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": 4.982278481012658,
+  "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": 6.532133039584261,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 0.8321,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 6.55724691498222,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 0.8372,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 6.424698354400342,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 0.8261,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 6.188993094740009,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 0.8122,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 5.753997955765257,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 0.8118,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 5.584034791221274,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 0.8003,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 4.380263136434547,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 0.7706,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 4.06378638849151,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 0.7662,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.25510941138677,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.7225,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.0704235775792146,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.7262,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 1.932013680460994,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.7427,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.8198996385134842,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.724,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.1400290386186915,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.7225,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 4.235322422433948,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.7214,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 4.028634213316277,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.7051,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 3.1627776235201597,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.6679,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.7806082427093552,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.6518,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.250739178388274,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.6675,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.4039236507494328,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.6405,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.230527635491119,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.6224,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.3230478105262506,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.6172,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.245524700871051,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.6138,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.0198058837875434,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.6084,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 0.8290929409703974,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.5856,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 0.8741671242896029,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.5896,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 0.8493828598848276,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.5795,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.6900389107412627,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.5557,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.7912922170533162,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.5712,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.8385065848719294,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.5681,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.5745776495969577,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.5593,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.5958409956699017,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.5541,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.8808606459230536,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.5529,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.641337275897733,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.5575,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.4644243145690973,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.5423,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.6470463313463611,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.5436,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.6250850655644297,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.5393,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.49653321540210393,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.5421,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.5918927916266972,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.5469,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.47320437472014154,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.5492,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.5159021650860737,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.5323,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.5058570501233153,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.5118,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.45824553617643554,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.5234,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.4730188811575589,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.5167,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.5305038120092757,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.5138,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.5360197658641483,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.5265,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.4698971999309097,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.5074,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.521564146211447,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.5178,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.4567565637249538,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.5139,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.476613335138623,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.5089,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.45435085032280775,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.504,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.47393210672462427,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.5055,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.8284949275937379,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.5515,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.4991002970040021,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.512,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.5730988591547069,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.504,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.44324636758298863,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.5023,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.4649876323014944,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.4914,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.5157048143830449,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.4866,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.49900276094945917,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.4955,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.5507643750914,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.5093,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.4855510936845392,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.4925,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.45179364347344964,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.5057,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.7006651261786269,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.4871,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.7165847700903264,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.4896,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.7050158404940791,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.4855,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.6234068322689708,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.4789,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.4123043202839664,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.4798,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.7738300383888161,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.5021,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.8233167414905349,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.5087,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.9821668466326946,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.4881,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.9428503915687323,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.4875,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.8846119488432975,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.4785,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.8425291836832073,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.4852,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.883766589876103,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.4854,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.783338443764314,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.4923,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.8311792893475278,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.4822,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.49064876036796806,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.4836,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.8295829965829589,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.5001,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.691313396236878,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.469,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6394127931817026,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.4851,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.7833380994909307,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.4704,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.6929181982045055,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.4879,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.113102772802165,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.4768,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.8784901377103049,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.5034,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.7046923478566811,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.4814,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.5951051101902664,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.4602,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.7701801705117475,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.4882,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.9501470076908978,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.4918,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.8049250488945074,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.498,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 0.8109297801915586,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.4801,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.6231098337782589,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.4744,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.7877895743723996,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.48,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.9126661571316452,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.4931,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.3065552814142638,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.4899,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 0.5999395617017904,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.4794,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.019522155844935,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.4807,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 0.8896725155340885,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.4695,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.7552872095957193,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.4746,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.793440449191414,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.4702,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.6170837959015021,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.4721,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.6225047240159131,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.4825,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.5605021883218029,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.467,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 0.7119112837241437,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.4722,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.0391216038160307,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.4769,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 0.7557277538241919,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.4636,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.8649919790972606,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.4618,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.4120481340867437,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.4776,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.9174699048755477,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.4929,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.2523112447242668,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.4748,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.8489251603614493,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.4704,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.2229015747030578,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.4712,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 0.7969920633577473,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.468,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 1.2848536760474871,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.4778,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.8449244872203884,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.4688,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.21666925461268,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.4767,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.9807464996425289,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.4622,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.063158690580962,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.4721,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.0396613840110422,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.48,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.899873241156355,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.4663,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.8948811837160393,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.4577,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.8166933237872561,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.4669,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.9209851462449821,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.4615,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.6617022292396986,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.4492,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 0.6937808171936135,
+      "learning_rate": 4e-05,
+      "loss": 0.4473,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.8215903562018791,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.4814,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.8293301243778226,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.4575,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.8459445195287912,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.4583,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.809642412971106,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.4655,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.5992590888360483,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.4677,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.8442726311994523,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.4665,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.596588963621378,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.4616,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.7081332607386057,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.4608,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.7762456810988156,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.4532,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.5866594627620179,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.4713,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.5340526098473829,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.461,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.6144108450903079,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.4641,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.5896282504012614,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.4446,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.9569824169940685,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.4839,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.5235710857192512,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.4413,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.6183472618528378,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.4537,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.6070765777611301,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.4624,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.8800062329885172,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.4541,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.7994359778420898,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.4632,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.523766961278588,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.4689,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.6276851405625623,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.4582,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.5795459149480725,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.4542,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.4846722103328143,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.4637,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.6477779733049581,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.4672,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.48592756960003713,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.4538,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.596471321637245,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.4626,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.5004585412936841,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.4553,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.5160036964114423,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.4507,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.3967294584842366,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.4651,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.5781220940929725,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.4606,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.6652903674694022,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.4541,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.7852026033004209,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.448,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.7253068614248434,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.4587,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.5411634655919866,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.4553,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5678631187010995,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.4639,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.6409570916623751,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.4593,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.7870375334405469,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.4485,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.6081565322790481,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.4683,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.6408578016133387,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.452,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.7148077009377766,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.4505,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.6518590631446205,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.476,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.6373074566650115,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.4686,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.7644102254653287,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.4498,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.6707968296312907,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.4569,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.7986830603171721,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.4489,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.8024760870709715,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.4597,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.5451184366914132,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.4537,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.6564503184648073,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.46,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.8019463253663605,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.4475,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.5918956963672852,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.4515,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.44816361709255015,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.4428,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.5879856892021745,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.4531,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.4449855971546745,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.4342,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.4956893615877283,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.4602,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.5454689276455801,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.447,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.5562425213676986,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.4508,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.5552186061779076,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.4616,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.5925713493034847,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.4542,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.6175425883267657,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.4562,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.48275910688585855,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.4681,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.549378323996798,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.4494,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.5670378280895085,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.4324,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.38079001461951395,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.455,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.6716782747035639,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.4514,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.6084584664532123,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.4491,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.5678704664431001,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.4453,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.4416789994091744,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.4477,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.4542128066438179,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.4494,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.575540176436905,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.4458,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.49821215068975166,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.4333,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.40952403347919597,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.4398,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.4694648677544673,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.456,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.5197449542476376,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.4437,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.4764595039299929,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.4469,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.4465752752018794,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.4479,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.4511005179603608,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.447,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.45846886879097287,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.4553,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.42496350357037765,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.4432,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.4441022546845547,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.458,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.4026316109191035,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.4521,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.45329504631714207,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.4501,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.584209338233781,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.4442,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.5327010601787499,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.4469,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.5701876097921064,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.4559,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.5365596401690965,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.4382,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.3368424218789634,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.4329,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.4344078624773968,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.4329,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.540566281834332,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.4373,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.5014457976288447,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.4311,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.9408306165983849,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.4897,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.501127503176562,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.4488,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.4280092501054334,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.4554,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.4176484999793951,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.4549,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.43155142280889164,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.4525,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.4239828450122875,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.4488,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.4233793185671727,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.455,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.4560550395115194,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.456,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.44516335548442604,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.4284,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.4133931624188761,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.4398,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.42388440035208486,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.4498,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.4131532808870014,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.4409,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.443922412207509,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.4547,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.39697835206344945,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.4458,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.6214823336988358,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.4573,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.378365533278279,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.432,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.3643614693385824,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.4458,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.39305872631656785,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.4371,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.34309461467325175,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.4308,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.4391473476875838,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.4469,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.4531342278565964,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.4329,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.3808275380141185,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.4527,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.4247317054325277,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.4447,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.44045080730268626,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.437,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5470172719966262,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.4337,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.45785131402567686,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.4236,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.4226547780987477,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.4462,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.497516196097743,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.457,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.49213288557503854,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.4341,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 2.7660389555633933,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.4526,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.6660713080867279,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.4458,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 1.4734375773456256,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.4491,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 1.1011620053066535,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.4391,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 1.1369344263037509,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.4403,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.8451325527897005,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.4489,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.8010386230668763,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.4174,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.5646912754741253,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.408,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.5780177476588978,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.4059,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 0.5532859551265328,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.4211,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.9552665350004305,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.4211,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.9089190497011683,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.3924,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 0.4305554365598635,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.3975,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.8351853691370429,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.4075,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.6427079324768665,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.4129,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.6721743724084044,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.4063,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.5262668205728211,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.3929,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.4988095330435612,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.4046,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.6416118200029007,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.421,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 0.6554035859893528,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.4185,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.44948042540478733,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.4009,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.5678831390696434,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.4002,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.5549843996498874,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.4056,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 0.4465789090715565,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.4058,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.40701300144261715,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.4087,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.5002390136874156,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.4011,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 2.1524818697966523,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.4083,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.7700002996867252,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.4123,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 1.0521706618307196,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.4191,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.6523743844982531,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.3981,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.7876202345812121,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.4018,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.7074650188293946,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.4199,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.4639747400629689,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.4124,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.9110533760437446,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.4035,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 1.7759678597681143,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.4781,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 1.3485619417598869,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.4227,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 1025.8582773333876,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.9272,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 37.12442181905226,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.5143,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 206.8669300400129,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.5988,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 59.87961221485313,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.7058,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 11.170925474923203,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.5681,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 7.81322584141767,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.5814,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 3.3329250910992076,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.5138,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 3.1346376407296845,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.4804,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 3.618962152376418,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.5001,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 3.2909922033123302,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.4884,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 2.3649850402275314,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.4502,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 3.421125134164473,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.4434,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 4.295415666039879,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.476,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 2.13422664138512,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.4584,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 1.0825926107579884,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.4324,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 1.5096591972900553,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.4232,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 1.444125464138491,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.4338,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 1.54379942044477,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.4269,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.979407507175933,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.4271,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.7757019970017879,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.43,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.8626187494271876,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.4154,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 5.361753699920954,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.436,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 5.2183470678160315,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.4255,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 1.6792254185211073,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.4446,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 8.788876895923698,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.4431,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 8.695989901198441,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.4572,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 4.037120199553027,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.4581,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 2.021836247734267,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.4579,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.8225882695006719,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.4304,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.8536213716644542,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.4148,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.838295215599383,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.4305,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.5630335062497537,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.4098,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 1.0496608007120687,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.4136,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.533717137478999,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.4102,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.4627834354909757,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.4002,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 1.3708291188426815,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.4126,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.4851894911383946,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.4208,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.4306709883463419,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.4048,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.43002171622737023,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.4115,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.42801922994575614,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.4062,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.39772064473287655,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.3928,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.42649337424600103,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.4068,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.3863634675666276,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.4053,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.39040012502333526,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.4118,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.3396375189141854,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.4002,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.4048375040291072,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.4077,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.3490139368794945,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.4125,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.34744194665467126,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.3981,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.4220163849294417,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.4028,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.2989400932731263,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.413,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.37455918883341144,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.4084,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.2739260040283174,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.3979,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.370463134516024,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.4047,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.3523210524573194,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.4105,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.2866522491834504,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.4058,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.32147288905455673,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.4032,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.32845486741908214,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.4047,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.4583252418384829,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.4026,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.29181086049630744,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.4015,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.7715088978928714,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.402,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.2857868988231324,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.3937,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.28203043432874036,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.4082,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.30531472948613186,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.4096,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.32533266546704115,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.4065,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.32550825831655933,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.4011,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.29921053459864366,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.4003,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.325317516084588,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.413,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.3268814973534259,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.4055,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.28355206826051976,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.4086,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.3273706329785133,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.4066,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.32161478219291123,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.392,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.30764246446217536,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.3945,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.27535328689517413,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.4018,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.3452148207824589,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.4007,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.31719274332292513,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.4205,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.3039195107183187,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.4109,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.32440279814216855,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.4019,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.30480599588079604,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.4023,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.30231958509996015,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.3871,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.27720738014291074,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.3907,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.30600216810204584,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.3918,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.3256026928940194,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.4118,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.2724807646500371,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.4018,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.30851238640219475,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.3896,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.2785604977457195,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.3966,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.32683739557158725,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.4049,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.28035980917918774,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.4019,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.2883933910596497,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.4001,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.28804507766136306,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.405,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.28040011173040685,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.3965,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.29845911310065343,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.3979,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.32818430066626814,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.3908,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.2800254440731786,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.3983,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.3574811630491918,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.3994,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.31727710306752976,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.401,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.3676189485089189,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.4096,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.3087604841372036,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.4047,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.337572529650528,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.3926,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.2918669052481138,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.4142,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.34798868755795564,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.3973,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.28807310331453817,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.4074,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.3483598345397016,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.4027,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.2638837605536445,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.395,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.34042543653585317,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.3945,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.32584423779467664,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.3913,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.28502745104056254,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.3967,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.3460616647522877,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.3982,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.3008249220434764,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.4016,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.37354436032841454,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.4036,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.3335600831735112,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.4075,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.324031760151563,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.4154,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.3319354235491296,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.3909,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.3065802822904757,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.4015,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.3330296041273453,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.4075,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.30839103767021186,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.4118,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.3029273483910999,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.3869,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.30957912699481827,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.3913,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.25478500689926403,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.3985,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.3675861189892056,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.4113,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.4036044434172424,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.3831,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.2638482854288027,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.398,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.43346726596832563,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.3988,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.2832121272098672,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.4089,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.35181082418939,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.388,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.32986150709633727,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.3939,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.3296541859166392,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.3958,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.28802435524420517,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.3953,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.309502390390023,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.4001,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.31075937995139896,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.402,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.3201855874091949,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.3907,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.2938281744019113,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.398,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.34974475106206376,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.4121,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.2950158217108298,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.384,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.3747715232791914,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.382,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.36029396769331645,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.4012,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.2950623183382536,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.3862,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.7917398294228565,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.4226,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.28813211729749766,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.4087,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 1.3790899866593584,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.4328,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.32552604863235257,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.4059,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 25.047680452896035,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.4257,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.7040052534391894,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.41,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.5691359298513559,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.3931,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 5.287753454251631,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.4037,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.7491872126611132,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.4066,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.7789862457046083,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.4106,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.5352882926475045,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.3981,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.6165033043032695,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.4003,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.47106511037182097,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.4019,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.5122553464206054,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.3946,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.44471293150828456,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.3917,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.4043625102393141,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.3929,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.43413576427840495,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.409,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.38796130777185817,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.387,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.41727669230764597,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.4107,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.39393018183413964,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.3901,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.38775464022002276,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.4052,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.41717334750090623,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.4051,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.3812159176626554,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.4045,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.3568899486563352,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.4028,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.35965881199830496,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.3995,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.3223927386779314,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.3911,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.41995646998742153,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.3987,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.3405015997443556,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.4016,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.34916357629685135,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.3962,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.8002756487104693,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.4007,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.33140064933393737,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.3894,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.338410215961819,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.395,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.3824724611806833,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.4097,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.26609034281035665,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.4017,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.3020311512126467,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.3992,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.30702642932478547,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.3986,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.27417173881822454,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.3915,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.30138586404416007,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.3962,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.264868795134402,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.407,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.31418743243392255,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.3998,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.27300053408900865,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.3949,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.2826105595565785,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.3851,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.23871233719103255,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.3842,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.33293927945303126,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.4025,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.24510549798980813,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.4066,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.30452732002360244,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.3939,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.27657240900958224,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.391,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.2321911642095816,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.3868,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.3769126029143979,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.3905,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.25253475710070616,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.3819,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.3403264489728991,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.3882,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.31517298617028905,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.3919,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.2997862023377442,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.3916,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.2897163596472061,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.4038,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.3296615360070917,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.3914,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.3012289060389266,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.4,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.6415224084034504,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.4213,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.342747326563051,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.4044,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.3240771241041766,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.3801,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.2954646819943863,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.3951,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2740642260929821,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.3903,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.33463974537218316,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.4012,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.2796904434257574,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.4027,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.2918463086443064,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.3997,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.2728297033492521,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.3921,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.31224082606931725,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.4022,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.2841728715046705,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.398,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.3240322372799053,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.3938,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.26938462770646276,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.3973,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.3306574349755594,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.3977,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.31767549020939095,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.3828,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.2867079678025377,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.3918,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.33844700648552706,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.4027,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.2992359730705919,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.3993,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.2902199793680572,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.3894,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.2726245869879467,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.3991,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.31262138340084594,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.4068,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.25315462144929207,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.4011,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.266987741726736,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.3912,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.26909466742801263,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.3908,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.30879795000919175,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.3887,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.2946187359300225,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.349,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 0.32648465971818075,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.3553,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.39985516294079787,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.3448,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.3600201841697603,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.3596,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.3898491339649574,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3421,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.365924967592496,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.3476,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.3400968163874117,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.353,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.31799606098655214,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.3443,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.3771441967578903,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.3563,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.2992536628168311,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.3553,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.47873945737448254,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.3489,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.35758701737051085,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.3469,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.36093617256696475,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.3446,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.3775953600001113,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.3513,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.43606380903664566,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.3533,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.34198819910356093,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.34,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.3704479976137312,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.355,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.3525613142318775,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.352,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.35652879695641804,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.3541,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.31763573339857665,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.3481,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.2861500646093877,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.3451,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.35561850176128906,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.345,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.2848127600098354,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.3603,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.3438392083533328,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.3477,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.33950116005963077,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.3553,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.3269154310659665,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.3524,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.36497665224039216,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.3477,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.45822720529269473,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.3579,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.34474378508686215,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.3525,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.32967712594868726,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.3367,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.2765082514915378,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.3517,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.3494647351629673,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.3513,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.2658225090074359,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.3378,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.4011910101793757,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.3677,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.28793687322521055,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.3421,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.33257681101722475,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.36,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.3083916525465513,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.3563,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.2781964128020093,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.3594,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.3259162995054383,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3569,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.2682013727461539,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.3501,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.3853955227996849,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.3581,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.29606722167574145,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.3531,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.36421732557881553,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.3449,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.3352323537708421,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.3469,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.3601310881623383,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.355,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.3744347037815094,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.3486,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.3222835426844154,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.3554,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.39248267956645816,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.3492,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.29455630743251515,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.3496,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.4026597223775707,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.3554,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.3145749058750984,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.3602,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.3511828423696781,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.358,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.33406009501855555,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.343,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.3427288272746871,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.3516,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.3950396570132591,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.3613,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.3555686988543008,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.3382,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.3338011211976489,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.3451,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.3334614397081832,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.3471,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.3133784945612427,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.3468,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.299580221767352,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.3547,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.31964428225190256,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.3441,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.2680421393033448,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.3546,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.32257787574765223,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.3598,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.2829082685659732,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.3413,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.2904139335458002,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.3519,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.3358152086252145,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.3521,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.282199872735539,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.3686,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.3569536732610532,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.3611,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.25373321379062586,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.3541,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.3211632036208442,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.356,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.30435910353626516,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.3541,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.2938569446373174,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.3408,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.3124658625110135,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.3568,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.3181000865273803,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.3448,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.3125684987715813,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.3527,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.3200722584783128,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.3462,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.3207652681634447,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.3603,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.3100150956271466,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.3556,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.2944305704283286,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.3514,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.3061569148938698,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.3437,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.2836147433348348,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.3505,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.2621933945735694,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.3481,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.2928638567740868,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.3634,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.28505672528904835,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.3567,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.29979618761029825,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.3607,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.24848615113665654,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.344,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.3409139611487724,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.3433,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.2711100533343054,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.3477,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.33873907945241616,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.3469,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.3535978198013254,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.3638,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.33862037067620127,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.3632,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.28738947723568814,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.3509,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.31622631055119743,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.3568,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.29800452610093947,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.357,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.2949434869261368,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.3427,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.29137532897174245,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.3432,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.3338096870608023,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.3519,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.24918612083653968,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.356,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.28775697916322546,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.354,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.2693943160676964,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.3499,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.2850520822271978,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.3495,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.3049102983111891,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.3622,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.3256617358604265,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.3528,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.2707861692211963,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.3418,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.3057904288216451,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.3539,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.24857318129773723,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.3442,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.28051883927131277,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.3445,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.24518298836703853,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.3572,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.25214404564487514,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.3522,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.2612710629152858,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.3588,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.2745395112668835,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.3596,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.27372649656787673,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.3495,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.26262629485376093,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.3567,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.2874101357556613,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.3604,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.22747795770769352,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.3499,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.272716424011408,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.3446,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.24344441084291274,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.35,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.24701470598486544,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.3569,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.25253002638510613,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.3472,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.2552850694997293,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.3426,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.24957090308393473,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3524,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.302094327836195,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.3512,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.2390273615654174,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.3543,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.32535853481739746,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.3494,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.24163505511293912,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.3529,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.32541491732652345,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.3355,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.2840198915138568,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.3523,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.31176496952398675,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.3537,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.2515296475319271,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.3469,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.3025467919206517,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.3635,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.2273953540292674,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.3472,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.3251099105972297,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.3631,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.26563284745705096,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.3385,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.29911707033810786,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.3515,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.2903177655835716,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.3524,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.2697726249256074,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.3451,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.3066348268898101,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.3386,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.2863077314421502,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.3425,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.2997092443569466,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.3512,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.2753445034708746,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.362,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.3504771300674148,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.3467,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.2579120538802542,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.3471,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.3452306782648512,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.3577,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.27844706508483824,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.3647,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.3819249965822982,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.3533,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.2898799926093991,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.3532,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.2951632251332297,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.3457,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.2834255406470996,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.3546,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.2902089546514115,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.3579,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.2232831753489167,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.3386,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.2724070337070466,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.3421,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.24641128792699102,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.3533,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.2631739422028684,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.355,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.2787236456158308,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.3379,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.24876699122165907,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.3511,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.26916531714766023,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.3477,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.25708761672059877,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.349,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.2559459122265582,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.35,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.32800674628258025,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.3382,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.24886550783462716,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.3575,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.30018355559561466,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.3556,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.2655379259371097,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3604,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.23687818980722178,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.349,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.2713725577637527,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.3479,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.24742371442978742,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.3528,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.26217375141113836,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.3478,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.2918627358161772,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.3517,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.23000645113987442,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.341,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.2521292184522108,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.3447,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.23854032995165828,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3471,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.2582041088588286,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.3483,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.27629971720486296,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.3501,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.2588433195251522,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.3662,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.2756634806357198,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.3373,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.27325073937438477,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.3666,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.26117895915527756,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.3456,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.2384615404403354,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.3429,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.2532875415303722,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.3502,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.27293135197520185,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.3537,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.2433938627758232,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.3521,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.29327257132470536,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.3502,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.25287478323732604,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.3509,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.29759092979528073,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.356,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.24895634066806752,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.3555,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.33407808275904777,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.3578,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.23414841626995647,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.3402,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.3206544996845211,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.3397,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.2324524755500536,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.3588,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.24292490920938947,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.3514,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.24399635588195465,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.3545,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.2518695136795285,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.347,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.24884336844115193,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.3556,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.22767924770178605,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.3541,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.23756267323715108,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.356,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.22700966862754857,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.3471,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.2554904523584583,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.3507,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.2332684394258549,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.3499,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.23809275238585556,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.3455,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.25767769074426655,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.3657,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.2525831677888879,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.3419,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.27913362335474345,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.3594,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.26057169001484665,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.3703,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.22540340864056851,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.3529,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.25107848604458244,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.3493,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.2420869861348637,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.3334,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.2495878524716229,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.3383,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.26181163415415315,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.3435,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.24029841151829276,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.3481,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.2566010219070237,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.3405,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.2518479644096767,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.3517,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.25896982646370514,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.3438,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.2711752887100306,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.3537,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.2803663104786842,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.359,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.3438390798626413,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.3487,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.2601181876318714,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.35,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.26228559851505423,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.3432,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.29943012322644696,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.3558,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.303117705651622,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.3548,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.2732932168841971,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.352,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.3414560347229972,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.3522,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.23790004848401577,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.3444,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.3139451660627059,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.3557,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.25704930128774794,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.3542,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.2929322859300731,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.344,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.2703593610776381,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.3475,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.25141117919144657,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.3527,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.264144348341108,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.3668,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.2435510774220007,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.3435,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.26150709054620613,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.3612,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.2316368620142563,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.3413,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.24591417313409167,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.36,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.26279267712360566,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.3542,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.24081788306947927,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.3423,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.24971164603584492,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.3432,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.24451034851133693,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.3588,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.2440384897603226,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.3416,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.2683793851688467,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.3584,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.22942579545059955,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.3517,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.25917158782113664,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.3461,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.24489816136732068,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.3505,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.2549287400495785,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.3477,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.2520820847533619,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.3578,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.24442003058734302,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.3422,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.23345655686926264,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3477,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.230115854931472,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.3506,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.22739159800615094,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.3473,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.26438385798509806,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.332,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.3154420065298799,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.3002,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.39335826646820427,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.2953,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.2902197773997057,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.2905,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.3567383700063536,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.3147,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.287396305320146,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.2957,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.29406007152584485,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.3065,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.29132649647118186,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.3037,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.27299131136893223,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.3047,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.2891036401083393,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.2974,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.27000414787459975,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.3043,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.256690104174489,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.2999,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.2460904098495158,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.3003,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.2443994477702688,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.3036,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.2571021283826004,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.2921,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.25290432114525147,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.32,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.25608373938214485,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.3015,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.2323512807673779,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.2867,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.2451203268890119,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.3046,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.23729253079849236,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.2952,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.25304828522074474,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.2917,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.24579874438074714,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.2996,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.2481131642398638,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.3008,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.22581606234338703,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.3068,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.26255573339413724,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.3131,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.2344982746969692,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.3059,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.26244595407019194,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.2952,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.2461496553952708,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.3081,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.2496981363876085,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.3045,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.2341645705664133,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.3032,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.24884896603841056,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.3096,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.24882877263678357,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.3061,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.22589104044623629,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.3009,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.26271523841066763,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.2997,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.24508035286059554,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.3153,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.2248936125408039,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.3085,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.21364189128258582,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.3037,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.25168297896525266,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.3028,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.2260084877517761,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2973,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.24080323046709481,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.2916,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.2352503293600071,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.2974,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.22060347816703538,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.2988,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.23828207948846084,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.3023,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.22259353079113836,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.309,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.2173614390542503,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.2976,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.2482984493922438,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.2989,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.2149361322261054,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.3021,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.22248714924064245,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.3015,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.23196202277296968,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.3114,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.21672630928153114,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.3045,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.21212684886858088,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.3097,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.24291055256497296,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.3014,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.21653363974667764,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.3099,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.23192727591774337,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.3038,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.2365593342751137,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.2993,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.20710469679345592,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.298,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.21555309590575647,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.3028,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.23192901338718347,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.3016,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.20687150525129452,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.3049,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.21256316702228997,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.2935,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.22226730961579103,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.3139,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.21484829303318065,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.3062,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.21126088351792308,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.2982,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.22300529165715596,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.3104,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.2077983173695733,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.2961,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.21054427544633794,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.3052,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.21109825254006861,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.3026,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.22118946635131692,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.3025,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.20462154571529967,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.3081,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.2529857078127118,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.3026,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.21463016934674486,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.3051,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.22295154871365744,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.2979,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.22476117666850645,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.2966,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.2223472814629082,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.3047,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.21594972224993225,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.3119,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.22236470404773323,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.2899,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.21997190285979087,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.3083,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.22826472890849273,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.304,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.2103487399569367,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.2924,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.20212386669151516,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.3094,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.2134701957512064,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.315,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.23097086458865795,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.2884,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.21953782944300468,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.2875,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.2412931905072053,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.3088,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.22695552301680602,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.2996,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.22379570401194795,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.3097,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.23497017396299524,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.2975,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.21425528506358746,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.3033,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.21054963340461108,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.2982,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.2275009302831448,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.3009,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.22700531522651518,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.3,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.21686916977685403,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.3136,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.23124434791008297,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.3008,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.19882468190189237,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.2836,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.22832289255650035,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.3055,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.2033672532765211,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.294,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.2223543917268838,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.3142,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.2123159749471159,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.2983,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.24542122307188552,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.3121,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.22233098261954815,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.2969,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.23338673639129157,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.3065,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.24508446033047493,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.2931,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.23140845315919012,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.2966,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.2436537707797995,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.2997,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.2145899116107727,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.3077,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.20925604747662918,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.3071,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.2210727726859272,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.3087,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.21850373668005083,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.304,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.22845231420461196,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.2975,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.37747324803244897,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.3043,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.24025667995909375,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.3066,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.2220497375144944,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.3072,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.21071350638192934,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.2914,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.4288025445668777,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.2977,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.2053111274592975,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.2969,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.22568204885994186,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.2987,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.2114505360922552,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.3025,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.2254098494377526,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.3025,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.204341856378612,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.3005,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.2120052358162623,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.3122,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.2107430608603274,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.3148,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.20897045075415976,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.2999,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.21406915408036475,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.3071,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.21183252163834873,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.3021,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.232323553724658,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.3044,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.23056750927154773,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.3105,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.22882769460896385,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.2938,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.22363250075428615,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.299,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.24207055084004905,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.301,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.2214012292725934,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.3055,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.23857773354380668,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.297,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.22792376047215704,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.2953,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.23354326343520118,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.3031,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.23337909872573454,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.3062,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.22684594580820613,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.3071,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.24705247626318178,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.3044,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.2134214878892756,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.3059,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.2338402066735803,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.2925,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.235460786212218,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.2956,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.20138747052071296,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.3121,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.22997273379701183,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.2921,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.2443654724474036,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.2977,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.218238527184597,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.2987,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.23633360978619455,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.3051,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.21323926627090756,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.2963,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.19151286079416865,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.2986,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.21489569341736786,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.2921,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.21632944657833997,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.2988,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.21710488446292492,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.3056,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.2055224419801305,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.3038,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.22704655123283377,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.3087,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.22916149870086128,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.3151,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.19764306127322392,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.3073,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.20808732619123263,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.3019,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.22710517564502702,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.2952,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.19822680815900015,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.3043,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.21617834800081834,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.3099,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.22971731296039288,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.2967,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.19085153411394593,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.3021,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.20637014762839026,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.3024,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.21256614939579038,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.2961,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.20491371299399733,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.3143,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.20673134215913366,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.3097,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.21386482389969716,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.3027,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.21363745428823797,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.3083,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.20632120604852322,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.3074,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.21738344289262684,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.294,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.2116751384258441,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.3025,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.20713717714823923,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.298,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.21849432553577436,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.3103,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.23108434409335799,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.2893,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.22680898335528035,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.2936,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.18742599819972328,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.293,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.20988877308076861,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.3016,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.22413872346285574,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.3102,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.18821503478777876,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.2984,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.20277930862192187,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.3014,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.22061724320723577,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.3031,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.20307876982994788,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.3048,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.2011726805087641,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.2987,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.2165230947658496,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.2893,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.19681151393888938,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.2998,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.19962996382093154,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.2894,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.22213374754218781,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.3156,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.1958278266186728,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.2966,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.20129571732217466,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.301,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.20344455830280977,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.3008,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.207166466997206,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.3078,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.19879263931573854,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.3084,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.1985356650983193,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.3059,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.1899548626242953,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.2845,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.200085100946873,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.3125,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.19094120383675653,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.3096,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.20346547525773723,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.3059,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.18591224956329608,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.2889,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.18714994602963464,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.3012,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.19151603633130612,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.306,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.20975541692554475,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.3036,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.19058113322519865,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.3077,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.20574063336393048,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.3106,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.20942869756026009,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.2973,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.19149660418052122,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.3008,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.20084118360879238,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2934,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.19723988684476282,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.3014,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.24171590583308064,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.3051,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.19256595756571265,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.3024,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.19491656153080195,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.3005,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.17920348526830776,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.3079,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.18410580357188783,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.3081,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.1922308012218206,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.3048,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.19417298872972594,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.3041,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.17947597096799528,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.2993,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.18499232216901276,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.2951,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.19265461535206335,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.2935,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.189022567659848,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.2966,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.18456117128539623,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.2919,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.20302324719319248,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.292,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.17960533399583267,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.3012,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.18277514124633043,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.3007,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.18703932696428766,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.2861,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.1878913451666577,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.3115,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.2030501578203081,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.3164,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.1853420622480249,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.306,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.19996506113586496,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.3036,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.20031872445198537,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.3135,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.1885090215503571,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.3043,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.21101758098185247,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.3067,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.19244637460547828,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.3038,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.18320571681850645,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.288,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.19727416708240164,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.2998,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.20143200442063874,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.3052,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.18334835515105263,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.3073,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.18642247581924598,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.2906,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.21477594190831334,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.2972,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.18617227210465642,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.3022,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.1884648937496575,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.3095,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.20439811927455742,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.2897,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.18650103221183445,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.2933,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.1864047841972926,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.3005,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.18238109929805987,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.2895,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.2005828225513602,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.3086,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.1852488953900081,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.2919,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.17746182587856754,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.3022,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.17841006268823623,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.2947,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.1758936554967002,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.3023,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.17596855422036348,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.2986,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.19493873328476455,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.307,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.23091789647145397,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.2909,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.2684848684335978,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.2651,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.20802336426973822,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.2592,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.2863998422497835,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.2579,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.3632605787832696,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.2702,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.20833210336768426,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.264,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.28878622213472965,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.2731,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.3292690906194357,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.2746,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.23005507899784294,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.2644,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.19467193807564695,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.2655,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.2800670217678115,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.2601,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.26178598233180045,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.2705,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.2096718045012149,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.2661,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.22034781726573469,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.2655,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.22898875648067166,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.2734,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.2420057156442431,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.2682,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.19686241327478346,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.2553,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.21112793838174654,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.2581,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.23933154782989635,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.2668,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.21318257334505805,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.2692,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.19253372183228779,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.2685,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.2144586493006519,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.2668,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.2128458720202371,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.2685,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.18019367220111576,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.2679,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.1920497891349743,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.263,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.20493469556081295,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.2703,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.18803016483530324,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.2599,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.19444728070823825,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.2655,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.18276436470341467,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.2603,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.18798785332491325,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.2775,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.18563491362934337,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.2694,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.17807695910415092,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.2799,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.19391728416395843,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.2561,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.18597875587415463,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.264,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.18174958733031835,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.2632,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.20690166705460802,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.2664,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.17691863465861843,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.2593,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.1828055950661491,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.2872,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.18369432476493913,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.2631,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.18345093182299,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.2657,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.18794709992749414,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.2662,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.1858639411631459,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.2542,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.18630253579964634,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.2754,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.189034068916359,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.2613,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.1914275473926052,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.2792,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.18025911974804243,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.2707,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.17934979890521108,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.2658,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.17709593264802015,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.2744,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.18917203046747533,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.2627,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.2095675936175753,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.2836,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.1795662018001616,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.2814,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.17820938471963463,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.2596,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.1743264817520895,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.2681,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.17872624903060363,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.2731,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.17035954550590332,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.2714,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.1784750821177353,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.2801,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.17752701766723342,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.2572,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.17684887853557255,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.2726,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.18049302370987538,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.271,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.17834511777560633,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.2644,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.17066484015849873,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.267,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.17544775734125798,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.2633,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.1799312352496095,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.268,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.16807208499468837,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.2695,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.17367882871027202,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.2682,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.1825465928324291,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.2627,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.17230799387287368,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.2705,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.17616748713645183,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.2638,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.17606017787701891,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.263,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.1813921578281895,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.2682,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.17255946701859723,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.2622,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.18493160354909638,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.2769,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.17511723825468428,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.2543,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.17386809392147595,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.2657,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.17660093200144314,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.2591,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.22223302505999118,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.2722,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.18377214162747504,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.2805,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.18691909404182294,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.2578,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.18165684989069852,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.2694,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.17904685475786217,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.2734,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.1670679132559021,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.2572,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.17061686533388282,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.277,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.16508965891048605,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.2651,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.22456843190829953,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.2645,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.18593936572595865,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.2592,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.17435492858437657,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.2692,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.174814986089018,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.2752,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.16605865827326716,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.2619,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.17493726009568253,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.2769,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.17516519474343514,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.2665,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.21254566481813106,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.2666,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.1685889832508063,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.2612,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.17352411409244017,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.2686,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.1688949043321581,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.2628,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.17389046688289364,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.2719,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.1692414042493534,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.2769,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.17138416002074836,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.2749,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.17764209984494997,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.2598,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.17612996453629745,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.2683,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.16283323313904408,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.2719,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.1712212612857408,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.263,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.17481185800617643,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.2606,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.1758026102644239,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.2581,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.17716800057843507,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.2647,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.17382876985415352,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.2713,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.17929541392095147,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.2579,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.17605931606910613,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.2569,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.17597378993711493,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.2661,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.17187911513554,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.2614,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.17307587522263623,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.2773,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.18505144556748512,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.258,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.17320488398699121,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.2764,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.1705344953709554,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.261,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.17177496657325148,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.2698,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.1735016835675083,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.263,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.16656288080396592,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.2632,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.170779811685624,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.2605,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.17014051757438456,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.2665,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.17405868481535416,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.2779,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.17275012543882937,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.2719,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.16891644499906922,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.2731,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.17049053113330498,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.271,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.17148998420091088,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.267,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.1704396821514694,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.2598,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.17764710671701883,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.2672,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.17634164944121702,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.2725,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.17389609221229582,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.2667,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.17439433213048347,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.2697,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.16800540241501274,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.2769,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.1662390384156525,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.2716,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.16989609876173642,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.2739,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.1693184176494399,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.2653,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.1689264255632609,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.273,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.16865987556782785,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.2607,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.16536180158342062,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.2707,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.17147306530420311,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.2712,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.17135109465650336,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.2793,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.17201247625120808,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.2692,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.17409788945880564,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.2649,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.2089059623036502,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.2698,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.17469976994774525,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.2795,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.1651410605426913,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.2685,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.1717583626218605,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.2606,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.16773998256779088,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.2711,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.1637088342784956,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.2723,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.16843911813068418,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.2661,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.1776799688108947,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.2645,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.16695036202273197,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.2737,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.1646004451975376,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.2692,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.16804752976655862,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.2671,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.16896559204782088,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.2882,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.16919945873940853,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.2659,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.17806085868187446,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.2723,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.16577959671856496,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.2665,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.1726559672052673,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.2785,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.16852259591101457,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.2723,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.16898942091081387,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.2709,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.16711580624262562,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.2691,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.1687202835151915,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.2794,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.16329964124996735,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.2621,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.16796522632831146,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.2692,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.16874277915777838,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.2503,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.16941504245071534,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.2708,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.17859908261959462,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.2756,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.206607281655657,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.2838,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.16535487076944994,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.2739,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.1648671720701431,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.2753,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.16368729343167193,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.2612,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.16689263597728868,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.2603,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.16610873138383006,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.2732,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.18025677742539942,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.2605,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.20632240898387566,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.2602,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.16252781046023226,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.2671,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.1721545869474773,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.2554,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.16624987839533586,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.2708,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.17255882346392518,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.2699,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.1715905515959846,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.2702,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.17514272598185834,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.2741,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.1671829776730169,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.2658,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.16830026850872726,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.2695,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.16797801223524192,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.2644,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.1721285001988442,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.2799,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.1657787223096309,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.2669,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.1711458999577188,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.2695,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.16618820614397914,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.2725,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.16755408564747187,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.2665,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.1629914504914103,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.2697,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.1686982943035203,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.2679,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.15640307973333226,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.2751,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.20216020878325422,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.2787,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.1732307212028107,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.2647,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.16570801265984597,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.2572,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.16169882126256505,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.2706,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.1713206249010911,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.2678,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.16664429071702386,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.2704,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.168220952616711,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.2672,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.1642894259854328,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.2804,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.16693614235063664,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.2636,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.16906129683486695,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.2645,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.17009320050882024,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.2744,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.20043483028432652,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.2687,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.16318528878091088,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.2687,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.16596209551008353,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.2657,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.16731737132924654,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.2748,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.16451478103445202,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.2653,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.1682891664020474,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.2737,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.16276103739221995,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.2703,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.16708566891982657,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.2546,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.1673290511996457,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.2593,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.16617627956281364,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.2634,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.1676017278681981,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.2631,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.16647198357843734,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.2615,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.1646560693259027,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.2727,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.16614494426252163,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.2629,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.16715414311279073,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.2725,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.15722657407071872,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.2672,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.16381173231546087,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.2635,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.16615353097064506,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.2679,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.16546887661055915,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.2571,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.168943034303157,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.2655,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.16879643982586806,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.2678,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.1673430794376305,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.2621,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.16590981818102368,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.2668,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.16288991489346852,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.2608,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.16976424922306468,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.2599,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.16258775180258933,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.2692,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.16109379242942304,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.2793,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.16351430533348252,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.2732,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.16863596293428681,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.2566,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.15965292078121168,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.2671,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.16526169456014564,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.2586,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.16138559390436627,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.2694,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.16465906501790353,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.2609,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.158822939724328,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.2682,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.164158124639341,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.2579,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.1626079871418947,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.2764,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.16969028129916977,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.2737,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.1775050801946383,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.2565,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.16629923718514666,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.2696,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.16985606222422325,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.2742,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.16988179435191922,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.2691,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.16238514912380572,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.2626,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.16913297468233524,
+      "learning_rate": 0.0,
+      "loss": 0.2776,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 5.900557800872346e+18,
+      "train_loss": 0.36615794705666177,
+      "train_runtime": 54580.8511,
+      "train_samples_per_second": 2.895,
+      "train_steps_per_second": 0.023
+    }
+  ],
+  "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": 5.900557800872346e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}