diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.988354430379747,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 8.545223204777095,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.1585,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 8.89267994930358,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.1431,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 8.66239468922594,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.1448,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 8.960855999582284,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.1598,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 8.09035366727078,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.1429,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 7.785508154954039,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.1368,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 5.991725162636349,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.0938,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 5.5619713095321295,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.0588,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 3.556825602134894,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.0125,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 3.217228685351757,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 1.0002,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 3.1464539477577977,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.9923,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 4.026205376637923,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.9545,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 3.944784951299204,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.9469,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 3.5802360711748973,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.9088,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.1543157666802157,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.9016,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.493578165383122,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.8678,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.5614398759312262,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.8482,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.1548259672648604,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.8327,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.7403594928661614,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.826,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.4531919627932277,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.7951,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.5177391039441823,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.7843,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.888276590226471,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.7751,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.567929511510444,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.7459,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.3522606003125452,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.7526,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.2224654088578348,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.74,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.3664220739417516,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.7395,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.2382576678727535,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.7365,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.042469054199596,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.7343,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.9677353021434165,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.71,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.9321540430013576,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.7216,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.8603281231870151,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.7128,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.8514792283000293,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.7057,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.7964227968271007,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.6834,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.80638246253113,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.6933,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.9120292092546093,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.6932,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.6962662994070467,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.6799,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.9692549534388211,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.6847,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.7799667464648206,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.6885,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.7005168866012748,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.6784,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.9694239893882967,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.666,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.9113757534949242,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.6809,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.7939347720068108,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.6744,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.7839269551714408,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.6577,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.8609784806198844,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.6695,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.8119502525620956,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.653,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.886788924931296,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.655,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 1.0090834018717683,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.6681,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.764809219348942,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.6562,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 1.1350844253649301,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.6591,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.9311131035609627,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.652,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.9588327206065632,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.6542,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.8397938656261769,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.6422,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 1.0006130376159277,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.6541,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.9382550508337522,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.645,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.9073086488302835,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.6429,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.8967460667194518,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.6602,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 1.0917843276307146,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.6326,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.7448970664573208,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.6509,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 1.2896651258575944,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.667,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.8540735911275511,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.6497,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 1.0914301338734063,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.6471,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.9063459918715201,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.6386,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.9286070313745329,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.6359,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.9252683352224531,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.653,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 1.1770192445472663,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.6363,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.8883705411644902,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.6488,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 1.0970953900439884,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.6294,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.9132646482046837,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.6341,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.8038706938774437,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.6326,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.81770960238598,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.6354,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.7190560069078902,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.6383,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.8155549392036447,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.6306,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.8471944095156886,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.6328,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 1.0609728377707128,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.6325,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.9523423828739331,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.638,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.1097557473329027,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.6252,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.9251009561891615,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.6254,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 1.3011094277004434,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.633,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9147215875675249,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.625,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 1.1103263768540894,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.6142,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.8745603609563959,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.6393,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.9543271979563033,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.6344,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 1.13135382387481,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.6314,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.9098264312084076,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.6275,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.0556987985276658,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.6521,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.0072268572315255,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.6371,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.1112229132743505,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.6161,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.7992811075178401,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.6252,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.0506180340890552,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.6411,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.880261085502058,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.6231,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.3362547895717327,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.6302,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.9388607281787134,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.6256,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 0.8920611743181106,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.6257,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 0.9716401386274649,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.6212,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.3875819017998197,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.6181,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.1277661311432348,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.6306,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 1.2217313482569472,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.6473,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.2228542282024004,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.6326,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.9029703632387598,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.6196,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.0903997204694325,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.6083,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 0.9129758958653195,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.623,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.0082010599272404,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.6265,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 8.105000432218803,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.6417,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 41.95707058185069,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.6752,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 2.604875113677778,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.6493,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.2773798021267888,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.646,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 1.858660641253015,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.6417,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.7811607997184946,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.6381,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.3521033892031797,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.6331,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.3598743382997094,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.639,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.244684645451096,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.6257,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 1.312869968223463,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.6172,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.9875562168356161,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.6248,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.2022245716502553,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.6405,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.9439969803319604,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.6177,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.9458279452166544,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.61,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.3768080669514482,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.6254,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.9937762411853861,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.6351,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.6949074099324906,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.6138,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.1710372329227758,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.6223,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 2.1881225287179404,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.6257,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.3070325827828595,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.6319,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.0998590897144278,
+      "learning_rate": 4e-05,
+      "loss": 0.6208,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.2100655649950087,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.613,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.9466345131904362,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.6247,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.0598246745485478,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.6063,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.1566826201494127,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.6113,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.7574837458778694,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.597,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 1.2255194114546801,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.6083,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.9231032448081452,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.6159,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.8172458999207178,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.6082,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 1.0963639784608832,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.6218,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.8672730020055115,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.6162,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.9058339286240319,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.626,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.0534284713332946,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.6321,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.7814288124331633,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.6155,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.8733004632175195,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.6334,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.7537153866813593,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.6124,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.9304854321505461,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.6104,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 1.1522601467391158,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.6172,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.9174565200481899,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.6155,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.7497641286906027,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.6209,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.9929145898442922,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.5997,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.039179620890967,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.6236,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.9727782571002714,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.617,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.676644910205238,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.6121,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 1.0282386256900378,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.621,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 1.1176780883299793,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.6209,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.7846848229132597,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.6083,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 1.0413186884582633,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.6122,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.7803673927315128,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.6076,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.7786284439374189,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.6132,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.8293427572101028,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.5993,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.7756600915430543,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.6128,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.6744382526326801,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.6193,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.7996628397308797,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.624,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 1.311940174273135,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.6102,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7758411649344755,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.6017,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 1.457581017318339,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.6139,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.8500704203271842,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.6098,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 1.4588474647111094,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.6199,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.8714948960372387,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.6215,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 1.4762565774419085,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.6101,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.8596617424030998,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.6128,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 1.448362098396025,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.6235,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 1.0297794402835465,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.6154,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 1.491484183123241,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.6128,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 1.2187401796482005,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.6086,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 1.2108036627862433,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.6172,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 1.0807373813947863,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.6063,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 1.044022585983875,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.6149,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.9661393358580027,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.6027,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.8775233571427684,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.6172,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.7401987475820295,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.6136,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.8564084497122588,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.6079,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.7484953082848005,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.6116,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.8574185607454958,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.6013,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.7612488482350002,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.6072,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.8337088230489047,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.6076,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.8041529186239261,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.6083,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.7886897442412163,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.6072,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.8377099502326635,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.6156,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.7157047381180172,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.5983,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.7737762650264838,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.6078,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.6591044767266601,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.6041,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.8360118048408077,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.6019,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.6864254852682384,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.615,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.6574751822619472,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.6056,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.6192695100332775,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.6134,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.7035680705765145,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.6158,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.6641020916944212,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.6018,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.5619388899768186,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.5989,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.6731351847430085,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.5985,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.5476292200104952,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.6125,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.8188976961425035,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.5753,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.7169474441361748,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.6032,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.6352640831859007,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.5981,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.6192442562065658,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.6063,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.5680608345270737,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.5902,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.742246948686346,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.6,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.9347461367282139,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.5933,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.7059876795524396,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.5939,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.8687082961633321,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.6136,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.6655452093503005,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.6056,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.8384495678853214,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.5915,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.6352681691431897,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.6109,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.6894007051377997,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.5995,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.6428368246240033,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.5973,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.6660490598514734,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.6081,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.5965986411751037,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.6161,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.6788434156085208,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.6215,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.6584642762792333,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.6116,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.7113383243764154,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.5842,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.6784371369439393,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.5973,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.7120675737349536,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.61,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.6998777448148116,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.6073,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.6518753113399325,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.5948,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.7313828387680844,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.599,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.6349616185821212,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.5826,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.6879606109912189,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.6061,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.6831458721165525,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.5873,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.5825290106983162,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.609,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.5981800032219312,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.6007,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.6070118916661601,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.6068,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.6495880650581808,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.6072,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.6960460021930341,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.5938,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.6210591475662117,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.5939,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.766316279938553,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.601,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.654976555036053,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.5857,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.6356448767891908,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.5802,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.6544311152024694,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.5948,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.5682844159303462,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.5821,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.5250593419654224,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.5989,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.6179388023460624,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.6,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.5328246658278336,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.5859,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.5847073495805489,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.5988,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5396968431156947,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.605,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.5887941891601726,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.5888,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.5544731013505054,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.577,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.5475649055064463,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.5856,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.6342987610976305,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.6089,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.6571127528379908,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.5952,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.5502731660241935,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.5974,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.750299969958319,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.5819,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.5773925614140548,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.5924,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.7431338659433179,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.5975,
+      "step": 246
+    },
+    {
+      "epoch": 1.0020253164556963,
+      "grad_norm": 0.5909783252014554,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.5741,
+      "step": 247
+    },
+    {
+      "epoch": 1.0060759493670886,
+      "grad_norm": 0.604466130717957,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.5255,
+      "step": 248
+    },
+    {
+      "epoch": 1.010126582278481,
+      "grad_norm": 0.6509525080620199,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.5286,
+      "step": 249
+    },
+    {
+      "epoch": 1.0141772151898734,
+      "grad_norm": 0.6578831312760169,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.5187,
+      "step": 250
+    },
+    {
+      "epoch": 1.0182278481012659,
+      "grad_norm": 0.5736730655137543,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.5127,
+      "step": 251
+    },
+    {
+      "epoch": 1.0222784810126582,
+      "grad_norm": 0.686689982831689,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.5107,
+      "step": 252
+    },
+    {
+      "epoch": 1.0263291139240507,
+      "grad_norm": 0.6802242579162129,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.4964,
+      "step": 253
+    },
+    {
+      "epoch": 1.030379746835443,
+      "grad_norm": 0.6479989535603095,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.5202,
+      "step": 254
+    },
+    {
+      "epoch": 1.0344303797468355,
+      "grad_norm": 0.5841672513342306,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.5071,
+      "step": 255
+    },
+    {
+      "epoch": 1.0384810126582278,
+      "grad_norm": 0.6390407225339245,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.5161,
+      "step": 256
+    },
+    {
+      "epoch": 1.0425316455696203,
+      "grad_norm": 0.6482248522746905,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.5179,
+      "step": 257
+    },
+    {
+      "epoch": 1.0465822784810126,
+      "grad_norm": 0.605314352317952,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.5342,
+      "step": 258
+    },
+    {
+      "epoch": 1.0506329113924051,
+      "grad_norm": 0.6148316399068005,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.5142,
+      "step": 259
+    },
+    {
+      "epoch": 1.0546835443037974,
+      "grad_norm": 0.733742722655509,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.5323,
+      "step": 260
+    },
+    {
+      "epoch": 1.05873417721519,
+      "grad_norm": 0.6304444266960922,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.5211,
+      "step": 261
+    },
+    {
+      "epoch": 1.0627848101265822,
+      "grad_norm": 0.6267749506085579,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.5049,
+      "step": 262
+    },
+    {
+      "epoch": 1.0668354430379747,
+      "grad_norm": 0.6719283334315643,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.5139,
+      "step": 263
+    },
+    {
+      "epoch": 1.070886075949367,
+      "grad_norm": 0.5307658394596948,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.4907,
+      "step": 264
+    },
+    {
+      "epoch": 1.0749367088607595,
+      "grad_norm": 0.6837416973228408,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.5058,
+      "step": 265
+    },
+    {
+      "epoch": 1.0789873417721518,
+      "grad_norm": 0.5826211173982611,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.5022,
+      "step": 266
+    },
+    {
+      "epoch": 1.0830379746835443,
+      "grad_norm": 0.7309463625802485,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.5065,
+      "step": 267
+    },
+    {
+      "epoch": 1.0870886075949366,
+      "grad_norm": 0.6820019277042998,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.5087,
+      "step": 268
+    },
+    {
+      "epoch": 1.0911392405063292,
+      "grad_norm": 0.6405704442917461,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.5299,
+      "step": 269
+    },
+    {
+      "epoch": 1.0951898734177214,
+      "grad_norm": 0.6467548987238786,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.506,
+      "step": 270
+    },
+    {
+      "epoch": 1.099240506329114,
+      "grad_norm": 0.5596319226932718,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.5049,
+      "step": 271
+    },
+    {
+      "epoch": 1.1032911392405063,
+      "grad_norm": 0.6269037757360619,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.5225,
+      "step": 272
+    },
+    {
+      "epoch": 1.1073417721518988,
+      "grad_norm": 0.6090562996767479,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.4843,
+      "step": 273
+    },
+    {
+      "epoch": 1.111392405063291,
+      "grad_norm": 0.5446992769776585,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.5115,
+      "step": 274
+    },
+    {
+      "epoch": 1.1154430379746836,
+      "grad_norm": 0.6414035484123298,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.5027,
+      "step": 275
+    },
+    {
+      "epoch": 1.1194936708860759,
+      "grad_norm": 0.5732848040583571,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.5173,
+      "step": 276
+    },
+    {
+      "epoch": 1.1235443037974684,
+      "grad_norm": 0.5870736462733124,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.5021,
+      "step": 277
+    },
+    {
+      "epoch": 1.1275949367088607,
+      "grad_norm": 0.706698529281176,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.516,
+      "step": 278
+    },
+    {
+      "epoch": 1.1316455696202532,
+      "grad_norm": 0.6312367100913827,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.4967,
+      "step": 279
+    },
+    {
+      "epoch": 1.1356962025316455,
+      "grad_norm": 0.6083136352373164,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.527,
+      "step": 280
+    },
+    {
+      "epoch": 1.139746835443038,
+      "grad_norm": 0.7254107368351129,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.5026,
+      "step": 281
+    },
+    {
+      "epoch": 1.1437974683544303,
+      "grad_norm": 0.54464390584546,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.5057,
+      "step": 282
+    },
+    {
+      "epoch": 1.1478481012658228,
+      "grad_norm": 0.549553884808906,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.4935,
+      "step": 283
+    },
+    {
+      "epoch": 1.1518987341772151,
+      "grad_norm": 0.5506896892012302,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.5171,
+      "step": 284
+    },
+    {
+      "epoch": 1.1559493670886076,
+      "grad_norm": 0.5744894315701029,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.4956,
+      "step": 285
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5326961059326032,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.5164,
+      "step": 286
+    },
+    {
+      "epoch": 1.1640506329113924,
+      "grad_norm": 0.5414564286594415,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.502,
+      "step": 287
+    },
+    {
+      "epoch": 1.1681012658227847,
+      "grad_norm": 0.626539754892579,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.5166,
+      "step": 288
+    },
+    {
+      "epoch": 1.1721518987341772,
+      "grad_norm": 0.5158311020876308,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.4938,
+      "step": 289
+    },
+    {
+      "epoch": 1.1762025316455695,
+      "grad_norm": 0.642614107126255,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.4825,
+      "step": 290
+    },
+    {
+      "epoch": 1.180253164556962,
+      "grad_norm": 0.6108862667183804,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.5124,
+      "step": 291
+    },
+    {
+      "epoch": 1.1843037974683543,
+      "grad_norm": 0.5703908703899113,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.5015,
+      "step": 292
+    },
+    {
+      "epoch": 1.1883544303797469,
+      "grad_norm": 0.7919521593418065,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.5157,
+      "step": 293
+    },
+    {
+      "epoch": 1.1924050632911392,
+      "grad_norm": 0.5829168189129895,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.5195,
+      "step": 294
+    },
+    {
+      "epoch": 1.1964556962025317,
+      "grad_norm": 0.8341166537139535,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.5144,
+      "step": 295
+    },
+    {
+      "epoch": 1.200506329113924,
+      "grad_norm": 0.6075981584337612,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.4937,
+      "step": 296
+    },
+    {
+      "epoch": 1.2045569620253165,
+      "grad_norm": 0.8455698393726788,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.5148,
+      "step": 297
+    },
+    {
+      "epoch": 1.2086075949367088,
+      "grad_norm": 0.6524733662264831,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.5129,
+      "step": 298
+    },
+    {
+      "epoch": 1.2126582278481013,
+      "grad_norm": 0.5821745764862113,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.5277,
+      "step": 299
+    },
+    {
+      "epoch": 1.2167088607594936,
+      "grad_norm": 0.7601380302013121,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.5128,
+      "step": 300
+    },
+    {
+      "epoch": 1.220759493670886,
+      "grad_norm": 0.6571477709478736,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.512,
+      "step": 301
+    },
+    {
+      "epoch": 1.2248101265822784,
+      "grad_norm": 0.609050754156057,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.5353,
+      "step": 302
+    },
+    {
+      "epoch": 1.228860759493671,
+      "grad_norm": 0.6080494548010308,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.5189,
+      "step": 303
+    },
+    {
+      "epoch": 1.2329113924050632,
+      "grad_norm": 0.6726268904107869,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.5393,
+      "step": 304
+    },
+    {
+      "epoch": 1.2369620253164557,
+      "grad_norm": 0.6972382776418119,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.5116,
+      "step": 305
+    },
+    {
+      "epoch": 1.241012658227848,
+      "grad_norm": 0.7352049549919802,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.5066,
+      "step": 306
+    },
+    {
+      "epoch": 1.2450632911392405,
+      "grad_norm": 0.6972677611505426,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.5115,
+      "step": 307
+    },
+    {
+      "epoch": 1.2491139240506328,
+      "grad_norm": 0.7113056720999439,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.5054,
+      "step": 308
+    },
+    {
+      "epoch": 1.2531645569620253,
+      "grad_norm": 0.6070343366280182,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.4889,
+      "step": 309
+    },
+    {
+      "epoch": 1.2572151898734178,
+      "grad_norm": 0.8024933657556995,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.4983,
+      "step": 310
+    },
+    {
+      "epoch": 1.2612658227848101,
+      "grad_norm": 0.5495220910003462,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.4989,
+      "step": 311
+    },
+    {
+      "epoch": 1.2653164556962024,
+      "grad_norm": 0.7137257774680018,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.5035,
+      "step": 312
+    },
+    {
+      "epoch": 1.269367088607595,
+      "grad_norm": 0.6075580694896626,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.5077,
+      "step": 313
+    },
+    {
+      "epoch": 1.2734177215189875,
+      "grad_norm": 0.6714715593440296,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.515,
+      "step": 314
+    },
+    {
+      "epoch": 1.2774683544303798,
+      "grad_norm": 0.5109679061853317,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.5094,
+      "step": 315
+    },
+    {
+      "epoch": 1.281518987341772,
+      "grad_norm": 0.6190794048545635,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.5127,
+      "step": 316
+    },
+    {
+      "epoch": 1.2855696202531646,
+      "grad_norm": 0.4854160213477782,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.4987,
+      "step": 317
+    },
+    {
+      "epoch": 1.289620253164557,
+      "grad_norm": 0.5208202130401384,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.5028,
+      "step": 318
+    },
+    {
+      "epoch": 1.2936708860759494,
+      "grad_norm": 0.5461748022266829,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.5236,
+      "step": 319
+    },
+    {
+      "epoch": 1.2977215189873417,
+      "grad_norm": 0.5848852404680638,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.5066,
+      "step": 320
+    },
+    {
+      "epoch": 1.3017721518987342,
+      "grad_norm": 0.5358595239129772,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.5248,
+      "step": 321
+    },
+    {
+      "epoch": 1.3058227848101267,
+      "grad_norm": 0.5039640312685854,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.528,
+      "step": 322
+    },
+    {
+      "epoch": 1.309873417721519,
+      "grad_norm": 0.576481192137766,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.5133,
+      "step": 323
+    },
+    {
+      "epoch": 1.3139240506329113,
+      "grad_norm": 0.5591223658271522,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.5019,
+      "step": 324
+    },
+    {
+      "epoch": 1.3179746835443038,
+      "grad_norm": 0.5692192677936383,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.5096,
+      "step": 325
+    },
+    {
+      "epoch": 1.3220253164556963,
+      "grad_norm": 0.6303357769648404,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.5106,
+      "step": 326
+    },
+    {
+      "epoch": 1.3260759493670886,
+      "grad_norm": 0.614256944660168,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.5109,
+      "step": 327
+    },
+    {
+      "epoch": 1.330126582278481,
+      "grad_norm": 0.7781567724839705,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5029,
+      "step": 328
+    },
+    {
+      "epoch": 1.3341772151898734,
+      "grad_norm": 0.6474998041642324,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.495,
+      "step": 329
+    },
+    {
+      "epoch": 1.338227848101266,
+      "grad_norm": 0.8177585902720443,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.5235,
+      "step": 330
+    },
+    {
+      "epoch": 1.3422784810126582,
+      "grad_norm": 0.6902041501637892,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.5145,
+      "step": 331
+    },
+    {
+      "epoch": 1.3463291139240505,
+      "grad_norm": 0.7950516545546872,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.5295,
+      "step": 332
+    },
+    {
+      "epoch": 1.350379746835443,
+      "grad_norm": 0.5633582329869898,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.5135,
+      "step": 333
+    },
+    {
+      "epoch": 1.3544303797468356,
+      "grad_norm": 0.6878066454227639,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.4752,
+      "step": 334
+    },
+    {
+      "epoch": 1.3584810126582278,
+      "grad_norm": 0.5806618094533973,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.4903,
+      "step": 335
+    },
+    {
+      "epoch": 1.3625316455696201,
+      "grad_norm": 0.5811831381256098,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.4974,
+      "step": 336
+    },
+    {
+      "epoch": 1.3665822784810127,
+      "grad_norm": 0.518109070185437,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.5366,
+      "step": 337
+    },
+    {
+      "epoch": 1.3706329113924052,
+      "grad_norm": 0.5530053874926594,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.515,
+      "step": 338
+    },
+    {
+      "epoch": 1.3746835443037975,
+      "grad_norm": 0.5292128227404245,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.4943,
+      "step": 339
+    },
+    {
+      "epoch": 1.3787341772151898,
+      "grad_norm": 0.6407911059593406,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.4981,
+      "step": 340
+    },
+    {
+      "epoch": 1.3827848101265823,
+      "grad_norm": 0.6692354346057942,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.5238,
+      "step": 341
+    },
+    {
+      "epoch": 1.3868354430379748,
+      "grad_norm": 0.5266894193790567,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.5406,
+      "step": 342
+    },
+    {
+      "epoch": 1.390886075949367,
+      "grad_norm": 0.6192385450116727,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.5178,
+      "step": 343
+    },
+    {
+      "epoch": 1.3949367088607594,
+      "grad_norm": 0.58537457771455,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.4951,
+      "step": 344
+    },
+    {
+      "epoch": 1.398987341772152,
+      "grad_norm": 0.5423201146340713,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.5161,
+      "step": 345
+    },
+    {
+      "epoch": 1.4030379746835444,
+      "grad_norm": 0.5513379668831782,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.5136,
+      "step": 346
+    },
+    {
+      "epoch": 1.4070886075949367,
+      "grad_norm": 0.6084795236130272,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.5174,
+      "step": 347
+    },
+    {
+      "epoch": 1.411139240506329,
+      "grad_norm": 0.5661650499107784,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.513,
+      "step": 348
+    },
+    {
+      "epoch": 1.4151898734177215,
+      "grad_norm": 0.4810995086771398,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.4915,
+      "step": 349
+    },
+    {
+      "epoch": 1.419240506329114,
+      "grad_norm": 0.5857420893396652,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.5182,
+      "step": 350
+    },
+    {
+      "epoch": 1.4232911392405063,
+      "grad_norm": 0.553938825475073,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.5134,
+      "step": 351
+    },
+    {
+      "epoch": 1.4273417721518986,
+      "grad_norm": 0.5405115854264892,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.516,
+      "step": 352
+    },
+    {
+      "epoch": 1.4313924050632911,
+      "grad_norm": 0.5414739355242181,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.4949,
+      "step": 353
+    },
+    {
+      "epoch": 1.4354430379746836,
+      "grad_norm": 0.5687957773681772,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.4962,
+      "step": 354
+    },
+    {
+      "epoch": 1.439493670886076,
+      "grad_norm": 0.4888388629569744,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.4993,
+      "step": 355
+    },
+    {
+      "epoch": 1.4435443037974682,
+      "grad_norm": 0.51112623735595,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.5287,
+      "step": 356
+    },
+    {
+      "epoch": 1.4475949367088607,
+      "grad_norm": 0.49761123296180465,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.494,
+      "step": 357
+    },
+    {
+      "epoch": 1.4516455696202533,
+      "grad_norm": 0.5641120241800798,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.5156,
+      "step": 358
+    },
+    {
+      "epoch": 1.4556962025316456,
+      "grad_norm": 0.48266207024806107,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.5006,
+      "step": 359
+    },
+    {
+      "epoch": 1.4597468354430378,
+      "grad_norm": 0.5419849696949761,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.516,
+      "step": 360
+    },
+    {
+      "epoch": 1.4637974683544304,
+      "grad_norm": 0.49449202156622885,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.522,
+      "step": 361
+    },
+    {
+      "epoch": 1.4678481012658229,
+      "grad_norm": 0.5052297949895234,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.508,
+      "step": 362
+    },
+    {
+      "epoch": 1.4718987341772152,
+      "grad_norm": 0.4821503586850275,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.523,
+      "step": 363
+    },
+    {
+      "epoch": 1.4759493670886075,
+      "grad_norm": 0.4667463806930707,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.5099,
+      "step": 364
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5116533050313411,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.5095,
+      "step": 365
+    },
+    {
+      "epoch": 1.4840506329113925,
+      "grad_norm": 0.46382218953679005,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.4837,
+      "step": 366
+    },
+    {
+      "epoch": 1.4881012658227848,
+      "grad_norm": 0.5152004941726875,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.5234,
+      "step": 367
+    },
+    {
+      "epoch": 1.492151898734177,
+      "grad_norm": 0.4716072747906056,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.5187,
+      "step": 368
+    },
+    {
+      "epoch": 1.4962025316455696,
+      "grad_norm": 0.5648079212481418,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5043,
+      "step": 369
+    },
+    {
+      "epoch": 1.5002531645569621,
+      "grad_norm": 0.4904371045186162,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.5084,
+      "step": 370
+    },
+    {
+      "epoch": 1.5043037974683544,
+      "grad_norm": 0.623015841893749,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.542,
+      "step": 371
+    },
+    {
+      "epoch": 1.5083544303797467,
+      "grad_norm": 0.48401957455628963,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.4999,
+      "step": 372
+    },
+    {
+      "epoch": 1.5124050632911392,
+      "grad_norm": 0.6325487630542175,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.5033,
+      "step": 373
+    },
+    {
+      "epoch": 1.5164556962025317,
+      "grad_norm": 0.4678396832330191,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.5176,
+      "step": 374
+    },
+    {
+      "epoch": 1.520506329113924,
+      "grad_norm": 0.6220117060014072,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.5082,
+      "step": 375
+    },
+    {
+      "epoch": 1.5245569620253163,
+      "grad_norm": 0.5006892139302578,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.5057,
+      "step": 376
+    },
+    {
+      "epoch": 1.5286075949367088,
+      "grad_norm": 0.6815567106573589,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.507,
+      "step": 377
+    },
+    {
+      "epoch": 1.5326582278481014,
+      "grad_norm": 0.5595073801550601,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.5131,
+      "step": 378
+    },
+    {
+      "epoch": 1.5367088607594936,
+      "grad_norm": 0.6505410779493931,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.4881,
+      "step": 379
+    },
+    {
+      "epoch": 1.540759493670886,
+      "grad_norm": 0.5158196402298248,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.5164,
+      "step": 380
+    },
+    {
+      "epoch": 1.5448101265822785,
+      "grad_norm": 0.5529933104808848,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.5081,
+      "step": 381
+    },
+    {
+      "epoch": 1.548860759493671,
+      "grad_norm": 0.5055183184971841,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.4952,
+      "step": 382
+    },
+    {
+      "epoch": 1.5529113924050633,
+      "grad_norm": 0.5419616563875198,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.516,
+      "step": 383
+    },
+    {
+      "epoch": 1.5569620253164556,
+      "grad_norm": 0.5787315331977662,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.5318,
+      "step": 384
+    },
+    {
+      "epoch": 1.561012658227848,
+      "grad_norm": 0.5601770531169437,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.518,
+      "step": 385
+    },
+    {
+      "epoch": 1.5650632911392406,
+      "grad_norm": 0.5647632806134155,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.4988,
+      "step": 386
+    },
+    {
+      "epoch": 1.5691139240506329,
+      "grad_norm": 0.6009828586016639,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.492,
+      "step": 387
+    },
+    {
+      "epoch": 1.5731645569620252,
+      "grad_norm": 0.6176465247944966,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.5127,
+      "step": 388
+    },
+    {
+      "epoch": 1.5772151898734177,
+      "grad_norm": 0.5653265531712323,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.5136,
+      "step": 389
+    },
+    {
+      "epoch": 1.5812658227848102,
+      "grad_norm": 0.5129684632370876,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.5264,
+      "step": 390
+    },
+    {
+      "epoch": 1.5853164556962025,
+      "grad_norm": 0.5573031268061416,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.5109,
+      "step": 391
+    },
+    {
+      "epoch": 1.5893670886075948,
+      "grad_norm": 0.5216139963373764,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.5221,
+      "step": 392
+    },
+    {
+      "epoch": 1.5934177215189873,
+      "grad_norm": 0.6474504333842164,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.5134,
+      "step": 393
+    },
+    {
+      "epoch": 1.5974683544303798,
+      "grad_norm": 0.4587088565302562,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.5299,
+      "step": 394
+    },
+    {
+      "epoch": 1.6015189873417721,
+      "grad_norm": 0.5642468965265649,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.5206,
+      "step": 395
+    },
+    {
+      "epoch": 1.6055696202531644,
+      "grad_norm": 0.5002828988052469,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.5152,
+      "step": 396
+    },
+    {
+      "epoch": 1.609620253164557,
+      "grad_norm": 0.6416071386849015,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.5205,
+      "step": 397
+    },
+    {
+      "epoch": 1.6136708860759494,
+      "grad_norm": 0.4919189855633578,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.5037,
+      "step": 398
+    },
+    {
+      "epoch": 1.6177215189873417,
+      "grad_norm": 0.5160240474090295,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.5279,
+      "step": 399
+    },
+    {
+      "epoch": 1.621772151898734,
+      "grad_norm": 0.5512498047179736,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.5264,
+      "step": 400
+    },
+    {
+      "epoch": 1.6258227848101265,
+      "grad_norm": 0.5155466517125822,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.5012,
+      "step": 401
+    },
+    {
+      "epoch": 1.629873417721519,
+      "grad_norm": 0.6460625213928356,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.516,
+      "step": 402
+    },
+    {
+      "epoch": 1.6339240506329114,
+      "grad_norm": 0.4546450532258875,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5067,
+      "step": 403
+    },
+    {
+      "epoch": 1.6379746835443036,
+      "grad_norm": 0.5234880448365632,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.5427,
+      "step": 404
+    },
+    {
+      "epoch": 1.6420253164556962,
+      "grad_norm": 0.4727244165886054,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.5227,
+      "step": 405
+    },
+    {
+      "epoch": 1.6460759493670887,
+      "grad_norm": 0.5342680297014191,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.5262,
+      "step": 406
+    },
+    {
+      "epoch": 1.650126582278481,
+      "grad_norm": 0.44324337491439486,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.5224,
+      "step": 407
+    },
+    {
+      "epoch": 1.6541772151898733,
+      "grad_norm": 0.4643214355910054,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.5079,
+      "step": 408
+    },
+    {
+      "epoch": 1.6582278481012658,
+      "grad_norm": 0.495726279565073,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.501,
+      "step": 409
+    },
+    {
+      "epoch": 1.6622784810126583,
+      "grad_norm": 0.559310252459632,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5231,
+      "step": 410
+    },
+    {
+      "epoch": 1.6663291139240506,
+      "grad_norm": 0.47528211811874005,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.5199,
+      "step": 411
+    },
+    {
+      "epoch": 1.6703797468354429,
+      "grad_norm": 0.6895560713383232,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.5262,
+      "step": 412
+    },
+    {
+      "epoch": 1.6744303797468354,
+      "grad_norm": 0.5426178819337256,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.509,
+      "step": 413
+    },
+    {
+      "epoch": 1.678481012658228,
+      "grad_norm": 0.6216381548173884,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.5052,
+      "step": 414
+    },
+    {
+      "epoch": 1.6825316455696202,
+      "grad_norm": 0.5452263498047871,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.5221,
+      "step": 415
+    },
+    {
+      "epoch": 1.6865822784810125,
+      "grad_norm": 0.500449622479967,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.5044,
+      "step": 416
+    },
+    {
+      "epoch": 1.690632911392405,
+      "grad_norm": 0.4644335088573735,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.5036,
+      "step": 417
+    },
+    {
+      "epoch": 1.6946835443037975,
+      "grad_norm": 0.48363955831768546,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.4947,
+      "step": 418
+    },
+    {
+      "epoch": 1.6987341772151898,
+      "grad_norm": 0.5063144597814654,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.5104,
+      "step": 419
+    },
+    {
+      "epoch": 1.7027848101265821,
+      "grad_norm": 0.47077836918811494,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5147,
+      "step": 420
+    },
+    {
+      "epoch": 1.7068354430379746,
+      "grad_norm": 0.4915503226300025,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.5026,
+      "step": 421
+    },
+    {
+      "epoch": 1.7108860759493671,
+      "grad_norm": 0.5097394458181134,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.5174,
+      "step": 422
+    },
+    {
+      "epoch": 1.7149367088607594,
+      "grad_norm": 0.4890283326129302,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.5161,
+      "step": 423
+    },
+    {
+      "epoch": 1.7189873417721517,
+      "grad_norm": 0.48527476269129394,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5099,
+      "step": 424
+    },
+    {
+      "epoch": 1.7230379746835442,
+      "grad_norm": 0.4810214825892005,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.5071,
+      "step": 425
+    },
+    {
+      "epoch": 1.7270886075949368,
+      "grad_norm": 0.4594236789175484,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.4989,
+      "step": 426
+    },
+    {
+      "epoch": 1.7311392405063293,
+      "grad_norm": 0.47318291145627656,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.5017,
+      "step": 427
+    },
+    {
+      "epoch": 1.7351898734177216,
+      "grad_norm": 0.5296565232117688,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.514,
+      "step": 428
+    },
+    {
+      "epoch": 1.7392405063291139,
+      "grad_norm": 0.44336103040781943,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.5016,
+      "step": 429
+    },
+    {
+      "epoch": 1.7432911392405064,
+      "grad_norm": 0.537966191673894,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.5047,
+      "step": 430
+    },
+    {
+      "epoch": 1.747341772151899,
+      "grad_norm": 0.5079926897635366,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.5041,
+      "step": 431
+    },
+    {
+      "epoch": 1.7513924050632912,
+      "grad_norm": 0.5000501377873933,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.5064,
+      "step": 432
+    },
+    {
+      "epoch": 1.7554430379746835,
+      "grad_norm": 0.5557119561911984,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.5032,
+      "step": 433
+    },
+    {
+      "epoch": 1.759493670886076,
+      "grad_norm": 0.43714967140822364,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.5137,
+      "step": 434
+    },
+    {
+      "epoch": 1.7635443037974685,
+      "grad_norm": 0.5718734683126129,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.5194,
+      "step": 435
+    },
+    {
+      "epoch": 1.7675949367088608,
+      "grad_norm": 0.459542328444994,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.5242,
+      "step": 436
+    },
+    {
+      "epoch": 1.771645569620253,
+      "grad_norm": 0.48576234762157655,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.5144,
+      "step": 437
+    },
+    {
+      "epoch": 1.7756962025316456,
+      "grad_norm": 0.4984786922324205,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.4976,
+      "step": 438
+    },
+    {
+      "epoch": 1.7797468354430381,
+      "grad_norm": 0.47419950946843437,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.4946,
+      "step": 439
+    },
+    {
+      "epoch": 1.7837974683544304,
+      "grad_norm": 0.5496396368562986,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.5089,
+      "step": 440
+    },
+    {
+      "epoch": 1.7878481012658227,
+      "grad_norm": 0.5023041028508904,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.5136,
+      "step": 441
+    },
+    {
+      "epoch": 1.7918987341772152,
+      "grad_norm": 0.447342879614259,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.5063,
+      "step": 442
+    },
+    {
+      "epoch": 1.7959493670886078,
+      "grad_norm": 0.5554238931601099,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.521,
+      "step": 443
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.4406075614422924,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.5218,
+      "step": 444
+    },
+    {
+      "epoch": 1.8040506329113923,
+      "grad_norm": 0.5241376660886581,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.499,
+      "step": 445
+    },
+    {
+      "epoch": 1.8081012658227849,
+      "grad_norm": 0.48022707028492456,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.5071,
+      "step": 446
+    },
+    {
+      "epoch": 1.8121518987341774,
+      "grad_norm": 0.4947791019399952,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.5095,
+      "step": 447
+    },
+    {
+      "epoch": 1.8162025316455697,
+      "grad_norm": 0.487467117410629,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.5055,
+      "step": 448
+    },
+    {
+      "epoch": 1.820253164556962,
+      "grad_norm": 0.49337708826398097,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.5185,
+      "step": 449
+    },
+    {
+      "epoch": 1.8243037974683545,
+      "grad_norm": 0.48240030178909293,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.503,
+      "step": 450
+    },
+    {
+      "epoch": 1.828354430379747,
+      "grad_norm": 0.5757216418910412,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5195,
+      "step": 451
+    },
+    {
+      "epoch": 1.8324050632911393,
+      "grad_norm": 0.47082532526913295,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.5041,
+      "step": 452
+    },
+    {
+      "epoch": 1.8364556962025316,
+      "grad_norm": 0.4694156486082637,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.5006,
+      "step": 453
+    },
+    {
+      "epoch": 1.840506329113924,
+      "grad_norm": 0.516311542306579,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.5235,
+      "step": 454
+    },
+    {
+      "epoch": 1.8445569620253166,
+      "grad_norm": 0.4808799539074152,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.5182,
+      "step": 455
+    },
+    {
+      "epoch": 1.848607594936709,
+      "grad_norm": 0.42941559213502406,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.5215,
+      "step": 456
+    },
+    {
+      "epoch": 1.8526582278481012,
+      "grad_norm": 0.4833830216248361,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.5086,
+      "step": 457
+    },
+    {
+      "epoch": 1.8567088607594937,
+      "grad_norm": 0.43469020027794264,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.5095,
+      "step": 458
+    },
+    {
+      "epoch": 1.8607594936708862,
+      "grad_norm": 0.43266810037169695,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.5012,
+      "step": 459
+    },
+    {
+      "epoch": 1.8648101265822785,
+      "grad_norm": 0.4698767546469313,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.5087,
+      "step": 460
+    },
+    {
+      "epoch": 1.8688607594936708,
+      "grad_norm": 0.4926156316696228,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.5151,
+      "step": 461
+    },
+    {
+      "epoch": 1.8729113924050633,
+      "grad_norm": 0.46325690054564445,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.4973,
+      "step": 462
+    },
+    {
+      "epoch": 1.8769620253164558,
+      "grad_norm": 0.5390193445421619,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.5012,
+      "step": 463
+    },
+    {
+      "epoch": 1.8810126582278481,
+      "grad_norm": 0.48007714317108185,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.4987,
+      "step": 464
+    },
+    {
+      "epoch": 1.8850632911392404,
+      "grad_norm": 0.4536215205175546,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.5113,
+      "step": 465
+    },
+    {
+      "epoch": 1.889113924050633,
+      "grad_norm": 0.5106974466322206,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.5133,
+      "step": 466
+    },
+    {
+      "epoch": 1.8931645569620255,
+      "grad_norm": 0.4481050557119131,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.5079,
+      "step": 467
+    },
+    {
+      "epoch": 1.8972151898734178,
+      "grad_norm": 0.4586047943757705,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.4973,
+      "step": 468
+    },
+    {
+      "epoch": 1.90126582278481,
+      "grad_norm": 0.49799511099435223,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.497,
+      "step": 469
+    },
+    {
+      "epoch": 1.9053164556962026,
+      "grad_norm": 0.48454732863021055,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.4958,
+      "step": 470
+    },
+    {
+      "epoch": 1.909367088607595,
+      "grad_norm": 0.5244874750345021,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5088,
+      "step": 471
+    },
+    {
+      "epoch": 1.9134177215189874,
+      "grad_norm": 0.4830919869583845,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.4988,
+      "step": 472
+    },
+    {
+      "epoch": 1.9174683544303797,
+      "grad_norm": 0.497627320481107,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.517,
+      "step": 473
+    },
+    {
+      "epoch": 1.9215189873417722,
+      "grad_norm": 0.4767513924446138,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.5306,
+      "step": 474
+    },
+    {
+      "epoch": 1.9255696202531647,
+      "grad_norm": 0.4639773479894447,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.5326,
+      "step": 475
+    },
+    {
+      "epoch": 1.929620253164557,
+      "grad_norm": 0.4495661169114423,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.517,
+      "step": 476
+    },
+    {
+      "epoch": 1.9336708860759493,
+      "grad_norm": 0.4454397111567179,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.5196,
+      "step": 477
+    },
+    {
+      "epoch": 1.9377215189873418,
+      "grad_norm": 0.46225623837431856,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.5189,
+      "step": 478
+    },
+    {
+      "epoch": 1.9417721518987343,
+      "grad_norm": 0.42383043566552797,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.4956,
+      "step": 479
+    },
+    {
+      "epoch": 1.9458227848101266,
+      "grad_norm": 0.46742054575818864,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.4991,
+      "step": 480
+    },
+    {
+      "epoch": 1.949873417721519,
+      "grad_norm": 0.49274974021935614,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.5034,
+      "step": 481
+    },
+    {
+      "epoch": 1.9539240506329114,
+      "grad_norm": 0.4803276838825387,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.5003,
+      "step": 482
+    },
+    {
+      "epoch": 1.957974683544304,
+      "grad_norm": 0.48049223825865695,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.5114,
+      "step": 483
+    },
+    {
+      "epoch": 1.9620253164556962,
+      "grad_norm": 0.45664805155776594,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.4893,
+      "step": 484
+    },
+    {
+      "epoch": 1.9660759493670885,
+      "grad_norm": 0.4680582576063018,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.5339,
+      "step": 485
+    },
+    {
+      "epoch": 1.970126582278481,
+      "grad_norm": 0.48504494657486463,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.4958,
+      "step": 486
+    },
+    {
+      "epoch": 1.9741772151898735,
+      "grad_norm": 0.5171812414985807,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.5015,
+      "step": 487
+    },
+    {
+      "epoch": 1.9782278481012658,
+      "grad_norm": 0.4705353744408362,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.5051,
+      "step": 488
+    },
+    {
+      "epoch": 1.9822784810126581,
+      "grad_norm": 0.4575980746005235,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.511,
+      "step": 489
+    },
+    {
+      "epoch": 1.9863291139240506,
+      "grad_norm": 0.437573444633656,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.4839,
+      "step": 490
+    },
+    {
+      "epoch": 1.9903797468354432,
+      "grad_norm": 0.48963479431677404,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.5094,
+      "step": 491
+    },
+    {
+      "epoch": 1.9944303797468355,
+      "grad_norm": 0.40293936886130316,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5072,
+      "step": 492
+    },
+    {
+      "epoch": 1.9984810126582278,
+      "grad_norm": 0.4314863236015621,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.4979,
+      "step": 493
+    },
+    {
+      "epoch": 2.0040506329113925,
+      "grad_norm": 0.6646767806956008,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.3934,
+      "step": 494
+    },
+    {
+      "epoch": 2.008101265822785,
+      "grad_norm": 0.7629436382850777,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.398,
+      "step": 495
+    },
+    {
+      "epoch": 2.012151898734177,
+      "grad_norm": 0.620095130554915,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.407,
+      "step": 496
+    },
+    {
+      "epoch": 2.0162025316455696,
+      "grad_norm": 0.5589213610287805,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.3902,
+      "step": 497
+    },
+    {
+      "epoch": 2.020253164556962,
+      "grad_norm": 0.6853231999065038,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.3878,
+      "step": 498
+    },
+    {
+      "epoch": 2.0243037974683546,
+      "grad_norm": 0.5913850099382937,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3967,
+      "step": 499
+    },
+    {
+      "epoch": 2.0283544303797467,
+      "grad_norm": 0.6069285981635788,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.3919,
+      "step": 500
+    },
+    {
+      "epoch": 2.0324050632911392,
+      "grad_norm": 0.6356311770384953,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.4067,
+      "step": 501
+    },
+    {
+      "epoch": 2.0364556962025318,
+      "grad_norm": 0.5038325551288915,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.379,
+      "step": 502
+    },
+    {
+      "epoch": 2.0405063291139243,
+      "grad_norm": 0.6865580995597887,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.4053,
+      "step": 503
+    },
+    {
+      "epoch": 2.0445569620253163,
+      "grad_norm": 0.5005474912737347,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.3754,
+      "step": 504
+    },
+    {
+      "epoch": 2.048607594936709,
+      "grad_norm": 0.681976159336387,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.3863,
+      "step": 505
+    },
+    {
+      "epoch": 2.0526582278481014,
+      "grad_norm": 0.5547414441097807,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.3982,
+      "step": 506
+    },
+    {
+      "epoch": 2.056708860759494,
+      "grad_norm": 0.6389921586058439,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.3908,
+      "step": 507
+    },
+    {
+      "epoch": 2.060759493670886,
+      "grad_norm": 0.5161075284347416,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.3924,
+      "step": 508
+    },
+    {
+      "epoch": 2.0648101265822785,
+      "grad_norm": 0.5441222894123788,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.4179,
+      "step": 509
+    },
+    {
+      "epoch": 2.068860759493671,
+      "grad_norm": 0.41563552280033367,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.3921,
+      "step": 510
+    },
+    {
+      "epoch": 2.0729113924050635,
+      "grad_norm": 0.4880882613326981,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.4013,
+      "step": 511
+    },
+    {
+      "epoch": 2.0769620253164556,
+      "grad_norm": 0.44397552579695876,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.3642,
+      "step": 512
+    },
+    {
+      "epoch": 2.081012658227848,
+      "grad_norm": 0.4797546483698974,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.4064,
+      "step": 513
+    },
+    {
+      "epoch": 2.0850632911392406,
+      "grad_norm": 0.5127815449192018,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.3676,
+      "step": 514
+    },
+    {
+      "epoch": 2.089113924050633,
+      "grad_norm": 0.4950957787934979,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.4093,
+      "step": 515
+    },
+    {
+      "epoch": 2.093164556962025,
+      "grad_norm": 0.4537263871687868,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.3993,
+      "step": 516
+    },
+    {
+      "epoch": 2.0972151898734177,
+      "grad_norm": 0.45104467487879774,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.3918,
+      "step": 517
+    },
+    {
+      "epoch": 2.1012658227848102,
+      "grad_norm": 0.44385147174944517,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.3898,
+      "step": 518
+    },
+    {
+      "epoch": 2.1053164556962027,
+      "grad_norm": 0.5598339088716956,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.371,
+      "step": 519
+    },
+    {
+      "epoch": 2.109367088607595,
+      "grad_norm": 0.4447506830454272,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.3996,
+      "step": 520
+    },
+    {
+      "epoch": 2.1134177215189873,
+      "grad_norm": 0.5113949869371613,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.3986,
+      "step": 521
+    },
+    {
+      "epoch": 2.11746835443038,
+      "grad_norm": 0.48227461313986875,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.3838,
+      "step": 522
+    },
+    {
+      "epoch": 2.1215189873417724,
+      "grad_norm": 0.4513279819999299,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.3697,
+      "step": 523
+    },
+    {
+      "epoch": 2.1255696202531644,
+      "grad_norm": 0.4675551361032135,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.3822,
+      "step": 524
+    },
+    {
+      "epoch": 2.129620253164557,
+      "grad_norm": 0.46778256562187137,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.3776,
+      "step": 525
+    },
+    {
+      "epoch": 2.1336708860759495,
+      "grad_norm": 0.5244347597469635,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.3984,
+      "step": 526
+    },
+    {
+      "epoch": 2.137721518987342,
+      "grad_norm": 0.45994970717067635,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.3775,
+      "step": 527
+    },
+    {
+      "epoch": 2.141772151898734,
+      "grad_norm": 0.5072623726124316,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.3965,
+      "step": 528
+    },
+    {
+      "epoch": 2.1458227848101266,
+      "grad_norm": 0.5024473113012162,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.3983,
+      "step": 529
+    },
+    {
+      "epoch": 2.149873417721519,
+      "grad_norm": 0.47315721586071147,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.3667,
+      "step": 530
+    },
+    {
+      "epoch": 2.1539240506329116,
+      "grad_norm": 0.4668806893379437,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.3775,
+      "step": 531
+    },
+    {
+      "epoch": 2.1579746835443037,
+      "grad_norm": 0.5108696702634385,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.3735,
+      "step": 532
+    },
+    {
+      "epoch": 2.162025316455696,
+      "grad_norm": 0.4738066287070088,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3868,
+      "step": 533
+    },
+    {
+      "epoch": 2.1660759493670887,
+      "grad_norm": 0.4809141496518958,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.3898,
+      "step": 534
+    },
+    {
+      "epoch": 2.170126582278481,
+      "grad_norm": 0.4835350924134697,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.39,
+      "step": 535
+    },
+    {
+      "epoch": 2.1741772151898733,
+      "grad_norm": 0.49149518118019836,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.3855,
+      "step": 536
+    },
+    {
+      "epoch": 2.178227848101266,
+      "grad_norm": 0.501047955794892,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.4131,
+      "step": 537
+    },
+    {
+      "epoch": 2.1822784810126583,
+      "grad_norm": 0.4711847175537035,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.3951,
+      "step": 538
+    },
+    {
+      "epoch": 2.186329113924051,
+      "grad_norm": 0.5139799274083784,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.3984,
+      "step": 539
+    },
+    {
+      "epoch": 2.190379746835443,
+      "grad_norm": 0.4384842960558496,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.3713,
+      "step": 540
+    },
+    {
+      "epoch": 2.1944303797468354,
+      "grad_norm": 0.481235730154064,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.4035,
+      "step": 541
+    },
+    {
+      "epoch": 2.198481012658228,
+      "grad_norm": 0.4410935979469798,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.3915,
+      "step": 542
+    },
+    {
+      "epoch": 2.2025316455696204,
+      "grad_norm": 0.4509206000580702,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.3778,
+      "step": 543
+    },
+    {
+      "epoch": 2.2065822784810125,
+      "grad_norm": 0.45627484601422313,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.374,
+      "step": 544
+    },
+    {
+      "epoch": 2.210632911392405,
+      "grad_norm": 0.4606147152500946,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.3846,
+      "step": 545
+    },
+    {
+      "epoch": 2.2146835443037975,
+      "grad_norm": 0.4555979084182703,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.3917,
+      "step": 546
+    },
+    {
+      "epoch": 2.21873417721519,
+      "grad_norm": 0.45133257490755146,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.4015,
+      "step": 547
+    },
+    {
+      "epoch": 2.222784810126582,
+      "grad_norm": 0.44660280498418903,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.4146,
+      "step": 548
+    },
+    {
+      "epoch": 2.2268354430379746,
+      "grad_norm": 0.45216945068927766,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.414,
+      "step": 549
+    },
+    {
+      "epoch": 2.230886075949367,
+      "grad_norm": 0.47708074734733863,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.3959,
+      "step": 550
+    },
+    {
+      "epoch": 2.2349367088607597,
+      "grad_norm": 0.4835095364901377,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.3922,
+      "step": 551
+    },
+    {
+      "epoch": 2.2389873417721518,
+      "grad_norm": 0.5311048950608456,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.4136,
+      "step": 552
+    },
+    {
+      "epoch": 2.2430379746835443,
+      "grad_norm": 0.5148565583032109,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.4116,
+      "step": 553
+    },
+    {
+      "epoch": 2.247088607594937,
+      "grad_norm": 0.48211019530558086,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.3911,
+      "step": 554
+    },
+    {
+      "epoch": 2.2511392405063293,
+      "grad_norm": 0.6212399090912746,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.3861,
+      "step": 555
+    },
+    {
+      "epoch": 2.2551898734177214,
+      "grad_norm": 0.45227007723325685,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.3879,
+      "step": 556
+    },
+    {
+      "epoch": 2.259240506329114,
+      "grad_norm": 0.5707629067117064,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.368,
+      "step": 557
+    },
+    {
+      "epoch": 2.2632911392405064,
+      "grad_norm": 0.4590355144959162,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.3877,
+      "step": 558
+    },
+    {
+      "epoch": 2.267341772151899,
+      "grad_norm": 0.5279066860249602,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.3863,
+      "step": 559
+    },
+    {
+      "epoch": 2.271392405063291,
+      "grad_norm": 0.45998964753787985,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.3935,
+      "step": 560
+    },
+    {
+      "epoch": 2.2754430379746835,
+      "grad_norm": 0.5172350524623377,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.4043,
+      "step": 561
+    },
+    {
+      "epoch": 2.279493670886076,
+      "grad_norm": 0.45425088281955867,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.3877,
+      "step": 562
+    },
+    {
+      "epoch": 2.2835443037974685,
+      "grad_norm": 0.48948963323680456,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.4095,
+      "step": 563
+    },
+    {
+      "epoch": 2.2875949367088606,
+      "grad_norm": 0.4565626404106166,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.412,
+      "step": 564
+    },
+    {
+      "epoch": 2.291645569620253,
+      "grad_norm": 0.47708231774260873,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.3758,
+      "step": 565
+    },
+    {
+      "epoch": 2.2956962025316456,
+      "grad_norm": 0.4069832349133339,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.4155,
+      "step": 566
+    },
+    {
+      "epoch": 2.299746835443038,
+      "grad_norm": 0.4790214167841527,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.3819,
+      "step": 567
+    },
+    {
+      "epoch": 2.3037974683544302,
+      "grad_norm": 0.44060812227486396,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.3725,
+      "step": 568
+    },
+    {
+      "epoch": 2.3078481012658227,
+      "grad_norm": 0.45717947584307406,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.3828,
+      "step": 569
+    },
+    {
+      "epoch": 2.3118987341772153,
+      "grad_norm": 0.4500685466139446,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.3942,
+      "step": 570
+    },
+    {
+      "epoch": 2.3159493670886078,
+      "grad_norm": 0.4357591591492838,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.4066,
+      "step": 571
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.4325944163342697,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.4028,
+      "step": 572
+    },
+    {
+      "epoch": 2.3240506329113924,
+      "grad_norm": 0.4659884491027717,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.4047,
+      "step": 573
+    },
+    {
+      "epoch": 2.328101265822785,
+      "grad_norm": 0.43358437418683665,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.4083,
+      "step": 574
+    },
+    {
+      "epoch": 2.3321518987341774,
+      "grad_norm": 0.5017241251469876,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.3711,
+      "step": 575
+    },
+    {
+      "epoch": 2.3362025316455695,
+      "grad_norm": 0.4484993698889143,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.3855,
+      "step": 576
+    },
+    {
+      "epoch": 2.340253164556962,
+      "grad_norm": 0.47139175349232665,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.3716,
+      "step": 577
+    },
+    {
+      "epoch": 2.3443037974683545,
+      "grad_norm": 0.43261151755724714,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.3833,
+      "step": 578
+    },
+    {
+      "epoch": 2.348354430379747,
+      "grad_norm": 0.4333069955717402,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.4226,
+      "step": 579
+    },
+    {
+      "epoch": 2.352405063291139,
+      "grad_norm": 0.4642767300874484,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.3772,
+      "step": 580
+    },
+    {
+      "epoch": 2.3564556962025316,
+      "grad_norm": 0.4537024188547904,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.367,
+      "step": 581
+    },
+    {
+      "epoch": 2.360506329113924,
+      "grad_norm": 0.45316836601649935,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.3779,
+      "step": 582
+    },
+    {
+      "epoch": 2.3645569620253166,
+      "grad_norm": 0.42031272774196715,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.3817,
+      "step": 583
+    },
+    {
+      "epoch": 2.3686075949367087,
+      "grad_norm": 0.4264855463434957,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.4083,
+      "step": 584
+    },
+    {
+      "epoch": 2.372658227848101,
+      "grad_norm": 0.4630647375280017,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.3854,
+      "step": 585
+    },
+    {
+      "epoch": 2.3767088607594937,
+      "grad_norm": 0.4702493207764152,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.3872,
+      "step": 586
+    },
+    {
+      "epoch": 2.3807594936708862,
+      "grad_norm": 0.4594459794366067,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.3983,
+      "step": 587
+    },
+    {
+      "epoch": 2.3848101265822783,
+      "grad_norm": 0.46043330319591,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.3897,
+      "step": 588
+    },
+    {
+      "epoch": 2.388860759493671,
+      "grad_norm": 0.434522577241602,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.4022,
+      "step": 589
+    },
+    {
+      "epoch": 2.3929113924050633,
+      "grad_norm": 0.4090545631989646,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.3885,
+      "step": 590
+    },
+    {
+      "epoch": 2.396962025316456,
+      "grad_norm": 0.4427847663530239,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.3973,
+      "step": 591
+    },
+    {
+      "epoch": 2.401012658227848,
+      "grad_norm": 0.4164015259537531,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.394,
+      "step": 592
+    },
+    {
+      "epoch": 2.4050632911392404,
+      "grad_norm": 0.43677726007050877,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.4093,
+      "step": 593
+    },
+    {
+      "epoch": 2.409113924050633,
+      "grad_norm": 0.41874787203549635,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.371,
+      "step": 594
+    },
+    {
+      "epoch": 2.4131645569620255,
+      "grad_norm": 0.4197181595179639,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.3953,
+      "step": 595
+    },
+    {
+      "epoch": 2.4172151898734175,
+      "grad_norm": 0.42843899838848026,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.388,
+      "step": 596
+    },
+    {
+      "epoch": 2.42126582278481,
+      "grad_norm": 0.4419014022889296,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.3871,
+      "step": 597
+    },
+    {
+      "epoch": 2.4253164556962026,
+      "grad_norm": 0.3954124179079982,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.4088,
+      "step": 598
+    },
+    {
+      "epoch": 2.429367088607595,
+      "grad_norm": 0.43015335502920293,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.4011,
+      "step": 599
+    },
+    {
+      "epoch": 2.433417721518987,
+      "grad_norm": 0.4884258639889432,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.3808,
+      "step": 600
+    },
+    {
+      "epoch": 2.4374683544303797,
+      "grad_norm": 0.43251139530572097,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.3847,
+      "step": 601
+    },
+    {
+      "epoch": 2.441518987341772,
+      "grad_norm": 0.4025618429030094,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.3984,
+      "step": 602
+    },
+    {
+      "epoch": 2.4455696202531647,
+      "grad_norm": 0.4416892463705104,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.3917,
+      "step": 603
+    },
+    {
+      "epoch": 2.449620253164557,
+      "grad_norm": 0.4075200930045322,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.411,
+      "step": 604
+    },
+    {
+      "epoch": 2.4536708860759493,
+      "grad_norm": 0.44198029419579876,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.4024,
+      "step": 605
+    },
+    {
+      "epoch": 2.457721518987342,
+      "grad_norm": 0.45774197296604835,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.3999,
+      "step": 606
+    },
+    {
+      "epoch": 2.4617721518987343,
+      "grad_norm": 0.4198534229547119,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.3822,
+      "step": 607
+    },
+    {
+      "epoch": 2.4658227848101264,
+      "grad_norm": 0.49557431444004674,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.3951,
+      "step": 608
+    },
+    {
+      "epoch": 2.469873417721519,
+      "grad_norm": 0.42516221789667136,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.3988,
+      "step": 609
+    },
+    {
+      "epoch": 2.4739240506329114,
+      "grad_norm": 0.4690613344343872,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.3978,
+      "step": 610
+    },
+    {
+      "epoch": 2.477974683544304,
+      "grad_norm": 0.4557621258509547,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.3921,
+      "step": 611
+    },
+    {
+      "epoch": 2.482025316455696,
+      "grad_norm": 0.44308848843740634,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.401,
+      "step": 612
+    },
+    {
+      "epoch": 2.4860759493670885,
+      "grad_norm": 0.4154064116087828,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.3856,
+      "step": 613
+    },
+    {
+      "epoch": 2.490126582278481,
+      "grad_norm": 0.42501076305642127,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.3793,
+      "step": 614
+    },
+    {
+      "epoch": 2.4941772151898736,
+      "grad_norm": 0.4407759356914818,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3821,
+      "step": 615
+    },
+    {
+      "epoch": 2.4982278481012656,
+      "grad_norm": 0.43284932252387687,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.3615,
+      "step": 616
+    },
+    {
+      "epoch": 2.502278481012658,
+      "grad_norm": 0.43012322851919943,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.3964,
+      "step": 617
+    },
+    {
+      "epoch": 2.5063291139240507,
+      "grad_norm": 0.4316828770107773,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.4057,
+      "step": 618
+    },
+    {
+      "epoch": 2.510379746835443,
+      "grad_norm": 0.4169905018452063,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.3959,
+      "step": 619
+    },
+    {
+      "epoch": 2.5144303797468357,
+      "grad_norm": 0.416521861511361,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.3632,
+      "step": 620
+    },
+    {
+      "epoch": 2.5184810126582278,
+      "grad_norm": 0.5235280029889751,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.4019,
+      "step": 621
+    },
+    {
+      "epoch": 2.5225316455696203,
+      "grad_norm": 0.41113672750317104,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.3933,
+      "step": 622
+    },
+    {
+      "epoch": 2.526582278481013,
+      "grad_norm": 0.43266531943131686,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.4139,
+      "step": 623
+    },
+    {
+      "epoch": 2.530632911392405,
+      "grad_norm": 0.41967147962784923,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.3917,
+      "step": 624
+    },
+    {
+      "epoch": 2.5346835443037974,
+      "grad_norm": 0.41221245645771754,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.4107,
+      "step": 625
+    },
+    {
+      "epoch": 2.53873417721519,
+      "grad_norm": 0.45001276706940246,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.4095,
+      "step": 626
+    },
+    {
+      "epoch": 2.5427848101265824,
+      "grad_norm": 0.42171039459370785,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.3864,
+      "step": 627
+    },
+    {
+      "epoch": 2.546835443037975,
+      "grad_norm": 0.4838750079018857,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.3925,
+      "step": 628
+    },
+    {
+      "epoch": 2.550886075949367,
+      "grad_norm": 0.44912649347226785,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.3802,
+      "step": 629
+    },
+    {
+      "epoch": 2.5549367088607595,
+      "grad_norm": 0.46709381051616566,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.3844,
+      "step": 630
+    },
+    {
+      "epoch": 2.558987341772152,
+      "grad_norm": 0.4558564857662176,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.3795,
+      "step": 631
+    },
+    {
+      "epoch": 2.563037974683544,
+      "grad_norm": 0.4449683903309659,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.4159,
+      "step": 632
+    },
+    {
+      "epoch": 2.5670886075949366,
+      "grad_norm": 0.4346247420678063,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.3852,
+      "step": 633
+    },
+    {
+      "epoch": 2.571139240506329,
+      "grad_norm": 0.4519576315283365,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.4102,
+      "step": 634
+    },
+    {
+      "epoch": 2.5751898734177217,
+      "grad_norm": 0.4674131866780374,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.3728,
+      "step": 635
+    },
+    {
+      "epoch": 2.579240506329114,
+      "grad_norm": 0.49843060446084286,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.381,
+      "step": 636
+    },
+    {
+      "epoch": 2.5832911392405062,
+      "grad_norm": 0.5478822380491116,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.4055,
+      "step": 637
+    },
+    {
+      "epoch": 2.5873417721518988,
+      "grad_norm": 0.4430776651722836,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.4078,
+      "step": 638
+    },
+    {
+      "epoch": 2.5913924050632913,
+      "grad_norm": 0.49112215841605866,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.4185,
+      "step": 639
+    },
+    {
+      "epoch": 2.5954430379746833,
+      "grad_norm": 0.43015744516585064,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.3901,
+      "step": 640
+    },
+    {
+      "epoch": 2.599493670886076,
+      "grad_norm": 0.4347448264648089,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.3771,
+      "step": 641
+    },
+    {
+      "epoch": 2.6035443037974684,
+      "grad_norm": 0.4611057603484489,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.3881,
+      "step": 642
+    },
+    {
+      "epoch": 2.607594936708861,
+      "grad_norm": 0.4276666115924056,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.383,
+      "step": 643
+    },
+    {
+      "epoch": 2.6116455696202534,
+      "grad_norm": 0.4344662277755495,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.3925,
+      "step": 644
+    },
+    {
+      "epoch": 2.6156962025316455,
+      "grad_norm": 0.42748559558485977,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.4089,
+      "step": 645
+    },
+    {
+      "epoch": 2.619746835443038,
+      "grad_norm": 0.4640040288733926,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.4087,
+      "step": 646
+    },
+    {
+      "epoch": 2.6237974683544305,
+      "grad_norm": 0.4351891553663807,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.3882,
+      "step": 647
+    },
+    {
+      "epoch": 2.6278481012658226,
+      "grad_norm": 0.4304207909454189,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.3963,
+      "step": 648
+    },
+    {
+      "epoch": 2.631898734177215,
+      "grad_norm": 0.48993269725500765,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.394,
+      "step": 649
+    },
+    {
+      "epoch": 2.6359493670886076,
+      "grad_norm": 0.4148070309081707,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.3942,
+      "step": 650
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.44452708491669735,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.3983,
+      "step": 651
+    },
+    {
+      "epoch": 2.6440506329113926,
+      "grad_norm": 0.46140875242814755,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.3969,
+      "step": 652
+    },
+    {
+      "epoch": 2.6481012658227847,
+      "grad_norm": 0.41283116959416644,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.4056,
+      "step": 653
+    },
+    {
+      "epoch": 2.6521518987341772,
+      "grad_norm": 0.4426783805435206,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.3954,
+      "step": 654
+    },
+    {
+      "epoch": 2.6562025316455697,
+      "grad_norm": 0.4699972659190643,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.3975,
+      "step": 655
+    },
+    {
+      "epoch": 2.660253164556962,
+      "grad_norm": 0.4301919902422754,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.393,
+      "step": 656
+    },
+    {
+      "epoch": 2.6643037974683543,
+      "grad_norm": 0.4288423711108351,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.3934,
+      "step": 657
+    },
+    {
+      "epoch": 2.668354430379747,
+      "grad_norm": 0.4402454693306987,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.4091,
+      "step": 658
+    },
+    {
+      "epoch": 2.6724050632911394,
+      "grad_norm": 0.41927787370059955,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.3707,
+      "step": 659
+    },
+    {
+      "epoch": 2.676455696202532,
+      "grad_norm": 0.43202073335735464,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.4005,
+      "step": 660
+    },
+    {
+      "epoch": 2.680506329113924,
+      "grad_norm": 0.48941085793629724,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.4079,
+      "step": 661
+    },
+    {
+      "epoch": 2.6845569620253165,
+      "grad_norm": 0.38903430653011306,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.4003,
+      "step": 662
+    },
+    {
+      "epoch": 2.688607594936709,
+      "grad_norm": 0.4873827725066845,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.3706,
+      "step": 663
+    },
+    {
+      "epoch": 2.692658227848101,
+      "grad_norm": 0.4302099280536406,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3955,
+      "step": 664
+    },
+    {
+      "epoch": 2.6967088607594936,
+      "grad_norm": 0.43768757159298916,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.3729,
+      "step": 665
+    },
+    {
+      "epoch": 2.700759493670886,
+      "grad_norm": 0.45788453313594873,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.3883,
+      "step": 666
+    },
+    {
+      "epoch": 2.7048101265822786,
+      "grad_norm": 0.40825387697994564,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.3849,
+      "step": 667
+    },
+    {
+      "epoch": 2.708860759493671,
+      "grad_norm": 0.43871075598491927,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.3868,
+      "step": 668
+    },
+    {
+      "epoch": 2.712911392405063,
+      "grad_norm": 0.47990382448165897,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.3961,
+      "step": 669
+    },
+    {
+      "epoch": 2.7169620253164557,
+      "grad_norm": 0.41279990140225203,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.3939,
+      "step": 670
+    },
+    {
+      "epoch": 2.721012658227848,
+      "grad_norm": 0.43627297533411563,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.3945,
+      "step": 671
+    },
+    {
+      "epoch": 2.7250632911392403,
+      "grad_norm": 0.4228589359901463,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.3932,
+      "step": 672
+    },
+    {
+      "epoch": 2.729113924050633,
+      "grad_norm": 0.44002109378746995,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.371,
+      "step": 673
+    },
+    {
+      "epoch": 2.7331645569620253,
+      "grad_norm": 0.42698845666188806,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.4043,
+      "step": 674
+    },
+    {
+      "epoch": 2.737215189873418,
+      "grad_norm": 0.4359263058812298,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.3905,
+      "step": 675
+    },
+    {
+      "epoch": 2.7412658227848103,
+      "grad_norm": 0.4455292125484589,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.3892,
+      "step": 676
+    },
+    {
+      "epoch": 2.7453164556962024,
+      "grad_norm": 0.4622918372101659,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.3996,
+      "step": 677
+    },
+    {
+      "epoch": 2.749367088607595,
+      "grad_norm": 0.4665212623030395,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.3666,
+      "step": 678
+    },
+    {
+      "epoch": 2.7534177215189874,
+      "grad_norm": 0.46764752809888743,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.3871,
+      "step": 679
+    },
+    {
+      "epoch": 2.7574683544303795,
+      "grad_norm": 0.4601642607079037,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.4016,
+      "step": 680
+    },
+    {
+      "epoch": 2.761518987341772,
+      "grad_norm": 0.46053791382746717,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.3822,
+      "step": 681
+    },
+    {
+      "epoch": 2.7655696202531646,
+      "grad_norm": 0.4299780511919403,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.3769,
+      "step": 682
+    },
+    {
+      "epoch": 2.769620253164557,
+      "grad_norm": 0.44187277390232976,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.4083,
+      "step": 683
+    },
+    {
+      "epoch": 2.7736708860759496,
+      "grad_norm": 0.43431157834181733,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.3852,
+      "step": 684
+    },
+    {
+      "epoch": 2.7777215189873417,
+      "grad_norm": 0.4558721533557596,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.4087,
+      "step": 685
+    },
+    {
+      "epoch": 2.781772151898734,
+      "grad_norm": 0.407672391045265,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.4051,
+      "step": 686
+    },
+    {
+      "epoch": 2.7858227848101267,
+      "grad_norm": 0.4337299088082281,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.4075,
+      "step": 687
+    },
+    {
+      "epoch": 2.7898734177215188,
+      "grad_norm": 0.45754185252346047,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.3996,
+      "step": 688
+    },
+    {
+      "epoch": 2.7939240506329113,
+      "grad_norm": 0.4004906822106136,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.3797,
+      "step": 689
+    },
+    {
+      "epoch": 2.797974683544304,
+      "grad_norm": 0.46415499303332136,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.3837,
+      "step": 690
+    },
+    {
+      "epoch": 2.8020253164556963,
+      "grad_norm": 0.4445560515617116,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.3906,
+      "step": 691
+    },
+    {
+      "epoch": 2.806075949367089,
+      "grad_norm": 0.39864905710932264,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.4166,
+      "step": 692
+    },
+    {
+      "epoch": 2.810126582278481,
+      "grad_norm": 0.46884525117995846,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.3812,
+      "step": 693
+    },
+    {
+      "epoch": 2.8141772151898734,
+      "grad_norm": 0.4293911709412662,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.4098,
+      "step": 694
+    },
+    {
+      "epoch": 2.818227848101266,
+      "grad_norm": 0.44890088114038396,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.3882,
+      "step": 695
+    },
+    {
+      "epoch": 2.822278481012658,
+      "grad_norm": 0.4440829884840836,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.3808,
+      "step": 696
+    },
+    {
+      "epoch": 2.8263291139240505,
+      "grad_norm": 0.44634982286735647,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.4044,
+      "step": 697
+    },
+    {
+      "epoch": 2.830379746835443,
+      "grad_norm": 0.4407962443031587,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.4021,
+      "step": 698
+    },
+    {
+      "epoch": 2.8344303797468355,
+      "grad_norm": 0.3954343057793035,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.4031,
+      "step": 699
+    },
+    {
+      "epoch": 2.838481012658228,
+      "grad_norm": 0.4568167592522591,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.3787,
+      "step": 700
+    },
+    {
+      "epoch": 2.84253164556962,
+      "grad_norm": 0.4457221628324438,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.3955,
+      "step": 701
+    },
+    {
+      "epoch": 2.8465822784810126,
+      "grad_norm": 0.45195533839458196,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.3963,
+      "step": 702
+    },
+    {
+      "epoch": 2.850632911392405,
+      "grad_norm": 0.4297751476556122,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.4051,
+      "step": 703
+    },
+    {
+      "epoch": 2.8546835443037972,
+      "grad_norm": 0.47639389087600453,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.3708,
+      "step": 704
+    },
+    {
+      "epoch": 2.8587341772151897,
+      "grad_norm": 0.45019669412014723,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.3842,
+      "step": 705
+    },
+    {
+      "epoch": 2.8627848101265823,
+      "grad_norm": 0.3988530886438395,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.3888,
+      "step": 706
+    },
+    {
+      "epoch": 2.8668354430379748,
+      "grad_norm": 0.443878050373567,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.3944,
+      "step": 707
+    },
+    {
+      "epoch": 2.8708860759493673,
+      "grad_norm": 0.41725456619941925,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.4117,
+      "step": 708
+    },
+    {
+      "epoch": 2.8749367088607594,
+      "grad_norm": 0.41761377177922965,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.4004,
+      "step": 709
+    },
+    {
+      "epoch": 2.878987341772152,
+      "grad_norm": 0.4195137408568746,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.3734,
+      "step": 710
+    },
+    {
+      "epoch": 2.8830379746835444,
+      "grad_norm": 0.43921828647058186,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.4189,
+      "step": 711
+    },
+    {
+      "epoch": 2.8870886075949365,
+      "grad_norm": 0.43126508371576333,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.3555,
+      "step": 712
+    },
+    {
+      "epoch": 2.891139240506329,
+      "grad_norm": 0.4943426565470045,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.4066,
+      "step": 713
+    },
+    {
+      "epoch": 2.8951898734177215,
+      "grad_norm": 0.40353845780824904,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.4084,
+      "step": 714
+    },
+    {
+      "epoch": 2.899240506329114,
+      "grad_norm": 0.4575407188347182,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.3853,
+      "step": 715
+    },
+    {
+      "epoch": 2.9032911392405065,
+      "grad_norm": 0.43453278180167887,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.3747,
+      "step": 716
+    },
+    {
+      "epoch": 2.9073417721518986,
+      "grad_norm": 0.39270456641834567,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.4083,
+      "step": 717
+    },
+    {
+      "epoch": 2.911392405063291,
+      "grad_norm": 0.4424618929886549,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.4088,
+      "step": 718
+    },
+    {
+      "epoch": 2.9154430379746836,
+      "grad_norm": 0.38925245283993193,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.412,
+      "step": 719
+    },
+    {
+      "epoch": 2.9194936708860757,
+      "grad_norm": 0.4170840843023451,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.3949,
+      "step": 720
+    },
+    {
+      "epoch": 2.923544303797468,
+      "grad_norm": 0.42133957945066003,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.419,
+      "step": 721
+    },
+    {
+      "epoch": 2.9275949367088607,
+      "grad_norm": 0.40700688312830846,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.3991,
+      "step": 722
+    },
+    {
+      "epoch": 2.9316455696202532,
+      "grad_norm": 0.42946595526049786,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.3997,
+      "step": 723
+    },
+    {
+      "epoch": 2.9356962025316458,
+      "grad_norm": 0.4085771533822731,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.3961,
+      "step": 724
+    },
+    {
+      "epoch": 2.939746835443038,
+      "grad_norm": 0.4065738367889194,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.3872,
+      "step": 725
+    },
+    {
+      "epoch": 2.9437974683544303,
+      "grad_norm": 0.40241848881277026,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.3601,
+      "step": 726
+    },
+    {
+      "epoch": 2.947848101265823,
+      "grad_norm": 0.40505956613616867,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.3791,
+      "step": 727
+    },
+    {
+      "epoch": 2.951898734177215,
+      "grad_norm": 0.4304995759277829,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.3652,
+      "step": 728
+    },
+    {
+      "epoch": 2.9559493670886074,
+      "grad_norm": 0.40774627731419544,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.3728,
+      "step": 729
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.4032891859295984,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.3912,
+      "step": 730
+    },
+    {
+      "epoch": 2.9640506329113925,
+      "grad_norm": 0.4036306719123583,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.4085,
+      "step": 731
+    },
+    {
+      "epoch": 2.968101265822785,
+      "grad_norm": 0.4075908502595599,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.3976,
+      "step": 732
+    },
+    {
+      "epoch": 2.972151898734177,
+      "grad_norm": 0.40402377661452554,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.3789,
+      "step": 733
+    },
+    {
+      "epoch": 2.9762025316455696,
+      "grad_norm": 0.46527230093525923,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.3864,
+      "step": 734
+    },
+    {
+      "epoch": 2.980253164556962,
+      "grad_norm": 0.4231334061498329,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.3705,
+      "step": 735
+    },
+    {
+      "epoch": 2.984303797468354,
+      "grad_norm": 0.41471171881953056,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.4048,
+      "step": 736
+    },
+    {
+      "epoch": 2.9883544303797467,
+      "grad_norm": 0.4198804954466373,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.3876,
+      "step": 737
+    },
+    {
+      "epoch": 2.992405063291139,
+      "grad_norm": 0.39298021885346673,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3988,
+      "step": 738
+    },
+    {
+      "epoch": 2.9964556962025317,
+      "grad_norm": 0.42419080433929407,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.3963,
+      "step": 739
+    },
+    {
+      "epoch": 3.002025316455696,
+      "grad_norm": 0.43908385085527163,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.3446,
+      "step": 740
+    },
+    {
+      "epoch": 3.0060759493670886,
+      "grad_norm": 0.5255044143129526,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.2796,
+      "step": 741
+    },
+    {
+      "epoch": 3.010126582278481,
+      "grad_norm": 0.5823567409360799,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.2844,
+      "step": 742
+    },
+    {
+      "epoch": 3.0141772151898736,
+      "grad_norm": 0.7124216882637842,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.2936,
+      "step": 743
+    },
+    {
+      "epoch": 3.0182278481012657,
+      "grad_norm": 0.4604440106840404,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.283,
+      "step": 744
+    },
+    {
+      "epoch": 3.022278481012658,
+      "grad_norm": 0.6920005018918701,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.3027,
+      "step": 745
+    },
+    {
+      "epoch": 3.0263291139240507,
+      "grad_norm": 0.507345739800152,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.2885,
+      "step": 746
+    },
+    {
+      "epoch": 3.030379746835443,
+      "grad_norm": 0.504197364616585,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.295,
+      "step": 747
+    },
+    {
+      "epoch": 3.0344303797468353,
+      "grad_norm": 0.4939188925092429,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.2745,
+      "step": 748
+    },
+    {
+      "epoch": 3.038481012658228,
+      "grad_norm": 0.4525730539659192,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.2603,
+      "step": 749
+    },
+    {
+      "epoch": 3.0425316455696203,
+      "grad_norm": 0.4451600677507954,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.3087,
+      "step": 750
+    },
+    {
+      "epoch": 3.046582278481013,
+      "grad_norm": 0.441789284939408,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.3016,
+      "step": 751
+    },
+    {
+      "epoch": 3.050632911392405,
+      "grad_norm": 0.4267012645923492,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.2782,
+      "step": 752
+    },
+    {
+      "epoch": 3.0546835443037974,
+      "grad_norm": 0.45730562221231846,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.3023,
+      "step": 753
+    },
+    {
+      "epoch": 3.05873417721519,
+      "grad_norm": 0.4429522989423884,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.3036,
+      "step": 754
+    },
+    {
+      "epoch": 3.0627848101265824,
+      "grad_norm": 0.418026955177608,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.2977,
+      "step": 755
+    },
+    {
+      "epoch": 3.0668354430379745,
+      "grad_norm": 0.4383016526980005,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.294,
+      "step": 756
+    },
+    {
+      "epoch": 3.070886075949367,
+      "grad_norm": 0.3825704043554266,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.3092,
+      "step": 757
+    },
+    {
+      "epoch": 3.0749367088607595,
+      "grad_norm": 0.4124252180504131,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.3062,
+      "step": 758
+    },
+    {
+      "epoch": 3.078987341772152,
+      "grad_norm": 0.4048492818291974,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.3035,
+      "step": 759
+    },
+    {
+      "epoch": 3.083037974683544,
+      "grad_norm": 0.42253564675559985,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.32,
+      "step": 760
+    },
+    {
+      "epoch": 3.0870886075949366,
+      "grad_norm": 0.41325827293567313,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.2822,
+      "step": 761
+    },
+    {
+      "epoch": 3.091139240506329,
+      "grad_norm": 0.4506809972640762,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.2772,
+      "step": 762
+    },
+    {
+      "epoch": 3.0951898734177217,
+      "grad_norm": 0.39057945463859384,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.2865,
+      "step": 763
+    },
+    {
+      "epoch": 3.0992405063291137,
+      "grad_norm": 0.4039411765843232,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.3096,
+      "step": 764
+    },
+    {
+      "epoch": 3.1032911392405063,
+      "grad_norm": 0.38919901860721473,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.2829,
+      "step": 765
+    },
+    {
+      "epoch": 3.1073417721518988,
+      "grad_norm": 0.3629369625391408,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.3189,
+      "step": 766
+    },
+    {
+      "epoch": 3.1113924050632913,
+      "grad_norm": 0.38208520565923804,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.2716,
+      "step": 767
+    },
+    {
+      "epoch": 3.1154430379746834,
+      "grad_norm": 0.4146829684642468,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.309,
+      "step": 768
+    },
+    {
+      "epoch": 3.119493670886076,
+      "grad_norm": 0.39004417220180737,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.298,
+      "step": 769
+    },
+    {
+      "epoch": 3.1235443037974684,
+      "grad_norm": 0.3972084834572719,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.2898,
+      "step": 770
+    },
+    {
+      "epoch": 3.127594936708861,
+      "grad_norm": 0.4242278983831727,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.2657,
+      "step": 771
+    },
+    {
+      "epoch": 3.131645569620253,
+      "grad_norm": 0.3896767103989836,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.2855,
+      "step": 772
+    },
+    {
+      "epoch": 3.1356962025316455,
+      "grad_norm": 0.3996051938192635,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.3044,
+      "step": 773
+    },
+    {
+      "epoch": 3.139746835443038,
+      "grad_norm": 0.3839386019780561,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.2893,
+      "step": 774
+    },
+    {
+      "epoch": 3.1437974683544305,
+      "grad_norm": 0.38103783584801754,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.2677,
+      "step": 775
+    },
+    {
+      "epoch": 3.1478481012658226,
+      "grad_norm": 0.3938989670635469,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.2748,
+      "step": 776
+    },
+    {
+      "epoch": 3.151898734177215,
+      "grad_norm": 0.38781799241153203,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.2742,
+      "step": 777
+    },
+    {
+      "epoch": 3.1559493670886076,
+      "grad_norm": 0.36274198061481183,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.2801,
+      "step": 778
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.37381405124476624,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2935,
+      "step": 779
+    },
+    {
+      "epoch": 3.164050632911392,
+      "grad_norm": 0.41457214471019627,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.2892,
+      "step": 780
+    },
+    {
+      "epoch": 3.1681012658227847,
+      "grad_norm": 0.3755518646973323,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.2954,
+      "step": 781
+    },
+    {
+      "epoch": 3.1721518987341772,
+      "grad_norm": 0.3823886973628527,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.2995,
+      "step": 782
+    },
+    {
+      "epoch": 3.1762025316455698,
+      "grad_norm": 0.37657339160028064,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.276,
+      "step": 783
+    },
+    {
+      "epoch": 3.180253164556962,
+      "grad_norm": 0.41849266492908427,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.2754,
+      "step": 784
+    },
+    {
+      "epoch": 3.1843037974683543,
+      "grad_norm": 0.37312365147493803,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.2853,
+      "step": 785
+    },
+    {
+      "epoch": 3.188354430379747,
+      "grad_norm": 0.3957800534745301,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.2993,
+      "step": 786
+    },
+    {
+      "epoch": 3.1924050632911394,
+      "grad_norm": 0.3846468711803092,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.3065,
+      "step": 787
+    },
+    {
+      "epoch": 3.1964556962025314,
+      "grad_norm": 0.40099604224171503,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.2773,
+      "step": 788
+    },
+    {
+      "epoch": 3.200506329113924,
+      "grad_norm": 0.3802194501318767,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.3048,
+      "step": 789
+    },
+    {
+      "epoch": 3.2045569620253165,
+      "grad_norm": 0.409727988020893,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.3095,
+      "step": 790
+    },
+    {
+      "epoch": 3.208607594936709,
+      "grad_norm": 0.3732831748715226,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.2729,
+      "step": 791
+    },
+    {
+      "epoch": 3.212658227848101,
+      "grad_norm": 0.3963404731922893,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.2814,
+      "step": 792
+    },
+    {
+      "epoch": 3.2167088607594936,
+      "grad_norm": 0.3969535152895088,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.2921,
+      "step": 793
+    },
+    {
+      "epoch": 3.220759493670886,
+      "grad_norm": 0.39194227422761047,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.285,
+      "step": 794
+    },
+    {
+      "epoch": 3.2248101265822786,
+      "grad_norm": 0.40182685082431846,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.2792,
+      "step": 795
+    },
+    {
+      "epoch": 3.2288607594936707,
+      "grad_norm": 0.40956223982806805,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.2818,
+      "step": 796
+    },
+    {
+      "epoch": 3.232911392405063,
+      "grad_norm": 0.3812857397076299,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.2913,
+      "step": 797
+    },
+    {
+      "epoch": 3.2369620253164557,
+      "grad_norm": 0.3881036640680446,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.2991,
+      "step": 798
+    },
+    {
+      "epoch": 3.2410126582278482,
+      "grad_norm": 0.40656685052522684,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.2988,
+      "step": 799
+    },
+    {
+      "epoch": 3.2450632911392403,
+      "grad_norm": 0.38476209803332445,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.2621,
+      "step": 800
+    },
+    {
+      "epoch": 3.249113924050633,
+      "grad_norm": 0.38029924631952505,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.2662,
+      "step": 801
+    },
+    {
+      "epoch": 3.2531645569620253,
+      "grad_norm": 0.4150589457331111,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.3109,
+      "step": 802
+    },
+    {
+      "epoch": 3.257215189873418,
+      "grad_norm": 0.39761031875215813,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.2767,
+      "step": 803
+    },
+    {
+      "epoch": 3.26126582278481,
+      "grad_norm": 0.3760672037876478,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.2834,
+      "step": 804
+    },
+    {
+      "epoch": 3.2653164556962024,
+      "grad_norm": 0.3947523734927867,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.2716,
+      "step": 805
+    },
+    {
+      "epoch": 3.269367088607595,
+      "grad_norm": 0.4199758350947253,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.2809,
+      "step": 806
+    },
+    {
+      "epoch": 3.2734177215189875,
+      "grad_norm": 0.4015960831778925,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.2723,
+      "step": 807
+    },
+    {
+      "epoch": 3.27746835443038,
+      "grad_norm": 0.4037780432872639,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.2937,
+      "step": 808
+    },
+    {
+      "epoch": 3.281518987341772,
+      "grad_norm": 0.3781314183698989,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.2977,
+      "step": 809
+    },
+    {
+      "epoch": 3.2855696202531646,
+      "grad_norm": 0.43365361962085824,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.2956,
+      "step": 810
+    },
+    {
+      "epoch": 3.289620253164557,
+      "grad_norm": 0.40397046970715794,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.2774,
+      "step": 811
+    },
+    {
+      "epoch": 3.293670886075949,
+      "grad_norm": 0.3542379401132839,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.305,
+      "step": 812
+    },
+    {
+      "epoch": 3.2977215189873417,
+      "grad_norm": 0.3853811157397989,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.2887,
+      "step": 813
+    },
+    {
+      "epoch": 3.301772151898734,
+      "grad_norm": 0.3963353734731669,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.27,
+      "step": 814
+    },
+    {
+      "epoch": 3.3058227848101267,
+      "grad_norm": 0.3566756774605569,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.2723,
+      "step": 815
+    },
+    {
+      "epoch": 3.309873417721519,
+      "grad_norm": 0.36443771962193067,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.3268,
+      "step": 816
+    },
+    {
+      "epoch": 3.3139240506329113,
+      "grad_norm": 0.4054473994352389,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.2895,
+      "step": 817
+    },
+    {
+      "epoch": 3.317974683544304,
+      "grad_norm": 0.38512043699988346,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.2741,
+      "step": 818
+    },
+    {
+      "epoch": 3.3220253164556963,
+      "grad_norm": 0.38762902585216386,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.2838,
+      "step": 819
+    },
+    {
+      "epoch": 3.3260759493670884,
+      "grad_norm": 0.39531224819394506,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.2988,
+      "step": 820
+    },
+    {
+      "epoch": 3.330126582278481,
+      "grad_norm": 0.37002553262323523,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.2991,
+      "step": 821
+    },
+    {
+      "epoch": 3.3341772151898734,
+      "grad_norm": 0.3881367787391865,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.2767,
+      "step": 822
+    },
+    {
+      "epoch": 3.338227848101266,
+      "grad_norm": 0.4188494003170869,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.2995,
+      "step": 823
+    },
+    {
+      "epoch": 3.3422784810126585,
+      "grad_norm": 0.37873556526187435,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.2637,
+      "step": 824
+    },
+    {
+      "epoch": 3.3463291139240505,
+      "grad_norm": 0.37704511075758074,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.2838,
+      "step": 825
+    },
+    {
+      "epoch": 3.350379746835443,
+      "grad_norm": 0.4613423735073576,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.2674,
+      "step": 826
+    },
+    {
+      "epoch": 3.3544303797468356,
+      "grad_norm": 0.4224113305108096,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.2902,
+      "step": 827
+    },
+    {
+      "epoch": 3.3584810126582276,
+      "grad_norm": 0.36172489284088544,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.2978,
+      "step": 828
+    },
+    {
+      "epoch": 3.36253164556962,
+      "grad_norm": 0.4262064375311769,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.2917,
+      "step": 829
+    },
+    {
+      "epoch": 3.3665822784810127,
+      "grad_norm": 0.41212986797151924,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.2638,
+      "step": 830
+    },
+    {
+      "epoch": 3.370632911392405,
+      "grad_norm": 0.38764612423076295,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.2583,
+      "step": 831
+    },
+    {
+      "epoch": 3.3746835443037977,
+      "grad_norm": 0.3905449669514155,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.276,
+      "step": 832
+    },
+    {
+      "epoch": 3.3787341772151898,
+      "grad_norm": 0.42312600395655475,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.2827,
+      "step": 833
+    },
+    {
+      "epoch": 3.3827848101265823,
+      "grad_norm": 0.40034065780040173,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.3124,
+      "step": 834
+    },
+    {
+      "epoch": 3.386835443037975,
+      "grad_norm": 0.37606187309087247,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.2763,
+      "step": 835
+    },
+    {
+      "epoch": 3.390886075949367,
+      "grad_norm": 0.43616009707563463,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.2705,
+      "step": 836
+    },
+    {
+      "epoch": 3.3949367088607594,
+      "grad_norm": 0.42174053470070827,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.3006,
+      "step": 837
+    },
+    {
+      "epoch": 3.398987341772152,
+      "grad_norm": 0.3838737941560011,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.2968,
+      "step": 838
+    },
+    {
+      "epoch": 3.4030379746835444,
+      "grad_norm": 0.41622266080497694,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.2822,
+      "step": 839
+    },
+    {
+      "epoch": 3.407088607594937,
+      "grad_norm": 0.4138895711240638,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.2781,
+      "step": 840
+    },
+    {
+      "epoch": 3.411139240506329,
+      "grad_norm": 0.38472475762791847,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.2753,
+      "step": 841
+    },
+    {
+      "epoch": 3.4151898734177215,
+      "grad_norm": 0.4196157490687972,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.2977,
+      "step": 842
+    },
+    {
+      "epoch": 3.419240506329114,
+      "grad_norm": 0.4019207906169507,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.3003,
+      "step": 843
+    },
+    {
+      "epoch": 3.423291139240506,
+      "grad_norm": 0.3755481190347852,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.2866,
+      "step": 844
+    },
+    {
+      "epoch": 3.4273417721518986,
+      "grad_norm": 0.3751822303527841,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.2841,
+      "step": 845
+    },
+    {
+      "epoch": 3.431392405063291,
+      "grad_norm": 0.4001120921377721,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.279,
+      "step": 846
+    },
+    {
+      "epoch": 3.4354430379746836,
+      "grad_norm": 0.37396326739241365,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.2742,
+      "step": 847
+    },
+    {
+      "epoch": 3.439493670886076,
+      "grad_norm": 0.37590570980222265,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.287,
+      "step": 848
+    },
+    {
+      "epoch": 3.4435443037974682,
+      "grad_norm": 0.4093222998203027,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.2924,
+      "step": 849
+    },
+    {
+      "epoch": 3.4475949367088607,
+      "grad_norm": 0.372678492086161,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.2928,
+      "step": 850
+    },
+    {
+      "epoch": 3.4516455696202533,
+      "grad_norm": 0.3678766808699414,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.2874,
+      "step": 851
+    },
+    {
+      "epoch": 3.4556962025316453,
+      "grad_norm": 0.36906289961118627,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.278,
+      "step": 852
+    },
+    {
+      "epoch": 3.459746835443038,
+      "grad_norm": 0.35991995105902364,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.2937,
+      "step": 853
+    },
+    {
+      "epoch": 3.4637974683544304,
+      "grad_norm": 0.3706100703254147,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.2974,
+      "step": 854
+    },
+    {
+      "epoch": 3.467848101265823,
+      "grad_norm": 0.3675100268892899,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.2742,
+      "step": 855
+    },
+    {
+      "epoch": 3.4718987341772154,
+      "grad_norm": 0.373126150191338,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.2783,
+      "step": 856
+    },
+    {
+      "epoch": 3.4759493670886075,
+      "grad_norm": 0.358364277495587,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.2881,
+      "step": 857
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.3769801372008218,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.2977,
+      "step": 858
+    },
+    {
+      "epoch": 3.4840506329113925,
+      "grad_norm": 0.380199486911561,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.2723,
+      "step": 859
+    },
+    {
+      "epoch": 3.4881012658227846,
+      "grad_norm": 0.36499130822470877,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.2923,
+      "step": 860
+    },
+    {
+      "epoch": 3.492151898734177,
+      "grad_norm": 0.3707455423819492,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.2719,
+      "step": 861
+    },
+    {
+      "epoch": 3.4962025316455696,
+      "grad_norm": 0.36879688226473517,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.3051,
+      "step": 862
+    },
+    {
+      "epoch": 3.500253164556962,
+      "grad_norm": 0.3506828864149294,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.3049,
+      "step": 863
+    },
+    {
+      "epoch": 3.5043037974683546,
+      "grad_norm": 0.3592597632997508,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.2733,
+      "step": 864
+    },
+    {
+      "epoch": 3.5083544303797467,
+      "grad_norm": 0.36734564394254293,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.2857,
+      "step": 865
+    },
+    {
+      "epoch": 3.512405063291139,
+      "grad_norm": 0.3583327482834697,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.3032,
+      "step": 866
+    },
+    {
+      "epoch": 3.5164556962025317,
+      "grad_norm": 0.3680547651341837,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.2728,
+      "step": 867
+    },
+    {
+      "epoch": 3.520506329113924,
+      "grad_norm": 0.37496263101773114,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.2805,
+      "step": 868
+    },
+    {
+      "epoch": 3.5245569620253163,
+      "grad_norm": 0.3649489626197383,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.279,
+      "step": 869
+    },
+    {
+      "epoch": 3.528607594936709,
+      "grad_norm": 0.369686854936383,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.2608,
+      "step": 870
+    },
+    {
+      "epoch": 3.5326582278481014,
+      "grad_norm": 0.3776418677736665,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.2783,
+      "step": 871
+    },
+    {
+      "epoch": 3.536708860759494,
+      "grad_norm": 0.3846354324139629,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.2758,
+      "step": 872
+    },
+    {
+      "epoch": 3.540759493670886,
+      "grad_norm": 0.36289036241643063,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.2971,
+      "step": 873
+    },
+    {
+      "epoch": 3.5448101265822785,
+      "grad_norm": 0.3498994637898824,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.3009,
+      "step": 874
+    },
+    {
+      "epoch": 3.548860759493671,
+      "grad_norm": 0.3642515974986083,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.2513,
+      "step": 875
+    },
+    {
+      "epoch": 3.552911392405063,
+      "grad_norm": 0.36400683548499135,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.2838,
+      "step": 876
+    },
+    {
+      "epoch": 3.5569620253164556,
+      "grad_norm": 0.38581473260781946,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.2655,
+      "step": 877
+    },
+    {
+      "epoch": 3.561012658227848,
+      "grad_norm": 0.3805755142298751,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.2622,
+      "step": 878
+    },
+    {
+      "epoch": 3.5650632911392406,
+      "grad_norm": 0.37069857431901265,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.2855,
+      "step": 879
+    },
+    {
+      "epoch": 3.569113924050633,
+      "grad_norm": 0.37669839910944997,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.2963,
+      "step": 880
+    },
+    {
+      "epoch": 3.573164556962025,
+      "grad_norm": 0.3596735184767526,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.2922,
+      "step": 881
+    },
+    {
+      "epoch": 3.5772151898734177,
+      "grad_norm": 0.3511915520661602,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.297,
+      "step": 882
+    },
+    {
+      "epoch": 3.58126582278481,
+      "grad_norm": 0.34845715380608694,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.2816,
+      "step": 883
+    },
+    {
+      "epoch": 3.5853164556962023,
+      "grad_norm": 0.36303371534653295,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.2837,
+      "step": 884
+    },
+    {
+      "epoch": 3.589367088607595,
+      "grad_norm": 0.3676020857411387,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.2798,
+      "step": 885
+    },
+    {
+      "epoch": 3.5934177215189873,
+      "grad_norm": 0.38291815781757665,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.284,
+      "step": 886
+    },
+    {
+      "epoch": 3.59746835443038,
+      "grad_norm": 0.3689492813969087,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.2921,
+      "step": 887
+    },
+    {
+      "epoch": 3.6015189873417723,
+      "grad_norm": 0.35087874970329913,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.2855,
+      "step": 888
+    },
+    {
+      "epoch": 3.6055696202531644,
+      "grad_norm": 0.3685474439014161,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.2982,
+      "step": 889
+    },
+    {
+      "epoch": 3.609620253164557,
+      "grad_norm": 0.3784803195730656,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.2965,
+      "step": 890
+    },
+    {
+      "epoch": 3.6136708860759494,
+      "grad_norm": 0.3623511053670477,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.2758,
+      "step": 891
+    },
+    {
+      "epoch": 3.6177215189873415,
+      "grad_norm": 0.37118931240745834,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.2921,
+      "step": 892
+    },
+    {
+      "epoch": 3.621772151898734,
+      "grad_norm": 0.37192162340571144,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.3001,
+      "step": 893
+    },
+    {
+      "epoch": 3.6258227848101265,
+      "grad_norm": 0.36930452490544086,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.2979,
+      "step": 894
+    },
+    {
+      "epoch": 3.629873417721519,
+      "grad_norm": 0.3736923871835405,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.2646,
+      "step": 895
+    },
+    {
+      "epoch": 3.6339240506329116,
+      "grad_norm": 0.3670809280632642,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.2875,
+      "step": 896
+    },
+    {
+      "epoch": 3.6379746835443036,
+      "grad_norm": 0.3621396844499297,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.296,
+      "step": 897
+    },
+    {
+      "epoch": 3.642025316455696,
+      "grad_norm": 0.3530939766718466,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.2754,
+      "step": 898
+    },
+    {
+      "epoch": 3.6460759493670887,
+      "grad_norm": 0.3865183636226944,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.2661,
+      "step": 899
+    },
+    {
+      "epoch": 3.6501265822784807,
+      "grad_norm": 0.39439380841132127,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.3062,
+      "step": 900
+    },
+    {
+      "epoch": 3.6541772151898733,
+      "grad_norm": 0.36597790203822145,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.2721,
+      "step": 901
+    },
+    {
+      "epoch": 3.6582278481012658,
+      "grad_norm": 0.3466983707293298,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.2789,
+      "step": 902
+    },
+    {
+      "epoch": 3.6622784810126583,
+      "grad_norm": 0.38031028617283785,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.293,
+      "step": 903
+    },
+    {
+      "epoch": 3.666329113924051,
+      "grad_norm": 0.37438017683797037,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.2798,
+      "step": 904
+    },
+    {
+      "epoch": 3.670379746835443,
+      "grad_norm": 0.36085236872113047,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.2837,
+      "step": 905
+    },
+    {
+      "epoch": 3.6744303797468354,
+      "grad_norm": 0.36938826653548346,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.281,
+      "step": 906
+    },
+    {
+      "epoch": 3.678481012658228,
+      "grad_norm": 0.39170509608671683,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.279,
+      "step": 907
+    },
+    {
+      "epoch": 3.68253164556962,
+      "grad_norm": 0.364684900178847,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.2747,
+      "step": 908
+    },
+    {
+      "epoch": 3.6865822784810125,
+      "grad_norm": 0.37875126329325987,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.2663,
+      "step": 909
+    },
+    {
+      "epoch": 3.690632911392405,
+      "grad_norm": 0.35631010021905835,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.2795,
+      "step": 910
+    },
+    {
+      "epoch": 3.6946835443037975,
+      "grad_norm": 0.3500181586129079,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.2893,
+      "step": 911
+    },
+    {
+      "epoch": 3.69873417721519,
+      "grad_norm": 0.36649999921400717,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.2861,
+      "step": 912
+    },
+    {
+      "epoch": 3.702784810126582,
+      "grad_norm": 0.3669405834362876,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.2715,
+      "step": 913
+    },
+    {
+      "epoch": 3.7068354430379746,
+      "grad_norm": 0.3521196532247426,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.2631,
+      "step": 914
+    },
+    {
+      "epoch": 3.710886075949367,
+      "grad_norm": 0.354279214484219,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.2933,
+      "step": 915
+    },
+    {
+      "epoch": 3.714936708860759,
+      "grad_norm": 0.3729518679035856,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.2806,
+      "step": 916
+    },
+    {
+      "epoch": 3.7189873417721517,
+      "grad_norm": 0.37096302320530034,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.2778,
+      "step": 917
+    },
+    {
+      "epoch": 3.7230379746835442,
+      "grad_norm": 0.3795094955440162,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.2828,
+      "step": 918
+    },
+    {
+      "epoch": 3.7270886075949368,
+      "grad_norm": 0.3769847584887711,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.3044,
+      "step": 919
+    },
+    {
+      "epoch": 3.7311392405063293,
+      "grad_norm": 0.36013719807662486,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.2899,
+      "step": 920
+    },
+    {
+      "epoch": 3.7351898734177214,
+      "grad_norm": 0.37123879933965653,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.284,
+      "step": 921
+    },
+    {
+      "epoch": 3.739240506329114,
+      "grad_norm": 0.3775554637052661,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.298,
+      "step": 922
+    },
+    {
+      "epoch": 3.7432911392405064,
+      "grad_norm": 0.3697630186361169,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.3052,
+      "step": 923
+    },
+    {
+      "epoch": 3.747341772151899,
+      "grad_norm": 0.3529349850012205,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.2884,
+      "step": 924
+    },
+    {
+      "epoch": 3.7513924050632914,
+      "grad_norm": 0.3512870925789448,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.2918,
+      "step": 925
+    },
+    {
+      "epoch": 3.7554430379746835,
+      "grad_norm": 0.37021714447043624,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.2606,
+      "step": 926
+    },
+    {
+      "epoch": 3.759493670886076,
+      "grad_norm": 0.3637500335928117,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.3066,
+      "step": 927
+    },
+    {
+      "epoch": 3.7635443037974685,
+      "grad_norm": 0.34760488186218785,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.2732,
+      "step": 928
+    },
+    {
+      "epoch": 3.7675949367088606,
+      "grad_norm": 0.36343465986006995,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.2825,
+      "step": 929
+    },
+    {
+      "epoch": 3.771645569620253,
+      "grad_norm": 0.3704373029428519,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.2869,
+      "step": 930
+    },
+    {
+      "epoch": 3.7756962025316456,
+      "grad_norm": 0.35745704638212017,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.2914,
+      "step": 931
+    },
+    {
+      "epoch": 3.779746835443038,
+      "grad_norm": 0.3589673328583172,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.2807,
+      "step": 932
+    },
+    {
+      "epoch": 3.7837974683544306,
+      "grad_norm": 0.36407296427080027,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.2673,
+      "step": 933
+    },
+    {
+      "epoch": 3.7878481012658227,
+      "grad_norm": 0.370084338451597,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.2842,
+      "step": 934
+    },
+    {
+      "epoch": 3.7918987341772152,
+      "grad_norm": 0.37346365303828893,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.2973,
+      "step": 935
+    },
+    {
+      "epoch": 3.7959493670886078,
+      "grad_norm": 0.37242977701707614,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.2871,
+      "step": 936
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.3577031036219551,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.2693,
+      "step": 937
+    },
+    {
+      "epoch": 3.8040506329113923,
+      "grad_norm": 0.35119963005041055,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.2969,
+      "step": 938
+    },
+    {
+      "epoch": 3.808101265822785,
+      "grad_norm": 0.3594831841142267,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.2778,
+      "step": 939
+    },
+    {
+      "epoch": 3.8121518987341774,
+      "grad_norm": 0.3533549664286567,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.2806,
+      "step": 940
+    },
+    {
+      "epoch": 3.81620253164557,
+      "grad_norm": 0.347818530106605,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.3014,
+      "step": 941
+    },
+    {
+      "epoch": 3.820253164556962,
+      "grad_norm": 0.36318051923628303,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.3081,
+      "step": 942
+    },
+    {
+      "epoch": 3.8243037974683545,
+      "grad_norm": 0.3605907126159513,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2738,
+      "step": 943
+    },
+    {
+      "epoch": 3.828354430379747,
+      "grad_norm": 0.37214897573659017,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.3064,
+      "step": 944
+    },
+    {
+      "epoch": 3.832405063291139,
+      "grad_norm": 0.3579829095990994,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.2865,
+      "step": 945
+    },
+    {
+      "epoch": 3.8364556962025316,
+      "grad_norm": 0.35077424122881923,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.2633,
+      "step": 946
+    },
+    {
+      "epoch": 3.840506329113924,
+      "grad_norm": 0.3555422811531313,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.2817,
+      "step": 947
+    },
+    {
+      "epoch": 3.8445569620253166,
+      "grad_norm": 0.37483312852257133,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.2953,
+      "step": 948
+    },
+    {
+      "epoch": 3.848607594936709,
+      "grad_norm": 0.34951061057608424,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.291,
+      "step": 949
+    },
+    {
+      "epoch": 3.852658227848101,
+      "grad_norm": 0.347014083245397,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.2956,
+      "step": 950
+    },
+    {
+      "epoch": 3.8567088607594937,
+      "grad_norm": 0.3736247598829423,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.2641,
+      "step": 951
+    },
+    {
+      "epoch": 3.8607594936708862,
+      "grad_norm": 0.3663491338168844,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.2733,
+      "step": 952
+    },
+    {
+      "epoch": 3.8648101265822783,
+      "grad_norm": 0.3769045900991832,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.3062,
+      "step": 953
+    },
+    {
+      "epoch": 3.868860759493671,
+      "grad_norm": 0.3571020143172679,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.294,
+      "step": 954
+    },
+    {
+      "epoch": 3.8729113924050633,
+      "grad_norm": 0.35758127477968976,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.2705,
+      "step": 955
+    },
+    {
+      "epoch": 3.876962025316456,
+      "grad_norm": 0.36670546661055314,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.2749,
+      "step": 956
+    },
+    {
+      "epoch": 3.8810126582278484,
+      "grad_norm": 0.34768157884153467,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.3029,
+      "step": 957
+    },
+    {
+      "epoch": 3.8850632911392404,
+      "grad_norm": 0.3693197285049912,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.2725,
+      "step": 958
+    },
+    {
+      "epoch": 3.889113924050633,
+      "grad_norm": 0.3405016355488147,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.3032,
+      "step": 959
+    },
+    {
+      "epoch": 3.8931645569620255,
+      "grad_norm": 0.34856504859636095,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.2535,
+      "step": 960
+    },
+    {
+      "epoch": 3.8972151898734175,
+      "grad_norm": 0.3541894816559125,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.2894,
+      "step": 961
+    },
+    {
+      "epoch": 3.90126582278481,
+      "grad_norm": 0.3520105800327801,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.2623,
+      "step": 962
+    },
+    {
+      "epoch": 3.9053164556962026,
+      "grad_norm": 0.36267700019268156,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.2863,
+      "step": 963
+    },
+    {
+      "epoch": 3.909367088607595,
+      "grad_norm": 0.3476430189358006,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.2907,
+      "step": 964
+    },
+    {
+      "epoch": 3.9134177215189876,
+      "grad_norm": 0.347110731655305,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.2902,
+      "step": 965
+    },
+    {
+      "epoch": 3.9174683544303797,
+      "grad_norm": 0.32830691580178906,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.3012,
+      "step": 966
+    },
+    {
+      "epoch": 3.921518987341772,
+      "grad_norm": 0.3557628562212323,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.3092,
+      "step": 967
+    },
+    {
+      "epoch": 3.9255696202531647,
+      "grad_norm": 0.3419596234792064,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.2763,
+      "step": 968
+    },
+    {
+      "epoch": 3.9296202531645568,
+      "grad_norm": 0.34248645061494665,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.2809,
+      "step": 969
+    },
+    {
+      "epoch": 3.9336708860759493,
+      "grad_norm": 0.34625064275131967,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.2819,
+      "step": 970
+    },
+    {
+      "epoch": 3.937721518987342,
+      "grad_norm": 0.3528053920879547,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.2767,
+      "step": 971
+    },
+    {
+      "epoch": 3.9417721518987343,
+      "grad_norm": 0.3396835109441269,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.2687,
+      "step": 972
+    },
+    {
+      "epoch": 3.945822784810127,
+      "grad_norm": 0.3585080235304461,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.2871,
+      "step": 973
+    },
+    {
+      "epoch": 3.949873417721519,
+      "grad_norm": 0.3472697014077317,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.2847,
+      "step": 974
+    },
+    {
+      "epoch": 3.9539240506329114,
+      "grad_norm": 0.3503965267131668,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.2828,
+      "step": 975
+    },
+    {
+      "epoch": 3.957974683544304,
+      "grad_norm": 0.3610493268440032,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.2816,
+      "step": 976
+    },
+    {
+      "epoch": 3.962025316455696,
+      "grad_norm": 0.36145801267170125,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.2852,
+      "step": 977
+    },
+    {
+      "epoch": 3.9660759493670885,
+      "grad_norm": 0.3350313656438796,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.2796,
+      "step": 978
+    },
+    {
+      "epoch": 3.970126582278481,
+      "grad_norm": 0.3479550576093787,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.2784,
+      "step": 979
+    },
+    {
+      "epoch": 3.9741772151898735,
+      "grad_norm": 0.3584143649440953,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.2791,
+      "step": 980
+    },
+    {
+      "epoch": 3.978227848101266,
+      "grad_norm": 0.36674463603713553,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.2664,
+      "step": 981
+    },
+    {
+      "epoch": 3.982278481012658,
+      "grad_norm": 0.33821811631852644,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.2934,
+      "step": 982
+    },
+    {
+      "epoch": 3.9863291139240506,
+      "grad_norm": 0.3488317895325462,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.2656,
+      "step": 983
+    },
+    {
+      "epoch": 3.990379746835443,
+      "grad_norm": 0.36066652142801886,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.2708,
+      "step": 984
+    },
+    {
+      "epoch": 3.9944303797468352,
+      "grad_norm": 0.3402334633083735,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.2797,
+      "step": 985
+    },
+    {
+      "epoch": 3.9984810126582278,
+      "grad_norm": 0.34617813927317204,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.2592,
+      "step": 986
+    },
+    {
+      "epoch": 4.004050632911392,
+      "grad_norm": 0.5397574059596091,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.2216,
+      "step": 987
+    },
+    {
+      "epoch": 4.008101265822785,
+      "grad_norm": 0.450502063698476,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.2227,
+      "step": 988
+    },
+    {
+      "epoch": 4.012151898734177,
+      "grad_norm": 0.3472417913583614,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.2162,
+      "step": 989
+    },
+    {
+      "epoch": 4.01620253164557,
+      "grad_norm": 0.35814743261732945,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.2244,
+      "step": 990
+    },
+    {
+      "epoch": 4.020253164556962,
+      "grad_norm": 0.49005651732327804,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.2107,
+      "step": 991
+    },
+    {
+      "epoch": 4.024303797468354,
+      "grad_norm": 0.6012061415523625,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.2363,
+      "step": 992
+    },
+    {
+      "epoch": 4.028354430379747,
+      "grad_norm": 0.46247031967022273,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.2309,
+      "step": 993
+    },
+    {
+      "epoch": 4.032405063291139,
+      "grad_norm": 0.38418055493046016,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.2103,
+      "step": 994
+    },
+    {
+      "epoch": 4.036455696202531,
+      "grad_norm": 0.44512880088160356,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.2169,
+      "step": 995
+    },
+    {
+      "epoch": 4.040506329113924,
+      "grad_norm": 0.48821454978211687,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.2024,
+      "step": 996
+    },
+    {
+      "epoch": 4.044556962025316,
+      "grad_norm": 0.44241895234192863,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.2152,
+      "step": 997
+    },
+    {
+      "epoch": 4.048607594936709,
+      "grad_norm": 0.4079082725007011,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.2317,
+      "step": 998
+    },
+    {
+      "epoch": 4.052658227848101,
+      "grad_norm": 0.35487926014887883,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.2021,
+      "step": 999
+    },
+    {
+      "epoch": 4.056708860759493,
+      "grad_norm": 0.3803213660607845,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.2189,
+      "step": 1000
+    },
+    {
+      "epoch": 4.060759493670886,
+      "grad_norm": 0.39415416075136667,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.2149,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0648101265822785,
+      "grad_norm": 0.40334829187278265,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.2072,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0688607594936705,
+      "grad_norm": 0.4213441202377213,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.2332,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0729113924050635,
+      "grad_norm": 0.3388917507347581,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.2234,
+      "step": 1004
+    },
+    {
+      "epoch": 4.076962025316456,
+      "grad_norm": 0.35505584738765145,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.2394,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0810126582278485,
+      "grad_norm": 0.33869038098960774,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.2317,
+      "step": 1006
+    },
+    {
+      "epoch": 4.085063291139241,
+      "grad_norm": 0.35476923165611096,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.2316,
+      "step": 1007
+    },
+    {
+      "epoch": 4.089113924050633,
+      "grad_norm": 0.3288011691897205,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.2418,
+      "step": 1008
+    },
+    {
+      "epoch": 4.093164556962026,
+      "grad_norm": 0.33789611748401094,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.2146,
+      "step": 1009
+    },
+    {
+      "epoch": 4.097215189873418,
+      "grad_norm": 0.34181386648628626,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.2236,
+      "step": 1010
+    },
+    {
+      "epoch": 4.10126582278481,
+      "grad_norm": 0.3309583445375729,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.222,
+      "step": 1011
+    },
+    {
+      "epoch": 4.105316455696203,
+      "grad_norm": 0.3257048940265136,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.2256,
+      "step": 1012
+    },
+    {
+      "epoch": 4.109367088607595,
+      "grad_norm": 0.3215466515847875,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.2168,
+      "step": 1013
+    },
+    {
+      "epoch": 4.113417721518988,
+      "grad_norm": 0.32241043364485616,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.1991,
+      "step": 1014
+    },
+    {
+      "epoch": 4.11746835443038,
+      "grad_norm": 0.32872026787634934,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.1995,
+      "step": 1015
+    },
+    {
+      "epoch": 4.121518987341772,
+      "grad_norm": 0.32284711581326786,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.2129,
+      "step": 1016
+    },
+    {
+      "epoch": 4.125569620253165,
+      "grad_norm": 0.316824215223744,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.2153,
+      "step": 1017
+    },
+    {
+      "epoch": 4.129620253164557,
+      "grad_norm": 0.31539494515935246,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.2273,
+      "step": 1018
+    },
+    {
+      "epoch": 4.133670886075949,
+      "grad_norm": 0.3114404115672428,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.2226,
+      "step": 1019
+    },
+    {
+      "epoch": 4.137721518987342,
+      "grad_norm": 0.3117681794079601,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.2352,
+      "step": 1020
+    },
+    {
+      "epoch": 4.141772151898734,
+      "grad_norm": 0.3186739219343768,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.2231,
+      "step": 1021
+    },
+    {
+      "epoch": 4.145822784810127,
+      "grad_norm": 0.29977462528681253,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.2317,
+      "step": 1022
+    },
+    {
+      "epoch": 4.149873417721519,
+      "grad_norm": 0.3138870879505504,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.2104,
+      "step": 1023
+    },
+    {
+      "epoch": 4.153924050632911,
+      "grad_norm": 0.293442468555658,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.2164,
+      "step": 1024
+    },
+    {
+      "epoch": 4.157974683544304,
+      "grad_norm": 0.3141014082882016,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.2206,
+      "step": 1025
+    },
+    {
+      "epoch": 4.162025316455696,
+      "grad_norm": 0.3040905292965723,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.2521,
+      "step": 1026
+    },
+    {
+      "epoch": 4.166075949367088,
+      "grad_norm": 0.3073566956505948,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.2063,
+      "step": 1027
+    },
+    {
+      "epoch": 4.170126582278481,
+      "grad_norm": 0.31693231568480695,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.2095,
+      "step": 1028
+    },
+    {
+      "epoch": 4.174177215189873,
+      "grad_norm": 0.30432674165972057,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.2261,
+      "step": 1029
+    },
+    {
+      "epoch": 4.178227848101266,
+      "grad_norm": 0.3077870843664819,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.2109,
+      "step": 1030
+    },
+    {
+      "epoch": 4.182278481012658,
+      "grad_norm": 0.32147134985629544,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.2092,
+      "step": 1031
+    },
+    {
+      "epoch": 4.18632911392405,
+      "grad_norm": 0.30834862427059945,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.2241,
+      "step": 1032
+    },
+    {
+      "epoch": 4.190379746835443,
+      "grad_norm": 0.3091370437106792,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.235,
+      "step": 1033
+    },
+    {
+      "epoch": 4.194430379746835,
+      "grad_norm": 0.2960638438586402,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.2347,
+      "step": 1034
+    },
+    {
+      "epoch": 4.1984810126582275,
+      "grad_norm": 0.30428708247076125,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.2158,
+      "step": 1035
+    },
+    {
+      "epoch": 4.2025316455696204,
+      "grad_norm": 0.30865887973252176,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1856,
+      "step": 1036
+    },
+    {
+      "epoch": 4.2065822784810125,
+      "grad_norm": 0.2942604559757917,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.2165,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2106329113924055,
+      "grad_norm": 0.2883729686083664,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.2221,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2146835443037975,
+      "grad_norm": 0.29163153269687514,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.2172,
+      "step": 1039
+    },
+    {
+      "epoch": 4.21873417721519,
+      "grad_norm": 0.30152418896035915,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.2143,
+      "step": 1040
+    },
+    {
+      "epoch": 4.222784810126583,
+      "grad_norm": 0.3332510914471555,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.2105,
+      "step": 1041
+    },
+    {
+      "epoch": 4.226835443037975,
+      "grad_norm": 0.3049223845473395,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.2165,
+      "step": 1042
+    },
+    {
+      "epoch": 4.230886075949367,
+      "grad_norm": 0.2994381016816319,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.244,
+      "step": 1043
+    },
+    {
+      "epoch": 4.23493670886076,
+      "grad_norm": 0.2995938443278436,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.2139,
+      "step": 1044
+    },
+    {
+      "epoch": 4.238987341772152,
+      "grad_norm": 0.29885157461857625,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.2344,
+      "step": 1045
+    },
+    {
+      "epoch": 4.243037974683545,
+      "grad_norm": 0.30266930225246935,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.2363,
+      "step": 1046
+    },
+    {
+      "epoch": 4.247088607594937,
+      "grad_norm": 0.302761913789544,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.2482,
+      "step": 1047
+    },
+    {
+      "epoch": 4.251139240506329,
+      "grad_norm": 0.30563282111840356,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.213,
+      "step": 1048
+    },
+    {
+      "epoch": 4.255189873417722,
+      "grad_norm": 0.3001293356696653,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.2104,
+      "step": 1049
+    },
+    {
+      "epoch": 4.259240506329114,
+      "grad_norm": 0.3066172952459019,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.1791,
+      "step": 1050
+    },
+    {
+      "epoch": 4.263291139240506,
+      "grad_norm": 0.30108072255682966,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.2146,
+      "step": 1051
+    },
+    {
+      "epoch": 4.267341772151899,
+      "grad_norm": 0.3001247971211386,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.2291,
+      "step": 1052
+    },
+    {
+      "epoch": 4.271392405063291,
+      "grad_norm": 0.34134649747436735,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.2347,
+      "step": 1053
+    },
+    {
+      "epoch": 4.275443037974684,
+      "grad_norm": 0.30680164864248755,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.186,
+      "step": 1054
+    },
+    {
+      "epoch": 4.279493670886076,
+      "grad_norm": 0.2926671350446585,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.2446,
+      "step": 1055
+    },
+    {
+      "epoch": 4.283544303797468,
+      "grad_norm": 0.3104844153366225,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.224,
+      "step": 1056
+    },
+    {
+      "epoch": 4.287594936708861,
+      "grad_norm": 0.3049706618437478,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.1806,
+      "step": 1057
+    },
+    {
+      "epoch": 4.291645569620253,
+      "grad_norm": 0.30072763164582633,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.226,
+      "step": 1058
+    },
+    {
+      "epoch": 4.295696202531645,
+      "grad_norm": 0.2848160000660899,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.2314,
+      "step": 1059
+    },
+    {
+      "epoch": 4.299746835443038,
+      "grad_norm": 0.3179781777040263,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.2222,
+      "step": 1060
+    },
+    {
+      "epoch": 4.30379746835443,
+      "grad_norm": 0.29934199022058045,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.1954,
+      "step": 1061
+    },
+    {
+      "epoch": 4.307848101265823,
+      "grad_norm": 0.31537354747576846,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.2167,
+      "step": 1062
+    },
+    {
+      "epoch": 4.311898734177215,
+      "grad_norm": 0.28629030680299733,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.2022,
+      "step": 1063
+    },
+    {
+      "epoch": 4.315949367088607,
+      "grad_norm": 0.30621618281842994,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.2249,
+      "step": 1064
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.30071076431191257,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.2057,
+      "step": 1065
+    },
+    {
+      "epoch": 4.324050632911392,
+      "grad_norm": 0.3158728226698913,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.2118,
+      "step": 1066
+    },
+    {
+      "epoch": 4.328101265822784,
+      "grad_norm": 0.2901595004488531,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.2152,
+      "step": 1067
+    },
+    {
+      "epoch": 4.332151898734177,
+      "grad_norm": 0.3056383279708546,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.2379,
+      "step": 1068
+    },
+    {
+      "epoch": 4.3362025316455695,
+      "grad_norm": 0.3069466434987273,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.2194,
+      "step": 1069
+    },
+    {
+      "epoch": 4.340253164556962,
+      "grad_norm": 0.284005133296267,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.2346,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3443037974683545,
+      "grad_norm": 0.2846472337371947,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.2456,
+      "step": 1071
+    },
+    {
+      "epoch": 4.348354430379747,
+      "grad_norm": 0.30118696814739776,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.2234,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3524050632911395,
+      "grad_norm": 0.2855672188339471,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.2397,
+      "step": 1073
+    },
+    {
+      "epoch": 4.356455696202532,
+      "grad_norm": 0.2851296100643242,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.2096,
+      "step": 1074
+    },
+    {
+      "epoch": 4.360506329113924,
+      "grad_norm": 0.29672566866513894,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.24,
+      "step": 1075
+    },
+    {
+      "epoch": 4.364556962025317,
+      "grad_norm": 0.3294022258268311,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.2113,
+      "step": 1076
+    },
+    {
+      "epoch": 4.368607594936709,
+      "grad_norm": 0.29374811248859517,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.2382,
+      "step": 1077
+    },
+    {
+      "epoch": 4.372658227848102,
+      "grad_norm": 0.29224959178773446,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.2246,
+      "step": 1078
+    },
+    {
+      "epoch": 4.376708860759494,
+      "grad_norm": 0.30613289563257723,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.2434,
+      "step": 1079
+    },
+    {
+      "epoch": 4.380759493670886,
+      "grad_norm": 0.2949995287137181,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.2395,
+      "step": 1080
+    },
+    {
+      "epoch": 4.384810126582279,
+      "grad_norm": 0.29266610446667624,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.2028,
+      "step": 1081
+    },
+    {
+      "epoch": 4.388860759493671,
+      "grad_norm": 0.29024879941662024,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.2221,
+      "step": 1082
+    },
+    {
+      "epoch": 4.392911392405063,
+      "grad_norm": 0.314245330135545,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.1809,
+      "step": 1083
+    },
+    {
+      "epoch": 4.396962025316456,
+      "grad_norm": 0.313401211166346,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.2237,
+      "step": 1084
+    },
+    {
+      "epoch": 4.401012658227848,
+      "grad_norm": 0.2860654491868347,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.2162,
+      "step": 1085
+    },
+    {
+      "epoch": 4.405063291139241,
+      "grad_norm": 0.2998824086695796,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.2232,
+      "step": 1086
+    },
+    {
+      "epoch": 4.409113924050633,
+      "grad_norm": 0.29743570840728484,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.2219,
+      "step": 1087
+    },
+    {
+      "epoch": 4.413164556962025,
+      "grad_norm": 0.2848782174293578,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.2155,
+      "step": 1088
+    },
+    {
+      "epoch": 4.417215189873418,
+      "grad_norm": 0.30158447822576057,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.2161,
+      "step": 1089
+    },
+    {
+      "epoch": 4.42126582278481,
+      "grad_norm": 0.2994592426184025,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.2166,
+      "step": 1090
+    },
+    {
+      "epoch": 4.425316455696202,
+      "grad_norm": 0.3007255129069184,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.1965,
+      "step": 1091
+    },
+    {
+      "epoch": 4.429367088607595,
+      "grad_norm": 0.3149103224698756,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.2257,
+      "step": 1092
+    },
+    {
+      "epoch": 4.433417721518987,
+      "grad_norm": 0.2990602435348281,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.206,
+      "step": 1093
+    },
+    {
+      "epoch": 4.43746835443038,
+      "grad_norm": 0.307751591158468,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.2251,
+      "step": 1094
+    },
+    {
+      "epoch": 4.441518987341772,
+      "grad_norm": 0.28895206140654983,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.2141,
+      "step": 1095
+    },
+    {
+      "epoch": 4.445569620253164,
+      "grad_norm": 0.2987116860071986,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.2214,
+      "step": 1096
+    },
+    {
+      "epoch": 4.449620253164557,
+      "grad_norm": 0.29459535183489055,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.2209,
+      "step": 1097
+    },
+    {
+      "epoch": 4.453670886075949,
+      "grad_norm": 0.29439709738085573,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.1833,
+      "step": 1098
+    },
+    {
+      "epoch": 4.457721518987341,
+      "grad_norm": 0.3110577888482061,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.204,
+      "step": 1099
+    },
+    {
+      "epoch": 4.461772151898734,
+      "grad_norm": 0.30021170962503835,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.1969,
+      "step": 1100
+    },
+    {
+      "epoch": 4.465822784810126,
+      "grad_norm": 0.2997033021380724,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.2157,
+      "step": 1101
+    },
+    {
+      "epoch": 4.469873417721519,
+      "grad_norm": 0.30753285767840327,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.2216,
+      "step": 1102
+    },
+    {
+      "epoch": 4.473924050632911,
+      "grad_norm": 0.3002362606477169,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.2148,
+      "step": 1103
+    },
+    {
+      "epoch": 4.4779746835443035,
+      "grad_norm": 0.307875423334184,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.2211,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4820253164556965,
+      "grad_norm": 0.2889388864238862,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.2145,
+      "step": 1105
+    },
+    {
+      "epoch": 4.4860759493670885,
+      "grad_norm": 0.3158833999488797,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.2066,
+      "step": 1106
+    },
+    {
+      "epoch": 4.490126582278481,
+      "grad_norm": 0.289080605797401,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.2131,
+      "step": 1107
+    },
+    {
+      "epoch": 4.494177215189874,
+      "grad_norm": 0.2888926080539976,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.2065,
+      "step": 1108
+    },
+    {
+      "epoch": 4.498227848101266,
+      "grad_norm": 0.290622457518494,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.1983,
+      "step": 1109
+    },
+    {
+      "epoch": 4.502278481012659,
+      "grad_norm": 0.2967702054753947,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.2503,
+      "step": 1110
+    },
+    {
+      "epoch": 4.506329113924051,
+      "grad_norm": 0.3015723388689254,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.2408,
+      "step": 1111
+    },
+    {
+      "epoch": 4.510379746835443,
+      "grad_norm": 0.29497246720667947,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.2179,
+      "step": 1112
+    },
+    {
+      "epoch": 4.514430379746836,
+      "grad_norm": 0.3064858570035279,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.2181,
+      "step": 1113
+    },
+    {
+      "epoch": 4.518481012658228,
+      "grad_norm": 0.2979206917372863,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.2299,
+      "step": 1114
+    },
+    {
+      "epoch": 4.52253164556962,
+      "grad_norm": 0.28956495732950177,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.2109,
+      "step": 1115
+    },
+    {
+      "epoch": 4.526582278481013,
+      "grad_norm": 0.31944392706527536,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.2267,
+      "step": 1116
+    },
+    {
+      "epoch": 4.530632911392405,
+      "grad_norm": 0.29324900388870556,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.2156,
+      "step": 1117
+    },
+    {
+      "epoch": 4.534683544303798,
+      "grad_norm": 0.3116898294587268,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.205,
+      "step": 1118
+    },
+    {
+      "epoch": 4.53873417721519,
+      "grad_norm": 0.2989751490017471,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.1887,
+      "step": 1119
+    },
+    {
+      "epoch": 4.542784810126582,
+      "grad_norm": 0.29676201288029397,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.196,
+      "step": 1120
+    },
+    {
+      "epoch": 4.546835443037975,
+      "grad_norm": 0.29914673098839517,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.2391,
+      "step": 1121
+    },
+    {
+      "epoch": 4.550886075949367,
+      "grad_norm": 0.28655757772456153,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.218,
+      "step": 1122
+    },
+    {
+      "epoch": 4.55493670886076,
+      "grad_norm": 0.29297163272203947,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.2425,
+      "step": 1123
+    },
+    {
+      "epoch": 4.558987341772152,
+      "grad_norm": 0.2990840698697507,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.2224,
+      "step": 1124
+    },
+    {
+      "epoch": 4.563037974683544,
+      "grad_norm": 0.28540725751154356,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.2116,
+      "step": 1125
+    },
+    {
+      "epoch": 4.567088607594937,
+      "grad_norm": 0.28621576408229343,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.2268,
+      "step": 1126
+    },
+    {
+      "epoch": 4.571139240506329,
+      "grad_norm": 0.2850688483018584,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.2084,
+      "step": 1127
+    },
+    {
+      "epoch": 4.575189873417721,
+      "grad_norm": 0.29218063813893436,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.1859,
+      "step": 1128
+    },
+    {
+      "epoch": 4.579240506329114,
+      "grad_norm": 0.2974626996526995,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.206,
+      "step": 1129
+    },
+    {
+      "epoch": 4.583291139240506,
+      "grad_norm": 0.32461990641172633,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.2298,
+      "step": 1130
+    },
+    {
+      "epoch": 4.587341772151898,
+      "grad_norm": 0.31735880112470016,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.2219,
+      "step": 1131
+    },
+    {
+      "epoch": 4.591392405063291,
+      "grad_norm": 0.2905731952115407,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.2223,
+      "step": 1132
+    },
+    {
+      "epoch": 4.595443037974683,
+      "grad_norm": 0.3018101251798338,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.1993,
+      "step": 1133
+    },
+    {
+      "epoch": 4.599493670886076,
+      "grad_norm": 0.2889572151919855,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.2355,
+      "step": 1134
+    },
+    {
+      "epoch": 4.603544303797468,
+      "grad_norm": 0.300462961314872,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.2004,
+      "step": 1135
+    },
+    {
+      "epoch": 4.6075949367088604,
+      "grad_norm": 0.2858558293644048,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.2313,
+      "step": 1136
+    },
+    {
+      "epoch": 4.611645569620253,
+      "grad_norm": 0.30071624301406996,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.2117,
+      "step": 1137
+    },
+    {
+      "epoch": 4.6156962025316455,
+      "grad_norm": 0.2872640423813283,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.2012,
+      "step": 1138
+    },
+    {
+      "epoch": 4.619746835443038,
+      "grad_norm": 0.30812522477258336,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.2171,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6237974683544305,
+      "grad_norm": 0.3060012625843342,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.2102,
+      "step": 1140
+    },
+    {
+      "epoch": 4.627848101265823,
+      "grad_norm": 0.29713190002804435,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.2267,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6318987341772155,
+      "grad_norm": 0.3011028441895585,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.2199,
+      "step": 1142
+    },
+    {
+      "epoch": 4.635949367088608,
+      "grad_norm": 0.29333420359676576,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.23,
+      "step": 1143
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.28318186964854775,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.2024,
+      "step": 1144
+    },
+    {
+      "epoch": 4.644050632911393,
+      "grad_norm": 0.2971928663147212,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.2269,
+      "step": 1145
+    },
+    {
+      "epoch": 4.648101265822785,
+      "grad_norm": 0.2803634956626488,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.2461,
+      "step": 1146
+    },
+    {
+      "epoch": 4.652151898734177,
+      "grad_norm": 0.29879957311318267,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.2139,
+      "step": 1147
+    },
+    {
+      "epoch": 4.65620253164557,
+      "grad_norm": 0.2866953778263888,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.2573,
+      "step": 1148
+    },
+    {
+      "epoch": 4.660253164556962,
+      "grad_norm": 0.29106854215351463,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.2323,
+      "step": 1149
+    },
+    {
+      "epoch": 4.664303797468355,
+      "grad_norm": 0.2932466315736546,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.212,
+      "step": 1150
+    },
+    {
+      "epoch": 4.668354430379747,
+      "grad_norm": 0.3047411557805243,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.2187,
+      "step": 1151
+    },
+    {
+      "epoch": 4.672405063291139,
+      "grad_norm": 0.29125790531758805,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.2146,
+      "step": 1152
+    },
+    {
+      "epoch": 4.676455696202532,
+      "grad_norm": 0.29897456729512556,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.2229,
+      "step": 1153
+    },
+    {
+      "epoch": 4.680506329113924,
+      "grad_norm": 0.2945866865441422,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.2176,
+      "step": 1154
+    },
+    {
+      "epoch": 4.684556962025317,
+      "grad_norm": 0.2996815824133542,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.2546,
+      "step": 1155
+    },
+    {
+      "epoch": 4.688607594936709,
+      "grad_norm": 0.28779074611154615,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.2366,
+      "step": 1156
+    },
+    {
+      "epoch": 4.692658227848101,
+      "grad_norm": 0.29086980134243623,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.2219,
+      "step": 1157
+    },
+    {
+      "epoch": 4.696708860759494,
+      "grad_norm": 0.283500723353293,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.2217,
+      "step": 1158
+    },
+    {
+      "epoch": 4.700759493670886,
+      "grad_norm": 0.29610916097006446,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.2087,
+      "step": 1159
+    },
+    {
+      "epoch": 4.704810126582278,
+      "grad_norm": 0.3018324073563392,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.2202,
+      "step": 1160
+    },
+    {
+      "epoch": 4.708860759493671,
+      "grad_norm": 0.2831923813098085,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.1978,
+      "step": 1161
+    },
+    {
+      "epoch": 4.712911392405063,
+      "grad_norm": 0.2960408738574354,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.2106,
+      "step": 1162
+    },
+    {
+      "epoch": 4.716962025316455,
+      "grad_norm": 0.31906508435295405,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.2044,
+      "step": 1163
+    },
+    {
+      "epoch": 4.721012658227848,
+      "grad_norm": 0.29314862954604426,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.1942,
+      "step": 1164
+    },
+    {
+      "epoch": 4.72506329113924,
+      "grad_norm": 0.28470085027824726,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.224,
+      "step": 1165
+    },
+    {
+      "epoch": 4.729113924050633,
+      "grad_norm": 0.30016976013607166,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.2399,
+      "step": 1166
+    },
+    {
+      "epoch": 4.733164556962025,
+      "grad_norm": 0.29722015471035146,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.2209,
+      "step": 1167
+    },
+    {
+      "epoch": 4.737215189873417,
+      "grad_norm": 0.2981379441926589,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.1848,
+      "step": 1168
+    },
+    {
+      "epoch": 4.74126582278481,
+      "grad_norm": 0.291593755395065,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.1989,
+      "step": 1169
+    },
+    {
+      "epoch": 4.745316455696202,
+      "grad_norm": 0.2999259400202548,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.1796,
+      "step": 1170
+    },
+    {
+      "epoch": 4.749367088607595,
+      "grad_norm": 0.28476927808142455,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.2128,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7534177215189874,
+      "grad_norm": 0.30078374074604913,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.209,
+      "step": 1172
+    },
+    {
+      "epoch": 4.7574683544303795,
+      "grad_norm": 0.283223419450551,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.2075,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7615189873417725,
+      "grad_norm": 0.29390413825540923,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.2106,
+      "step": 1174
+    },
+    {
+      "epoch": 4.7655696202531646,
+      "grad_norm": 0.2877217735640784,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.2251,
+      "step": 1175
+    },
+    {
+      "epoch": 4.769620253164557,
+      "grad_norm": 0.298502297885446,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.2322,
+      "step": 1176
+    },
+    {
+      "epoch": 4.77367088607595,
+      "grad_norm": 0.2939573728813273,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.2149,
+      "step": 1177
+    },
+    {
+      "epoch": 4.777721518987342,
+      "grad_norm": 0.30401691592609853,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.2188,
+      "step": 1178
+    },
+    {
+      "epoch": 4.781772151898734,
+      "grad_norm": 0.2930520971534313,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.181,
+      "step": 1179
+    },
+    {
+      "epoch": 4.785822784810127,
+      "grad_norm": 0.3013052053965951,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.1948,
+      "step": 1180
+    },
+    {
+      "epoch": 4.789873417721519,
+      "grad_norm": 0.2928353534861373,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.2325,
+      "step": 1181
+    },
+    {
+      "epoch": 4.793924050632912,
+      "grad_norm": 0.2950069407249688,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.2085,
+      "step": 1182
+    },
+    {
+      "epoch": 4.797974683544304,
+      "grad_norm": 0.28697540257935683,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.2151,
+      "step": 1183
+    },
+    {
+      "epoch": 4.802025316455696,
+      "grad_norm": 0.3023145826388531,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.2,
+      "step": 1184
+    },
+    {
+      "epoch": 4.806075949367089,
+      "grad_norm": 0.28852198497331666,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.2415,
+      "step": 1185
+    },
+    {
+      "epoch": 4.810126582278481,
+      "grad_norm": 0.284856039390243,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.2377,
+      "step": 1186
+    },
+    {
+      "epoch": 4.814177215189874,
+      "grad_norm": 0.29139851126095445,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.2219,
+      "step": 1187
+    },
+    {
+      "epoch": 4.818227848101266,
+      "grad_norm": 0.30505180691239164,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.2153,
+      "step": 1188
+    },
+    {
+      "epoch": 4.822278481012658,
+      "grad_norm": 0.29139871434196046,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.2114,
+      "step": 1189
+    },
+    {
+      "epoch": 4.826329113924051,
+      "grad_norm": 0.2931763861694432,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.2305,
+      "step": 1190
+    },
+    {
+      "epoch": 4.830379746835443,
+      "grad_norm": 0.2938835724485945,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.2295,
+      "step": 1191
+    },
+    {
+      "epoch": 4.834430379746835,
+      "grad_norm": 0.2961491666053218,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.206,
+      "step": 1192
+    },
+    {
+      "epoch": 4.838481012658228,
+      "grad_norm": 0.3031700645030975,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.1905,
+      "step": 1193
+    },
+    {
+      "epoch": 4.84253164556962,
+      "grad_norm": 0.29487001692862064,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.2026,
+      "step": 1194
+    },
+    {
+      "epoch": 4.846582278481012,
+      "grad_norm": 0.30699137998545106,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.2133,
+      "step": 1195
+    },
+    {
+      "epoch": 4.850632911392405,
+      "grad_norm": 0.305355136512352,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.2126,
+      "step": 1196
+    },
+    {
+      "epoch": 4.854683544303797,
+      "grad_norm": 0.28985171902846996,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.2432,
+      "step": 1197
+    },
+    {
+      "epoch": 4.85873417721519,
+      "grad_norm": 0.30562392328606963,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.2094,
+      "step": 1198
+    },
+    {
+      "epoch": 4.862784810126582,
+      "grad_norm": 0.2917103716200464,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.2485,
+      "step": 1199
+    },
+    {
+      "epoch": 4.866835443037974,
+      "grad_norm": 0.29448461985734586,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.2019,
+      "step": 1200
+    },
+    {
+      "epoch": 4.870886075949367,
+      "grad_norm": 0.28812751433569755,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.2342,
+      "step": 1201
+    },
+    {
+      "epoch": 4.874936708860759,
+      "grad_norm": 0.31541933313972825,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.1971,
+      "step": 1202
+    },
+    {
+      "epoch": 4.878987341772152,
+      "grad_norm": 0.29697838654056685,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.2108,
+      "step": 1203
+    },
+    {
+      "epoch": 4.883037974683544,
+      "grad_norm": 0.30700848371858497,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.217,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8870886075949365,
+      "grad_norm": 0.3001013441994226,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.1986,
+      "step": 1205
+    },
+    {
+      "epoch": 4.891139240506329,
+      "grad_norm": 0.29218711734802105,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.239,
+      "step": 1206
+    },
+    {
+      "epoch": 4.8951898734177215,
+      "grad_norm": 0.2975811701094111,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.2232,
+      "step": 1207
+    },
+    {
+      "epoch": 4.899240506329114,
+      "grad_norm": 0.3034825290353758,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.2113,
+      "step": 1208
+    },
+    {
+      "epoch": 4.9032911392405065,
+      "grad_norm": 0.2916460063497647,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.2258,
+      "step": 1209
+    },
+    {
+      "epoch": 4.907341772151899,
+      "grad_norm": 0.30360893085509005,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.1986,
+      "step": 1210
+    },
+    {
+      "epoch": 4.911392405063291,
+      "grad_norm": 0.2917487594596834,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.2168,
+      "step": 1211
+    },
+    {
+      "epoch": 4.915443037974684,
+      "grad_norm": 0.3036692021409533,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.2513,
+      "step": 1212
+    },
+    {
+      "epoch": 4.919493670886076,
+      "grad_norm": 0.2967396253154479,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.2086,
+      "step": 1213
+    },
+    {
+      "epoch": 4.923544303797469,
+      "grad_norm": 0.2879407851749776,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.226,
+      "step": 1214
+    },
+    {
+      "epoch": 4.927594936708861,
+      "grad_norm": 0.2921974064183767,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.2263,
+      "step": 1215
+    },
+    {
+      "epoch": 4.931645569620253,
+      "grad_norm": 0.29858681921749836,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.2174,
+      "step": 1216
+    },
+    {
+      "epoch": 4.935696202531646,
+      "grad_norm": 0.29758514414501985,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.2155,
+      "step": 1217
+    },
+    {
+      "epoch": 4.939746835443038,
+      "grad_norm": 0.28844848578794113,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.2186,
+      "step": 1218
+    },
+    {
+      "epoch": 4.943797468354431,
+      "grad_norm": 0.3083413480399838,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.2097,
+      "step": 1219
+    },
+    {
+      "epoch": 4.947848101265823,
+      "grad_norm": 0.29025390250975297,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.2347,
+      "step": 1220
+    },
+    {
+      "epoch": 4.951898734177215,
+      "grad_norm": 0.2981592568653464,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.1937,
+      "step": 1221
+    },
+    {
+      "epoch": 4.955949367088608,
+      "grad_norm": 0.27246481032057357,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.2353,
+      "step": 1222
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.30802919983785987,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.1915,
+      "step": 1223
+    },
+    {
+      "epoch": 4.964050632911392,
+      "grad_norm": 0.29762709860939635,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.2323,
+      "step": 1224
+    },
+    {
+      "epoch": 4.968101265822785,
+      "grad_norm": 0.2977661070987578,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.2181,
+      "step": 1225
+    },
+    {
+      "epoch": 4.972151898734177,
+      "grad_norm": 0.30514672042721513,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.2216,
+      "step": 1226
+    },
+    {
+      "epoch": 4.976202531645569,
+      "grad_norm": 0.2901228591699796,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.2193,
+      "step": 1227
+    },
+    {
+      "epoch": 4.980253164556962,
+      "grad_norm": 0.29602406924291996,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.215,
+      "step": 1228
+    },
+    {
+      "epoch": 4.984303797468354,
+      "grad_norm": 0.29892050652614455,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.2104,
+      "step": 1229
+    },
+    {
+      "epoch": 4.988354430379747,
+      "grad_norm": 0.3003697770664795,
+      "learning_rate": 0.0,
+      "loss": 0.2365,
+      "step": 1230
+    },
+    {
+      "epoch": 4.988354430379747,
+      "step": 1230,
+      "total_flos": 1.9047998106831421e+18,
+      "train_loss": 0.41321635669082163,
+      "train_runtime": 69555.1098,
+      "train_samples_per_second": 2.272,
+      "train_steps_per_second": 0.018
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1230,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.9047998106831421e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}