diff --git "a/20_23/trainer_state.json" "b/20_23/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/20_23/trainer_state.json"
@@ -0,0 +1,5343 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 759,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4626648426055908,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 0.3776,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4091928005218506,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 0.3922,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.363326072692871,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 0.3599,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1503901481628418,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 0.318,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7365157604217529,
+      "learning_rate": 2.173913043478261e-05,
+      "loss": 0.2927,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6607586145401001,
+      "learning_rate": 2.608695652173913e-05,
+      "loss": 0.2404,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5922650098800659,
+      "learning_rate": 3.0434782608695656e-05,
+      "loss": 0.1948,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3794650733470917,
+      "learning_rate": 3.478260869565218e-05,
+      "loss": 0.15,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3533678948879242,
+      "learning_rate": 3.91304347826087e-05,
+      "loss": 0.1375,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.28944671154022217,
+      "learning_rate": 4.347826086956522e-05,
+      "loss": 0.1285,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.22203072905540466,
+      "learning_rate": 4.782608695652174e-05,
+      "loss": 0.1202,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.20047953724861145,
+      "learning_rate": 5.217391304347826e-05,
+      "loss": 0.1136,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.16146406531333923,
+      "learning_rate": 5.652173913043478e-05,
+      "loss": 0.0967,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.22107285261154175,
+      "learning_rate": 6.086956521739131e-05,
+      "loss": 0.1121,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.22693370282649994,
+      "learning_rate": 6.521739130434783e-05,
+      "loss": 0.1157,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2420782446861267,
+      "learning_rate": 6.956521739130436e-05,
+      "loss": 0.1019,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2712867259979248,
+      "learning_rate": 7.391304347826086e-05,
+      "loss": 0.0927,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.16125057637691498,
+      "learning_rate": 7.82608695652174e-05,
+      "loss": 0.0956,
+      "step": 18
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1539854109287262,
+      "learning_rate": 8.260869565217392e-05,
+      "loss": 0.0887,
+      "step": 19
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17831459641456604,
+      "learning_rate": 8.695652173913044e-05,
+      "loss": 0.0993,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17226484417915344,
+      "learning_rate": 9.130434782608696e-05,
+      "loss": 0.0974,
+      "step": 21
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17678162455558777,
+      "learning_rate": 9.565217391304348e-05,
+      "loss": 0.0938,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18184365332126617,
+      "learning_rate": 0.0001,
+      "loss": 0.0952,
+      "step": 23
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.20284950733184814,
+      "learning_rate": 9.99995445051553e-05,
+      "loss": 0.0948,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1573178917169571,
+      "learning_rate": 9.99981780289202e-05,
+      "loss": 0.0927,
+      "step": 25
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.19211460649967194,
+      "learning_rate": 9.999590059619164e-05,
+      "loss": 0.0909,
+      "step": 26
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13000018894672394,
+      "learning_rate": 9.999271224846396e-05,
+      "loss": 0.0935,
+      "step": 27
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.10931742191314697,
+      "learning_rate": 9.998861304382818e-05,
+      "loss": 0.0875,
+      "step": 28
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15605789422988892,
+      "learning_rate": 9.998360305697101e-05,
+      "loss": 0.1009,
+      "step": 29
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.09958553314208984,
+      "learning_rate": 9.997768237917333e-05,
+      "loss": 0.0859,
+      "step": 30
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.16153687238693237,
+      "learning_rate": 9.99708511183087e-05,
+      "loss": 0.0928,
+      "step": 31
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13266044855117798,
+      "learning_rate": 9.996310939884128e-05,
+      "loss": 0.0884,
+      "step": 32
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14477793872356415,
+      "learning_rate": 9.995445736182358e-05,
+      "loss": 0.1003,
+      "step": 33
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.16262026131153107,
+      "learning_rate": 9.994489516489396e-05,
+      "loss": 0.0959,
+      "step": 34
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1339733898639679,
+      "learning_rate": 9.993442298227365e-05,
+      "loss": 0.0834,
+      "step": 35
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1648261696100235,
+      "learning_rate": 9.992304100476367e-05,
+      "loss": 0.091,
+      "step": 36
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.10690194368362427,
+      "learning_rate": 9.991074943974131e-05,
+      "loss": 0.0917,
+      "step": 37
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13092245161533356,
+      "learning_rate": 9.989754851115635e-05,
+      "loss": 0.0875,
+      "step": 38
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.11928656697273254,
+      "learning_rate": 9.988343845952697e-05,
+      "loss": 0.0908,
+      "step": 39
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14026062190532684,
+      "learning_rate": 9.98684195419354e-05,
+      "loss": 0.0885,
+      "step": 40
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.16105924546718597,
+      "learning_rate": 9.985249203202324e-05,
+      "loss": 0.0836,
+      "step": 41
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1550360769033432,
+      "learning_rate": 9.983565621998644e-05,
+      "loss": 0.0865,
+      "step": 42
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1308027058839798,
+      "learning_rate": 9.981791241257e-05,
+      "loss": 0.0924,
+      "step": 43
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15918859839439392,
+      "learning_rate": 9.979926093306245e-05,
+      "loss": 0.0935,
+      "step": 44
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1623181700706482,
+      "learning_rate": 9.97797021212899e-05,
+      "loss": 0.0873,
+      "step": 45
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11046040058135986,
+      "learning_rate": 9.975923633360985e-05,
+      "loss": 0.0887,
+      "step": 46
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14533914625644684,
+      "learning_rate": 9.973786394290474e-05,
+      "loss": 0.0928,
+      "step": 47
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.10755528509616852,
+      "learning_rate": 9.971558533857515e-05,
+      "loss": 0.0811,
+      "step": 48
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13999031484127045,
+      "learning_rate": 9.969240092653261e-05,
+      "loss": 0.0872,
+      "step": 49
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11199218034744263,
+      "learning_rate": 9.966831112919235e-05,
+      "loss": 0.083,
+      "step": 50
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11688799411058426,
+      "learning_rate": 9.96433163854655e-05,
+      "loss": 0.0916,
+      "step": 51
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1127270981669426,
+      "learning_rate": 9.961741715075115e-05,
+      "loss": 0.0908,
+      "step": 52
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1565464586019516,
+      "learning_rate": 9.9590613896928e-05,
+      "loss": 0.0869,
+      "step": 53
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10714145749807358,
+      "learning_rate": 9.956290711234584e-05,
+      "loss": 0.086,
+      "step": 54
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.09617923945188522,
+      "learning_rate": 9.953429730181653e-05,
+      "loss": 0.0911,
+      "step": 55
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1050075963139534,
+      "learning_rate": 9.950478498660496e-05,
+      "loss": 0.0869,
+      "step": 56
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11093985289335251,
+      "learning_rate": 9.947437070441938e-05,
+      "loss": 0.0896,
+      "step": 57
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10569228231906891,
+      "learning_rate": 9.944305500940178e-05,
+      "loss": 0.0954,
+      "step": 58
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10598912090063095,
+      "learning_rate": 9.941083847211765e-05,
+      "loss": 0.0909,
+      "step": 59
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1131819486618042,
+      "learning_rate": 9.937772167954564e-05,
+      "loss": 0.0962,
+      "step": 60
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1235487088561058,
+      "learning_rate": 9.93437052350669e-05,
+      "loss": 0.096,
+      "step": 61
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13566148281097412,
+      "learning_rate": 9.930878975845406e-05,
+      "loss": 0.0917,
+      "step": 62
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10908336937427521,
+      "learning_rate": 9.927297588585984e-05,
+      "loss": 0.0853,
+      "step": 63
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10724148899316788,
+      "learning_rate": 9.923626426980567e-05,
+      "loss": 0.0853,
+      "step": 64
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1097191870212555,
+      "learning_rate": 9.91986555791696e-05,
+      "loss": 0.0821,
+      "step": 65
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.09474969655275345,
+      "learning_rate": 9.91601504991742e-05,
+      "loss": 0.079,
+      "step": 66
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1255607008934021,
+      "learning_rate": 9.912074973137412e-05,
+      "loss": 0.0877,
+      "step": 67
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.08756373077630997,
+      "learning_rate": 9.908045399364322e-05,
+      "loss": 0.0859,
+      "step": 68
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.13312621414661407,
+      "learning_rate": 9.903926402016153e-05,
+      "loss": 0.0924,
+      "step": 69
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10092984139919281,
+      "learning_rate": 9.899718056140186e-05,
+      "loss": 0.0791,
+      "step": 70
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11156034469604492,
+      "learning_rate": 9.895420438411616e-05,
+      "loss": 0.0794,
+      "step": 71
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.09780649840831757,
+      "learning_rate": 9.891033627132148e-05,
+      "loss": 0.0881,
+      "step": 72
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10802754014730453,
+      "learning_rate": 9.886557702228587e-05,
+      "loss": 0.0863,
+      "step": 73
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09974031895399094,
+      "learning_rate": 9.881992745251356e-05,
+      "loss": 0.0781,
+      "step": 74
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09702256321907043,
+      "learning_rate": 9.877338839373032e-05,
+      "loss": 0.0831,
+      "step": 75
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11543964594602585,
+      "learning_rate": 9.872596069386817e-05,
+      "loss": 0.0894,
+      "step": 76
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09120846539735794,
+      "learning_rate": 9.867764521705005e-05,
+      "loss": 0.0875,
+      "step": 77
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09430748224258423,
+      "learning_rate": 9.8628442843574e-05,
+      "loss": 0.0743,
+      "step": 78
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09428028017282486,
+      "learning_rate": 9.857835446989707e-05,
+      "loss": 0.0807,
+      "step": 79
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.08958426862955093,
+      "learning_rate": 9.852738100861916e-05,
+      "loss": 0.092,
+      "step": 80
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10480056703090668,
+      "learning_rate": 9.847552338846617e-05,
+      "loss": 0.0864,
+      "step": 81
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1178659051656723,
+      "learning_rate": 9.842278255427327e-05,
+      "loss": 0.079,
+      "step": 82
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15165001153945923,
+      "learning_rate": 9.836915946696759e-05,
+      "loss": 0.0948,
+      "step": 83
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12370532751083374,
+      "learning_rate": 9.831465510355069e-05,
+      "loss": 0.0782,
+      "step": 84
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1311461627483368,
+      "learning_rate": 9.825927045708088e-05,
+      "loss": 0.0865,
+      "step": 85
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10418644547462463,
+      "learning_rate": 9.820300653665495e-05,
+      "loss": 0.0789,
+      "step": 86
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11643680185079575,
+      "learning_rate": 9.814586436738998e-05,
+      "loss": 0.0846,
+      "step": 87
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12964823842048645,
+      "learning_rate": 9.808784499040445e-05,
+      "loss": 0.0789,
+      "step": 88
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1188376173377037,
+      "learning_rate": 9.80289494627995e-05,
+      "loss": 0.0903,
+      "step": 89
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.08869314193725586,
+      "learning_rate": 9.796917885763946e-05,
+      "loss": 0.0791,
+      "step": 90
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09060724079608917,
+      "learning_rate": 9.790853426393245e-05,
+      "loss": 0.0795,
+      "step": 91
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09172531962394714,
+      "learning_rate": 9.784701678661045e-05,
+      "loss": 0.0826,
+      "step": 92
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.08181829005479813,
+      "learning_rate": 9.778462754650921e-05,
+      "loss": 0.0847,
+      "step": 93
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0949179083108902,
+      "learning_rate": 9.772136768034785e-05,
+      "loss": 0.0813,
+      "step": 94
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0956520140171051,
+      "learning_rate": 9.765723834070804e-05,
+      "loss": 0.086,
+      "step": 95
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1323326826095581,
+      "learning_rate": 9.759224069601317e-05,
+      "loss": 0.0838,
+      "step": 96
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1346251666545868,
+      "learning_rate": 9.752637593050689e-05,
+      "loss": 0.0841,
+      "step": 97
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09667576849460602,
+      "learning_rate": 9.745964524423165e-05,
+      "loss": 0.0802,
+      "step": 98
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.13813923299312592,
+      "learning_rate": 9.739204985300679e-05,
+      "loss": 0.0814,
+      "step": 99
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11109049618244171,
+      "learning_rate": 9.732359098840642e-05,
+      "loss": 0.0923,
+      "step": 100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10362949222326279,
+      "learning_rate": 9.725426989773693e-05,
+      "loss": 0.0843,
+      "step": 101
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11478770524263382,
+      "learning_rate": 9.718408784401427e-05,
+      "loss": 0.0864,
+      "step": 102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.11098852008581161,
+      "learning_rate": 9.711304610594104e-05,
+      "loss": 0.0906,
+      "step": 103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1379733830690384,
+      "learning_rate": 9.7041145977883e-05,
+      "loss": 0.0802,
+      "step": 104
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.12902683019638062,
+      "learning_rate": 9.696838876984569e-05,
+      "loss": 0.0805,
+      "step": 105
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.097101591527462,
+      "learning_rate": 9.689477580745042e-05,
+      "loss": 0.0841,
+      "step": 106
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.12172501534223557,
+      "learning_rate": 9.682030843191022e-05,
+      "loss": 0.0837,
+      "step": 107
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.11309538781642914,
+      "learning_rate": 9.674498800000526e-05,
+      "loss": 0.0857,
+      "step": 108
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10910997539758682,
+      "learning_rate": 9.666881588405832e-05,
+      "loss": 0.0939,
+      "step": 109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.12583476305007935,
+      "learning_rate": 9.659179347190963e-05,
+      "loss": 0.0914,
+      "step": 110
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10188722610473633,
+      "learning_rate": 9.651392216689165e-05,
+      "loss": 0.0838,
+      "step": 111
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08564438670873642,
+      "learning_rate": 9.643520338780354e-05,
+      "loss": 0.0773,
+      "step": 112
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08970914781093597,
+      "learning_rate": 9.635563856888517e-05,
+      "loss": 0.0916,
+      "step": 113
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09510532021522522,
+      "learning_rate": 9.627522915979114e-05,
+      "loss": 0.0861,
+      "step": 114
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10671927779912949,
+      "learning_rate": 9.619397662556435e-05,
+      "loss": 0.087,
+      "step": 115
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10593091696500778,
+      "learning_rate": 9.611188244660915e-05,
+      "loss": 0.0815,
+      "step": 116
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.14427386224269867,
+      "learning_rate": 9.60289481186646e-05,
+      "loss": 0.0847,
+      "step": 117
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0999630019068718,
+      "learning_rate": 9.594517515277705e-05,
+      "loss": 0.081,
+      "step": 118
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10400121659040451,
+      "learning_rate": 9.586056507527266e-05,
+      "loss": 0.0778,
+      "step": 119
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09455764293670654,
+      "learning_rate": 9.577511942772957e-05,
+      "loss": 0.0838,
+      "step": 120
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10111963003873825,
+      "learning_rate": 9.568883976694989e-05,
+      "loss": 0.0807,
+      "step": 121
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10892399400472641,
+      "learning_rate": 9.560172766493122e-05,
+      "loss": 0.0731,
+      "step": 122
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1193283423781395,
+      "learning_rate": 9.551378470883812e-05,
+      "loss": 0.0787,
+      "step": 123
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.11868227273225784,
+      "learning_rate": 9.542501250097309e-05,
+      "loss": 0.0794,
+      "step": 124
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.11394625902175903,
+      "learning_rate": 9.533541265874749e-05,
+      "loss": 0.0806,
+      "step": 125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.11192644387483597,
+      "learning_rate": 9.524498681465191e-05,
+      "loss": 0.084,
+      "step": 126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1109413430094719,
+      "learning_rate": 9.515373661622664e-05,
+      "loss": 0.0783,
+      "step": 127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.12146401405334473,
+      "learning_rate": 9.506166372603144e-05,
+      "loss": 0.0799,
+      "step": 128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09204913675785065,
+      "learning_rate": 9.496876982161542e-05,
+      "loss": 0.0795,
+      "step": 129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13367091119289398,
+      "learning_rate": 9.487505659548633e-05,
+      "loss": 0.0865,
+      "step": 130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.11176060140132904,
+      "learning_rate": 9.478052575507982e-05,
+      "loss": 0.0832,
+      "step": 131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0970437228679657,
+      "learning_rate": 9.468517902272836e-05,
+      "loss": 0.0853,
+      "step": 132
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10132851451635361,
+      "learning_rate": 9.45890181356297e-05,
+      "loss": 0.0906,
+      "step": 133
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.12170128524303436,
+      "learning_rate": 9.449204484581539e-05,
+      "loss": 0.0869,
+      "step": 134
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10223282128572464,
+      "learning_rate": 9.439426092011875e-05,
+      "loss": 0.0772,
+      "step": 135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.11369918286800385,
+      "learning_rate": 9.429566814014281e-05,
+      "loss": 0.0819,
+      "step": 136
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.12759996950626373,
+      "learning_rate": 9.419626830222762e-05,
+      "loss": 0.0837,
+      "step": 137
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.11163049191236496,
+      "learning_rate": 9.409606321741775e-05,
+      "loss": 0.0868,
+      "step": 138
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10981953144073486,
+      "learning_rate": 9.39950547114292e-05,
+      "loss": 0.079,
+      "step": 139
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09947255253791809,
+      "learning_rate": 9.38932446246161e-05,
+      "loss": 0.0803,
+      "step": 140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10793828219175339,
+      "learning_rate": 9.379063481193725e-05,
+      "loss": 0.0888,
+      "step": 141
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10283903777599335,
+      "learning_rate": 9.368722714292227e-05,
+      "loss": 0.082,
+      "step": 142
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.11633747816085815,
+      "learning_rate": 9.358302350163757e-05,
+      "loss": 0.0737,
+      "step": 143
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.11395063251256943,
+      "learning_rate": 9.347802578665199e-05,
+      "loss": 0.0781,
+      "step": 144
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.11062236875295639,
+      "learning_rate": 9.337223591100228e-05,
+      "loss": 0.0848,
+      "step": 145
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1253620684146881,
+      "learning_rate": 9.326565580215811e-05,
+      "loss": 0.0907,
+      "step": 146
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1102418303489685,
+      "learning_rate": 9.315828740198714e-05,
+      "loss": 0.0814,
+      "step": 147
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.11026754230260849,
+      "learning_rate": 9.305013266671945e-05,
+      "loss": 0.0759,
+      "step": 148
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.12312046438455582,
+      "learning_rate": 9.2941193566912e-05,
+      "loss": 0.0853,
+      "step": 149
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1312018185853958,
+      "learning_rate": 9.283147208741276e-05,
+      "loss": 0.0862,
+      "step": 150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14230753481388092,
+      "learning_rate": 9.272097022732443e-05,
+      "loss": 0.0913,
+      "step": 151
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1690019816160202,
+      "learning_rate": 9.260968999996813e-05,
+      "loss": 0.0892,
+      "step": 152
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08823404461145401,
+      "learning_rate": 9.249763343284665e-05,
+      "loss": 0.0788,
+      "step": 153
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09784504026174545,
+      "learning_rate": 9.238480256760755e-05,
+      "loss": 0.0799,
+      "step": 154
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.11563264578580856,
+      "learning_rate": 9.22711994600059e-05,
+      "loss": 0.0839,
+      "step": 155
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09026098251342773,
+      "learning_rate": 9.21568261798669e-05,
+      "loss": 0.0812,
+      "step": 156
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09114386886358261,
+      "learning_rate": 9.204168481104815e-05,
+      "loss": 0.0818,
+      "step": 157
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09261880069971085,
+      "learning_rate": 9.192577745140164e-05,
+      "loss": 0.0775,
+      "step": 158
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.11318907141685486,
+      "learning_rate": 9.180910621273555e-05,
+      "loss": 0.0804,
+      "step": 159
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1209554672241211,
+      "learning_rate": 9.169167322077578e-05,
+      "loss": 0.0854,
+      "step": 160
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.12757565081119537,
+      "learning_rate": 9.157348061512727e-05,
+      "loss": 0.0886,
+      "step": 161
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.11294035613536835,
+      "learning_rate": 9.145453054923487e-05,
+      "loss": 0.0941,
+      "step": 162
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.11256886273622513,
+      "learning_rate": 9.133482519034428e-05,
+      "loss": 0.0771,
+      "step": 163
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.11136170476675034,
+      "learning_rate": 9.121436671946247e-05,
+      "loss": 0.0873,
+      "step": 164
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.12420880794525146,
+      "learning_rate": 9.109315733131791e-05,
+      "loss": 0.0808,
+      "step": 165
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.11222650110721588,
+      "learning_rate": 9.097119923432066e-05,
+      "loss": 0.0855,
+      "step": 166
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09688053280115128,
+      "learning_rate": 9.08484946505221e-05,
+      "loss": 0.0795,
+      "step": 167
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.10351915657520294,
+      "learning_rate": 9.072504581557446e-05,
+      "loss": 0.0864,
+      "step": 168
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.10438606888055801,
+      "learning_rate": 9.060085497869004e-05,
+      "loss": 0.0805,
+      "step": 169
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.11572164297103882,
+      "learning_rate": 9.047592440260029e-05,
+      "loss": 0.0786,
+      "step": 170
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.10311150550842285,
+      "learning_rate": 9.035025636351452e-05,
+      "loss": 0.0813,
+      "step": 171
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0920066237449646,
+      "learning_rate": 9.022385315107853e-05,
+      "loss": 0.0836,
+      "step": 172
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.12571555376052856,
+      "learning_rate": 9.009671706833276e-05,
+      "loss": 0.0908,
+      "step": 173
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.12330035865306854,
+      "learning_rate": 8.99688504316704e-05,
+      "loss": 0.0832,
+      "step": 174
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.11306159198284149,
+      "learning_rate": 8.984025557079523e-05,
+      "loss": 0.0804,
+      "step": 175
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0854019969701767,
+      "learning_rate": 8.97109348286791e-05,
+      "loss": 0.0878,
+      "step": 176
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09134180843830109,
+      "learning_rate": 8.958089056151924e-05,
+      "loss": 0.086,
+      "step": 177
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1045374944806099,
+      "learning_rate": 8.945012513869542e-05,
+      "loss": 0.0812,
+      "step": 178
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.10817347466945648,
+      "learning_rate": 8.931864094272663e-05,
+      "loss": 0.0733,
+      "step": 179
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.11727076768875122,
+      "learning_rate": 8.918644036922783e-05,
+      "loss": 0.0791,
+      "step": 180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.09995684027671814,
+      "learning_rate": 8.905352582686622e-05,
+      "loss": 0.0784,
+      "step": 181
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.10814005136489868,
+      "learning_rate": 8.891989973731735e-05,
+      "loss": 0.0801,
+      "step": 182
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.11438959091901779,
+      "learning_rate": 8.8785564535221e-05,
+      "loss": 0.0825,
+      "step": 183
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.09477780759334564,
+      "learning_rate": 8.865052266813685e-05,
+      "loss": 0.0781,
+      "step": 184
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12783315777778625,
+      "learning_rate": 8.851477659649989e-05,
+      "loss": 0.0852,
+      "step": 185
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13230524957180023,
+      "learning_rate": 8.837832879357555e-05,
+      "loss": 0.0855,
+      "step": 186
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.11520255357027054,
+      "learning_rate": 8.824118174541464e-05,
+      "loss": 0.0799,
+      "step": 187
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.11846806108951569,
+      "learning_rate": 8.810333795080812e-05,
+      "loss": 0.0837,
+      "step": 188
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1257033497095108,
+      "learning_rate": 8.79647999212415e-05,
+      "loss": 0.0812,
+      "step": 189
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1521902084350586,
+      "learning_rate": 8.782557018084908e-05,
+      "loss": 0.0767,
+      "step": 190
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.11576004326343536,
+      "learning_rate": 8.768565126636806e-05,
+      "loss": 0.0792,
+      "step": 191
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1163419857621193,
+      "learning_rate": 8.754504572709219e-05,
+      "loss": 0.0762,
+      "step": 192
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.11233483254909515,
+      "learning_rate": 8.740375612482541e-05,
+      "loss": 0.0874,
+      "step": 193
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.11611688137054443,
+      "learning_rate": 8.726178503383513e-05,
+      "loss": 0.0858,
+      "step": 194
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.10915149748325348,
+      "learning_rate": 8.711913504080534e-05,
+      "loss": 0.0814,
+      "step": 195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.11281640082597733,
+      "learning_rate": 8.697580874478952e-05,
+      "loss": 0.0827,
+      "step": 196
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08252295106649399,
+      "learning_rate": 8.683180875716322e-05,
+      "loss": 0.0791,
+      "step": 197
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09433229267597198,
+      "learning_rate": 8.668713770157652e-05,
+      "loss": 0.0769,
+      "step": 198
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.11094626039266586,
+      "learning_rate": 8.654179821390621e-05,
+      "loss": 0.0786,
+      "step": 199
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.11672339588403702,
+      "learning_rate": 8.639579294220779e-05,
+      "loss": 0.0744,
+      "step": 200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0986831933259964,
+      "learning_rate": 8.624912454666723e-05,
+      "loss": 0.0759,
+      "step": 201
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13550534844398499,
+      "learning_rate": 8.610179569955238e-05,
+      "loss": 0.0746,
+      "step": 202
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0924728587269783,
+      "learning_rate": 8.595380908516454e-05,
+      "loss": 0.0794,
+      "step": 203
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.09090446680784225,
+      "learning_rate": 8.580516739978926e-05,
+      "loss": 0.0822,
+      "step": 204
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1201312392950058,
+      "learning_rate": 8.56558733516474e-05,
+      "loss": 0.0856,
+      "step": 205
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.10289566963911057,
+      "learning_rate": 8.550592966084573e-05,
+      "loss": 0.0798,
+      "step": 206
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1051202043890953,
+      "learning_rate": 8.535533905932738e-05,
+      "loss": 0.0688,
+      "step": 207
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.09235143661499023,
+      "learning_rate": 8.520410429082206e-05,
+      "loss": 0.0793,
+      "step": 208
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.11254550516605377,
+      "learning_rate": 8.505222811079608e-05,
+      "loss": 0.079,
+      "step": 209
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.11825918406248093,
+      "learning_rate": 8.489971328640207e-05,
+      "loss": 0.0744,
+      "step": 210
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13841897249221802,
+      "learning_rate": 8.474656259642873e-05,
+      "loss": 0.089,
+      "step": 211
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.109561987221241,
+      "learning_rate": 8.459277883125004e-05,
+      "loss": 0.0777,
+      "step": 212
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16192643344402313,
+      "learning_rate": 8.443836479277449e-05,
+      "loss": 0.0894,
+      "step": 213
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.10433942824602127,
+      "learning_rate": 8.428332329439399e-05,
+      "loss": 0.079,
+      "step": 214
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.10524985939264297,
+      "learning_rate": 8.412765716093272e-05,
+      "loss": 0.0675,
+      "step": 215
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08769605308771133,
+      "learning_rate": 8.397136922859548e-05,
+      "loss": 0.0823,
+      "step": 216
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.10349096357822418,
+      "learning_rate": 8.381446234491618e-05,
+      "loss": 0.0847,
+      "step": 217
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.11507713049650192,
+      "learning_rate": 8.365693936870594e-05,
+      "loss": 0.0777,
+      "step": 218
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.120513916015625,
+      "learning_rate": 8.349880317000082e-05,
+      "loss": 0.0887,
+      "step": 219
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.09667017310857773,
+      "learning_rate": 8.334005663000982e-05,
+      "loss": 0.0702,
+      "step": 220
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.09578403830528259,
+      "learning_rate": 8.318070264106213e-05,
+      "loss": 0.0796,
+      "step": 221
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08959231525659561,
+      "learning_rate": 8.302074410655455e-05,
+      "loss": 0.0857,
+      "step": 222
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.10325565189123154,
+      "learning_rate": 8.286018394089863e-05,
+      "loss": 0.0794,
+      "step": 223
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1005704402923584,
+      "learning_rate": 8.269902506946746e-05,
+      "loss": 0.0821,
+      "step": 224
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.09767957776784897,
+      "learning_rate": 8.253727042854245e-05,
+      "loss": 0.0776,
+      "step": 225
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.11237328499555588,
+      "learning_rate": 8.237492296525981e-05,
+      "loss": 0.0735,
+      "step": 226
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13158953189849854,
+      "learning_rate": 8.221198563755682e-05,
+      "loss": 0.0896,
+      "step": 227
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.10914699733257294,
+      "learning_rate": 8.204846141411801e-05,
+      "loss": 0.0753,
+      "step": 228
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1349821537733078,
+      "learning_rate": 8.1884353274321e-05,
+      "loss": 0.0806,
+      "step": 229
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.09405103325843811,
+      "learning_rate": 8.171966420818228e-05,
+      "loss": 0.083,
+      "step": 230
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.09107044339179993,
+      "learning_rate": 8.155439721630264e-05,
+      "loss": 0.076,
+      "step": 231
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1389627307653427,
+      "learning_rate": 8.138855530981262e-05,
+      "loss": 0.078,
+      "step": 232
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12448082119226456,
+      "learning_rate": 8.122214151031753e-05,
+      "loss": 0.0859,
+      "step": 233
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08903878182172775,
+      "learning_rate": 8.10551588498425e-05,
+      "loss": 0.0805,
+      "step": 234
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12410246580839157,
+      "learning_rate": 8.088761037077718e-05,
+      "loss": 0.0823,
+      "step": 235
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.09901540726423264,
+      "learning_rate": 8.071949912582029e-05,
+      "loss": 0.0835,
+      "step": 236
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.10345863550901413,
+      "learning_rate": 8.055082817792403e-05,
+      "loss": 0.0723,
+      "step": 237
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11597669124603271,
+      "learning_rate": 8.038160060023834e-05,
+      "loss": 0.0772,
+      "step": 238
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12257598340511322,
+      "learning_rate": 8.021181947605473e-05,
+      "loss": 0.079,
+      "step": 239
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12233386933803558,
+      "learning_rate": 8.00414878987503e-05,
+      "loss": 0.084,
+      "step": 240
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.09252316504716873,
+      "learning_rate": 7.987060897173128e-05,
+      "loss": 0.0812,
+      "step": 241
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1188056468963623,
+      "learning_rate": 7.969918580837648e-05,
+      "loss": 0.0797,
+      "step": 242
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.10174442082643509,
+      "learning_rate": 7.952722153198054e-05,
+      "loss": 0.0736,
+      "step": 243
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.10943493247032166,
+      "learning_rate": 7.935471927569718e-05,
+      "loss": 0.0831,
+      "step": 244
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.10414671897888184,
+      "learning_rate": 7.918168218248188e-05,
+      "loss": 0.0842,
+      "step": 245
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.11411801725625992,
+      "learning_rate": 7.900811340503483e-05,
+      "loss": 0.0782,
+      "step": 246
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.10560184717178345,
+      "learning_rate": 7.883401610574336e-05,
+      "loss": 0.0774,
+      "step": 247
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13932396471500397,
+      "learning_rate": 7.865939345662436e-05,
+      "loss": 0.0871,
+      "step": 248
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.10407272726297379,
+      "learning_rate": 7.848424863926649e-05,
+      "loss": 0.0754,
+      "step": 249
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.09210895001888275,
+      "learning_rate": 7.830858484477218e-05,
+      "loss": 0.0744,
+      "step": 250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1021440252661705,
+      "learning_rate": 7.813240527369959e-05,
+      "loss": 0.0852,
+      "step": 251
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11797258257865906,
+      "learning_rate": 7.795571313600411e-05,
+      "loss": 0.076,
+      "step": 252
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1269652247428894,
+      "learning_rate": 7.777851165098012e-05,
+      "loss": 0.0838,
+      "step": 253
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1156778559088707,
+      "learning_rate": 7.760080404720209e-05,
+      "loss": 0.081,
+      "step": 254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.10935825854539871,
+      "learning_rate": 7.742259356246593e-05,
+      "loss": 0.0821,
+      "step": 255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.10187061131000519,
+      "learning_rate": 7.724388344372995e-05,
+      "loss": 0.0782,
+      "step": 256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.10703828930854797,
+      "learning_rate": 7.706467694705561e-05,
+      "loss": 0.0779,
+      "step": 257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.09682856500148773,
+      "learning_rate": 7.688497733754835e-05,
+      "loss": 0.0798,
+      "step": 258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.098836250603199,
+      "learning_rate": 7.670478788929802e-05,
+      "loss": 0.0852,
+      "step": 259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08768448978662491,
+      "learning_rate": 7.652411188531917e-05,
+      "loss": 0.0725,
+      "step": 260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.10361611843109131,
+      "learning_rate": 7.634295261749135e-05,
+      "loss": 0.0749,
+      "step": 261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08791149407625198,
+      "learning_rate": 7.616131338649907e-05,
+      "loss": 0.075,
+      "step": 262
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08269870281219482,
+      "learning_rate": 7.597919750177168e-05,
+      "loss": 0.0706,
+      "step": 263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.09004026651382446,
+      "learning_rate": 7.579660828142301e-05,
+      "loss": 0.083,
+      "step": 264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12535926699638367,
+      "learning_rate": 7.561354905219102e-05,
+      "loss": 0.0891,
+      "step": 265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.10626240819692612,
+      "learning_rate": 7.543002314937712e-05,
+      "loss": 0.0793,
+      "step": 266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.10296738147735596,
+      "learning_rate": 7.524603391678541e-05,
+      "loss": 0.0873,
+      "step": 267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12124089896678925,
+      "learning_rate": 7.506158470666174e-05,
+      "loss": 0.0764,
+      "step": 268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.09938234090805054,
+      "learning_rate": 7.487667887963273e-05,
+      "loss": 0.066,
+      "step": 269
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1229834258556366,
+      "learning_rate": 7.469131980464439e-05,
+      "loss": 0.0764,
+      "step": 270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12650369107723236,
+      "learning_rate": 7.450551085890087e-05,
+      "loss": 0.0837,
+      "step": 271
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12816506624221802,
+      "learning_rate": 7.431925542780281e-05,
+      "loss": 0.0804,
+      "step": 272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.11941610276699066,
+      "learning_rate": 7.413255690488577e-05,
+      "loss": 0.087,
+      "step": 273
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.11334256082773209,
+      "learning_rate": 7.394541869175834e-05,
+      "loss": 0.0742,
+      "step": 274
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.14086034893989563,
+      "learning_rate": 7.375784419804019e-05,
+      "loss": 0.0859,
+      "step": 275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.09905703365802765,
+      "learning_rate": 7.35698368412999e-05,
+      "loss": 0.0745,
+      "step": 276
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.10592259466648102,
+      "learning_rate": 7.338140004699272e-05,
+      "loss": 0.0762,
+      "step": 277
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08352551609277725,
+      "learning_rate": 7.319253724839821e-05,
+      "loss": 0.0711,
+      "step": 278
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.09426555037498474,
+      "learning_rate": 7.300325188655761e-05,
+      "loss": 0.0623,
+      "step": 279
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1022004559636116,
+      "learning_rate": 7.281354741021118e-05,
+      "loss": 0.0815,
+      "step": 280
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11823340505361557,
+      "learning_rate": 7.262342727573535e-05,
+      "loss": 0.0753,
+      "step": 281
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.10613666474819183,
+      "learning_rate": 7.24328949470798e-05,
+      "loss": 0.0805,
+      "step": 282
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08761774003505707,
+      "learning_rate": 7.224195389570422e-05,
+      "loss": 0.0763,
+      "step": 283
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.09059295803308487,
+      "learning_rate": 7.205060760051522e-05,
+      "loss": 0.0736,
+      "step": 284
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.09126615524291992,
+      "learning_rate": 7.185885954780282e-05,
+      "loss": 0.0745,
+      "step": 285
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.09419333934783936,
+      "learning_rate": 7.166671323117703e-05,
+      "loss": 0.0706,
+      "step": 286
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10592428594827652,
+      "learning_rate": 7.14741721515041e-05,
+      "loss": 0.0804,
+      "step": 287
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10718122124671936,
+      "learning_rate": 7.128123981684279e-05,
+      "loss": 0.076,
+      "step": 288
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.09602747857570648,
+      "learning_rate": 7.108791974238047e-05,
+      "loss": 0.0722,
+      "step": 289
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10073014348745346,
+      "learning_rate": 7.0894215450369e-05,
+      "loss": 0.0704,
+      "step": 290
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10062811523675919,
+      "learning_rate": 7.070013047006068e-05,
+      "loss": 0.074,
+      "step": 291
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10420974344015121,
+      "learning_rate": 7.05056683376438e-05,
+      "loss": 0.0811,
+      "step": 292
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10123053938150406,
+      "learning_rate": 7.031083259617832e-05,
+      "loss": 0.0737,
+      "step": 293
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1051117554306984,
+      "learning_rate": 7.011562679553126e-05,
+      "loss": 0.0795,
+      "step": 294
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10867037624120712,
+      "learning_rate": 6.992005449231208e-05,
+      "loss": 0.0779,
+      "step": 295
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10807811468839645,
+      "learning_rate": 6.972411924980778e-05,
+      "loss": 0.0796,
+      "step": 296
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11884191632270813,
+      "learning_rate": 6.952782463791813e-05,
+      "loss": 0.0843,
+      "step": 297
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1229209154844284,
+      "learning_rate": 6.933117423309042e-05,
+      "loss": 0.0839,
+      "step": 298
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11353997886180878,
+      "learning_rate": 6.91341716182545e-05,
+      "loss": 0.0703,
+      "step": 299
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.10744056850671768,
+      "learning_rate": 6.893682038275738e-05,
+      "loss": 0.0769,
+      "step": 300
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1081230416893959,
+      "learning_rate": 6.873912412229788e-05,
+      "loss": 0.0848,
+      "step": 301
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11869249492883682,
+      "learning_rate": 6.854108643886113e-05,
+      "loss": 0.0857,
+      "step": 302
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.10422538965940475,
+      "learning_rate": 6.834271094065283e-05,
+      "loss": 0.0773,
+      "step": 303
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.09786193817853928,
+      "learning_rate": 6.814400124203369e-05,
+      "loss": 0.0701,
+      "step": 304
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.10703827440738678,
+      "learning_rate": 6.794496096345341e-05,
+      "loss": 0.0838,
+      "step": 305
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.10062588006258011,
+      "learning_rate": 6.774559373138484e-05,
+      "loss": 0.0772,
+      "step": 306
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.09322860091924667,
+      "learning_rate": 6.754590317825785e-05,
+      "loss": 0.0706,
+      "step": 307
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11063066124916077,
+      "learning_rate": 6.734589294239311e-05,
+      "loss": 0.0791,
+      "step": 308
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10088267177343369,
+      "learning_rate": 6.71455666679359e-05,
+      "loss": 0.0751,
+      "step": 309
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10203511267900467,
+      "learning_rate": 6.69449280047896e-05,
+      "loss": 0.0763,
+      "step": 310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.09792084246873856,
+      "learning_rate": 6.674398060854931e-05,
+      "loss": 0.0785,
+      "step": 311
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.09468895196914673,
+      "learning_rate": 6.654272814043514e-05,
+      "loss": 0.0788,
+      "step": 312
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.09165459871292114,
+      "learning_rate": 6.634117426722556e-05,
+      "loss": 0.0767,
+      "step": 313
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10596252977848053,
+      "learning_rate": 6.613932266119056e-05,
+      "loss": 0.0796,
+      "step": 314
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1063828319311142,
+      "learning_rate": 6.59371770000248e-05,
+      "loss": 0.0755,
+      "step": 315
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11581625789403915,
+      "learning_rate": 6.573474096678052e-05,
+      "loss": 0.0732,
+      "step": 316
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11985045671463013,
+      "learning_rate": 6.553201824980054e-05,
+      "loss": 0.0781,
+      "step": 317
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12919990718364716,
+      "learning_rate": 6.532901254265093e-05,
+      "loss": 0.0855,
+      "step": 318
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.14869804680347443,
+      "learning_rate": 6.51257275440538e-05,
+      "loss": 0.0772,
+      "step": 319
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10909485816955566,
+      "learning_rate": 6.492216695781992e-05,
+      "loss": 0.0676,
+      "step": 320
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.14061373472213745,
+      "learning_rate": 6.47183344927812e-05,
+      "loss": 0.0804,
+      "step": 321
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12443160265684128,
+      "learning_rate": 6.451423386272312e-05,
+      "loss": 0.0708,
+      "step": 322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1312866508960724,
+      "learning_rate": 6.430986878631707e-05,
+      "loss": 0.0776,
+      "step": 323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11382008343935013,
+      "learning_rate": 6.41052429870526e-05,
+      "loss": 0.0709,
+      "step": 324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11452754586935043,
+      "learning_rate": 6.390036019316956e-05,
+      "loss": 0.0797,
+      "step": 325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09669509530067444,
+      "learning_rate": 6.369522413759022e-05,
+      "loss": 0.0661,
+      "step": 326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09662206470966339,
+      "learning_rate": 6.348983855785121e-05,
+      "loss": 0.0709,
+      "step": 327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10699667036533356,
+      "learning_rate": 6.328420719603546e-05,
+      "loss": 0.0809,
+      "step": 328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1173032596707344,
+      "learning_rate": 6.307833379870394e-05,
+      "loss": 0.0792,
+      "step": 329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1062864363193512,
+      "learning_rate": 6.287222211682751e-05,
+      "loss": 0.0801,
+      "step": 330
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10499227792024612,
+      "learning_rate": 6.266587590571852e-05,
+      "loss": 0.076,
+      "step": 331
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08642653375864029,
+      "learning_rate": 6.245929892496239e-05,
+      "loss": 0.0778,
+      "step": 332
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09800557792186737,
+      "learning_rate": 6.225249493834911e-05,
+      "loss": 0.0802,
+      "step": 333
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11565027385950089,
+      "learning_rate": 6.204546771380462e-05,
+      "loss": 0.0788,
+      "step": 334
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09418751299381256,
+      "learning_rate": 6.183822102332234e-05,
+      "loss": 0.0722,
+      "step": 335
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11058859527111053,
+      "learning_rate": 6.16307586428942e-05,
+      "loss": 0.0788,
+      "step": 336
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10711957514286041,
+      "learning_rate": 6.142308435244195e-05,
+      "loss": 0.074,
+      "step": 337
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1091398298740387,
+      "learning_rate": 6.121520193574841e-05,
+      "loss": 0.0718,
+      "step": 338
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.11050156503915787,
+      "learning_rate": 6.1007115180388285e-05,
+      "loss": 0.0827,
+      "step": 339
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10380526632070541,
+      "learning_rate": 6.079882787765938e-05,
+      "loss": 0.0808,
+      "step": 340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10426554083824158,
+      "learning_rate": 6.059034382251338e-05,
+      "loss": 0.0693,
+      "step": 341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1098361536860466,
+      "learning_rate": 6.0381666813486795e-05,
+      "loss": 0.0741,
+      "step": 342
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1143401488661766,
+      "learning_rate": 6.01728006526317e-05,
+      "loss": 0.0738,
+      "step": 343
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.104662224650383,
+      "learning_rate": 5.9963749145446444e-05,
+      "loss": 0.0764,
+      "step": 344
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10660408437252045,
+      "learning_rate": 5.9754516100806423e-05,
+      "loss": 0.0762,
+      "step": 345
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1031242236495018,
+      "learning_rate": 5.9545105330894526e-05,
+      "loss": 0.0793,
+      "step": 346
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11632685363292694,
+      "learning_rate": 5.9335520651131814e-05,
+      "loss": 0.0683,
+      "step": 347
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11623501777648926,
+      "learning_rate": 5.912576588010795e-05,
+      "loss": 0.0699,
+      "step": 348
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12420337647199631,
+      "learning_rate": 5.891584483951157e-05,
+      "loss": 0.0801,
+      "step": 349
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11674680560827255,
+      "learning_rate": 5.870576135406077e-05,
+      "loss": 0.0755,
+      "step": 350
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12480311095714569,
+      "learning_rate": 5.849551925143334e-05,
+      "loss": 0.0642,
+      "step": 351
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12658356130123138,
+      "learning_rate": 5.828512236219701e-05,
+      "loss": 0.0763,
+      "step": 352
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10800620913505554,
+      "learning_rate": 5.807457451973975e-05,
+      "loss": 0.0688,
+      "step": 353
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.12109076231718063,
+      "learning_rate": 5.78638795601998e-05,
+      "loss": 0.0744,
+      "step": 354
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1361728012561798,
+      "learning_rate": 5.7653041322395895e-05,
+      "loss": 0.0715,
+      "step": 355
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11848083138465881,
+      "learning_rate": 5.744206364775724e-05,
+      "loss": 0.0803,
+      "step": 356
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.12507352232933044,
+      "learning_rate": 5.723095038025356e-05,
+      "loss": 0.072,
+      "step": 357
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1153586283326149,
+      "learning_rate": 5.701970536632507e-05,
+      "loss": 0.071,
+      "step": 358
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.13595929741859436,
+      "learning_rate": 5.680833245481234e-05,
+      "loss": 0.0762,
+      "step": 359
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1186838150024414,
+      "learning_rate": 5.659683549688623e-05,
+      "loss": 0.0746,
+      "step": 360
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.14524149894714355,
+      "learning_rate": 5.638521834597774e-05,
+      "loss": 0.0819,
+      "step": 361
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10339554399251938,
+      "learning_rate": 5.617348485770767e-05,
+      "loss": 0.0745,
+      "step": 362
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.12478932738304138,
+      "learning_rate": 5.596163888981656e-05,
+      "loss": 0.0847,
+      "step": 363
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0925443097949028,
+      "learning_rate": 5.574968430209423e-05,
+      "loss": 0.0697,
+      "step": 364
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11409781873226166,
+      "learning_rate": 5.553762495630956e-05,
+      "loss": 0.0693,
+      "step": 365
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.08425464481115341,
+      "learning_rate": 5.532546471614012e-05,
+      "loss": 0.076,
+      "step": 366
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.09573796391487122,
+      "learning_rate": 5.511320744710171e-05,
+      "loss": 0.0663,
+      "step": 367
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10230566561222076,
+      "learning_rate": 5.490085701647805e-05,
+      "loss": 0.0789,
+      "step": 368
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10512962937355042,
+      "learning_rate": 5.468841729325014e-05,
+      "loss": 0.0795,
+      "step": 369
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09100647270679474,
+      "learning_rate": 5.447589214802594e-05,
+      "loss": 0.0724,
+      "step": 370
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08053728193044662,
+      "learning_rate": 5.4263285452969806e-05,
+      "loss": 0.0759,
+      "step": 371
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08965244889259338,
+      "learning_rate": 5.4050601081731845e-05,
+      "loss": 0.0757,
+      "step": 372
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08649550378322601,
+      "learning_rate": 5.383784290937747e-05,
+      "loss": 0.0725,
+      "step": 373
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09037112444639206,
+      "learning_rate": 5.36250148123167e-05,
+      "loss": 0.075,
+      "step": 374
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10361938178539276,
+      "learning_rate": 5.341212066823355e-05,
+      "loss": 0.0747,
+      "step": 375
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09907560050487518,
+      "learning_rate": 5.319916435601546e-05,
+      "loss": 0.0739,
+      "step": 376
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10159469395875931,
+      "learning_rate": 5.298614975568249e-05,
+      "loss": 0.0695,
+      "step": 377
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.11597546190023422,
+      "learning_rate": 5.277308074831677e-05,
+      "loss": 0.0702,
+      "step": 378
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.11498542129993439,
+      "learning_rate": 5.255996121599167e-05,
+      "loss": 0.0749,
+      "step": 379
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.11171206086874008,
+      "learning_rate": 5.2346795041701077e-05,
+      "loss": 0.0732,
+      "step": 380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.11158174276351929,
+      "learning_rate": 5.213358610928878e-05,
+      "loss": 0.0811,
+      "step": 381
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.11858642846345901,
+      "learning_rate": 5.192033830337754e-05,
+      "loss": 0.0758,
+      "step": 382
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.12075002491474152,
+      "learning_rate": 5.1707055509298396e-05,
+      "loss": 0.0752,
+      "step": 383
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08938009291887283,
+      "learning_rate": 5.1493741613019906e-05,
+      "loss": 0.0653,
+      "step": 384
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09382277727127075,
+      "learning_rate": 5.128040050107724e-05,
+      "loss": 0.076,
+      "step": 385
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.12058842927217484,
+      "learning_rate": 5.1067036060501486e-05,
+      "loss": 0.0753,
+      "step": 386
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10002080351114273,
+      "learning_rate": 5.0853652178748746e-05,
+      "loss": 0.0803,
+      "step": 387
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10245774686336517,
+      "learning_rate": 5.0640252743629326e-05,
+      "loss": 0.0729,
+      "step": 388
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.11456061154603958,
+      "learning_rate": 5.042684164323698e-05,
+      "loss": 0.0675,
+      "step": 389
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.11435601860284805,
+      "learning_rate": 5.021342276587787e-05,
+      "loss": 0.0727,
+      "step": 390
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10582521557807922,
+      "learning_rate": 5e-05,
+      "loss": 0.0694,
+      "step": 391
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10909827798604965,
+      "learning_rate": 4.978657723412212e-05,
+      "loss": 0.0724,
+      "step": 392
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.12171333283185959,
+      "learning_rate": 4.957315835676305e-05,
+      "loss": 0.0817,
+      "step": 393
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10484538972377777,
+      "learning_rate": 4.935974725637068e-05,
+      "loss": 0.072,
+      "step": 394
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.13289706408977509,
+      "learning_rate": 4.9146347821251266e-05,
+      "loss": 0.0682,
+      "step": 395
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.11009448766708374,
+      "learning_rate": 4.893296393949854e-05,
+      "loss": 0.0785,
+      "step": 396
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.13469281792640686,
+      "learning_rate": 4.871959949892277e-05,
+      "loss": 0.0715,
+      "step": 397
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.11217113584280014,
+      "learning_rate": 4.8506258386980106e-05,
+      "loss": 0.0668,
+      "step": 398
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.12297201156616211,
+      "learning_rate": 4.829294449070161e-05,
+      "loss": 0.0669,
+      "step": 399
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.12945881485939026,
+      "learning_rate": 4.807966169662249e-05,
+      "loss": 0.0813,
+      "step": 400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1016998440027237,
+      "learning_rate": 4.786641389071123e-05,
+      "loss": 0.072,
+      "step": 401
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.13026472926139832,
+      "learning_rate": 4.765320495829893e-05,
+      "loss": 0.0798,
+      "step": 402
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.12311934679746628,
+      "learning_rate": 4.744003878400835e-05,
+      "loss": 0.0777,
+      "step": 403
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.12245490401983261,
+      "learning_rate": 4.7226919251683235e-05,
+      "loss": 0.0721,
+      "step": 404
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10729845613241196,
+      "learning_rate": 4.70138502443175e-05,
+      "loss": 0.0658,
+      "step": 405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10610393434762955,
+      "learning_rate": 4.6800835643984564e-05,
+      "loss": 0.0688,
+      "step": 406
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10243082046508789,
+      "learning_rate": 4.658787933176646e-05,
+      "loss": 0.0767,
+      "step": 407
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.12154947221279144,
+      "learning_rate": 4.637498518768331e-05,
+      "loss": 0.0792,
+      "step": 408
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.12794026732444763,
+      "learning_rate": 4.616215709062253e-05,
+      "loss": 0.0787,
+      "step": 409
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08822939544916153,
+      "learning_rate": 4.5949398918268166e-05,
+      "loss": 0.0697,
+      "step": 410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09519528597593307,
+      "learning_rate": 4.57367145470302e-05,
+      "loss": 0.0668,
+      "step": 411
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10057291388511658,
+      "learning_rate": 4.5524107851974056e-05,
+      "loss": 0.0707,
+      "step": 412
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09450813382863998,
+      "learning_rate": 4.531158270674989e-05,
+      "loss": 0.0809,
+      "step": 413
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.08685977756977081,
+      "learning_rate": 4.509914298352197e-05,
+      "loss": 0.0754,
+      "step": 414
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09042025357484818,
+      "learning_rate": 4.4886792552898286e-05,
+      "loss": 0.0689,
+      "step": 415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10866349190473557,
+      "learning_rate": 4.46745352838599e-05,
+      "loss": 0.0746,
+      "step": 416
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09778650850057602,
+      "learning_rate": 4.446237504369045e-05,
+      "loss": 0.0698,
+      "step": 417
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09621346741914749,
+      "learning_rate": 4.4250315697905777e-05,
+      "loss": 0.0726,
+      "step": 418
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09709804505109787,
+      "learning_rate": 4.403836111018346e-05,
+      "loss": 0.0663,
+      "step": 419
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10436049848794937,
+      "learning_rate": 4.382651514229234e-05,
+      "loss": 0.075,
+      "step": 420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.11697012186050415,
+      "learning_rate": 4.3614781654022276e-05,
+      "loss": 0.0724,
+      "step": 421
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10285527259111404,
+      "learning_rate": 4.340316450311376e-05,
+      "loss": 0.0714,
+      "step": 422
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1052049994468689,
+      "learning_rate": 4.319166754518768e-05,
+      "loss": 0.0709,
+      "step": 423
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10365312546491623,
+      "learning_rate": 4.2980294633674946e-05,
+      "loss": 0.083,
+      "step": 424
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.11218774318695068,
+      "learning_rate": 4.2769049619746446e-05,
+      "loss": 0.069,
+      "step": 425
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.11706428229808807,
+      "learning_rate": 4.2557936352242784e-05,
+      "loss": 0.0758,
+      "step": 426
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1439225971698761,
+      "learning_rate": 4.234695867760412e-05,
+      "loss": 0.0836,
+      "step": 427
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.12518368661403656,
+      "learning_rate": 4.21361204398002e-05,
+      "loss": 0.0754,
+      "step": 428
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1182391494512558,
+      "learning_rate": 4.1925425480260266e-05,
+      "loss": 0.0768,
+      "step": 429
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10238274186849594,
+      "learning_rate": 4.1714877637803e-05,
+      "loss": 0.0655,
+      "step": 430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09843271970748901,
+      "learning_rate": 4.150448074856667e-05,
+      "loss": 0.0665,
+      "step": 431
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10684695839881897,
+      "learning_rate": 4.129423864593923e-05,
+      "loss": 0.0771,
+      "step": 432
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.11525686830282211,
+      "learning_rate": 4.108415516048845e-05,
+      "loss": 0.0716,
+      "step": 433
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.11386000365018845,
+      "learning_rate": 4.087423411989206e-05,
+      "loss": 0.0698,
+      "step": 434
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.12380651384592056,
+      "learning_rate": 4.066447934886819e-05,
+      "loss": 0.0723,
+      "step": 435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1446354240179062,
+      "learning_rate": 4.045489466910549e-05,
+      "loss": 0.0731,
+      "step": 436
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10570847988128662,
+      "learning_rate": 4.0245483899193595e-05,
+      "loss": 0.0714,
+      "step": 437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.12854741513729095,
+      "learning_rate": 4.0036250854553554e-05,
+      "loss": 0.0702,
+      "step": 438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.12250388413667679,
+      "learning_rate": 3.982719934736832e-05,
+      "loss": 0.0735,
+      "step": 439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.13413843512535095,
+      "learning_rate": 3.9618333186513216e-05,
+      "loss": 0.0647,
+      "step": 440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10849061608314514,
+      "learning_rate": 3.940965617748662e-05,
+      "loss": 0.0603,
+      "step": 441
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.12001483887434006,
+      "learning_rate": 3.9201172122340645e-05,
+      "loss": 0.0749,
+      "step": 442
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.11212868988513947,
+      "learning_rate": 3.899288481961173e-05,
+      "loss": 0.0718,
+      "step": 443
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1401149332523346,
+      "learning_rate": 3.87847980642516e-05,
+      "loss": 0.0736,
+      "step": 444
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.12652656435966492,
+      "learning_rate": 3.857691564755805e-05,
+      "loss": 0.063,
+      "step": 445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.11927732825279236,
+      "learning_rate": 3.8369241357105835e-05,
+      "loss": 0.0686,
+      "step": 446
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1603834480047226,
+      "learning_rate": 3.8161778976677666e-05,
+      "loss": 0.0698,
+      "step": 447
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.13165497779846191,
+      "learning_rate": 3.795453228619537e-05,
+      "loss": 0.0706,
+      "step": 448
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1464444249868393,
+      "learning_rate": 3.774750506165092e-05,
+      "loss": 0.0828,
+      "step": 449
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1456499546766281,
+      "learning_rate": 3.7540701075037626e-05,
+      "loss": 0.0767,
+      "step": 450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.10524341464042664,
+      "learning_rate": 3.733412409428148e-05,
+      "loss": 0.0692,
+      "step": 451
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.12726537883281708,
+      "learning_rate": 3.712777788317251e-05,
+      "loss": 0.0687,
+      "step": 452
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.12855078279972076,
+      "learning_rate": 3.692166620129608e-05,
+      "loss": 0.0725,
+      "step": 453
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.10352146625518799,
+      "learning_rate": 3.671579280396455e-05,
+      "loss": 0.0694,
+      "step": 454
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.10154904425144196,
+      "learning_rate": 3.651016144214878e-05,
+      "loss": 0.0682,
+      "step": 455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09369635581970215,
+      "learning_rate": 3.63047758624098e-05,
+      "loss": 0.0659,
+      "step": 456
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1108255609869957,
+      "learning_rate": 3.6099639806830455e-05,
+      "loss": 0.0672,
+      "step": 457
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.11742108315229416,
+      "learning_rate": 3.589475701294741e-05,
+      "loss": 0.0607,
+      "step": 458
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1176430806517601,
+      "learning_rate": 3.5690131213682944e-05,
+      "loss": 0.0854,
+      "step": 459
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1096600741147995,
+      "learning_rate": 3.5485766137276894e-05,
+      "loss": 0.0711,
+      "step": 460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.12446652352809906,
+      "learning_rate": 3.52816655072188e-05,
+      "loss": 0.0744,
+      "step": 461
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.11120939254760742,
+      "learning_rate": 3.5077833042180094e-05,
+      "loss": 0.0739,
+      "step": 462
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.14443209767341614,
+      "learning_rate": 3.487427245594622e-05,
+      "loss": 0.0698,
+      "step": 463
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.13654518127441406,
+      "learning_rate": 3.467098745734909e-05,
+      "loss": 0.0658,
+      "step": 464
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.10497821122407913,
+      "learning_rate": 3.446798175019948e-05,
+      "loss": 0.0683,
+      "step": 465
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.11521188169717789,
+      "learning_rate": 3.426525903321949e-05,
+      "loss": 0.0602,
+      "step": 466
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1282206028699875,
+      "learning_rate": 3.406282299997521e-05,
+      "loss": 0.0694,
+      "step": 467
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.14426809549331665,
+      "learning_rate": 3.386067733880944e-05,
+      "loss": 0.0652,
+      "step": 468
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17491450905799866,
+      "learning_rate": 3.3658825732774456e-05,
+      "loss": 0.0718,
+      "step": 469
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.12897461652755737,
+      "learning_rate": 3.3457271859564864e-05,
+      "loss": 0.0716,
+      "step": 470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.12860888242721558,
+      "learning_rate": 3.325601939145069e-05,
+      "loss": 0.0792,
+      "step": 471
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.14141874015331268,
+      "learning_rate": 3.305507199521041e-05,
+      "loss": 0.0739,
+      "step": 472
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.13235457241535187,
+      "learning_rate": 3.2854433332064114e-05,
+      "loss": 0.0735,
+      "step": 473
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.12121675908565521,
+      "learning_rate": 3.265410705760689e-05,
+      "loss": 0.0639,
+      "step": 474
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.12805382907390594,
+      "learning_rate": 3.245409682174217e-05,
+      "loss": 0.0741,
+      "step": 475
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.11965460330247879,
+      "learning_rate": 3.225440626861517e-05,
+      "loss": 0.0717,
+      "step": 476
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.12619951367378235,
+      "learning_rate": 3.2055039036546596e-05,
+      "loss": 0.0646,
+      "step": 477
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.11961790174245834,
+      "learning_rate": 3.1855998757966324e-05,
+      "loss": 0.0686,
+      "step": 478
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1319866180419922,
+      "learning_rate": 3.165728905934718e-05,
+      "loss": 0.065,
+      "step": 479
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.12313670665025711,
+      "learning_rate": 3.1458913561138884e-05,
+      "loss": 0.0676,
+      "step": 480
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.13332146406173706,
+      "learning_rate": 3.126087587770211e-05,
+      "loss": 0.0716,
+      "step": 481
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.11296407878398895,
+      "learning_rate": 3.106317961724263e-05,
+      "loss": 0.0658,
+      "step": 482
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.15619413554668427,
+      "learning_rate": 3.086582838174551e-05,
+      "loss": 0.0726,
+      "step": 483
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.12882710993289948,
+      "learning_rate": 3.0668825766909595e-05,
+      "loss": 0.0691,
+      "step": 484
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.12875738739967346,
+      "learning_rate": 3.0472175362081897e-05,
+      "loss": 0.0658,
+      "step": 485
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.12345566600561142,
+      "learning_rate": 3.0275880750192227e-05,
+      "loss": 0.0678,
+      "step": 486
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1387387365102768,
+      "learning_rate": 3.007994550768793e-05,
+      "loss": 0.0734,
+      "step": 487
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.14710651338100433,
+      "learning_rate": 2.9884373204468753e-05,
+      "loss": 0.0784,
+      "step": 488
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.11002452671527863,
+      "learning_rate": 2.9689167403821694e-05,
+      "loss": 0.0646,
+      "step": 489
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.12409970909357071,
+      "learning_rate": 2.94943316623562e-05,
+      "loss": 0.0677,
+      "step": 490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1445932388305664,
+      "learning_rate": 2.929986952993933e-05,
+      "loss": 0.0614,
+      "step": 491
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.11631357669830322,
+      "learning_rate": 2.910578454963101e-05,
+      "loss": 0.0692,
+      "step": 492
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1650647670030594,
+      "learning_rate": 2.8912080257619544e-05,
+      "loss": 0.0692,
+      "step": 493
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.14293146133422852,
+      "learning_rate": 2.871876018315721e-05,
+      "loss": 0.068,
+      "step": 494
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.12101910263299942,
+      "learning_rate": 2.8525827848495913e-05,
+      "loss": 0.0622,
+      "step": 495
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.12193334102630615,
+      "learning_rate": 2.833328676882297e-05,
+      "loss": 0.0691,
+      "step": 496
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.11649112403392792,
+      "learning_rate": 2.814114045219717e-05,
+      "loss": 0.0751,
+      "step": 497
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.13954710960388184,
+      "learning_rate": 2.7949392399484792e-05,
+      "loss": 0.0636,
+      "step": 498
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.137295663356781,
+      "learning_rate": 2.7758046104295797e-05,
+      "loss": 0.0785,
+      "step": 499
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.13648007810115814,
+      "learning_rate": 2.7567105052920216e-05,
+      "loss": 0.0732,
+      "step": 500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1289539486169815,
+      "learning_rate": 2.7376572724264644e-05,
+      "loss": 0.0754,
+      "step": 501
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1078784316778183,
+      "learning_rate": 2.7186452589788835e-05,
+      "loss": 0.0641,
+      "step": 502
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.13168664276599884,
+      "learning_rate": 2.6996748113442394e-05,
+      "loss": 0.0677,
+      "step": 503
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.12153467535972595,
+      "learning_rate": 2.680746275160179e-05,
+      "loss": 0.0708,
+      "step": 504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.13471120595932007,
+      "learning_rate": 2.6618599953007296e-05,
+      "loss": 0.0766,
+      "step": 505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.14139395952224731,
+      "learning_rate": 2.6430163158700115e-05,
+      "loss": 0.0715,
+      "step": 506
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1328204870223999,
+      "learning_rate": 2.624215580195981e-05,
+      "loss": 0.0723,
+      "step": 507
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.11794441193342209,
+      "learning_rate": 2.605458130824166e-05,
+      "loss": 0.0671,
+      "step": 508
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.20007918775081635,
+      "learning_rate": 2.5867443095114248e-05,
+      "loss": 0.076,
+      "step": 509
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.12216662615537643,
+      "learning_rate": 2.56807445721972e-05,
+      "loss": 0.0656,
+      "step": 510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.12289852648973465,
+      "learning_rate": 2.5494489141099153e-05,
+      "loss": 0.0692,
+      "step": 511
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.16199322044849396,
+      "learning_rate": 2.5308680195355616e-05,
+      "loss": 0.0749,
+      "step": 512
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09642549604177475,
+      "learning_rate": 2.5123321120367273e-05,
+      "loss": 0.071,
+      "step": 513
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.10821285098791122,
+      "learning_rate": 2.4938415293338253e-05,
+      "loss": 0.0724,
+      "step": 514
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.10590888559818268,
+      "learning_rate": 2.4753966083214615e-05,
+      "loss": 0.0695,
+      "step": 515
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09137357026338577,
+      "learning_rate": 2.4569976850622888e-05,
+      "loss": 0.0553,
+      "step": 516
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1135525181889534,
+      "learning_rate": 2.4386450947808974e-05,
+      "loss": 0.0787,
+      "step": 517
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.12290289998054504,
+      "learning_rate": 2.4203391718576994e-05,
+      "loss": 0.0666,
+      "step": 518
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.11229589581489563,
+      "learning_rate": 2.4020802498228335e-05,
+      "loss": 0.0669,
+      "step": 519
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1318112462759018,
+      "learning_rate": 2.3838686613500927e-05,
+      "loss": 0.0706,
+      "step": 520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.12568672001361847,
+      "learning_rate": 2.3657047382508664e-05,
+      "loss": 0.0767,
+      "step": 521
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.11129189282655716,
+      "learning_rate": 2.3475888114680845e-05,
+      "loss": 0.0647,
+      "step": 522
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.14194755256175995,
+      "learning_rate": 2.329521211070199e-05,
+      "loss": 0.0646,
+      "step": 523
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1168050616979599,
+      "learning_rate": 2.3115022662451642e-05,
+      "loss": 0.0718,
+      "step": 524
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.109759122133255,
+      "learning_rate": 2.29353230529444e-05,
+      "loss": 0.0703,
+      "step": 525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.11648988723754883,
+      "learning_rate": 2.275611655627006e-05,
+      "loss": 0.0715,
+      "step": 526
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.15601898729801178,
+      "learning_rate": 2.2577406437534054e-05,
+      "loss": 0.0734,
+      "step": 527
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.13575918972492218,
+      "learning_rate": 2.2399195952797913e-05,
+      "loss": 0.067,
+      "step": 528
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.16341762244701385,
+      "learning_rate": 2.2221488349019903e-05,
+      "loss": 0.0682,
+      "step": 529
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.12766920030117035,
+      "learning_rate": 2.2044286863995893e-05,
+      "loss": 0.0696,
+      "step": 530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.12745466828346252,
+      "learning_rate": 2.186759472630045e-05,
+      "loss": 0.0754,
+      "step": 531
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.15519979596138,
+      "learning_rate": 2.1691415155227835e-05,
+      "loss": 0.0718,
+      "step": 532
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.12283770740032196,
+      "learning_rate": 2.151575136073353e-05,
+      "loss": 0.0637,
+      "step": 533
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.11813485622406006,
+      "learning_rate": 2.1340606543375664e-05,
+      "loss": 0.0627,
+      "step": 534
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.13709497451782227,
+      "learning_rate": 2.1165983894256647e-05,
+      "loss": 0.0744,
+      "step": 535
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1242271438241005,
+      "learning_rate": 2.099188659496517e-05,
+      "loss": 0.0674,
+      "step": 536
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.11997799575328827,
+      "learning_rate": 2.0818317817518118e-05,
+      "loss": 0.0621,
+      "step": 537
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.13129732012748718,
+      "learning_rate": 2.0645280724302836e-05,
+      "loss": 0.0721,
+      "step": 538
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.14182625710964203,
+      "learning_rate": 2.0472778468019454e-05,
+      "loss": 0.0718,
+      "step": 539
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.12494273483753204,
+      "learning_rate": 2.030081419162354e-05,
+      "loss": 0.0611,
+      "step": 540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1131092980504036,
+      "learning_rate": 2.0129391028268735e-05,
+      "loss": 0.0685,
+      "step": 541
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.11226702481508255,
+      "learning_rate": 1.9958512101249705e-05,
+      "loss": 0.0622,
+      "step": 542
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.11921816319227219,
+      "learning_rate": 1.9788180523945277e-05,
+      "loss": 0.0666,
+      "step": 543
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1638924926519394,
+      "learning_rate": 1.9618399399761688e-05,
+      "loss": 0.0757,
+      "step": 544
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.13155657052993774,
+      "learning_rate": 1.9449171822075973e-05,
+      "loss": 0.0651,
+      "step": 545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1587192714214325,
+      "learning_rate": 1.9280500874179726e-05,
+      "loss": 0.0711,
+      "step": 546
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.13725396990776062,
+      "learning_rate": 1.9112389629222823e-05,
+      "loss": 0.0629,
+      "step": 547
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.13107489049434662,
+      "learning_rate": 1.89448411501575e-05,
+      "loss": 0.0667,
+      "step": 548
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.12860113382339478,
+      "learning_rate": 1.8777858489682466e-05,
+      "loss": 0.0696,
+      "step": 549
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1347755342721939,
+      "learning_rate": 1.8611444690187392e-05,
+      "loss": 0.063,
+      "step": 550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.14524155855178833,
+      "learning_rate": 1.8445602783697374e-05,
+      "loss": 0.0709,
+      "step": 551
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.12448085844516754,
+      "learning_rate": 1.8280335791817733e-05,
+      "loss": 0.0657,
+      "step": 552
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.12951478362083435,
+      "learning_rate": 1.8115646725678997e-05,
+      "loss": 0.069,
+      "step": 553
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1278984248638153,
+      "learning_rate": 1.7951538585882007e-05,
+      "loss": 0.0753,
+      "step": 554
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.12768521904945374,
+      "learning_rate": 1.778801436244319e-05,
+      "loss": 0.0677,
+      "step": 555
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1416548192501068,
+      "learning_rate": 1.7625077034740195e-05,
+      "loss": 0.0666,
+      "step": 556
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.11569023132324219,
+      "learning_rate": 1.7462729571457558e-05,
+      "loss": 0.0662,
+      "step": 557
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.14115728437900543,
+      "learning_rate": 1.7300974930532542e-05,
+      "loss": 0.0741,
+      "step": 558
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.12663856148719788,
+      "learning_rate": 1.713981605910137e-05,
+      "loss": 0.0694,
+      "step": 559
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.14533084630966187,
+      "learning_rate": 1.6979255893445462e-05,
+      "loss": 0.0771,
+      "step": 560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1276610940694809,
+      "learning_rate": 1.68192973589379e-05,
+      "loss": 0.0685,
+      "step": 561
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.12586858868598938,
+      "learning_rate": 1.66599433699902e-05,
+      "loss": 0.0695,
+      "step": 562
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.11848849058151245,
+      "learning_rate": 1.650119682999918e-05,
+      "loss": 0.0754,
+      "step": 563
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.11971968412399292,
+      "learning_rate": 1.6343060631294083e-05,
+      "loss": 0.0664,
+      "step": 564
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.10469724982976913,
+      "learning_rate": 1.6185537655083817e-05,
+      "loss": 0.0643,
+      "step": 565
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.11540230363607407,
+      "learning_rate": 1.6028630771404523e-05,
+      "loss": 0.0651,
+      "step": 566
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.09816712141036987,
+      "learning_rate": 1.5872342839067306e-05,
+      "loss": 0.0634,
+      "step": 567
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1148160770535469,
+      "learning_rate": 1.571667670560601e-05,
+      "loss": 0.0705,
+      "step": 568
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.11048528552055359,
+      "learning_rate": 1.5561635207225515e-05,
+      "loss": 0.0741,
+      "step": 569
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1149090901017189,
+      "learning_rate": 1.5407221168749967e-05,
+      "loss": 0.0694,
+      "step": 570
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.12481275945901871,
+      "learning_rate": 1.525343740357128e-05,
+      "loss": 0.063,
+      "step": 571
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.12050320208072662,
+      "learning_rate": 1.5100286713597938e-05,
+      "loss": 0.0684,
+      "step": 572
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.11793509125709534,
+      "learning_rate": 1.4947771889203938e-05,
+      "loss": 0.0673,
+      "step": 573
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.14838404953479767,
+      "learning_rate": 1.4795895709177954e-05,
+      "loss": 0.068,
+      "step": 574
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.12349526584148407,
+      "learning_rate": 1.4644660940672627e-05,
+      "loss": 0.0691,
+      "step": 575
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.152385875582695,
+      "learning_rate": 1.4494070339154276e-05,
+      "loss": 0.0737,
+      "step": 576
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.11645952612161636,
+      "learning_rate": 1.4344126648352618e-05,
+      "loss": 0.0729,
+      "step": 577
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.11911028623580933,
+      "learning_rate": 1.419483260021075e-05,
+      "loss": 0.067,
+      "step": 578
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.11739668995141983,
+      "learning_rate": 1.404619091483546e-05,
+      "loss": 0.0703,
+      "step": 579
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.11539848148822784,
+      "learning_rate": 1.3898204300447616e-05,
+      "loss": 0.0684,
+      "step": 580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1286860853433609,
+      "learning_rate": 1.3750875453332801e-05,
+      "loss": 0.0694,
+      "step": 581
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.12313937395811081,
+      "learning_rate": 1.3604207057792218e-05,
+      "loss": 0.0647,
+      "step": 582
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.11173125356435776,
+      "learning_rate": 1.3458201786093794e-05,
+      "loss": 0.0675,
+      "step": 583
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.12815962731838226,
+      "learning_rate": 1.3312862298423496e-05,
+      "loss": 0.0685,
+      "step": 584
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.11466740816831589,
+      "learning_rate": 1.3168191242836786e-05,
+      "loss": 0.0646,
+      "step": 585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1391080915927887,
+      "learning_rate": 1.3024191255210477e-05,
+      "loss": 0.0674,
+      "step": 586
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.13418692350387573,
+      "learning_rate": 1.2880864959194665e-05,
+      "loss": 0.0683,
+      "step": 587
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.10707702487707138,
+      "learning_rate": 1.273821496616488e-05,
+      "loss": 0.0645,
+      "step": 588
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.13759590685367584,
+      "learning_rate": 1.2596243875174591e-05,
+      "loss": 0.0764,
+      "step": 589
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1846940666437149,
+      "learning_rate": 1.2454954272907815e-05,
+      "loss": 0.0679,
+      "step": 590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.11831660568714142,
+      "learning_rate": 1.2314348733631959e-05,
+      "loss": 0.0655,
+      "step": 591
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.11042218655347824,
+      "learning_rate": 1.2174429819150929e-05,
+      "loss": 0.0616,
+      "step": 592
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.12103278189897537,
+      "learning_rate": 1.2035200078758519e-05,
+      "loss": 0.0724,
+      "step": 593
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1172623485326767,
+      "learning_rate": 1.1896662049191898e-05,
+      "loss": 0.0653,
+      "step": 594
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.11154913157224655,
+      "learning_rate": 1.1758818254585369e-05,
+      "loss": 0.0621,
+      "step": 595
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.17942464351654053,
+      "learning_rate": 1.1621671206424462e-05,
+      "loss": 0.0735,
+      "step": 596
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.12177997827529907,
+      "learning_rate": 1.1485223403500117e-05,
+      "loss": 0.0718,
+      "step": 597
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.12012594193220139,
+      "learning_rate": 1.134947733186315e-05,
+      "loss": 0.0682,
+      "step": 598
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.11858519166707993,
+      "learning_rate": 1.1214435464779006e-05,
+      "loss": 0.0674,
+      "step": 599
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.13078168034553528,
+      "learning_rate": 1.1080100262682663e-05,
+      "loss": 0.0636,
+      "step": 600
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.14218783378601074,
+      "learning_rate": 1.0946474173133792e-05,
+      "loss": 0.0654,
+      "step": 601
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.11505080759525299,
+      "learning_rate": 1.0813559630772173e-05,
+      "loss": 0.0615,
+      "step": 602
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.12753449380397797,
+      "learning_rate": 1.0681359057273388e-05,
+      "loss": 0.0754,
+      "step": 603
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.13185451924800873,
+      "learning_rate": 1.0549874861304605e-05,
+      "loss": 0.06,
+      "step": 604
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.13252925872802734,
+      "learning_rate": 1.0419109438480762e-05,
+      "loss": 0.0668,
+      "step": 605
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.12701675295829773,
+      "learning_rate": 1.0289065171320905e-05,
+      "loss": 0.0724,
+      "step": 606
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.13044829666614532,
+      "learning_rate": 1.0159744429204777e-05,
+      "loss": 0.0659,
+      "step": 607
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1393629014492035,
+      "learning_rate": 1.00311495683296e-05,
+      "loss": 0.0566,
+      "step": 608
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.12651897966861725,
+      "learning_rate": 9.903282931667246e-06,
+      "loss": 0.0659,
+      "step": 609
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.11385196447372437,
+      "learning_rate": 9.776146848921475e-06,
+      "loss": 0.0696,
+      "step": 610
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.13774777948856354,
+      "learning_rate": 9.64974363648548e-06,
+      "loss": 0.0725,
+      "step": 611
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.14081576466560364,
+      "learning_rate": 9.524075597399718e-06,
+      "loss": 0.0696,
+      "step": 612
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.13455650210380554,
+      "learning_rate": 9.399145021309974e-06,
+      "loss": 0.068,
+      "step": 613
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1522856503725052,
+      "learning_rate": 9.27495418442555e-06,
+      "loss": 0.0649,
+      "step": 614
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.13210617005825043,
+      "learning_rate": 9.151505349477902e-06,
+      "loss": 0.0699,
+      "step": 615
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.13013535737991333,
+      "learning_rate": 9.028800765679347e-06,
+      "loss": 0.0626,
+      "step": 616
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.230327770113945,
+      "learning_rate": 8.906842668682102e-06,
+      "loss": 0.0625,
+      "step": 617
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.16013884544372559,
+      "learning_rate": 8.785633280537537e-06,
+      "loss": 0.0706,
+      "step": 618
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.14744172990322113,
+      "learning_rate": 8.665174809655708e-06,
+      "loss": 0.065,
+      "step": 619
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.11363288015127182,
+      "learning_rate": 8.54546945076513e-06,
+      "loss": 0.0553,
+      "step": 620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30628418922424316,
+      "learning_rate": 8.426519384872733e-06,
+      "loss": 0.0682,
+      "step": 621
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18932583928108215,
+      "learning_rate": 8.308326779224218e-06,
+      "loss": 0.0712,
+      "step": 622
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.14569301903247833,
+      "learning_rate": 8.19089378726447e-06,
+      "loss": 0.0648,
+      "step": 623
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.14561575651168823,
+      "learning_rate": 8.074222548598371e-06,
+      "loss": 0.0643,
+      "step": 624
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.13920505344867706,
+      "learning_rate": 7.958315188951848e-06,
+      "loss": 0.0624,
+      "step": 625
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.12166207283735275,
+      "learning_rate": 7.843173820133104e-06,
+      "loss": 0.0705,
+      "step": 626
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1481250822544098,
+      "learning_rate": 7.728800539994113e-06,
+      "loss": 0.0646,
+      "step": 627
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1408376544713974,
+      "learning_rate": 7.615197432392462e-06,
+      "loss": 0.0573,
+      "step": 628
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1506175696849823,
+      "learning_rate": 7.502366567153346e-06,
+      "loss": 0.066,
+      "step": 629
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.15556824207305908,
+      "learning_rate": 7.390310000031875e-06,
+      "loss": 0.0761,
+      "step": 630
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.11223191022872925,
+      "learning_rate": 7.2790297726755716e-06,
+      "loss": 0.0588,
+      "step": 631
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1281258761882782,
+      "learning_rate": 7.168527912587253e-06,
+      "loss": 0.0634,
+      "step": 632
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1385982483625412,
+      "learning_rate": 7.058806433088011e-06,
+      "loss": 0.0584,
+      "step": 633
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1930031180381775,
+      "learning_rate": 6.949867333280569e-06,
+      "loss": 0.0723,
+      "step": 634
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.16700227558612823,
+      "learning_rate": 6.8417125980128675e-06,
+      "loss": 0.0725,
+      "step": 635
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.13950631022453308,
+      "learning_rate": 6.734344197841891e-06,
+      "loss": 0.0663,
+      "step": 636
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17406101524829865,
+      "learning_rate": 6.627764088997734e-06,
+      "loss": 0.0795,
+      "step": 637
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.15877845883369446,
+      "learning_rate": 6.5219742133480085e-06,
+      "loss": 0.0649,
+      "step": 638
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1270444542169571,
+      "learning_rate": 6.416976498362432e-06,
+      "loss": 0.0652,
+      "step": 639
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.12452571839094162,
+      "learning_rate": 6.3127728570777365e-06,
+      "loss": 0.0664,
+      "step": 640
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.16081424057483673,
+      "learning_rate": 6.209365188062749e-06,
+      "loss": 0.0652,
+      "step": 641
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1544044464826584,
+      "learning_rate": 6.106755375383904e-06,
+      "loss": 0.0643,
+      "step": 642
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.12766015529632568,
+      "learning_rate": 6.004945288570813e-06,
+      "loss": 0.0672,
+      "step": 643
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1414952278137207,
+      "learning_rate": 5.903936782582253e-06,
+      "loss": 0.0659,
+      "step": 644
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.12395791709423065,
+      "learning_rate": 5.803731697772391e-06,
+      "loss": 0.0717,
+      "step": 645
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.12516911327838898,
+      "learning_rate": 5.7043318598572094e-06,
+      "loss": 0.064,
+      "step": 646
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1448909342288971,
+      "learning_rate": 5.605739079881239e-06,
+      "loss": 0.0734,
+      "step": 647
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1468668133020401,
+      "learning_rate": 5.507955154184619e-06,
+      "loss": 0.0647,
+      "step": 648
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.13697059452533722,
+      "learning_rate": 5.4109818643703124e-06,
+      "loss": 0.0696,
+      "step": 649
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.13059355318546295,
+      "learning_rate": 5.314820977271645e-06,
+      "loss": 0.0587,
+      "step": 650
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.12498347461223602,
+      "learning_rate": 5.219474244920164e-06,
+      "loss": 0.0648,
+      "step": 651
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.11734464764595032,
+      "learning_rate": 5.1249434045136765e-06,
+      "loss": 0.0602,
+      "step": 652
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.14948828518390656,
+      "learning_rate": 5.031230178384594e-06,
+      "loss": 0.0765,
+      "step": 653
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1507301926612854,
+      "learning_rate": 4.938336273968558e-06,
+      "loss": 0.0672,
+      "step": 654
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1803879290819168,
+      "learning_rate": 4.846263383773364e-06,
+      "loss": 0.0604,
+      "step": 655
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1707804799079895,
+      "learning_rate": 4.755013185348095e-06,
+      "loss": 0.0734,
+      "step": 656
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.15067826211452484,
+      "learning_rate": 4.66458734125253e-06,
+      "loss": 0.0646,
+      "step": 657
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.12395960837602615,
+      "learning_rate": 4.574987499026912e-06,
+      "loss": 0.0595,
+      "step": 658
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.13461753726005554,
+      "learning_rate": 4.4862152911618934e-06,
+      "loss": 0.06,
+      "step": 659
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.14375466108322144,
+      "learning_rate": 4.398272335068787e-06,
+      "loss": 0.0613,
+      "step": 660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.11830610781908035,
+      "learning_rate": 4.311160233050121e-06,
+      "loss": 0.0637,
+      "step": 661
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.13027916848659515,
+      "learning_rate": 4.224880572270434e-06,
+      "loss": 0.0647,
+      "step": 662
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.13159357011318207,
+      "learning_rate": 4.139434924727359e-06,
+      "loss": 0.0663,
+      "step": 663
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.12546558678150177,
+      "learning_rate": 4.054824847222949e-06,
+      "loss": 0.0603,
+      "step": 664
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.13789647817611694,
+      "learning_rate": 3.971051881335391e-06,
+      "loss": 0.0561,
+      "step": 665
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.142785906791687,
+      "learning_rate": 3.888117553390852e-06,
+      "loss": 0.0667,
+      "step": 666
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1499502956867218,
+      "learning_rate": 3.8060233744356633e-06,
+      "loss": 0.0675,
+      "step": 667
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1378251016139984,
+      "learning_rate": 3.724770840208852e-06,
+      "loss": 0.0628,
+      "step": 668
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.12548185884952545,
+      "learning_rate": 3.6443614311148456e-06,
+      "loss": 0.0689,
+      "step": 669
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.13463036715984344,
+      "learning_rate": 3.564796612196475e-06,
+      "loss": 0.0667,
+      "step": 670
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.14040209352970123,
+      "learning_rate": 3.486077833108342e-06,
+      "loss": 0.071,
+      "step": 671
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.11872732639312744,
+      "learning_rate": 3.408206528090374e-06,
+      "loss": 0.0589,
+      "step": 672
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.13672368228435516,
+      "learning_rate": 3.3311841159416936e-06,
+      "loss": 0.0605,
+      "step": 673
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.16072207689285278,
+      "learning_rate": 3.25501199999475e-06,
+      "loss": 0.0637,
+      "step": 674
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.12611547112464905,
+      "learning_rate": 3.1796915680897988e-06,
+      "loss": 0.0614,
+      "step": 675
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.15802842378616333,
+      "learning_rate": 3.1052241925495885e-06,
+      "loss": 0.0651,
+      "step": 676
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.15073926746845245,
+      "learning_rate": 3.031611230154324e-06,
+      "loss": 0.0657,
+      "step": 677
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.16130346059799194,
+      "learning_rate": 2.958854022117008e-06,
+      "loss": 0.0745,
+      "step": 678
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.15508459508419037,
+      "learning_rate": 2.88695389405898e-06,
+      "loss": 0.0749,
+      "step": 679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1521391123533249,
+      "learning_rate": 2.8159121559857284e-06,
+      "loss": 0.0599,
+      "step": 680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.12179268896579742,
+      "learning_rate": 2.745730102263078e-06,
+      "loss": 0.0588,
+      "step": 681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1358071267604828,
+      "learning_rate": 2.6764090115935834e-06,
+      "loss": 0.0648,
+      "step": 682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.17397892475128174,
+      "learning_rate": 2.6079501469932156e-06,
+      "loss": 0.0618,
+      "step": 683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.15012788772583008,
+      "learning_rate": 2.540354755768365e-06,
+      "loss": 0.0605,
+      "step": 684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.15713779628276825,
+      "learning_rate": 2.4736240694931247e-06,
+      "loss": 0.0675,
+      "step": 685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1717958003282547,
+      "learning_rate": 2.407759303986845e-06,
+      "loss": 0.0679,
+      "step": 686
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.15394960343837738,
+      "learning_rate": 2.342761659291959e-06,
+      "loss": 0.071,
+      "step": 687
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.17352795600891113,
+      "learning_rate": 2.2786323196521575e-06,
+      "loss": 0.0674,
+      "step": 688
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.16719239950180054,
+      "learning_rate": 2.215372453490788e-06,
+      "loss": 0.071,
+      "step": 689
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1454174518585205,
+      "learning_rate": 2.152983213389559e-06,
+      "loss": 0.0736,
+      "step": 690
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1298518180847168,
+      "learning_rate": 2.0914657360675606e-06,
+      "loss": 0.0617,
+      "step": 691
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.14243261516094208,
+      "learning_rate": 2.0308211423605505e-06,
+      "loss": 0.0622,
+      "step": 692
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1646115630865097,
+      "learning_rate": 1.971050537200514e-06,
+      "loss": 0.0685,
+      "step": 693
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.127439484000206,
+      "learning_rate": 1.912155009595551e-06,
+      "loss": 0.0667,
+      "step": 694
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.14966559410095215,
+      "learning_rate": 1.8541356326100433e-06,
+      "loss": 0.066,
+      "step": 695
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1566605120897293,
+      "learning_rate": 1.7969934633450535e-06,
+      "loss": 0.0646,
+      "step": 696
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.14571478962898254,
+      "learning_rate": 1.740729542919134e-06,
+      "loss": 0.0679,
+      "step": 697
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1392918974161148,
+      "learning_rate": 1.685344896449309e-06,
+      "loss": 0.0674,
+      "step": 698
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1348922848701477,
+      "learning_rate": 1.6308405330324294e-06,
+      "loss": 0.0638,
+      "step": 699
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.16134820878505707,
+      "learning_rate": 1.5772174457267362e-06,
+      "loss": 0.0681,
+      "step": 700
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1294383555650711,
+      "learning_rate": 1.5244766115338372e-06,
+      "loss": 0.0581,
+      "step": 701
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.17935150861740112,
+      "learning_rate": 1.472618991380853e-06,
+      "loss": 0.074,
+      "step": 702
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.13329242169857025,
+      "learning_rate": 1.4216455301029275e-06,
+      "loss": 0.0617,
+      "step": 703
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.171311616897583,
+      "learning_rate": 1.3715571564260122e-06,
+      "loss": 0.0694,
+      "step": 704
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1370270550251007,
+      "learning_rate": 1.3223547829499527e-06,
+      "loss": 0.06,
+      "step": 705
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1376296430826187,
+      "learning_rate": 1.2740393061318357e-06,
+      "loss": 0.0654,
+      "step": 706
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.14175420999526978,
+      "learning_rate": 1.2266116062696953e-06,
+      "loss": 0.0737,
+      "step": 707
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.16615577042102814,
+      "learning_rate": 1.1800725474864438e-06,
+      "loss": 0.0682,
+      "step": 708
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1245865449309349,
+      "learning_rate": 1.1344229777141336e-06,
+      "loss": 0.0624,
+      "step": 709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1388980895280838,
+      "learning_rate": 1.0896637286785084e-06,
+      "loss": 0.0683,
+      "step": 710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.14993278682231903,
+      "learning_rate": 1.0457956158838544e-06,
+      "loss": 0.0679,
+      "step": 711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.13806217908859253,
+      "learning_rate": 1.0028194385981515e-06,
+      "loss": 0.0599,
+      "step": 712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.17131811380386353,
+      "learning_rate": 9.607359798384785e-07,
+      "loss": 0.0624,
+      "step": 713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.14887043833732605,
+      "learning_rate": 9.195460063567807e-07,
+      "loss": 0.0652,
+      "step": 714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1282467544078827,
+      "learning_rate": 8.792502686258752e-07,
+      "loss": 0.056,
+      "step": 715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1668287068605423,
+      "learning_rate": 8.398495008257956e-07,
+      "loss": 0.0629,
+      "step": 716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.14125637710094452,
+      "learning_rate": 8.013444208304133e-07,
+      "loss": 0.0719,
+      "step": 717
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.14389286935329437,
+      "learning_rate": 7.637357301943371e-07,
+      "loss": 0.0634,
+      "step": 718
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.15357066690921783,
+      "learning_rate": 7.270241141401569e-07,
+      "loss": 0.063,
+      "step": 719
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.127652108669281,
+      "learning_rate": 6.912102415459476e-07,
+      "loss": 0.0592,
+      "step": 720
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.15074937045574188,
+      "learning_rate": 6.562947649330853e-07,
+      "loss": 0.0665,
+      "step": 721
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.14497500658035278,
+      "learning_rate": 6.222783204543559e-07,
+      "loss": 0.0679,
+      "step": 722
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.16830851137638092,
+      "learning_rate": 5.891615278823537e-07,
+      "loss": 0.0662,
+      "step": 723
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.11228172481060028,
+      "learning_rate": 5.569449905982239e-07,
+      "loss": 0.0621,
+      "step": 724
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.15398009121418,
+      "learning_rate": 5.256292955806208e-07,
+      "loss": 0.0641,
+      "step": 725
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.14682018756866455,
+      "learning_rate": 4.952150133950506e-07,
+      "loss": 0.0676,
+      "step": 726
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.13635100424289703,
+      "learning_rate": 4.6570269818346224e-07,
+      "loss": 0.0661,
+      "step": 727
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1787886917591095,
+      "learning_rate": 4.370928876541613e-07,
+      "loss": 0.0606,
+      "step": 728
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.16212987899780273,
+      "learning_rate": 4.093861030719903e-07,
+      "loss": 0.0672,
+      "step": 729
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.13791510462760925,
+      "learning_rate": 3.8258284924885255e-07,
+      "loss": 0.064,
+      "step": 730
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1542445421218872,
+      "learning_rate": 3.5668361453450317e-07,
+      "loss": 0.0647,
+      "step": 731
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.14709891378879547,
+      "learning_rate": 3.3168887080766154e-07,
+      "loss": 0.0615,
+      "step": 732
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.16037517786026,
+      "learning_rate": 3.075990734674017e-07,
+      "loss": 0.0605,
+      "step": 733
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.14704789221286774,
+      "learning_rate": 2.8441466142486994e-07,
+      "loss": 0.0702,
+      "step": 734
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.14369893074035645,
+      "learning_rate": 2.62136057095258e-07,
+      "loss": 0.0671,
+      "step": 735
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.14050166308879852,
+      "learning_rate": 2.407636663901591e-07,
+      "loss": 0.0585,
+      "step": 736
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.13352462649345398,
+      "learning_rate": 2.2029787871010733e-07,
+      "loss": 0.0657,
+      "step": 737
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.14897920191287994,
+      "learning_rate": 2.007390669375553e-07,
+      "loss": 0.0675,
+      "step": 738
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.12918329238891602,
+      "learning_rate": 1.820875874300021e-07,
+      "loss": 0.0559,
+      "step": 739
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.15355056524276733,
+      "learning_rate": 1.64343780013565e-07,
+      "loss": 0.0653,
+      "step": 740
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.15641310811042786,
+      "learning_rate": 1.4750796797675658e-07,
+      "loss": 0.0674,
+      "step": 741
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1416761726140976,
+      "learning_rate": 1.3158045806460073e-07,
+      "loss": 0.0593,
+      "step": 742
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17423132061958313,
+      "learning_rate": 1.1656154047303691e-07,
+      "loss": 0.0657,
+      "step": 743
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.14585049450397491,
+      "learning_rate": 1.024514888436634e-07,
+      "loss": 0.0667,
+      "step": 744
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.12190530449151993,
+      "learning_rate": 8.925056025869128e-08,
+      "loss": 0.0575,
+      "step": 745
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1758907437324524,
+      "learning_rate": 7.695899523633143e-08,
+      "loss": 0.0665,
+      "step": 746
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.11777140945196152,
+      "learning_rate": 6.557701772635372e-08,
+      "loss": 0.0658,
+      "step": 747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.14637833833694458,
+      "learning_rate": 5.5104835106051245e-08,
+      "loss": 0.0643,
+      "step": 748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.13879886269569397,
+      "learning_rate": 4.5542638176421237e-08,
+      "loss": 0.0643,
+      "step": 749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.15284717082977295,
+      "learning_rate": 3.689060115872889e-08,
+      "loss": 0.0705,
+      "step": 750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.14697931706905365,
+      "learning_rate": 2.9148881691298813e-08,
+      "loss": 0.0698,
+      "step": 751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.17319975793361664,
+      "learning_rate": 2.231762082666733e-08,
+      "loss": 0.0683,
+      "step": 752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.13041572272777557,
+      "learning_rate": 1.6396943028995638e-08,
+      "loss": 0.0646,
+      "step": 753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.16726787388324738,
+      "learning_rate": 1.1386956171816066e-08,
+      "loss": 0.0589,
+      "step": 754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.15242543816566467,
+      "learning_rate": 7.287751536050324e-09,
+      "loss": 0.0652,
+      "step": 755
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1539144217967987,
+      "learning_rate": 4.099403808366376e-09,
+      "loss": 0.0694,
+      "step": 756
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.13503475487232208,
+      "learning_rate": 1.821971079796203e-09,
+      "loss": 0.0682,
+      "step": 757
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.16742639243602753,
+      "learning_rate": 4.5549484470330807e-10,
+      "loss": 0.072,
+      "step": 758
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.12691205739974976,
+      "learning_rate": 0.0,
+      "loss": 0.0598,
+      "step": 759
+    },
+    {
+      "epoch": 1.0,
+      "step": 759,
+      "total_flos": 1.6072338418759107e+18,
+      "train_loss": 0.07730967932833513,
+      "train_runtime": 10109.3463,
+      "train_samples_per_second": 9.612,
+      "train_steps_per_second": 0.075
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 759,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 1.6072338418759107e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}