437,
+      "step": 1670
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.215442033801701e-06,
+      "loss": 0.0467,
+      "step": 1671
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.2048330060243085e-06,
+      "loss": 0.0608,
+      "step": 1672
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.1942381681923186e-06,
+      "loss": 0.0311,
+      "step": 1673
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.183657542430304e-06,
+      "loss": 0.0494,
+      "step": 1674
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.1730911508331654e-06,
+      "loss": 0.0476,
+      "step": 1675
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.625,
+      "learning_rate": 3.162539015466065e-06,
+      "loss": 0.0406,
+      "step": 1676
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.152001158364404e-06,
+      "loss": 0.0377,
+      "step": 1677
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.1414776015337755e-06,
+      "loss": 0.0401,
+      "step": 1678
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.1309683669498915e-06,
+      "loss": 0.0452,
+      "step": 1679
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.1204734765585678e-06,
+      "loss": 0.0566,
+      "step": 1680
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.1099929522756645e-06,
+      "loss": 0.0569,
+      "step": 1681
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.0995268159870396e-06,
+      "loss": 0.0524,
+      "step": 1682
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.08907508954851e-06,
+      "loss": 0.034,
+      "step": 1683
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.0786377947857913e-06,
+      "loss": 0.0463,
+      "step": 1684
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.0682149534944737e-06,
+      "loss": 0.0498,
+      "step": 1685
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.0578065874399587e-06,
+      "loss": 0.0424,
+      "step": 1686
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.0474127183574233e-06,
+      "loss": 0.0449,
+      "step": 1687
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.037033367951767e-06,
+      "loss": 0.0607,
+      "step": 1688
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.0266685578975753e-06,
+      "loss": 0.0377,
+      "step": 1689
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.0163183098390713e-06,
+      "loss": 0.049,
+      "step": 1690
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.0059826453900597e-06,
+      "loss": 0.0516,
+      "step": 1691
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.9956615861338966e-06,
+      "loss": 0.0641,
+      "step": 1692
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.9853551536234494e-06,
+      "loss": 0.0392,
+      "step": 1693
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.9750633693810226e-06,
+      "loss": 0.0597,
+      "step": 1694
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.964786254898343e-06,
+      "loss": 0.032,
+      "step": 1695
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.954523831636502e-06,
+      "loss": 0.0489,
+      "step": 1696
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.944276121025912e-06,
+      "loss": 0.04,
+      "step": 1697
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.934043144466262e-06,
+      "loss": 0.0465,
+      "step": 1698
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.9238249233264628e-06,
+      "loss": 0.038,
+      "step": 1699
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.913621478944635e-06,
+      "loss": 0.0426,
+      "step": 1700
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.9034328326280183e-06,
+      "loss": 0.0571,
+      "step": 1701
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.8932590056529643e-06,
+      "loss": 0.0565,
+      "step": 1702
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.8831000192648716e-06,
+      "loss": 0.0523,
+      "step": 1703
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.8729558946781534e-06,
+      "loss": 0.0421,
+      "step": 1704
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.862826653076187e-06,
+      "loss": 0.0432,
+      "step": 1705
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.8527123156112614e-06,
+      "loss": 0.041,
+      "step": 1706
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.842612903404555e-06,
+      "loss": 0.0403,
+      "step": 1707
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.832528437546076e-06,
+      "loss": 0.0437,
+      "step": 1708
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.822458939094611e-06,
+      "loss": 0.0415,
+      "step": 1709
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.8124044290777007e-06,
+      "loss": 0.0337,
+      "step": 1710
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.8023649284915856e-06,
+      "loss": 0.0499,
+      "step": 1711
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.7923404583011583e-06,
+      "loss": 0.0533,
+      "step": 1712
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.7823310394399295e-06,
+      "loss": 0.0501,
+      "step": 1713
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.772336692809976e-06,
+      "loss": 0.0476,
+      "step": 1714
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.762357439281903e-06,
+      "loss": 0.0417,
+      "step": 1715
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.752393299694791e-06,
+      "loss": 0.0439,
+      "step": 1716
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.742444294856166e-06,
+      "loss": 0.0454,
+      "step": 1717
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.875,
+      "learning_rate": 2.732510445541946e-06,
+      "loss": 0.0575,
+      "step": 1718
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.7225917724964013e-06,
+      "loss": 0.0435,
+      "step": 1719
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.7126882964321124e-06,
+      "loss": 0.0398,
+      "step": 1720
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.7028000380299156e-06,
+      "loss": 0.046,
+      "step": 1721
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.6929270179388832e-06,
+      "loss": 0.049,
+      "step": 1722
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.6830692567762604e-06,
+      "loss": 0.0359,
+      "step": 1723
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.673226775127422e-06,
+      "loss": 0.0499,
+      "step": 1724
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.663399593545841e-06,
+      "loss": 0.0521,
+      "step": 1725
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.6535877325530412e-06,
+      "loss": 0.0472,
+      "step": 1726
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.64379121263855e-06,
+      "loss": 0.043,
+      "step": 1727
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.6340100542598633e-06,
+      "loss": 0.0409,
+      "step": 1728
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.6242442778423927e-06,
+      "loss": 0.0433,
+      "step": 1729
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.614493903779437e-06,
+      "loss": 0.0561,
+      "step": 1730
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.604758952432119e-06,
+      "loss": 0.0529,
+      "step": 1731
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.5950394441293615e-06,
+      "loss": 0.0574,
+      "step": 1732
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.5853353991678487e-06,
+      "loss": 0.0426,
+      "step": 1733
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.575646837811954e-06,
+      "loss": 0.0585,
+      "step": 1734
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.56597378029373e-06,
+      "loss": 0.0578,
+      "step": 1735
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.556316246812851e-06,
+      "loss": 0.0438,
+      "step": 1736
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.5466742575365744e-06,
+      "loss": 0.0547,
+      "step": 1737
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.537047832599697e-06,
+      "loss": 0.0466,
+      "step": 1738
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.527436992104506e-06,
+      "loss": 0.0678,
+      "step": 1739
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.517841756120759e-06,
+      "loss": 0.0537,
+      "step": 1740
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.508262144685619e-06,
+      "loss": 0.0529,
+      "step": 1741
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.498698177803618e-06,
+      "loss": 0.0395,
+      "step": 1742
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.4891498754466246e-06,
+      "loss": 0.0502,
+      "step": 1743
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.4796172575537937e-06,
+      "loss": 0.0384,
+      "step": 1744
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.4701003440315285e-06,
+      "loss": 0.0676,
+      "step": 1745
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.460599154753435e-06,
+      "loss": 0.0416,
+      "step": 1746
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.4511137095602867e-06,
+      "loss": 0.0368,
+      "step": 1747
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.4416440282599786e-06,
+      "loss": 0.0416,
+      "step": 1748
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.432190130627483e-06,
+      "loss": 0.0452,
+      "step": 1749
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.422752036404816e-06,
+      "loss": 0.0537,
+      "step": 1750
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.4133297653009937e-06,
+      "loss": 0.042,
+      "step": 1751
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.403923336991987e-06,
+      "loss": 0.0359,
+      "step": 1752
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.3945327711206845e-06,
+      "loss": 0.0484,
+      "step": 1753
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.3851580872968437e-06,
+      "loss": 0.0495,
+      "step": 1754
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.3757993050970705e-06,
+      "loss": 0.06,
+      "step": 1755
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.366456444064756e-06,
+      "loss": 0.0552,
+      "step": 1756
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.357129523710039e-06,
+      "loss": 0.0415,
+      "step": 1757
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.3478185635097773e-06,
+      "loss": 0.0403,
+      "step": 1758
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.3385235829075004e-06,
+      "loss": 0.059,
+      "step": 1759
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.3292446013133653e-06,
+      "loss": 0.0509,
+      "step": 1760
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.3199816381041206e-06,
+      "loss": 0.043,
+      "step": 1761
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.310734712623065e-06,
+      "loss": 0.0455,
+      "step": 1762
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.3015038441800096e-06,
+      "loss": 0.0459,
+      "step": 1763
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.292289052051224e-06,
+      "loss": 0.0477,
+      "step": 1764
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.2830903554794148e-06,
+      "loss": 0.0375,
+      "step": 1765
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.2739077736736848e-06,
+      "loss": 0.0522,
+      "step": 1766
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.2647413258094674e-06,
+      "loss": 0.0605,
+      "step": 1767
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.2555910310285156e-06,
+      "loss": 0.0527,
+      "step": 1768
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.246456908438851e-06,
+      "loss": 0.0511,
+      "step": 1769
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.2373389771147204e-06,
+      "loss": 0.0489,
+      "step": 1770
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.2282372560965636e-06,
+      "loss": 0.0507,
+      "step": 1771
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.2191517643909576e-06,
+      "loss": 0.0588,
+      "step": 1772
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.210082520970609e-06,
+      "loss": 0.0512,
+      "step": 1773
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.2010295447742747e-06,
+      "loss": 0.0314,
+      "step": 1774
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.1919928547067516e-06,
+      "loss": 0.0671,
+      "step": 1775
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.1829724696388243e-06,
+      "loss": 0.0592,
+      "step": 1776
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.1739684084072323e-06,
+      "loss": 0.0374,
+      "step": 1777
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.164980689814623e-06,
+      "loss": 0.0604,
+      "step": 1778
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.1560093326295163e-06,
+      "loss": 0.0533,
+      "step": 1779
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.1470543555862644e-06,
+      "loss": 0.047,
+      "step": 1780
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.1381157773850246e-06,
+      "loss": 0.0452,
+      "step": 1781
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.129193616691694e-06,
+      "loss": 0.0508,
+      "step": 1782
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.1202878921378957e-06,
+      "loss": 0.0487,
+      "step": 1783
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.111398622320927e-06,
+      "loss": 0.0565,
+      "step": 1784
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.102525825803724e-06,
+      "loss": 0.0553,
+      "step": 1785
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.0936695211148274e-06,
+      "loss": 0.0417,
+      "step": 1786
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.084829726748325e-06,
+      "loss": 0.0382,
+      "step": 1787
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.0760064611638485e-06,
+      "loss": 0.0409,
+      "step": 1788
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.067199742786492e-06,
+      "loss": 0.0659,
+      "step": 1789
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.058409590006809e-06,
+      "loss": 0.0446,
+      "step": 1790
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.0496360211807555e-06,
+      "loss": 0.0535,
+      "step": 1791
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.0408790546296553e-06,
+      "loss": 0.0633,
+      "step": 1792
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.0321387086401665e-06,
+      "loss": 0.0679,
+      "step": 1793
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.0234150014642305e-06,
+      "loss": 0.0493,
+      "step": 1794
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.014707951319057e-06,
+      "loss": 0.0459,
+      "step": 1795
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.0060175763870616e-06,
+      "loss": 0.0436,
+      "step": 1796
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9973438948158373e-06,
+      "loss": 0.0386,
+      "step": 1797
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.988686924718122e-06,
+      "loss": 0.0406,
+      "step": 1798
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9800466841717546e-06,
+      "loss": 0.0504,
+      "step": 1799
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9714231912196393e-06,
+      "loss": 0.0409,
+      "step": 1800
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9628164638697044e-06,
+      "loss": 0.0383,
+      "step": 1801
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.954226520094872e-06,
+      "loss": 0.0517,
+      "step": 1802
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9456533778330154e-06,
+      "loss": 0.0332,
+      "step": 1803
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.937097054986915e-06,
+      "loss": 0.0436,
+      "step": 1804
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.928557569424234e-06,
+      "loss": 0.041,
+      "step": 1805
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.9200349389774842e-06,
+      "loss": 0.0318,
+      "step": 1806
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.9115291814439606e-06,
+      "loss": 0.0653,
+      "step": 1807
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.9030403145857378e-06,
+      "loss": 0.0633,
+      "step": 1808
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.8945683561296147e-06,
+      "loss": 0.0514,
+      "step": 1809
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8861133237670815e-06,
+      "loss": 0.0448,
+      "step": 1810
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.8776752351542827e-06,
+      "loss": 0.05,
+      "step": 1811
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.8692541079119763e-06,
+      "loss": 0.0491,
+      "step": 1812
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.860849959625507e-06,
+      "loss": 0.0573,
+      "step": 1813
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.8524628078447605e-06,
+      "loss": 0.0487,
+      "step": 1814
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.844092670084129e-06,
+      "loss": 0.0314,
+      "step": 1815
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8357395638224763e-06,
+      "loss": 0.0439,
+      "step": 1816
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.8274035065031027e-06,
+      "loss": 0.0619,
+      "step": 1817
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.8190845155337e-06,
+      "loss": 0.0483,
+      "step": 1818
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8107826082863311e-06,
+      "loss": 0.0424,
+      "step": 1819
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.802497802097367e-06,
+      "loss": 0.0486,
+      "step": 1820
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.7942301142674922e-06,
+      "loss": 0.0531,
+      "step": 1821
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7859795620616205e-06,
+      "loss": 0.0364,
+      "step": 1822
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.7777461627088955e-06,
+      "loss": 0.0497,
+      "step": 1823
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.7695299334026373e-06,
+      "loss": 0.0711,
+      "step": 1824
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.7613308913003125e-06,
+      "loss": 0.047,
+      "step": 1825
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7531490535234974e-06,
+      "loss": 0.0425,
+      "step": 1826
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.744984437157834e-06,
+      "loss": 0.068,
+      "step": 1827
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7368370592530148e-06,
+      "loss": 0.0487,
+      "step": 1828
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.728706936822726e-06,
+      "loss": 0.0407,
+      "step": 1829
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.720594086844618e-06,
+      "loss": 0.0493,
+      "step": 1830
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.7124985262602789e-06,
+      "loss": 0.0471,
+      "step": 1831
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.7044202719751868e-06,
+      "loss": 0.0559,
+      "step": 1832
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6963593408586831e-06,
+      "loss": 0.0526,
+      "step": 1833
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.6883157497439351e-06,
+      "loss": 0.0424,
+      "step": 1834
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.6802895154278975e-06,
+      "loss": 0.0504,
+      "step": 1835
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.6722806546712823e-06,
+      "loss": 0.0609,
+      "step": 1836
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.6642891841985164e-06,
+      "loss": 0.0376,
+      "step": 1837
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.6563151206977145e-06,
+      "loss": 0.0445,
+      "step": 1838
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.6483584808206488e-06,
+      "loss": 0.053,
+      "step": 1839
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.6404192811826925e-06,
+      "loss": 0.0453,
+      "step": 1840
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.6324975383628094e-06,
+      "loss": 0.0631,
+      "step": 1841
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.6245932689035072e-06,
+      "loss": 0.0574,
+      "step": 1842
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.6167064893108042e-06,
+      "loss": 0.0431,
+      "step": 1843
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6088372160541966e-06,
+      "loss": 0.0556,
+      "step": 1844
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6009854655666203e-06,
+      "loss": 0.0471,
+      "step": 1845
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.5931512542444216e-06,
+      "loss": 0.0385,
+      "step": 1846
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.585334598447321e-06,
+      "loss": 0.0374,
+      "step": 1847
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5775355144983806e-06,
+      "loss": 0.0558,
+      "step": 1848
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.5697540186839643e-06,
+      "loss": 0.0486,
+      "step": 1849
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.5619901272537108e-06,
+      "loss": 0.0868,
+      "step": 1850
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5542438564204965e-06,
+      "loss": 0.0422,
+      "step": 1851
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.5465152223603963e-06,
+      "loss": 0.0655,
+      "step": 1852
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.5388042412126603e-06,
+      "loss": 0.0717,
+      "step": 1853
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.75,
+      "learning_rate": 1.531110929079681e-06,
+      "loss": 0.0537,
+      "step": 1854
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.625,
+      "learning_rate": 1.523435302026941e-06,
+      "loss": 0.0343,
+      "step": 1855
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.5157773760829997e-06,
+      "loss": 0.0535,
+      "step": 1856
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.5081371672394518e-06,
+      "loss": 0.0469,
+      "step": 1857
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.500514691450894e-06,
+      "loss": 0.0491,
+      "step": 1858
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4929099646348921e-06,
+      "loss": 0.0422,
+      "step": 1859
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4853230026719413e-06,
+      "loss": 0.041,
+      "step": 1860
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4777538214054566e-06,
+      "loss": 0.0473,
+      "step": 1861
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4702024366417023e-06,
+      "loss": 0.0456,
+      "step": 1862
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.4626688641497922e-06,
+      "loss": 0.0438,
+      "step": 1863
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4551531196616398e-06,
+      "loss": 0.0443,
+      "step": 1864
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.4476552188719305e-06,
+      "loss": 0.0524,
+      "step": 1865
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.440175177438089e-06,
+      "loss": 0.0522,
+      "step": 1866
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4327130109802367e-06,
+      "loss": 0.0513,
+      "step": 1867
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.425268735081181e-06,
+      "loss": 0.046,
+      "step": 1868
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4178423652863637e-06,
+      "loss": 0.0485,
+      "step": 1869
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.4104339171038284e-06,
+      "loss": 0.0438,
+      "step": 1870
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.403043406004203e-06,
+      "loss": 0.0403,
+      "step": 1871
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.3956708474206527e-06,
+      "loss": 0.0505,
+      "step": 1872
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.3883162567488574e-06,
+      "loss": 0.0593,
+      "step": 1873
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3809796493469728e-06,
+      "loss": 0.0493,
+      "step": 1874
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.3736610405356033e-06,
+      "loss": 0.0387,
+      "step": 1875
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.3663604455977687e-06,
+      "loss": 0.0458,
+      "step": 1876
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3590778797788652e-06,
+      "loss": 0.0428,
+      "step": 1877
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3518133582866456e-06,
+      "loss": 0.0418,
+      "step": 1878
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.625,
+      "learning_rate": 1.3445668962911818e-06,
+      "loss": 0.0473,
+      "step": 1879
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.3373385089248315e-06,
+      "loss": 0.0357,
+      "step": 1880
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.3301282112822101e-06,
+      "loss": 0.0615,
+      "step": 1881
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.32293601842015e-06,
+      "loss": 0.0409,
+      "step": 1882
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3157619453576898e-06,
+      "loss": 0.0403,
+      "step": 1883
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.3086060070760199e-06,
+      "loss": 0.0407,
+      "step": 1884
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.3014682185184603e-06,
+      "loss": 0.0369,
+      "step": 1885
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2943485945904344e-06,
+      "loss": 0.0489,
+      "step": 1886
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2872471501594307e-06,
+      "loss": 0.0414,
+      "step": 1887
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.2801639000549781e-06,
+      "loss": 0.044,
+      "step": 1888
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.2730988590686066e-06,
+      "loss": 0.0419,
+      "step": 1889
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2660520419538248e-06,
+      "loss": 0.0529,
+      "step": 1890
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2590234634260855e-06,
+      "loss": 0.0533,
+      "step": 1891
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2520131381627509e-06,
+      "loss": 0.0429,
+      "step": 1892
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2450210808030672e-06,
+      "loss": 0.0399,
+      "step": 1893
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2380473059481401e-06,
+      "loss": 0.0417,
+      "step": 1894
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2310918281608875e-06,
+      "loss": 0.0373,
+      "step": 1895
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2241546619660216e-06,
+      "loss": 0.0365,
+      "step": 1896
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2172358218500179e-06,
+      "loss": 0.043,
+      "step": 1897
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.2103353222610802e-06,
+      "loss": 0.0305,
+      "step": 1898
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2034531776091174e-06,
+      "loss": 0.0472,
+      "step": 1899
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.1965894022656977e-06,
+      "loss": 0.0476,
+      "step": 1900
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.1897440105640435e-06,
+      "loss": 0.0573,
+      "step": 1901
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.1829170167989812e-06,
+      "loss": 0.048,
+      "step": 1902
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.1761084352269147e-06,
+      "loss": 0.046,
+      "step": 1903
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.1693182800658043e-06,
+      "loss": 0.046,
+      "step": 1904
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.1625465654951296e-06,
+      "loss": 0.035,
+      "step": 1905
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.1557933056558612e-06,
+      "loss": 0.0569,
+      "step": 1906
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1490585146504351e-06,
+      "loss": 0.0418,
+      "step": 1907
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.1423422065427093e-06,
+      "loss": 0.0599,
+      "step": 1908
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.1356443953579633e-06,
+      "loss": 0.0473,
+      "step": 1909
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.1289650950828324e-06,
+      "loss": 0.0385,
+      "step": 1910
+    },
+    {
+      "epoch": 1.7,
+      "eval_loss": 0.04961521923542023,
+      "eval_runtime": 228.7063,
+      "eval_samples_per_second": 34.857,
+      "eval_steps_per_second": 4.359,
+      "step": 1910
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.1223043196653073e-06,
+      "loss": 0.0378,
+      "step": 1911
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.1156620830146903e-06,
+      "loss": 0.0459,
+      "step": 1912
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.1090383990015718e-06,
+      "loss": 0.0546,
+      "step": 1913
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.102433281457802e-06,
+      "loss": 0.0447,
+      "step": 1914
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.095846744176452e-06,
+      "loss": 0.0435,
+      "step": 1915
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0892788009118016e-06,
+      "loss": 0.0437,
+      "step": 1916
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.75,
+      "learning_rate": 1.0827294653793008e-06,
+      "loss": 0.0508,
+      "step": 1917
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.076198751255535e-06,
+      "loss": 0.0449,
+      "step": 1918
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.0696866721782096e-06,
+      "loss": 0.0461,
+      "step": 1919
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.0631932417461155e-06,
+      "loss": 0.0451,
+      "step": 1920
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.0567184735190971e-06,
+      "loss": 0.0404,
+      "step": 1921
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.0502623810180335e-06,
+      "loss": 0.0576,
+      "step": 1922
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.0438249777247978e-06,
+      "loss": 0.048,
+      "step": 1923
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0374062770822412e-06,
+      "loss": 0.041,
+      "step": 1924
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.0310062924941545e-06,
+      "loss": 0.0666,
+      "step": 1925
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0246250373252454e-06,
+      "loss": 0.0436,
+      "step": 1926
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.018262524901118e-06,
+      "loss": 0.0518,
+      "step": 1927
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.0119187685082254e-06,
+      "loss": 0.0397,
+      "step": 1928
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0055937813938599e-06,
+      "loss": 0.0531,
+      "step": 1929
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.992875767661192e-07,
+      "loss": 0.0547,
+      "step": 1930
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.930001677938771e-07,
+      "loss": 0.0579,
+      "step": 1931
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.86731567606759e-07,
+      "loss": 0.047,
+      "step": 1932
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.804817892951068e-07,
+      "loss": 0.0431,
+      "step": 1933
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.74250845909971e-07,
+      "loss": 0.0412,
+      "step": 1934
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.680387504630562e-07,
+      "loss": 0.0401,
+      "step": 1935
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.61845515926717e-07,
+      "loss": 0.0347,
+      "step": 1936
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.556711552339159e-07,
+      "loss": 0.0447,
+      "step": 1937
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.495156812782081e-07,
+      "loss": 0.051,
+      "step": 1938
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.828125,
+      "learning_rate": 9.433791069137066e-07,
+      "loss": 0.0519,
+      "step": 1939
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.372614449550533e-07,
+      "loss": 0.0451,
+      "step": 1940
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.311627081773999e-07,
+      "loss": 0.0526,
+      "step": 1941
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.250829093163827e-07,
+      "loss": 0.0383,
+      "step": 1942
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.190220610680811e-07,
+      "loss": 0.0432,
+      "step": 1943
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.129801760890078e-07,
+      "loss": 0.0396,
+      "step": 1944
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.069572669960735e-07,
+      "loss": 0.0651,
+      "step": 1945
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.009533463665643e-07,
+      "loss": 0.0467,
+      "step": 1946
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.949684267381143e-07,
+      "loss": 0.0623,
+      "step": 1947
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.890025206086694e-07,
+      "loss": 0.054,
+      "step": 1948
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.830556404364898e-07,
+      "loss": 0.0514,
+      "step": 1949
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.77127798640085e-07,
+      "loss": 0.0445,
+      "step": 1950
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.712190075982185e-07,
+      "loss": 0.0546,
+      "step": 1951
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.653292796498713e-07,
+      "loss": 0.0412,
+      "step": 1952
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.594586270942129e-07,
+      "loss": 0.0375,
+      "step": 1953
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.536070621905812e-07,
+      "loss": 0.0396,
+      "step": 1954
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.477745971584483e-07,
+      "loss": 0.0459,
+      "step": 1955
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.419612441774106e-07,
+      "loss": 0.0638,
+      "step": 1956
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.361670153871504e-07,
+      "loss": 0.0488,
+      "step": 1957
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.303919228874092e-07,
+      "loss": 0.0543,
+      "step": 1958
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.34375,
+      "learning_rate": 8.246359787379732e-07,
+      "loss": 0.0447,
+      "step": 1959
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.188991949586389e-07,
+      "loss": 0.0398,
+      "step": 1960
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.131815835291945e-07,
+      "loss": 0.048,
+      "step": 1961
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.074831563893881e-07,
+      "loss": 0.06,
+      "step": 1962
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.018039254389099e-07,
+      "loss": 0.0391,
+      "step": 1963
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.961439025373618e-07,
+      "loss": 0.038,
+      "step": 1964
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.905030995042329e-07,
+      "loss": 0.0451,
+      "step": 1965
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.848815281188782e-07,
+      "loss": 0.0628,
+      "step": 1966
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.792792001204974e-07,
+      "loss": 0.0485,
+      "step": 1967
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.736961272080968e-07,
+      "loss": 0.0418,
+      "step": 1968
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.681323210404778e-07,
+      "loss": 0.0334,
+      "step": 1969
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.62587793236208e-07,
+      "loss": 0.0405,
+      "step": 1970
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.5,
+      "learning_rate": 7.570625553735988e-07,
+      "loss": 0.0321,
+      "step": 1971
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.515566189906764e-07,
+      "loss": 0.0492,
+      "step": 1972
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.460699955851613e-07,
+      "loss": 0.0511,
+      "step": 1973
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.406026966144441e-07,
+      "loss": 0.0419,
+      "step": 1974
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.351547334955655e-07,
+      "loss": 0.0523,
+      "step": 1975
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.297261176051817e-07,
+      "loss": 0.0461,
+      "step": 1976
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.243168602795514e-07,
+      "loss": 0.048,
+      "step": 1977
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.189269728145054e-07,
+      "loss": 0.0477,
+      "step": 1978
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.135564664654282e-07,
+      "loss": 0.0454,
+      "step": 1979
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.082053524472332e-07,
+      "loss": 0.0454,
+      "step": 1980
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.028736419343284e-07,
+      "loss": 0.0452,
+      "step": 1981
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.975613460606167e-07,
+      "loss": 0.0427,
+      "step": 1982
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.922684759194476e-07,
+      "loss": 0.0418,
+      "step": 1983
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.869950425636096e-07,
+      "loss": 0.0361,
+      "step": 1984
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.817410570053007e-07,
+      "loss": 0.0372,
+      "step": 1985
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5,
+      "learning_rate": 6.765065302161111e-07,
+      "loss": 0.0343,
+      "step": 1986
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.712914731269926e-07,
+      "loss": 0.0491,
+      "step": 1987
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.660958966282361e-07,
+      "loss": 0.039,
+      "step": 1988
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.609198115694627e-07,
+      "loss": 0.0527,
+      "step": 1989
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.557632287595838e-07,
+      "loss": 0.0452,
+      "step": 1990
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.506261589667839e-07,
+      "loss": 0.0365,
+      "step": 1991
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.455086129185028e-07,
+      "loss": 0.0372,
+      "step": 1992
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.4041060130141e-07,
+      "loss": 0.0566,
+      "step": 1993
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.353321347613817e-07,
+      "loss": 0.0451,
+      "step": 1994
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.302732239034781e-07,
+      "loss": 0.0498,
+      "step": 1995
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.252338792919255e-07,
+      "loss": 0.0509,
+      "step": 1996
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.202141114500882e-07,
+      "loss": 0.0478,
+      "step": 1997
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.152139308604499e-07,
+      "loss": 0.0398,
+      "step": 1998
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.102333479645916e-07,
+      "loss": 0.0563,
+      "step": 1999
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.052723731631738e-07,
+      "loss": 0.0421,
+      "step": 2000
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.003310168159026e-07,
+      "loss": 0.049,
+      "step": 2001
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.954092892415208e-07,
+      "loss": 0.039,
+      "step": 2002
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.905072007177836e-07,
+      "loss": 0.0405,
+      "step": 2003
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.856247614814293e-07,
+      "loss": 0.0428,
+      "step": 2004
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.807619817281717e-07,
+      "loss": 0.043,
+      "step": 2005
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.759188716126606e-07,
+      "loss": 0.0504,
+      "step": 2006
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.875,
+      "learning_rate": 5.710954412484804e-07,
+      "loss": 0.0522,
+      "step": 2007
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.662917007081137e-07,
+      "loss": 0.0516,
+      "step": 2008
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.615076600229275e-07,
+      "loss": 0.0472,
+      "step": 2009
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.82421875,
+      "learning_rate": 5.567433291831525e-07,
+      "loss": 0.046,
+      "step": 2010
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.519987181378583e-07,
+      "loss": 0.0435,
+      "step": 2011
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.47273836794937e-07,
+      "loss": 0.0521,
+      "step": 2012
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.80859375,
+      "learning_rate": 5.425686950210762e-07,
+      "loss": 0.0521,
+      "step": 2013
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.378833026417452e-07,
+      "loss": 0.0573,
+      "step": 2014
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.332176694411761e-07,
+      "loss": 0.042,
+      "step": 2015
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.285718051623312e-07,
+      "loss": 0.0349,
+      "step": 2016
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.239457195068942e-07,
+      "loss": 0.0552,
+      "step": 2017
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.193394221352466e-07,
+      "loss": 0.0396,
+      "step": 2018
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.875,
+      "learning_rate": 5.147529226664471e-07,
+      "loss": 0.0537,
+      "step": 2019
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.101862306782135e-07,
+      "loss": 0.0329,
+      "step": 2020
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.056393557068928e-07,
+      "loss": 0.0329,
+      "step": 2021
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.88671875,
+      "learning_rate": 5.011123072474622e-07,
+      "loss": 0.0418,
+      "step": 2022
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.96605094753485e-07,
+      "loss": 0.0405,
+      "step": 2023
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.92117727637107e-07,
+      "loss": 0.061,
+      "step": 2024
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.87650215269031e-07,
+      "loss": 0.043,
+      "step": 2025
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.832025669785012e-07,
+      "loss": 0.0552,
+      "step": 2026
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.787747920532782e-07,
+      "loss": 0.0505,
+      "step": 2027
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.743668997396178e-07,
+      "loss": 0.0433,
+      "step": 2028
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.6997889924226715e-07,
+      "loss": 0.0423,
+      "step": 2029
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.6561079972442634e-07,
+      "loss": 0.0618,
+      "step": 2030
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.612626103077378e-07,
+      "loss": 0.0593,
+      "step": 2031
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.5693434007226924e-07,
+      "loss": 0.0555,
+      "step": 2032
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.526259980564907e-07,
+      "loss": 0.0462,
+      "step": 2033
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.483375932572598e-07,
+      "loss": 0.039,
+      "step": 2034
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.440691346297976e-07,
+      "loss": 0.055,
+      "step": 2035
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.398206310876718e-07,
+      "loss": 0.0453,
+      "step": 2036
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.355920915027856e-07,
+      "loss": 0.0577,
+      "step": 2037
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.313835247053444e-07,
+      "loss": 0.039,
+      "step": 2038
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.271949394838515e-07,
+      "loss": 0.0374,
+      "step": 2039
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.230263445850802e-07,
+      "loss": 0.0497,
+      "step": 2040
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.188777487140616e-07,
+      "loss": 0.0415,
+      "step": 2041
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.1474916053406365e-07,
+      "loss": 0.0691,
+      "step": 2042
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.1064058866656987e-07,
+      "loss": 0.0448,
+      "step": 2043
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.065520416912716e-07,
+      "loss": 0.0614,
+      "step": 2044
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.024835281460404e-07,
+      "loss": 0.0606,
+      "step": 2045
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.9843505652690793e-07,
+      "loss": 0.0449,
+      "step": 2046
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.944066352880616e-07,
+      "loss": 0.0453,
+      "step": 2047
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.903982728418143e-07,
+      "loss": 0.046,
+      "step": 2048
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.8640997755859165e-07,
+      "loss": 0.0529,
+      "step": 2049
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.8244175776691596e-07,
+      "loss": 0.043,
+      "step": 2050
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.784936217533852e-07,
+      "loss": 0.0516,
+      "step": 2051
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.745655777626589e-07,
+      "loss": 0.0507,
+      "step": 2052
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.706576339974399e-07,
+      "loss": 0.0531,
+      "step": 2053
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.6676979861845264e-07,
+      "loss": 0.0474,
+      "step": 2054
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.629020797444383e-07,
+      "loss": 0.0356,
+      "step": 2055
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.590544854521216e-07,
+      "loss": 0.0597,
+      "step": 2056
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.75,
+      "learning_rate": 3.55227023776209e-07,
+      "loss": 0.0605,
+      "step": 2057
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.514197027093602e-07,
+      "loss": 0.0616,
+      "step": 2058
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.476325302021799e-07,
+      "loss": 0.0475,
+      "step": 2059
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.438655141631986e-07,
+      "loss": 0.0467,
+      "step": 2060
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.401186624588471e-07,
+      "loss": 0.0654,
+      "step": 2061
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.3639198291345986e-07,
+      "loss": 0.0367,
+      "step": 2062
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.75,
+      "learning_rate": 3.326854833092408e-07,
+      "loss": 0.0505,
+      "step": 2063
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.289991713862506e-07,
+      "loss": 0.0537,
+      "step": 2064
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.253330548423994e-07,
+      "loss": 0.0561,
+      "step": 2065
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.2168714133341993e-07,
+      "loss": 0.0381,
+      "step": 2066
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.1806143847285864e-07,
+      "loss": 0.0435,
+      "step": 2067
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.1445595383205683e-07,
+      "loss": 0.0412,
+      "step": 2068
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.10870694940133e-07,
+      "loss": 0.0419,
+      "step": 2069
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.0730566928397487e-07,
+      "loss": 0.0515,
+      "step": 2070
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.037608843082107e-07,
+      "loss": 0.0578,
+      "step": 2071
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.00236347415207e-07,
+      "loss": 0.049,
+      "step": 2072
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.9673206596504524e-07,
+      "loss": 0.0527,
+      "step": 2073
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.932480472755106e-07,
+      "loss": 0.0486,
+      "step": 2074
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.8978429862207335e-07,
+      "loss": 0.0559,
+      "step": 2075
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.8634082723787093e-07,
+      "loss": 0.0582,
+      "step": 2076
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.829176403137057e-07,
+      "loss": 0.0415,
+      "step": 2077
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.7951474499801625e-07,
+      "loss": 0.0377,
+      "step": 2078
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.761321483968649e-07,
+      "loss": 0.0433,
+      "step": 2079
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.72769857573929e-07,
+      "loss": 0.0413,
+      "step": 2080
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.6942787955048235e-07,
+      "loss": 0.0431,
+      "step": 2081
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.75,
+      "learning_rate": 2.6610622130538e-07,
+      "loss": 0.0592,
+      "step": 2082
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.6280488977504346e-07,
+      "loss": 0.0475,
+      "step": 2083
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.595238918534493e-07,
+      "loss": 0.0534,
+      "step": 2084
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.5626323439211233e-07,
+      "loss": 0.0511,
+      "step": 2085
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.530229242000681e-07,
+      "loss": 0.0527,
+      "step": 2086
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.498029680438652e-07,
+      "loss": 0.0568,
+      "step": 2087
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.466033726475503e-07,
+      "loss": 0.0343,
+      "step": 2088
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.434241446926466e-07,
+      "loss": 0.0404,
+      "step": 2089
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.402652908181491e-07,
+      "loss": 0.0527,
+      "step": 2090
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.371268176205044e-07,
+      "loss": 0.0648,
+      "step": 2091
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.3400873165360217e-07,
+      "loss": 0.0511,
+      "step": 2092
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.3091103942875614e-07,
+      "loss": 0.0377,
+      "step": 2093
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.2783374741469188e-07,
+      "loss": 0.046,
+      "step": 2094
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.2477686203754013e-07,
+      "loss": 0.0432,
+      "step": 2095
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.2174038968081124e-07,
+      "loss": 0.058,
+      "step": 2096
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.1872433668539084e-07,
+      "loss": 0.0558,
+      "step": 2097
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.1572870934952528e-07,
+      "loss": 0.0458,
+      "step": 2098
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.1275351392880507e-07,
+      "loss": 0.06,
+      "step": 2099
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.75,
+      "learning_rate": 2.09798756636157e-07,
+      "loss": 0.0447,
+      "step": 2100
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.0686444364182213e-07,
+      "loss": 0.049,
+      "step": 2101
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.0395058107335442e-07,
+      "loss": 0.054,
+      "step": 2102
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.010571750156032e-07,
+      "loss": 0.045,
+      "step": 2103
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.981842315106941e-07,
+      "loss": 0.058,
+      "step": 2104
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9533175655802595e-07,
+      "loss": 0.0496,
+      "step": 2105
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9249975611425498e-07,
+      "loss": 0.0466,
+      "step": 2106
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.8968823609327835e-07,
+      "loss": 0.0583,
+      "step": 2107
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8689720236622854e-07,
+      "loss": 0.0365,
+      "step": 2108
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8412666076145335e-07,
+      "loss": 0.0382,
+      "step": 2109
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.813766170645148e-07,
+      "loss": 0.0426,
+      "step": 2110
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.7864707701816474e-07,
+      "loss": 0.0523,
+      "step": 2111
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.7593804632234035e-07,
+      "loss": 0.0728,
+      "step": 2112
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.0,
+      "learning_rate": 1.7324953063414973e-07,
+      "loss": 0.0523,
+      "step": 2113
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7058153556786194e-07,
+      "loss": 0.0493,
+      "step": 2114
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.6793406669489477e-07,
+      "loss": 0.0633,
+      "step": 2115
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.653071295437958e-07,
+      "loss": 0.0571,
+      "step": 2116
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.6270072960024806e-07,
+      "loss": 0.0529,
+      "step": 2117
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6011487230704115e-07,
+      "loss": 0.0452,
+      "step": 2118
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5754956306406556e-07,
+      "loss": 0.0517,
+      "step": 2119
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5500480722830724e-07,
+      "loss": 0.0426,
+      "step": 2120
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.5248061011382764e-07,
+      "loss": 0.0473,
+      "step": 2121
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.499769769917603e-07,
+      "loss": 0.0459,
+      "step": 2122
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.4749391309029192e-07,
+      "loss": 0.055,
+      "step": 2123
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4503142359465926e-07,
+      "loss": 0.0624,
+      "step": 2124
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4258951364713557e-07,
+      "loss": 0.0375,
+      "step": 2125
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4016818834701297e-07,
+      "loss": 0.049,
+      "step": 2126
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3776745275060342e-07,
+      "loss": 0.0457,
+      "step": 2127
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.3538731187122234e-07,
+      "loss": 0.0515,
+      "step": 2128
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.3302777067917606e-07,
+      "loss": 0.0543,
+      "step": 2129
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.3068883410175536e-07,
+      "loss": 0.0512,
+      "step": 2130
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.2837050702322328e-07,
+      "loss": 0.0461,
+      "step": 2131
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.260727942848039e-07,
+      "loss": 0.0449,
+      "step": 2132
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2379570068467794e-07,
+      "loss": 0.0447,
+      "step": 2133
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.2153923097796172e-07,
+      "loss": 0.0491,
+      "step": 2134
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.193033898767082e-07,
+      "loss": 0.043,
+      "step": 2135
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.170881820498948e-07,
+      "loss": 0.0433,
+      "step": 2136
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.1489361212340566e-07,
+      "loss": 0.0448,
+      "step": 2137
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.1271968468003381e-07,
+      "loss": 0.0381,
+      "step": 2138
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.105664042594612e-07,
+      "loss": 0.038,
+      "step": 2139
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0843377535825872e-07,
+      "loss": 0.0402,
+      "step": 2140
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.0632180242986734e-07,
+      "loss": 0.0328,
+      "step": 2141
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.0423048988459472e-07,
+      "loss": 0.0404,
+      "step": 2142
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.021598420896075e-07,
+      "loss": 0.0635,
+      "step": 2143
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0010986336891459e-07,
+      "loss": 0.0543,
+      "step": 2144
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.808055800336614e-08,
+      "loss": 0.0458,
+      "step": 2145
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.607193023064121e-08,
+      "loss": 0.0659,
+      "step": 2146
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.408398424523791e-08,
+      "loss": 0.0449,
+      "step": 2147
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.211672419846662e-08,
+      "loss": 0.0368,
+      "step": 2148
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.0170154198439e-08,
+      "loss": 0.0637,
+      "step": 2149
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.824427831006343e-08,
+      "loss": 0.0357,
+      "step": 2150
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.633910055503181e-08,
+      "loss": 0.049,
+      "step": 2151
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.4454624911815e-08,
+      "loss": 0.0488,
+      "step": 2152
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.259085531565181e-08,
+      "loss": 0.0473,
+      "step": 2153
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.074779565854118e-08,
+      "loss": 0.0408,
+      "step": 2154
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.892544978923555e-08,
+      "loss": 0.0762,
+      "step": 2155
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.712382151323416e-08,
+      "loss": 0.039,
+      "step": 2156
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.534291459276866e-08,
+      "loss": 0.0635,
+      "step": 2157
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.358273274680195e-08,
+      "loss": 0.049,
+      "step": 2158
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.184327965101934e-08,
+      "loss": 0.0357,
+      "step": 2159
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.012455893781522e-08,
+      "loss": 0.0479,
+      "step": 2160
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.842657419629417e-08,
+      "loss": 0.0389,
+      "step": 2161
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.67493289722576e-08,
+      "loss": 0.0487,
+      "step": 2162
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.509282676819717e-08,
+      "loss": 0.044,
+      "step": 2163
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.345707104328691e-08,
+      "loss": 0.047,
+      "step": 2164
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.184206521337999e-08,
+      "loss": 0.0544,
+      "step": 2165
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.024781265099866e-08,
+      "loss": 0.049,
+      "step": 2166
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.867431668532542e-08,
+      "loss": 0.0425,
+      "step": 2167
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.712158060219741e-08,
+      "loss": 0.0418,
+      "step": 2168
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.558960764410426e-08,
+      "loss": 0.0552,
+      "step": 2169
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.407840101017248e-08,
+      "loss": 0.0424,
+      "step": 2170
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.258796385616549e-08,
+      "loss": 0.0503,
+      "step": 2171
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.111829929447698e-08,
+      "loss": 0.0392,
+      "step": 2172
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.966941039411866e-08,
+      "loss": 0.0476,
+      "step": 2173
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.824130018072026e-08,
+      "loss": 0.0359,
+      "step": 2174
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.683397163651959e-08,
+      "loss": 0.0553,
+      "step": 2175
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.544742770035804e-08,
+      "loss": 0.0467,
+      "step": 2176
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.408167126767282e-08,
+      "loss": 0.0561,
+      "step": 2177
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.273670519049256e-08,
+      "loss": 0.0578,
+      "step": 2178
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.1412532277430585e-08,
+      "loss": 0.0542,
+      "step": 2179
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.010915529367831e-08,
+      "loss": 0.0357,
+      "step": 2180
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.88265769610019e-08,
+      "loss": 0.0568,
+      "step": 2181
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.756479995773332e-08,
+      "loss": 0.0456,
+      "step": 2182
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.6323826918766016e-08,
+      "loss": 0.0531,
+      "step": 2183
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.510366043555147e-08,
+      "loss": 0.0488,
+      "step": 2184
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.390430305609149e-08,
+      "loss": 0.0397,
+      "step": 2185
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.2725757284932655e-08,
+      "loss": 0.0551,
+      "step": 2186
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.1568025583161855e-08,
+      "loss": 0.0446,
+      "step": 2187
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.043111036840296e-08,
+      "loss": 0.0577,
+      "step": 2188
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.931501401480796e-08,
+      "loss": 0.0433,
+      "step": 2189
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.8219738853054736e-08,
+      "loss": 0.0395,
+      "step": 2190
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.7145287170342592e-08,
+      "loss": 0.0515,
+      "step": 2191
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.6091661210384532e-08,
+      "loss": 0.0484,
+      "step": 2192
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.5058863173406113e-08,
+      "loss": 0.0403,
+      "step": 2193
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.4046895216136566e-08,
+      "loss": 0.0483,
+      "step": 2194
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.3055759451811044e-08,
+      "loss": 0.0556,
+      "step": 2195
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.2085457950159482e-08,
+      "loss": 0.0502,
+      "step": 2196
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.1135992737403298e-08,
+      "loss": 0.0403,
+      "step": 2197
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.0207365796256483e-08,
+      "loss": 0.0562,
+      "step": 2198
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9299579065914508e-08,
+      "loss": 0.0572,
+      "step": 2199
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8412634442055434e-08,
+      "loss": 0.0483,
+      "step": 2200
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.754653377683324e-08,
+      "loss": 0.0394,
+      "step": 2201
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6701278878873406e-08,
+      "loss": 0.0636,
+      "step": 2202
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5876871513273993e-08,
+      "loss": 0.0402,
+      "step": 2203
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.507331340159457e-08,
+      "loss": 0.039,
+      "step": 2204
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.4290606221858405e-08,
+      "loss": 0.0405,
+      "step": 2205
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3528751608548052e-08,
+      "loss": 0.041,
+      "step": 2206
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.278775115259645e-08,
+      "loss": 0.0365,
+      "step": 2207
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2067606401392484e-08,
+      "loss": 0.0348,
+      "step": 2208
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.1368318858773208e-08,
+      "loss": 0.0586,
+      "step": 2209
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.0689889985018298e-08,
+      "loss": 0.0396,
+      "step": 2210
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0032321196850048e-08,
+      "loss": 0.0452,
+      "step": 2211
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.395613867430043e-09,
+      "loss": 0.0433,
+      "step": 2212
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.779769326359155e-09,
+      "loss": 0.0667,
+      "step": 2213
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.184788859667559e-09,
+      "loss": 0.0618,
+      "step": 2214
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.610673709816941e-09,
+      "loss": 0.0407,
+      "step": 2215
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.057425075698288e-09,
+      "loss": 0.0598,
+      "step": 2216
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.525044112629664e-09,
+      "loss": 0.0448,
+      "step": 2217
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.013531932349548e-09,
+      "loss": 0.0597,
+      "step": 2218
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.522889603017945e-09,
+      "loss": 0.0425,
+      "step": 2219
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.053118149215275e-09,
+      "loss": 0.0401,
+      "step": 2220
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.604218551937933e-09,
+      "loss": 0.0518,
+      "step": 2221
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.17619174859496e-09,
+      "loss": 0.0565,
+      "step": 2222
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.769038633009147e-09,
+      "loss": 0.0468,
+      "step": 2223
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.382760055417045e-09,
+      "loss": 0.0486,
+      "step": 2224
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.017356822457851e-09,
+      "loss": 0.0452,
+      "step": 2225
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.67282969718341e-09,
+      "loss": 0.0349,
+      "step": 2226
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.3491793990504386e-09,
+      "loss": 0.0493,
+      "step": 2227
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.0464066039160847e-09,
+      "loss": 0.0418,
+      "step": 2228
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.764511944044589e-09,
+      "loss": 0.0472,
+      "step": 2229
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.625,
+      "learning_rate": 1.5034960080995143e-09,
+      "loss": 0.0423,
+      "step": 2230
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.75,
+      "learning_rate": 1.2633593411448542e-09,
+      "loss": 0.0514,
+      "step": 2231
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0441024446439242e-09,
+      "loss": 0.0405,
+      "step": 2232
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.457257764582505e-10,
+      "loss": 0.0436,
+      "step": 2233
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.682297508464608e-10,
+      "loss": 0.0443,
+      "step": 2234
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.116147384609527e-10,
+      "loss": 0.0567,
+      "step": 2235
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.758810663534451e-10,
+      "loss": 0.0468,
+      "step": 2236
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.6102901796831724e-10,
+      "loss": 0.0415,
+      "step": 2237
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.670588331437184e-10,
+      "loss": 0.037,
+      "step": 2238
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.397070811267838e-11,
+      "loss": 0.0378,
+      "step": 2239
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.176479550088708e-11,
+      "loss": 0.0359,
+      "step": 2240
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.0441204326694376e-11,
+      "loss": 0.0473,
+      "step": 2241
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0,
+      "loss": 0.0484,
+      "step": 2242
+    },
+    {
+      "epoch": 2.0,
+      "step": 2242,
+      "total_flos": 2.764018965901476e+17,
+      "train_loss": 0.11786383062926761,
+      "train_runtime": 12105.9407,
+      "train_samples_per_second": 11.853,
+      "train_steps_per_second": 0.185
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2242,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "total_flos": 2.764018965901476e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}